Mercurial > libavcodec.hg
annotate snow.c @ 5664:1f9e8321c2fb libavcodec
unneeded and impossible during init ...
author | michael |
---|---|
date | Sun, 09 Sep 2007 12:46:25 +0000 |
parents | 4264b8bb3659 |
children | 834f359e7257 |
rev | line source |
---|---|
2138 | 1 /* |
2 * Copyright (C) 2004 Michael Niedermayer <michaelni@gmx.at> | |
3 * | |
3947
c8c591fe26f8
Change license headers to say 'FFmpeg' instead of 'this program/this library'
diego
parents:
3920
diff
changeset
|
4 * This file is part of FFmpeg. |
c8c591fe26f8
Change license headers to say 'FFmpeg' instead of 'this program/this library'
diego
parents:
3920
diff
changeset
|
5 * |
c8c591fe26f8
Change license headers to say 'FFmpeg' instead of 'this program/this library'
diego
parents:
3920
diff
changeset
|
6 * FFmpeg is free software; you can redistribute it and/or |
2138 | 7 * modify it under the terms of the GNU Lesser General Public |
8 * License as published by the Free Software Foundation; either | |
3947
c8c591fe26f8
Change license headers to say 'FFmpeg' instead of 'this program/this library'
diego
parents:
3920
diff
changeset
|
9 * version 2.1 of the License, or (at your option) any later version. |
2138 | 10 * |
3947
c8c591fe26f8
Change license headers to say 'FFmpeg' instead of 'this program/this library'
diego
parents:
3920
diff
changeset
|
11 * FFmpeg is distributed in the hope that it will be useful, |
2138 | 12 * but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
14 * Lesser General Public License for more details. | |
15 * | |
16 * You should have received a copy of the GNU Lesser General Public | |
3947
c8c591fe26f8
Change license headers to say 'FFmpeg' instead of 'this program/this library'
diego
parents:
3920
diff
changeset
|
17 * License along with FFmpeg; if not, write to the Free Software |
3036
0b546eab515d
Update licensing information: The FSF changed postal address.
diego
parents:
3035
diff
changeset
|
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
2138 | 19 */ |
20 | |
21 #include "avcodec.h" | |
22 #include "dsputil.h" | |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
23 #include "snow.h" |
2335 | 24 |
25 #include "rangecoder.h" | |
2138 | 26 |
27 #include "mpegvideo.h" | |
28 | |
29 #undef NDEBUG | |
30 #include <assert.h> | |
31 | |
32 static const int8_t quant3[256]={ | |
33 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
34 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
35 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
36 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
37 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
38 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
39 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
40 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
41 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
42 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
43 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
44 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
45 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
46 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
47 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
48 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, 0, | |
49 }; | |
50 static const int8_t quant3b[256]={ | |
51 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
52 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
53 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
54 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
55 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
56 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
57 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
58 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, | |
59 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
60 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
61 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
62 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
63 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
64 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
65 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
66 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, | |
67 }; | |
2596 | 68 static const int8_t quant3bA[256]={ |
69 0, 0, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
70 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
71 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
72 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
73 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
74 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
75 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
76 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
77 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
78 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
79 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
80 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
81 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
82 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
83 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
84 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, 1,-1, | |
85 }; | |
2138 | 86 static const int8_t quant5[256]={ |
87 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
88 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
89 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
90 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
91 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
92 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
93 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
94 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
95 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
96 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
97 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
98 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
99 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
100 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
101 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
102 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-1,-1,-1, | |
103 }; | |
104 static const int8_t quant7[256]={ | |
105 0, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
106 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, | |
107 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, | |
108 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, | |
109 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, | |
110 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, | |
111 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, | |
112 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, | |
113 -3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3, | |
114 -3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3, | |
115 -3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3, | |
116 -3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3, | |
117 -3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-3, | |
118 -3,-3,-3,-3,-3,-3,-3,-3,-3,-2,-2,-2,-2,-2,-2,-2, | |
119 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2, | |
120 -2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-2,-1,-1, | |
121 }; | |
122 static const int8_t quant9[256]={ | |
123 0, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3, | |
124 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
125 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
126 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
127 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
128 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
129 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
130 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
131 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
132 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
133 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
134 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
135 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
136 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
137 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-3,-3,-3,-3, | |
138 -3,-3,-3,-3,-3,-3,-3,-3,-3,-3,-2,-2,-2,-2,-1,-1, | |
139 }; | |
140 static const int8_t quant11[256]={ | |
141 0, 1, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, | |
142 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, | |
143 4, 4, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
144 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
145 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
146 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
147 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
148 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
149 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
150 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
151 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
152 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
153 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
154 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-4,-4, | |
155 -4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, | |
156 -4,-4,-4,-4,-4,-3,-3,-3,-3,-3,-3,-3,-2,-2,-2,-1, | |
157 }; | |
158 static const int8_t quant13[256]={ | |
159 0, 1, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, | |
160 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
161 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, | |
162 5, 5, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, | |
163 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, | |
164 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, | |
165 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, | |
166 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, | |
167 -6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6, | |
168 -6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6, | |
169 -6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6, | |
170 -6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6, | |
171 -6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-6,-5, | |
172 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
173 -5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5,-5, | |
174 -4,-4,-4,-4,-4,-4,-4,-4,-4,-3,-3,-3,-3,-2,-2,-1, | |
175 }; | |
176 | |
177 #if 0 //64*cubic | |
178 static const uint8_t obmc32[1024]={ | |
5659
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
179 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
180 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 8, 8, 8, 8, 8, 8, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
181 0, 0, 0, 4, 4, 4, 4, 8, 8, 12, 12, 12, 16, 16, 16, 16, 16, 16, 16, 16, 12, 12, 12, 8, 8, 4, 4, 4, 4, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
182 0, 0, 4, 4, 8, 8, 12, 16, 16, 20, 24, 24, 28, 28, 32, 32, 32, 32, 28, 28, 24, 24, 20, 16, 16, 12, 8, 8, 4, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
183 0, 0, 4, 8, 8, 12, 16, 24, 28, 32, 36, 40, 44, 48, 48, 48, 48, 48, 48, 44, 40, 36, 32, 28, 24, 16, 12, 8, 8, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
184 0, 4, 4, 8, 12, 20, 24, 32, 40, 44, 52, 56, 60, 64, 68, 72, 72, 68, 64, 60, 56, 52, 44, 40, 32, 24, 20, 12, 8, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
185 0, 4, 4, 12, 16, 24, 32, 40, 52, 60, 68, 76, 80, 88, 88, 92, 92, 88, 88, 80, 76, 68, 60, 52, 40, 32, 24, 16, 12, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
186 0, 4, 8, 16, 24, 32, 40, 52, 64, 76, 84, 92,100,108,112,116,116,112,108,100, 92, 84, 76, 64, 52, 40, 32, 24, 16, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
187 0, 4, 8, 16, 28, 40, 52, 64, 76, 88,100,112,124,132,136,140,140,136,132,124,112,100, 88, 76, 64, 52, 40, 28, 16, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
188 0, 4, 12, 20, 32, 44, 60, 76, 88,104,120,132,144,152,160,164,164,160,152,144,132,120,104, 88, 76, 60, 44, 32, 20, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
189 0, 4, 12, 24, 36, 48, 68, 84,100,120,136,152,164,176,180,184,184,180,176,164,152,136,120,100, 84, 68, 48, 36, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
190 0, 4, 12, 24, 40, 56, 76, 92,112,132,152,168,180,192,204,208,208,204,192,180,168,152,132,112, 92, 76, 56, 40, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
191 0, 4, 16, 28, 44, 60, 80,100,124,144,164,180,196,208,220,224,224,220,208,196,180,164,144,124,100, 80, 60, 44, 28, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
192 0, 8, 16, 28, 48, 64, 88,108,132,152,176,192,208,224,232,240,240,232,224,208,192,176,152,132,108, 88, 64, 48, 28, 16, 8, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
193 0, 4, 16, 32, 48, 68, 88,112,136,160,180,204,220,232,244,248,248,244,232,220,204,180,160,136,112, 88, 68, 48, 32, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
194 1, 8, 16, 32, 48, 72, 92,116,140,164,184,208,224,240,248,255,255,248,240,224,208,184,164,140,116, 92, 72, 48, 32, 16, 8, 1, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
195 1, 8, 16, 32, 48, 72, 92,116,140,164,184,208,224,240,248,255,255,248,240,224,208,184,164,140,116, 92, 72, 48, 32, 16, 8, 1, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
196 0, 4, 16, 32, 48, 68, 88,112,136,160,180,204,220,232,244,248,248,244,232,220,204,180,160,136,112, 88, 68, 48, 32, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
197 0, 8, 16, 28, 48, 64, 88,108,132,152,176,192,208,224,232,240,240,232,224,208,192,176,152,132,108, 88, 64, 48, 28, 16, 8, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
198 0, 4, 16, 28, 44, 60, 80,100,124,144,164,180,196,208,220,224,224,220,208,196,180,164,144,124,100, 80, 60, 44, 28, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
199 0, 4, 12, 24, 40, 56, 76, 92,112,132,152,168,180,192,204,208,208,204,192,180,168,152,132,112, 92, 76, 56, 40, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
200 0, 4, 12, 24, 36, 48, 68, 84,100,120,136,152,164,176,180,184,184,180,176,164,152,136,120,100, 84, 68, 48, 36, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
201 0, 4, 12, 20, 32, 44, 60, 76, 88,104,120,132,144,152,160,164,164,160,152,144,132,120,104, 88, 76, 60, 44, 32, 20, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
202 0, 4, 8, 16, 28, 40, 52, 64, 76, 88,100,112,124,132,136,140,140,136,132,124,112,100, 88, 76, 64, 52, 40, 28, 16, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
203 0, 4, 8, 16, 24, 32, 40, 52, 64, 76, 84, 92,100,108,112,116,116,112,108,100, 92, 84, 76, 64, 52, 40, 32, 24, 16, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
204 0, 4, 4, 12, 16, 24, 32, 40, 52, 60, 68, 76, 80, 88, 88, 92, 92, 88, 88, 80, 76, 68, 60, 52, 40, 32, 24, 16, 12, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
205 0, 4, 4, 8, 12, 20, 24, 32, 40, 44, 52, 56, 60, 64, 68, 72, 72, 68, 64, 60, 56, 52, 44, 40, 32, 24, 20, 12, 8, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
206 0, 0, 4, 8, 8, 12, 16, 24, 28, 32, 36, 40, 44, 48, 48, 48, 48, 48, 48, 44, 40, 36, 32, 28, 24, 16, 12, 8, 8, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
207 0, 0, 4, 4, 8, 8, 12, 16, 16, 20, 24, 24, 28, 28, 32, 32, 32, 32, 28, 28, 24, 24, 20, 16, 16, 12, 8, 8, 4, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
208 0, 0, 0, 4, 4, 4, 4, 8, 8, 12, 12, 12, 16, 16, 16, 16, 16, 16, 16, 16, 12, 12, 12, 8, 8, 4, 4, 4, 4, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
209 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 8, 8, 8, 8, 8, 8, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
210 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
2138 | 211 //error:0.000022 |
212 }; | |
213 static const uint8_t obmc16[256]={ | |
5659
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
214 0, 0, 0, 0, 0, 0, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
215 0, 4, 4, 8, 16, 20, 20, 24, 24, 20, 20, 16, 8, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
216 0, 4, 16, 24, 36, 44, 52, 60, 60, 52, 44, 36, 24, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
217 0, 8, 24, 44, 60, 80, 96,104,104, 96, 80, 60, 44, 24, 8, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
218 0, 16, 36, 60, 92,116,136,152,152,136,116, 92, 60, 36, 16, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
219 0, 20, 44, 80,116,152,180,196,196,180,152,116, 80, 44, 20, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
220 4, 20, 52, 96,136,180,212,228,228,212,180,136, 96, 52, 20, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
221 4, 24, 60,104,152,196,228,248,248,228,196,152,104, 60, 24, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
222 4, 24, 60,104,152,196,228,248,248,228,196,152,104, 60, 24, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
223 4, 20, 52, 96,136,180,212,228,228,212,180,136, 96, 52, 20, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
224 0, 20, 44, 80,116,152,180,196,196,180,152,116, 80, 44, 20, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
225 0, 16, 36, 60, 92,116,136,152,152,136,116, 92, 60, 36, 16, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
226 0, 8, 24, 44, 60, 80, 96,104,104, 96, 80, 60, 44, 24, 8, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
227 0, 4, 16, 24, 36, 44, 52, 60, 60, 52, 44, 36, 24, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
228 0, 4, 4, 8, 16, 20, 20, 24, 24, 20, 20, 16, 8, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
229 0, 0, 0, 0, 0, 0, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, |
2138 | 230 //error:0.000033 |
231 }; | |
232 #elif 1 // 64*linear | |
233 static const uint8_t obmc32[1024]={ | |
3206 | 234 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 8, 8, 8, 8, 8, 8, 8, 8, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, |
235 0, 4, 4, 4, 8, 8, 8, 12, 12, 16, 16, 16, 20, 20, 20, 24, 24, 20, 20, 20, 16, 16, 16, 12, 12, 8, 8, 8, 4, 4, 4, 0, | |
236 0, 4, 8, 8, 12, 12, 16, 20, 20, 24, 28, 28, 32, 32, 36, 40, 40, 36, 32, 32, 28, 28, 24, 20, 20, 16, 12, 12, 8, 8, 4, 0, | |
237 0, 4, 8, 12, 16, 20, 24, 28, 28, 32, 36, 40, 44, 48, 52, 56, 56, 52, 48, 44, 40, 36, 32, 28, 28, 24, 20, 16, 12, 8, 4, 0, | |
238 4, 8, 12, 16, 20, 24, 28, 32, 40, 44, 48, 52, 56, 60, 64, 68, 68, 64, 60, 56, 52, 48, 44, 40, 32, 28, 24, 20, 16, 12, 8, 4, | |
239 4, 8, 12, 20, 24, 32, 36, 40, 48, 52, 56, 64, 68, 76, 80, 84, 84, 80, 76, 68, 64, 56, 52, 48, 40, 36, 32, 24, 20, 12, 8, 4, | |
240 4, 8, 16, 24, 28, 36, 44, 48, 56, 60, 68, 76, 80, 88, 96,100,100, 96, 88, 80, 76, 68, 60, 56, 48, 44, 36, 28, 24, 16, 8, 4, | |
241 4, 12, 20, 28, 32, 40, 48, 56, 64, 72, 80, 88, 92,100,108,116,116,108,100, 92, 88, 80, 72, 64, 56, 48, 40, 32, 28, 20, 12, 4, | |
242 4, 12, 20, 28, 40, 48, 56, 64, 72, 80, 88, 96,108,116,124,132,132,124,116,108, 96, 88, 80, 72, 64, 56, 48, 40, 28, 20, 12, 4, | |
243 4, 16, 24, 32, 44, 52, 60, 72, 80, 92,100,108,120,128,136,148,148,136,128,120,108,100, 92, 80, 72, 60, 52, 44, 32, 24, 16, 4, | |
244 4, 16, 28, 36, 48, 56, 68, 80, 88,100,112,120,132,140,152,164,164,152,140,132,120,112,100, 88, 80, 68, 56, 48, 36, 28, 16, 4, | |
245 4, 16, 28, 40, 52, 64, 76, 88, 96,108,120,132,144,156,168,180,180,168,156,144,132,120,108, 96, 88, 76, 64, 52, 40, 28, 16, 4, | |
246 8, 20, 32, 44, 56, 68, 80, 92,108,120,132,144,156,168,180,192,192,180,168,156,144,132,120,108, 92, 80, 68, 56, 44, 32, 20, 8, | |
247 8, 20, 32, 48, 60, 76, 88,100,116,128,140,156,168,184,196,208,208,196,184,168,156,140,128,116,100, 88, 76, 60, 48, 32, 20, 8, | |
248 8, 20, 36, 52, 64, 80, 96,108,124,136,152,168,180,196,212,224,224,212,196,180,168,152,136,124,108, 96, 80, 64, 52, 36, 20, 8, | |
249 8, 24, 40, 56, 68, 84,100,116,132,148,164,180,192,208,224,240,240,224,208,192,180,164,148,132,116,100, 84, 68, 56, 40, 24, 8, | |
250 8, 24, 40, 56, 68, 84,100,116,132,148,164,180,192,208,224,240,240,224,208,192,180,164,148,132,116,100, 84, 68, 56, 40, 24, 8, | |
251 8, 20, 36, 52, 64, 80, 96,108,124,136,152,168,180,196,212,224,224,212,196,180,168,152,136,124,108, 96, 80, 64, 52, 36, 20, 8, | |
252 8, 20, 32, 48, 60, 76, 88,100,116,128,140,156,168,184,196,208,208,196,184,168,156,140,128,116,100, 88, 76, 60, 48, 32, 20, 8, | |
253 8, 20, 32, 44, 56, 68, 80, 92,108,120,132,144,156,168,180,192,192,180,168,156,144,132,120,108, 92, 80, 68, 56, 44, 32, 20, 8, | |
254 4, 16, 28, 40, 52, 64, 76, 88, 96,108,120,132,144,156,168,180,180,168,156,144,132,120,108, 96, 88, 76, 64, 52, 40, 28, 16, 4, | |
255 4, 16, 28, 36, 48, 56, 68, 80, 88,100,112,120,132,140,152,164,164,152,140,132,120,112,100, 88, 80, 68, 56, 48, 36, 28, 16, 4, | |
256 4, 16, 24, 32, 44, 52, 60, 72, 80, 92,100,108,120,128,136,148,148,136,128,120,108,100, 92, 80, 72, 60, 52, 44, 32, 24, 16, 4, | |
257 4, 12, 20, 28, 40, 48, 56, 64, 72, 80, 88, 96,108,116,124,132,132,124,116,108, 96, 88, 80, 72, 64, 56, 48, 40, 28, 20, 12, 4, | |
258 4, 12, 20, 28, 32, 40, 48, 56, 64, 72, 80, 88, 92,100,108,116,116,108,100, 92, 88, 80, 72, 64, 56, 48, 40, 32, 28, 20, 12, 4, | |
259 4, 8, 16, 24, 28, 36, 44, 48, 56, 60, 68, 76, 80, 88, 96,100,100, 96, 88, 80, 76, 68, 60, 56, 48, 44, 36, 28, 24, 16, 8, 4, | |
260 4, 8, 12, 20, 24, 32, 36, 40, 48, 52, 56, 64, 68, 76, 80, 84, 84, 80, 76, 68, 64, 56, 52, 48, 40, 36, 32, 24, 20, 12, 8, 4, | |
261 4, 8, 12, 16, 20, 24, 28, 32, 40, 44, 48, 52, 56, 60, 64, 68, 68, 64, 60, 56, 52, 48, 44, 40, 32, 28, 24, 20, 16, 12, 8, 4, | |
262 0, 4, 8, 12, 16, 20, 24, 28, 28, 32, 36, 40, 44, 48, 52, 56, 56, 52, 48, 44, 40, 36, 32, 28, 28, 24, 20, 16, 12, 8, 4, 0, | |
263 0, 4, 8, 8, 12, 12, 16, 20, 20, 24, 28, 28, 32, 32, 36, 40, 40, 36, 32, 32, 28, 28, 24, 20, 20, 16, 12, 12, 8, 8, 4, 0, | |
264 0, 4, 4, 4, 8, 8, 8, 12, 12, 16, 16, 16, 20, 20, 20, 24, 24, 20, 20, 20, 16, 16, 16, 12, 12, 8, 8, 8, 4, 4, 4, 0, | |
265 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 8, 8, 8, 8, 8, 8, 8, 8, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, | |
2138 | 266 //error:0.000020 |
267 }; | |
268 static const uint8_t obmc16[256]={ | |
3206 | 269 0, 4, 4, 8, 8, 12, 12, 16, 16, 12, 12, 8, 8, 4, 4, 0, |
270 4, 8, 16, 20, 28, 32, 40, 44, 44, 40, 32, 28, 20, 16, 8, 4, | |
271 4, 16, 24, 36, 44, 56, 64, 76, 76, 64, 56, 44, 36, 24, 16, 4, | |
272 8, 20, 36, 48, 64, 76, 92,104,104, 92, 76, 64, 48, 36, 20, 8, | |
273 8, 28, 44, 64, 80,100,116,136,136,116,100, 80, 64, 44, 28, 8, | |
274 12, 32, 56, 76,100,120,144,164,164,144,120,100, 76, 56, 32, 12, | |
275 12, 40, 64, 92,116,144,168,196,196,168,144,116, 92, 64, 40, 12, | |
276 16, 44, 76,104,136,164,196,224,224,196,164,136,104, 76, 44, 16, | |
277 16, 44, 76,104,136,164,196,224,224,196,164,136,104, 76, 44, 16, | |
278 12, 40, 64, 92,116,144,168,196,196,168,144,116, 92, 64, 40, 12, | |
279 12, 32, 56, 76,100,120,144,164,164,144,120,100, 76, 56, 32, 12, | |
280 8, 28, 44, 64, 80,100,116,136,136,116,100, 80, 64, 44, 28, 8, | |
281 8, 20, 36, 48, 64, 76, 92,104,104, 92, 76, 64, 48, 36, 20, 8, | |
282 4, 16, 24, 36, 44, 56, 64, 76, 76, 64, 56, 44, 36, 24, 16, 4, | |
283 4, 8, 16, 20, 28, 32, 40, 44, 44, 40, 32, 28, 20, 16, 8, 4, | |
284 0, 4, 4, 8, 8, 12, 12, 16, 16, 12, 12, 8, 8, 4, 4, 0, | |
2138 | 285 //error:0.000015 |
286 }; | |
287 #else //64*cos | |
288 static const uint8_t obmc32[1024]={ | |
5659
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
289 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
290 0, 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 8, 4, 4, 8, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
291 0, 0, 0, 4, 4, 4, 4, 8, 8, 12, 12, 12, 12, 16, 16, 16, 16, 16, 16, 12, 12, 12, 12, 8, 8, 4, 4, 4, 4, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
292 0, 0, 4, 4, 4, 8, 8, 12, 16, 20, 20, 24, 28, 28, 28, 28, 28, 28, 28, 28, 24, 20, 20, 16, 12, 8, 8, 4, 4, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
293 0, 0, 4, 4, 8, 12, 16, 20, 24, 28, 36, 40, 44, 44, 48, 48, 48, 48, 44, 44, 40, 36, 28, 24, 20, 16, 12, 8, 4, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
294 0, 0, 4, 8, 12, 20, 24, 32, 36, 44, 48, 56, 60, 64, 68, 68, 68, 68, 64, 60, 56, 48, 44, 36, 32, 24, 20, 12, 8, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
295 0, 4, 4, 8, 16, 24, 32, 40, 48, 60, 68, 76, 80, 84, 88, 92, 92, 88, 84, 80, 76, 68, 60, 48, 40, 32, 24, 16, 8, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
296 0, 4, 8, 12, 20, 32, 40, 52, 64, 76, 84, 96,104,108,112,116,116,112,108,104, 96, 84, 76, 64, 52, 40, 32, 20, 12, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
297 0, 4, 8, 16, 24, 36, 48, 64, 76, 92,104,116,124,132,136,140,140,136,132,124,116,104, 92, 76, 64, 48, 36, 24, 16, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
298 0, 4, 12, 20, 28, 44, 60, 76, 92,104,120,136,148,156,160,164,164,160,156,148,136,120,104, 92, 76, 60, 44, 28, 20, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
299 0, 4, 12, 20, 36, 48, 68, 84,104,120,140,152,168,176,184,188,188,184,176,168,152,140,120,104, 84, 68, 48, 36, 20, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
300 0, 4, 12, 24, 36, 56, 76, 96,116,136,152,172,184,196,204,208,208,204,196,184,172,152,136,116, 96, 76, 56, 36, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
301 0, 4, 12, 24, 44, 60, 80,104,124,148,168,184,200,212,224,228,228,224,212,200,184,168,148,124,104, 80, 60, 44, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
302 0, 4, 12, 28, 44, 64, 84,108,132,156,176,196,212,228,236,240,240,236,228,212,196,176,156,132,108, 84, 64, 44, 28, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
303 0, 4, 16, 28, 48, 68, 88,112,136,160,184,204,224,236,244,252,252,244,236,224,204,184,160,136,112, 88, 68, 48, 28, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
304 1, 4, 16, 28, 48, 68, 92,116,140,164,188,208,228,240,252,255,255,252,240,228,208,188,164,140,116, 92, 68, 48, 28, 16, 4, 1, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
305 1, 4, 16, 28, 48, 68, 92,116,140,164,188,208,228,240,252,255,255,252,240,228,208,188,164,140,116, 92, 68, 48, 28, 16, 4, 1, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
306 0, 4, 16, 28, 48, 68, 88,112,136,160,184,204,224,236,244,252,252,244,236,224,204,184,160,136,112, 88, 68, 48, 28, 16, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
307 0, 4, 12, 28, 44, 64, 84,108,132,156,176,196,212,228,236,240,240,236,228,212,196,176,156,132,108, 84, 64, 44, 28, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
308 0, 4, 12, 24, 44, 60, 80,104,124,148,168,184,200,212,224,228,228,224,212,200,184,168,148,124,104, 80, 60, 44, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
309 0, 4, 12, 24, 36, 56, 76, 96,116,136,152,172,184,196,204,208,208,204,196,184,172,152,136,116, 96, 76, 56, 36, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
310 0, 4, 12, 20, 36, 48, 68, 84,104,120,140,152,168,176,184,188,188,184,176,168,152,140,120,104, 84, 68, 48, 36, 20, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
311 0, 4, 12, 20, 28, 44, 60, 76, 92,104,120,136,148,156,160,164,164,160,156,148,136,120,104, 92, 76, 60, 44, 28, 20, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
312 0, 4, 8, 16, 24, 36, 48, 64, 76, 92,104,116,124,132,136,140,140,136,132,124,116,104, 92, 76, 64, 48, 36, 24, 16, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
313 0, 4, 8, 12, 20, 32, 40, 52, 64, 76, 84, 96,104,108,112,116,116,112,108,104, 96, 84, 76, 64, 52, 40, 32, 20, 12, 8, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
314 0, 4, 4, 8, 16, 24, 32, 40, 48, 60, 68, 76, 80, 84, 88, 92, 92, 88, 84, 80, 76, 68, 60, 48, 40, 32, 24, 16, 8, 4, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
315 0, 0, 4, 8, 12, 20, 24, 32, 36, 44, 48, 56, 60, 64, 68, 68, 68, 68, 64, 60, 56, 48, 44, 36, 32, 24, 20, 12, 8, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
316 0, 0, 4, 4, 8, 12, 16, 20, 24, 28, 36, 40, 44, 44, 48, 48, 48, 48, 44, 44, 40, 36, 28, 24, 20, 16, 12, 8, 4, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
317 0, 0, 4, 4, 4, 8, 8, 12, 16, 20, 20, 24, 28, 28, 28, 28, 28, 28, 28, 28, 24, 20, 20, 16, 12, 8, 8, 4, 4, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
318 0, 0, 0, 4, 4, 4, 4, 8, 8, 12, 12, 12, 12, 16, 16, 16, 16, 16, 16, 12, 12, 12, 12, 8, 8, 4, 4, 4, 4, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
319 0, 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 8, 4, 4, 8, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
320 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
2138 | 321 //error:0.000022 |
322 }; | |
323 static const uint8_t obmc16[256]={ | |
5659
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
324 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
325 0, 0, 4, 8, 12, 16, 20, 20, 20, 20, 16, 12, 8, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
326 0, 4, 12, 24, 32, 44, 52, 56, 56, 52, 44, 32, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
327 0, 8, 24, 40, 60, 80, 96,104,104, 96, 80, 60, 40, 24, 8, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
328 0, 12, 32, 64, 92,120,140,152,152,140,120, 92, 64, 32, 12, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
329 4, 16, 44, 80,120,156,184,196,196,184,156,120, 80, 44, 16, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
330 4, 20, 52, 96,140,184,216,232,232,216,184,140, 96, 52, 20, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
331 0, 20, 56,104,152,196,232,252,252,232,196,152,104, 56, 20, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
332 0, 20, 56,104,152,196,232,252,252,232,196,152,104, 56, 20, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
333 4, 20, 52, 96,140,184,216,232,232,216,184,140, 96, 52, 20, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
334 4, 16, 44, 80,120,156,184,196,196,184,156,120, 80, 44, 16, 4, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
335 0, 12, 32, 64, 92,120,140,152,152,140,120, 92, 64, 32, 12, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
336 0, 8, 24, 40, 60, 80, 96,104,104, 96, 80, 60, 40, 24, 8, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
337 0, 4, 12, 24, 32, 44, 52, 56, 56, 52, 44, 32, 24, 12, 4, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
338 0, 0, 4, 8, 12, 16, 20, 20, 20, 20, 16, 12, 8, 4, 0, 0, |
ff44e77914ca
scaling obmc tables under #if 0 also by 4 (yes they where forgotten apparently)
michael
parents:
5652
diff
changeset
|
339 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, |
2138 | 340 //error:0.000022 |
341 }; | |
342 #endif | |
343 | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
344 //linear *64 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
345 static const uint8_t obmc8[64]={ |
3206 | 346 4, 12, 20, 28, 28, 20, 12, 4, |
347 12, 36, 60, 84, 84, 60, 36, 12, | |
348 20, 60,100,140,140,100, 60, 20, | |
349 28, 84,140,196,196,140, 84, 28, | |
350 28, 84,140,196,196,140, 84, 28, | |
351 20, 60,100,140,140,100, 60, 20, | |
352 12, 36, 60, 84, 84, 60, 36, 12, | |
353 4, 12, 20, 28, 28, 20, 12, 4, | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
354 //error:0.000000 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
355 }; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
356 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
357 //linear *64 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
358 static const uint8_t obmc4[16]={ |
3206 | 359 16, 48, 48, 16, |
360 48,144,144, 48, | |
361 48,144,144, 48, | |
362 16, 48, 48, 16, | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
363 //error:0.000000 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
364 }; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
365 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
366 static const uint8_t *obmc_tab[4]={ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
367 obmc32, obmc16, obmc8, obmc4 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
368 }; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
369 |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
370 static int scale_mv_ref[MAX_REF_FRAMES][MAX_REF_FRAMES]; |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
371 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
372 typedef struct BlockNode{ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
373 int16_t mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
374 int16_t my; |
3314 | 375 uint8_t ref; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
376 uint8_t color[3]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
377 uint8_t type; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
378 //#define TYPE_SPLIT 1 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
379 #define BLOCK_INTRA 1 |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
380 #define BLOCK_OPT 2 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
381 //#define TYPE_NOCOLOR 4 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
382 uint8_t level; //FIXME merge into type? |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
383 }BlockNode; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
384 |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
385 static const BlockNode null_block= { //FIXME add border maybe |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
386 .color= {128,128,128}, |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
387 .mx= 0, |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
388 .my= 0, |
3314 | 389 .ref= 0, |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
390 .type= 0, |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
391 .level= 0, |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
392 }; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
393 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
394 #define LOG2_MB_SIZE 4 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
395 #define MB_SIZE (1<<LOG2_MB_SIZE) |
5575 | 396 #define ENCODER_EXTRA_BITS 4 |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
397 #define HTAPS_MAX 8 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
398 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
399 typedef struct x_and_coeff{ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
400 int16_t x; |
2596 | 401 uint16_t coeff; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
402 } x_and_coeff; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
403 |
2138 | 404 typedef struct SubBand{ |
405 int level; | |
406 int stride; | |
407 int width; | |
408 int height; | |
409 int qlog; ///< log(qscale)/log[2^(1/6)] | |
410 DWTELEM *buf; | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
411 IDWTELEM *ibuf; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
412 int buf_x_offset; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
413 int buf_y_offset; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
414 int stride_line; ///< Stride measured in lines, not pixels. |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
415 x_and_coeff * x_coeff; |
2138 | 416 struct SubBand *parent; |
417 uint8_t state[/*7*2*/ 7 + 512][32]; | |
418 }SubBand; | |
419 | |
420 typedef struct Plane{ | |
421 int width; | |
422 int height; | |
423 SubBand band[MAX_DECOMPOSITIONS][4]; | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
424 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
425 int htaps; |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
426 int8_t hcoeff[HTAPS_MAX/2]; |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
427 int diag_mc; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
428 int fast_mc; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
429 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
430 int last_htaps; |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
431 int8_t last_hcoeff[HTAPS_MAX/2]; |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
432 int last_diag_mc; |
2138 | 433 }Plane; |
434 | |
435 typedef struct SnowContext{ | |
4588
fc155ff94878
cosmetics: Fix another common typo, dependAnt --> dependEnt.
diego
parents:
4494
diff
changeset
|
436 // MpegEncContext m; // needed for motion estimation, should not be used for anything else, the idea is to make the motion estimation eventually independent of MpegEncContext, so this will be removed then (FIXME/XXX) |
2138 | 437 |
438 AVCodecContext *avctx; | |
2335 | 439 RangeCoder c; |
2138 | 440 DSPContext dsp; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
441 AVFrame new_picture; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
442 AVFrame input_picture; ///< new_picture with the internal linesizes |
2138 | 443 AVFrame current_picture; |
3314 | 444 AVFrame last_picture[MAX_REF_FRAMES]; |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
445 uint8_t *halfpel_plane[MAX_REF_FRAMES][4][4]; |
2138 | 446 AVFrame mconly_picture; |
447 // uint8_t q_context[16]; | |
448 uint8_t header_state[32]; | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
449 uint8_t block_state[128 + 32*128]; |
2138 | 450 int keyframe; |
2199 | 451 int always_reset; |
2138 | 452 int version; |
453 int spatial_decomposition_type; | |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
454 int last_spatial_decomposition_type; |
2138 | 455 int temporal_decomposition_type; |
456 int spatial_decomposition_count; | |
457 int temporal_decomposition_count; | |
3314 | 458 int max_ref_frames; |
459 int ref_frames; | |
460 int16_t (*ref_mvs[MAX_REF_FRAMES])[2]; | |
461 uint32_t *ref_scores[MAX_REF_FRAMES]; | |
2138 | 462 DWTELEM *spatial_dwt_buffer; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
463 IDWTELEM *spatial_idwt_buffer; |
2138 | 464 int colorspace_type; |
465 int chroma_h_shift; | |
466 int chroma_v_shift; | |
467 int spatial_scalability; | |
468 int qlog; | |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
469 int last_qlog; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
470 int lambda; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
471 int lambda2; |
3313 | 472 int pass1_rc; |
2138 | 473 int mv_scale; |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
474 int last_mv_scale; |
2138 | 475 int qbias; |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
476 int last_qbias; |
2138 | 477 #define QBIAS_SHIFT 3 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
478 int b_width; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
479 int b_height; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
480 int block_max_depth; |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
481 int last_block_max_depth; |
2138 | 482 Plane plane[MAX_PLANES]; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
483 BlockNode *block; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
484 #define ME_CACHE_SIZE 1024 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
485 int me_cache[ME_CACHE_SIZE]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
486 int me_cache_generation; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
487 slice_buffer sb; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
488 |
4588
fc155ff94878
cosmetics: Fix another common typo, dependAnt --> dependEnt.
diego
parents:
4494
diff
changeset
|
489 MpegEncContext m; // needed for motion estimation, should not be used for anything else, the idea is to make the motion estimation eventually independent of MpegEncContext, so this will be removed then (FIXME/XXX) |
2138 | 490 }SnowContext; |
491 | |
2562 | 492 typedef struct { |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
493 IDWTELEM *b0; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
494 IDWTELEM *b1; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
495 IDWTELEM *b2; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
496 IDWTELEM *b3; |
2562 | 497 int y; |
498 } dwt_compose_t; | |
499 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
500 #define slice_buffer_get_line(slice_buf, line_num) ((slice_buf)->line[line_num] ? (slice_buf)->line[line_num] : slice_buffer_load_line((slice_buf), (line_num))) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
501 //#define slice_buffer_get_line(slice_buf, line_num) (slice_buffer_load_line((slice_buf), (line_num))) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
502 |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
503 static void iterative_me(SnowContext *s); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
504 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
505 static void slice_buffer_init(slice_buffer * buf, int line_count, int max_allocated_lines, int line_width, IDWTELEM * base_buffer) |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
506 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
507 int i; |
2967 | 508 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
509 buf->base_buffer = base_buffer; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
510 buf->line_count = line_count; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
511 buf->line_width = line_width; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
512 buf->data_count = max_allocated_lines; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
513 buf->line = av_mallocz (sizeof(IDWTELEM *) * line_count); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
514 buf->data_stack = av_malloc (sizeof(IDWTELEM *) * max_allocated_lines); |
2967 | 515 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
516 for (i = 0; i < max_allocated_lines; i++) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
517 { |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
518 buf->data_stack[i] = av_malloc (sizeof(IDWTELEM) * line_width); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
519 } |
2967 | 520 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
521 buf->data_stack_top = max_allocated_lines - 1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
522 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
523 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
524 static IDWTELEM * slice_buffer_load_line(slice_buffer * buf, int line) |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
525 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
526 int offset; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
527 IDWTELEM * buffer; |
2967 | 528 |
529 // av_log(NULL, AV_LOG_DEBUG, "Cache hit: %d\n", line); | |
530 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
531 assert(buf->data_stack_top >= 0); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
532 // assert(!buf->line[line]); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
533 if (buf->line[line]) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
534 return buf->line[line]; |
2967 | 535 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
536 offset = buf->line_width * line; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
537 buffer = buf->data_stack[buf->data_stack_top]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
538 buf->data_stack_top--; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
539 buf->line[line] = buffer; |
2967 | 540 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
541 // av_log(NULL, AV_LOG_DEBUG, "slice_buffer_load_line: line: %d remaining: %d\n", line, buf->data_stack_top + 1); |
2967 | 542 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
543 return buffer; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
544 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
545 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
546 static void slice_buffer_release(slice_buffer * buf, int line) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
547 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
548 int offset; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
549 IDWTELEM * buffer; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
550 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
551 assert(line >= 0 && line < buf->line_count); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
552 assert(buf->line[line]); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
553 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
554 offset = buf->line_width * line; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
555 buffer = buf->line[line]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
556 buf->data_stack_top++; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
557 buf->data_stack[buf->data_stack_top] = buffer; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
558 buf->line[line] = NULL; |
2967 | 559 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
560 // av_log(NULL, AV_LOG_DEBUG, "slice_buffer_release: line: %d remaining: %d\n", line, buf->data_stack_top + 1); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
561 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
562 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
563 static void slice_buffer_flush(slice_buffer * buf) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
564 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
565 int i; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
566 for (i = 0; i < buf->line_count; i++) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
567 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
568 if (buf->line[i]) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
569 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
570 // av_log(NULL, AV_LOG_DEBUG, "slice_buffer_flush: line: %d \n", i); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
571 slice_buffer_release(buf, i); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
572 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
573 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
574 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
575 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
576 static void slice_buffer_destroy(slice_buffer * buf) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
577 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
578 int i; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
579 slice_buffer_flush(buf); |
2967 | 580 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
581 for (i = buf->data_count - 1; i >= 0; i--) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
582 { |
3190 | 583 av_freep(&buf->data_stack[i]); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
584 } |
3190 | 585 av_freep(&buf->data_stack); |
586 av_freep(&buf->line); | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
587 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
588 |
2979 | 589 #ifdef __sgi |
2368
a7ac68734a91
fix for build on IRIX by (Michel Bardiaux {mbardiaux peaktime be})
michael
parents:
2335
diff
changeset
|
590 // Avoid a name clash on SGI IRIX |
2979 | 591 #undef qexp |
2368
a7ac68734a91
fix for build on IRIX by (Michel Bardiaux {mbardiaux peaktime be})
michael
parents:
2335
diff
changeset
|
592 #endif |
2246 | 593 #define QEXPSHIFT (7-FRAC_BITS+8) //FIXME try to change this to 0 |
2600 | 594 static uint8_t qexp[QROOT]; |
2138 | 595 |
596 static inline int mirror(int v, int m){ | |
2998 | 597 while((unsigned)v > (unsigned)m){ |
598 v=-v; | |
599 if(v<0) v+= 2*m; | |
600 } | |
601 return v; | |
2138 | 602 } |
603 | |
2335 | 604 static inline void put_symbol(RangeCoder *c, uint8_t *state, int v, int is_signed){ |
2138 | 605 int i; |
606 | |
607 if(v){ | |
4001 | 608 const int a= FFABS(v); |
2138 | 609 const int e= av_log2(a); |
610 #if 1 | |
2967 | 611 const int el= FFMIN(e, 10); |
2335 | 612 put_rac(c, state+0, 0); |
2138 | 613 |
614 for(i=0; i<el; i++){ | |
2335 | 615 put_rac(c, state+1+i, 1); //1..10 |
2138 | 616 } |
617 for(; i<e; i++){ | |
2335 | 618 put_rac(c, state+1+9, 1); //1..10 |
2138 | 619 } |
2335 | 620 put_rac(c, state+1+FFMIN(i,9), 0); |
2138 | 621 |
622 for(i=e-1; i>=el; i--){ | |
2335 | 623 put_rac(c, state+22+9, (a>>i)&1); //22..31 |
2138 | 624 } |
625 for(; i>=0; i--){ | |
2335 | 626 put_rac(c, state+22+i, (a>>i)&1); //22..31 |
2138 | 627 } |
628 | |
629 if(is_signed) | |
2335 | 630 put_rac(c, state+11 + el, v < 0); //11..21 |
2138 | 631 #else |
2967 | 632 |
2335 | 633 put_rac(c, state+0, 0); |
2138 | 634 if(e<=9){ |
635 for(i=0; i<e; i++){ | |
2335 | 636 put_rac(c, state+1+i, 1); //1..10 |
2138 | 637 } |
2335 | 638 put_rac(c, state+1+i, 0); |
2138 | 639 |
640 for(i=e-1; i>=0; i--){ | |
2335 | 641 put_rac(c, state+22+i, (a>>i)&1); //22..31 |
2138 | 642 } |
643 | |
644 if(is_signed) | |
2335 | 645 put_rac(c, state+11 + e, v < 0); //11..21 |
2138 | 646 }else{ |
647 for(i=0; i<e; i++){ | |
2335 | 648 put_rac(c, state+1+FFMIN(i,9), 1); //1..10 |
2138 | 649 } |
2335 | 650 put_rac(c, state+1+FFMIN(i,9), 0); |
2138 | 651 |
652 for(i=e-1; i>=0; i--){ | |
2335 | 653 put_rac(c, state+22+FFMIN(i,9), (a>>i)&1); //22..31 |
2138 | 654 } |
655 | |
656 if(is_signed) | |
2335 | 657 put_rac(c, state+11 + FFMIN(e,10), v < 0); //11..21 |
2138 | 658 } |
659 #endif | |
660 }else{ | |
2335 | 661 put_rac(c, state+0, 1); |
2138 | 662 } |
663 } | |
664 | |
2335 | 665 static inline int get_symbol(RangeCoder *c, uint8_t *state, int is_signed){ |
666 if(get_rac(c, state+0)) | |
2138 | 667 return 0; |
668 else{ | |
2240
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
669 int i, e, a; |
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
670 e= 0; |
2335 | 671 while(get_rac(c, state+1 + FFMIN(e,9))){ //1..10 |
2240
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
672 e++; |
2138 | 673 } |
674 | |
2240
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
675 a= 1; |
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
676 for(i=e-1; i>=0; i--){ |
2335 | 677 a += a + get_rac(c, state+22 + FFMIN(i,9)); //22..31 |
2240
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
678 } |
c46fed9b7575
simplify getsymbol patch by (Loren Merritt <lorenm at u dot washington dot edu>)
michael
parents:
2232
diff
changeset
|
679 |
2335 | 680 if(is_signed && get_rac(c, state+11 + FFMIN(e,10))) //11..21 |
2138 | 681 return -a; |
682 else | |
683 return a; | |
684 } | |
685 } | |
686 | |
2335 | 687 static inline void put_symbol2(RangeCoder *c, uint8_t *state, int v, int log2){ |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
688 int i; |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
689 int r= log2>=0 ? 1<<log2 : 1; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
690 |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
691 assert(v>=0); |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
692 assert(log2>=-4); |
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
693 |
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
694 while(v >= r){ |
2335 | 695 put_rac(c, state+4+log2, 1); |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
696 v -= r; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
697 log2++; |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
698 if(log2>0) r+=r; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
699 } |
2335 | 700 put_rac(c, state+4+log2, 0); |
2967 | 701 |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
702 for(i=log2-1; i>=0; i--){ |
2335 | 703 put_rac(c, state+31-i, (v>>i)&1); |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
704 } |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
705 } |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
706 |
2335 | 707 static inline int get_symbol2(RangeCoder *c, uint8_t *state, int log2){ |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
708 int i; |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
709 int r= log2>=0 ? 1<<log2 : 1; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
710 int v=0; |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
711 |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
712 assert(log2>=-4); |
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
713 |
2335 | 714 while(get_rac(c, state+4+log2)){ |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
715 v+= r; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
716 log2++; |
2159
7f42295c1517
improved magnitude coding, 0.2% lower bitrate (foreman@352x288 qscale 1 and 8)
michael
parents:
2156
diff
changeset
|
717 if(log2>0) r+=r; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
718 } |
2967 | 719 |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
720 for(i=log2-1; i>=0; i--){ |
2335 | 721 v+= get_rac(c, state+31-i)<<i; |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
722 } |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
723 |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
724 return v; |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
725 } |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
726 |
5627 | 727 static av_always_inline void |
728 lift(DWTELEM *dst, DWTELEM *src, DWTELEM *ref, | |
729 int dst_step, int src_step, int ref_step, | |
730 int width, int mul, int add, int shift, | |
731 int highpass, int inverse){ | |
2138 | 732 const int mirror_left= !highpass; |
733 const int mirror_right= (width&1) ^ highpass; | |
734 const int w= (width>>1) - 1 + (highpass & width); | |
735 int i; | |
736 | |
737 #define LIFT(src, ref, inv) ((src) + ((inv) ? - (ref) : + (ref))) | |
738 if(mirror_left){ | |
739 dst[0] = LIFT(src[0], ((mul*2*ref[0]+add)>>shift), inverse); | |
740 dst += dst_step; | |
741 src += src_step; | |
742 } | |
2967 | 743 |
2138 | 744 for(i=0; i<w; i++){ |
5627 | 745 dst[i*dst_step] = |
746 LIFT(src[i*src_step], | |
747 ((mul*(ref[i*ref_step] + ref[(i+1)*ref_step])+add)>>shift), | |
748 inverse); | |
2138 | 749 } |
2967 | 750 |
2138 | 751 if(mirror_right){ |
5627 | 752 dst[w*dst_step] = |
753 LIFT(src[w*src_step], | |
754 ((mul*2*ref[w*ref_step]+add)>>shift), | |
755 inverse); | |
2138 | 756 } |
757 } | |
758 | |
5627 | 759 static av_always_inline void |
760 inv_lift(IDWTELEM *dst, IDWTELEM *src, IDWTELEM *ref, | |
761 int dst_step, int src_step, int ref_step, | |
762 int width, int mul, int add, int shift, | |
763 int highpass, int inverse){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
764 const int mirror_left= !highpass; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
765 const int mirror_right= (width&1) ^ highpass; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
766 const int w= (width>>1) - 1 + (highpass & width); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
767 int i; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
768 |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
769 #define LIFT(src, ref, inv) ((src) + ((inv) ? - (ref) : + (ref))) |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
770 if(mirror_left){ |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
771 dst[0] = LIFT(src[0], ((mul*2*ref[0]+add)>>shift), inverse); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
772 dst += dst_step; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
773 src += src_step; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
774 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
775 |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
776 for(i=0; i<w; i++){ |
5627 | 777 dst[i*dst_step] = |
778 LIFT(src[i*src_step], | |
779 ((mul*(ref[i*ref_step] + ref[(i+1)*ref_step])+add)>>shift), | |
780 inverse); | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
781 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
782 |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
783 if(mirror_right){ |
5627 | 784 dst[w*dst_step] = |
785 LIFT(src[w*src_step], | |
786 ((mul*2*ref[w*ref_step]+add)>>shift), | |
787 inverse); | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
788 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
789 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
790 |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
791 #ifndef liftS |
5627 | 792 static av_always_inline void |
793 liftS(DWTELEM *dst, DWTELEM *src, DWTELEM *ref, | |
794 int dst_step, int src_step, int ref_step, | |
795 int width, int mul, int add, int shift, | |
796 int highpass, int inverse){ | |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
797 const int mirror_left= !highpass; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
798 const int mirror_right= (width&1) ^ highpass; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
799 const int w= (width>>1) - 1 + (highpass & width); |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
800 int i; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
801 |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
802 assert(shift == 4); |
5627 | 803 #define LIFTS(src, ref, inv) \ |
804 ((inv) ? \ | |
805 (src) + (((ref) + 4*(src))>>shift): \ | |
806 -((-16*(src) + (ref) + add/4 + 1 + (5<<25))/(5*4) - (1<<23))) | |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
807 if(mirror_left){ |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
808 dst[0] = LIFTS(src[0], mul*2*ref[0]+add, inverse); |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
809 dst += dst_step; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
810 src += src_step; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
811 } |
2967 | 812 |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
813 for(i=0; i<w; i++){ |
5627 | 814 dst[i*dst_step] = |
815 LIFTS(src[i*src_step], | |
816 mul*(ref[i*ref_step] + ref[(i+1)*ref_step])+add, | |
817 inverse); | |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
818 } |
2967 | 819 |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
820 if(mirror_right){ |
5627 | 821 dst[w*dst_step] = |
822 LIFTS(src[w*src_step], mul*2*ref[w*ref_step]+add, inverse); | |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
823 } |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
824 } |
5627 | 825 static av_always_inline void |
826 inv_liftS(IDWTELEM *dst, IDWTELEM *src, IDWTELEM *ref, | |
827 int dst_step, int src_step, int ref_step, | |
828 int width, int mul, int add, int shift, | |
829 int highpass, int inverse){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
830 const int mirror_left= !highpass; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
831 const int mirror_right= (width&1) ^ highpass; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
832 const int w= (width>>1) - 1 + (highpass & width); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
833 int i; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
834 |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
835 assert(shift == 4); |
5627 | 836 #define LIFTS(src, ref, inv) \ |
837 ((inv) ? \ | |
838 (src) + (((ref) + 4*(src))>>shift): \ | |
839 -((-16*(src) + (ref) + add/4 + 1 + (5<<25))/(5*4) - (1<<23))) | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
840 if(mirror_left){ |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
841 dst[0] = LIFTS(src[0], mul*2*ref[0]+add, inverse); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
842 dst += dst_step; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
843 src += src_step; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
844 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
845 |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
846 for(i=0; i<w; i++){ |
5627 | 847 dst[i*dst_step] = |
848 LIFTS(src[i*src_step], | |
849 mul*(ref[i*ref_step] + ref[(i+1)*ref_step])+add, | |
850 inverse); | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
851 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
852 |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
853 if(mirror_right){ |
5627 | 854 dst[w*dst_step] = |
855 LIFTS(src[w*src_step], mul*2*ref[w*ref_step]+add, inverse); | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
856 } |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
857 } |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
858 #endif |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
859 |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
860 static void horizontal_decompose53i(DWTELEM *b, int width){ |
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
861 DWTELEM temp[width]; |
2138 | 862 const int width2= width>>1; |
2893 | 863 int x; |
2138 | 864 const int w2= (width+1)>>1; |
865 | |
866 for(x=0; x<width2; x++){ | |
867 temp[x ]= b[2*x ]; | |
868 temp[x+w2]= b[2*x + 1]; | |
869 } | |
870 if(width&1) | |
871 temp[x ]= b[2*x ]; | |
872 #if 0 | |
2893 | 873 { |
874 int A1,A2,A3,A4; | |
2138 | 875 A2= temp[1 ]; |
876 A4= temp[0 ]; | |
877 A1= temp[0+width2]; | |
878 A1 -= (A2 + A4)>>1; | |
879 A4 += (A1 + 1)>>1; | |
880 b[0+width2] = A1; | |
881 b[0 ] = A4; | |
882 for(x=1; x+1<width2; x+=2){ | |
883 A3= temp[x+width2]; | |
884 A4= temp[x+1 ]; | |
885 A3 -= (A2 + A4)>>1; | |
886 A2 += (A1 + A3 + 2)>>2; | |
887 b[x+width2] = A3; | |
888 b[x ] = A2; | |
889 | |
890 A1= temp[x+1+width2]; | |
891 A2= temp[x+2 ]; | |
892 A1 -= (A2 + A4)>>1; | |
893 A4 += (A1 + A3 + 2)>>2; | |
894 b[x+1+width2] = A1; | |
895 b[x+1 ] = A4; | |
896 } | |
897 A3= temp[width-1]; | |
898 A3 -= A2; | |
899 A2 += (A1 + A3 + 2)>>2; | |
900 b[width -1] = A3; | |
901 b[width2-1] = A2; | |
2893 | 902 } |
2967 | 903 #else |
2138 | 904 lift(b+w2, temp+w2, temp, 1, 1, 1, width, -1, 0, 1, 1, 0); |
905 lift(b , temp , b+w2, 1, 1, 1, width, 1, 2, 2, 0, 0); | |
906 #endif | |
907 } | |
908 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
909 static void vertical_decompose53iH0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, int width){ |
2138 | 910 int i; |
2967 | 911 |
2138 | 912 for(i=0; i<width; i++){ |
913 b1[i] -= (b0[i] + b2[i])>>1; | |
914 } | |
915 } | |
916 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
917 static void vertical_decompose53iL0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, int width){ |
2138 | 918 int i; |
2967 | 919 |
2138 | 920 for(i=0; i<width; i++){ |
921 b1[i] += (b0[i] + b2[i] + 2)>>2; | |
922 } | |
923 } | |
924 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
925 static void spatial_decompose53i(DWTELEM *buffer, int width, int height, int stride){ |
2198 | 926 int y; |
2138 | 927 DWTELEM *b0= buffer + mirror(-2-1, height-1)*stride; |
928 DWTELEM *b1= buffer + mirror(-2 , height-1)*stride; | |
2967 | 929 |
2138 | 930 for(y=-2; y<height; y+=2){ |
931 DWTELEM *b2= buffer + mirror(y+1, height-1)*stride; | |
932 DWTELEM *b3= buffer + mirror(y+2, height-1)*stride; | |
933 | |
934 {START_TIMER | |
2998 | 935 if(y+1<(unsigned)height) horizontal_decompose53i(b2, width); |
936 if(y+2<(unsigned)height) horizontal_decompose53i(b3, width); | |
2138 | 937 STOP_TIMER("horizontal_decompose53i")} |
2967 | 938 |
2138 | 939 {START_TIMER |
2998 | 940 if(y+1<(unsigned)height) vertical_decompose53iH0(b1, b2, b3, width); |
941 if(y+0<(unsigned)height) vertical_decompose53iL0(b0, b1, b2, width); | |
2138 | 942 STOP_TIMER("vertical_decompose53i*")} |
2967 | 943 |
2138 | 944 b0=b2; |
945 b1=b3; | |
946 } | |
947 } | |
948 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
949 static void horizontal_decompose97i(DWTELEM *b, int width){ |
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
950 DWTELEM temp[width]; |
2138 | 951 const int w2= (width+1)>>1; |
952 | |
5565
93082c591c8b
Change rounding of the horizontal DWT to match the vertical one.
michael
parents:
5551
diff
changeset
|
953 lift (temp+w2, b +1, b , 1, 2, 2, width, W_AM, W_AO, W_AS, 1, 1); |
93082c591c8b
Change rounding of the horizontal DWT to match the vertical one.
michael
parents:
5551
diff
changeset
|
954 liftS(temp , b , temp+w2, 1, 2, 1, width, W_BM, W_BO, W_BS, 0, 0); |
5589
946c2db0a093
cleanup (remove some old experimentation related code)
michael
parents:
5588
diff
changeset
|
955 lift (b +w2, temp+w2, temp , 1, 1, 1, width, W_CM, W_CO, W_CS, 1, 0); |
2138 | 956 lift (b , temp , b +w2, 1, 1, 1, width, W_DM, W_DO, W_DS, 0, 0); |
957 } | |
958 | |
959 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
960 static void vertical_decompose97iH0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, int width){ |
2138 | 961 int i; |
2967 | 962 |
2138 | 963 for(i=0; i<width; i++){ |
964 b1[i] -= (W_AM*(b0[i] + b2[i])+W_AO)>>W_AS; | |
965 } | |
966 } | |
967 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
968 static void vertical_decompose97iH1(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, int width){ |
2138 | 969 int i; |
2967 | 970 |
2138 | 971 for(i=0; i<width; i++){ |
972 b1[i] += (W_CM*(b0[i] + b2[i])+W_CO)>>W_CS; | |
973 } | |
974 } | |
975 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
976 static void vertical_decompose97iL0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, int width){ |
2138 | 977 int i; |
2967 | 978 |
2138 | 979 for(i=0; i<width; i++){ |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
980 #ifdef liftS |
2138 | 981 b1[i] -= (W_BM*(b0[i] + b2[i])+W_BO)>>W_BS; |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
982 #else |
5565
93082c591c8b
Change rounding of the horizontal DWT to match the vertical one.
michael
parents:
5551
diff
changeset
|
983 b1[i] = (16*4*b1[i] - 4*(b0[i] + b2[i]) + W_BO*5 + (5<<27)) / (5*16) - (1<<23); |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
984 #endif |
2138 | 985 } |
986 } | |
987 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
988 static void vertical_decompose97iL1(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, int width){ |
2138 | 989 int i; |
2967 | 990 |
2138 | 991 for(i=0; i<width; i++){ |
992 b1[i] += (W_DM*(b0[i] + b2[i])+W_DO)>>W_DS; | |
993 } | |
994 } | |
995 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
996 static void spatial_decompose97i(DWTELEM *buffer, int width, int height, int stride){ |
2198 | 997 int y; |
2138 | 998 DWTELEM *b0= buffer + mirror(-4-1, height-1)*stride; |
999 DWTELEM *b1= buffer + mirror(-4 , height-1)*stride; | |
1000 DWTELEM *b2= buffer + mirror(-4+1, height-1)*stride; | |
1001 DWTELEM *b3= buffer + mirror(-4+2, height-1)*stride; | |
2967 | 1002 |
2138 | 1003 for(y=-4; y<height; y+=2){ |
1004 DWTELEM *b4= buffer + mirror(y+3, height-1)*stride; | |
1005 DWTELEM *b5= buffer + mirror(y+4, height-1)*stride; | |
1006 | |
1007 {START_TIMER | |
2998 | 1008 if(y+3<(unsigned)height) horizontal_decompose97i(b4, width); |
1009 if(y+4<(unsigned)height) horizontal_decompose97i(b5, width); | |
2138 | 1010 if(width>400){ |
1011 STOP_TIMER("horizontal_decompose97i") | |
1012 }} | |
2967 | 1013 |
2138 | 1014 {START_TIMER |
2998 | 1015 if(y+3<(unsigned)height) vertical_decompose97iH0(b3, b4, b5, width); |
1016 if(y+2<(unsigned)height) vertical_decompose97iL0(b2, b3, b4, width); | |
1017 if(y+1<(unsigned)height) vertical_decompose97iH1(b1, b2, b3, width); | |
1018 if(y+0<(unsigned)height) vertical_decompose97iL1(b0, b1, b2, width); | |
2138 | 1019 |
1020 if(width>400){ | |
1021 STOP_TIMER("vertical_decompose97i") | |
1022 }} | |
2967 | 1023 |
2138 | 1024 b0=b2; |
1025 b1=b3; | |
1026 b2=b4; | |
1027 b3=b5; | |
1028 } | |
1029 } | |
1030 | |
2241
c26038875ebc
consistent use of types patch by (D Richard Felker III <dalias at aerifal dot cx>)
michael
parents:
2240
diff
changeset
|
1031 void ff_spatial_dwt(DWTELEM *buffer, int width, int height, int stride, int type, int decomposition_count){ |
2138 | 1032 int level; |
2967 | 1033 |
2164 | 1034 for(level=0; level<decomposition_count; level++){ |
1035 switch(type){ | |
3326 | 1036 case DWT_97: spatial_decompose97i(buffer, width>>level, height>>level, stride<<level); break; |
1037 case DWT_53: spatial_decompose53i(buffer, width>>level, height>>level, stride<<level); break; | |
2138 | 1038 } |
1039 } | |
1040 } | |
1041 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1042 static void horizontal_compose53i(IDWTELEM *b, int width){ |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1043 IDWTELEM temp[width]; |
2138 | 1044 const int width2= width>>1; |
1045 const int w2= (width+1)>>1; | |
2893 | 1046 int x; |
2138 | 1047 |
1048 #if 0 | |
2893 | 1049 int A1,A2,A3,A4; |
2138 | 1050 A2= temp[1 ]; |
1051 A4= temp[0 ]; | |
1052 A1= temp[0+width2]; | |
1053 A1 -= (A2 + A4)>>1; | |
1054 A4 += (A1 + 1)>>1; | |
1055 b[0+width2] = A1; | |
1056 b[0 ] = A4; | |
1057 for(x=1; x+1<width2; x+=2){ | |
1058 A3= temp[x+width2]; | |
1059 A4= temp[x+1 ]; | |
1060 A3 -= (A2 + A4)>>1; | |
1061 A2 += (A1 + A3 + 2)>>2; | |
1062 b[x+width2] = A3; | |
1063 b[x ] = A2; | |
1064 | |
1065 A1= temp[x+1+width2]; | |
1066 A2= temp[x+2 ]; | |
1067 A1 -= (A2 + A4)>>1; | |
1068 A4 += (A1 + A3 + 2)>>2; | |
1069 b[x+1+width2] = A1; | |
1070 b[x+1 ] = A4; | |
1071 } | |
1072 A3= temp[width-1]; | |
1073 A3 -= A2; | |
1074 A2 += (A1 + A3 + 2)>>2; | |
1075 b[width -1] = A3; | |
1076 b[width2-1] = A2; | |
2967 | 1077 #else |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1078 inv_lift(temp , b , b+w2, 1, 1, 1, width, 1, 2, 2, 0, 1); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1079 inv_lift(temp+w2, b+w2, temp, 1, 1, 1, width, -1, 0, 1, 1, 1); |
2138 | 1080 #endif |
1081 for(x=0; x<width2; x++){ | |
1082 b[2*x ]= temp[x ]; | |
1083 b[2*x + 1]= temp[x+w2]; | |
1084 } | |
1085 if(width&1) | |
1086 b[2*x ]= temp[x ]; | |
1087 } | |
1088 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1089 static void vertical_compose53iH0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, int width){ |
2138 | 1090 int i; |
2967 | 1091 |
2138 | 1092 for(i=0; i<width; i++){ |
1093 b1[i] += (b0[i] + b2[i])>>1; | |
1094 } | |
1095 } | |
1096 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1097 static void vertical_compose53iL0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, int width){ |
2138 | 1098 int i; |
2967 | 1099 |
2138 | 1100 for(i=0; i<width; i++){ |
1101 b1[i] -= (b0[i] + b2[i] + 2)>>2; | |
1102 } | |
1103 } | |
1104 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1105 static void spatial_compose53i_buffered_init(dwt_compose_t *cs, slice_buffer * sb, int height, int stride_line){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1106 cs->b0 = slice_buffer_get_line(sb, mirror(-1-1, height-1) * stride_line); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1107 cs->b1 = slice_buffer_get_line(sb, mirror(-1 , height-1) * stride_line); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1108 cs->y = -1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1109 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1110 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1111 static void spatial_compose53i_init(dwt_compose_t *cs, IDWTELEM *buffer, int height, int stride){ |
2562 | 1112 cs->b0 = buffer + mirror(-1-1, height-1)*stride; |
1113 cs->b1 = buffer + mirror(-1 , height-1)*stride; | |
1114 cs->y = -1; | |
1115 } | |
1116 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1117 static void spatial_compose53i_dy_buffered(dwt_compose_t *cs, slice_buffer * sb, int width, int height, int stride_line){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1118 int y= cs->y; |
2967 | 1119 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1120 IDWTELEM *b0= cs->b0; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1121 IDWTELEM *b1= cs->b1; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1122 IDWTELEM *b2= slice_buffer_get_line(sb, mirror(y+1, height-1) * stride_line); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1123 IDWTELEM *b3= slice_buffer_get_line(sb, mirror(y+2, height-1) * stride_line); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1124 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1125 {START_TIMER |
2998 | 1126 if(y+1<(unsigned)height) vertical_compose53iL0(b1, b2, b3, width); |
1127 if(y+0<(unsigned)height) vertical_compose53iH0(b0, b1, b2, width); | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1128 STOP_TIMER("vertical_compose53i*")} |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1129 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1130 {START_TIMER |
2998 | 1131 if(y-1<(unsigned)height) horizontal_compose53i(b0, width); |
1132 if(y+0<(unsigned)height) horizontal_compose53i(b1, width); | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1133 STOP_TIMER("horizontal_compose53i")} |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1134 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1135 cs->b0 = b2; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1136 cs->b1 = b3; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1137 cs->y += 2; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1138 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1139 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1140 static void spatial_compose53i_dy(dwt_compose_t *cs, IDWTELEM *buffer, int width, int height, int stride){ |
2562 | 1141 int y= cs->y; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1142 IDWTELEM *b0= cs->b0; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1143 IDWTELEM *b1= cs->b1; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1144 IDWTELEM *b2= buffer + mirror(y+1, height-1)*stride; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1145 IDWTELEM *b3= buffer + mirror(y+2, height-1)*stride; |
2138 | 1146 |
1147 {START_TIMER | |
2998 | 1148 if(y+1<(unsigned)height) vertical_compose53iL0(b1, b2, b3, width); |
1149 if(y+0<(unsigned)height) vertical_compose53iH0(b0, b1, b2, width); | |
2138 | 1150 STOP_TIMER("vertical_compose53i*")} |
1151 | |
1152 {START_TIMER | |
2998 | 1153 if(y-1<(unsigned)height) horizontal_compose53i(b0, width); |
1154 if(y+0<(unsigned)height) horizontal_compose53i(b1, width); | |
2138 | 1155 STOP_TIMER("horizontal_compose53i")} |
1156 | |
2562 | 1157 cs->b0 = b2; |
1158 cs->b1 = b3; | |
1159 cs->y += 2; | |
1160 } | |
1161 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1162 static void spatial_compose53i(IDWTELEM *buffer, int width, int height, int stride){ |
2562 | 1163 dwt_compose_t cs; |
1164 spatial_compose53i_init(&cs, buffer, height, stride); | |
1165 while(cs.y <= height) | |
1166 spatial_compose53i_dy(&cs, buffer, width, height, stride); | |
2967 | 1167 } |
1168 | |
1169 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1170 void ff_snow_horizontal_compose97i(IDWTELEM *b, int width){ |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1171 IDWTELEM temp[width]; |
2138 | 1172 const int w2= (width+1)>>1; |
1173 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1174 inv_lift (temp , b , b +w2, 1, 1, 1, width, W_DM, W_DO, W_DS, 0, 1); |
5589
946c2db0a093
cleanup (remove some old experimentation related code)
michael
parents:
5588
diff
changeset
|
1175 inv_lift (temp+w2, b +w2, temp , 1, 1, 1, width, W_CM, W_CO, W_CS, 1, 1); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1176 inv_liftS(b , temp , temp+w2, 2, 1, 1, width, W_BM, W_BO, W_BS, 0, 1); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1177 inv_lift (b+1 , temp+w2, b , 2, 1, 2, width, W_AM, W_AO, W_AS, 1, 0); |
2138 | 1178 } |
1179 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1180 static void vertical_compose97iH0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, int width){ |
2138 | 1181 int i; |
2967 | 1182 |
2138 | 1183 for(i=0; i<width; i++){ |
1184 b1[i] += (W_AM*(b0[i] + b2[i])+W_AO)>>W_AS; | |
1185 } | |
1186 } | |
1187 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1188 static void vertical_compose97iH1(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, int width){ |
2138 | 1189 int i; |
2967 | 1190 |
2138 | 1191 for(i=0; i<width; i++){ |
1192 b1[i] -= (W_CM*(b0[i] + b2[i])+W_CO)>>W_CS; | |
1193 } | |
1194 } | |
1195 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1196 static void vertical_compose97iL0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, int width){ |
2138 | 1197 int i; |
2967 | 1198 |
2138 | 1199 for(i=0; i<width; i++){ |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1200 #ifdef liftS |
2138 | 1201 b1[i] += (W_BM*(b0[i] + b2[i])+W_BO)>>W_BS; |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1202 #else |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1203 b1[i] += (W_BM*(b0[i] + b2[i])+4*b1[i]+W_BO)>>W_BS; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1204 #endif |
2138 | 1205 } |
1206 } | |
1207 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1208 static void vertical_compose97iL1(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, int width){ |
2138 | 1209 int i; |
2967 | 1210 |
2138 | 1211 for(i=0; i<width; i++){ |
1212 b1[i] -= (W_DM*(b0[i] + b2[i])+W_DO)>>W_DS; | |
1213 } | |
1214 } | |
1215 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1216 void ff_snow_vertical_compose97i(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2, IDWTELEM *b3, IDWTELEM *b4, IDWTELEM *b5, int width){ |
2592 | 1217 int i; |
2967 | 1218 |
2592 | 1219 for(i=0; i<width; i++){ |
1220 b4[i] -= (W_DM*(b3[i] + b5[i])+W_DO)>>W_DS; | |
1221 b3[i] -= (W_CM*(b2[i] + b4[i])+W_CO)>>W_CS; | |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1222 #ifdef liftS |
2592 | 1223 b2[i] += (W_BM*(b1[i] + b3[i])+W_BO)>>W_BS; |
2602
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1224 #else |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1225 b2[i] += (W_BM*(b1[i] + b3[i])+4*b2[i]+W_BO)>>W_BS; |
5ec55feb6fdd
rescale coefficients during IDWT, that way the lifting steps are much simpler and faster
michael
parents:
2601
diff
changeset
|
1226 #endif |
2592 | 1227 b1[i] += (W_AM*(b0[i] + b2[i])+W_AO)>>W_AS; |
1228 } | |
1229 } | |
1230 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1231 static void spatial_compose97i_buffered_init(dwt_compose_t *cs, slice_buffer * sb, int height, int stride_line){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1232 cs->b0 = slice_buffer_get_line(sb, mirror(-3-1, height-1) * stride_line); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1233 cs->b1 = slice_buffer_get_line(sb, mirror(-3 , height-1) * stride_line); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1234 cs->b2 = slice_buffer_get_line(sb, mirror(-3+1, height-1) * stride_line); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1235 cs->b3 = slice_buffer_get_line(sb, mirror(-3+2, height-1) * stride_line); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1236 cs->y = -3; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1237 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1238 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1239 static void spatial_compose97i_init(dwt_compose_t *cs, IDWTELEM *buffer, int height, int stride){ |
2562 | 1240 cs->b0 = buffer + mirror(-3-1, height-1)*stride; |
1241 cs->b1 = buffer + mirror(-3 , height-1)*stride; | |
1242 cs->b2 = buffer + mirror(-3+1, height-1)*stride; | |
1243 cs->b3 = buffer + mirror(-3+2, height-1)*stride; | |
1244 cs->y = -3; | |
1245 } | |
2138 | 1246 |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1247 static void spatial_compose97i_dy_buffered(DSPContext *dsp, dwt_compose_t *cs, slice_buffer * sb, int width, int height, int stride_line){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1248 int y = cs->y; |
2967 | 1249 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1250 IDWTELEM *b0= cs->b0; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1251 IDWTELEM *b1= cs->b1; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1252 IDWTELEM *b2= cs->b2; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1253 IDWTELEM *b3= cs->b3; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1254 IDWTELEM *b4= slice_buffer_get_line(sb, mirror(y + 3, height - 1) * stride_line); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1255 IDWTELEM *b5= slice_buffer_get_line(sb, mirror(y + 4, height - 1) * stride_line); |
2967 | 1256 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1257 {START_TIMER |
2592 | 1258 if(y>0 && y+4<height){ |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1259 dsp->vertical_compose97i(b0, b1, b2, b3, b4, b5, width); |
2592 | 1260 }else{ |
2998 | 1261 if(y+3<(unsigned)height) vertical_compose97iL1(b3, b4, b5, width); |
1262 if(y+2<(unsigned)height) vertical_compose97iH1(b2, b3, b4, width); | |
1263 if(y+1<(unsigned)height) vertical_compose97iL0(b1, b2, b3, width); | |
1264 if(y+0<(unsigned)height) vertical_compose97iH0(b0, b1, b2, width); | |
2592 | 1265 } |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1266 if(width>400){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1267 STOP_TIMER("vertical_compose97i")}} |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1268 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1269 {START_TIMER |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1270 if(y-1<(unsigned)height) dsp->horizontal_compose97i(b0, width); |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1271 if(y+0<(unsigned)height) dsp->horizontal_compose97i(b1, width); |
3012 | 1272 if(width>400 && y+0<(unsigned)height){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1273 STOP_TIMER("horizontal_compose97i")}} |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1274 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1275 cs->b0=b2; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1276 cs->b1=b3; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1277 cs->b2=b4; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1278 cs->b3=b5; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1279 cs->y += 2; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1280 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1281 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1282 static void spatial_compose97i_dy(dwt_compose_t *cs, IDWTELEM *buffer, int width, int height, int stride){ |
2562 | 1283 int y = cs->y; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1284 IDWTELEM *b0= cs->b0; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1285 IDWTELEM *b1= cs->b1; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1286 IDWTELEM *b2= cs->b2; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1287 IDWTELEM *b3= cs->b3; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1288 IDWTELEM *b4= buffer + mirror(y+3, height-1)*stride; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1289 IDWTELEM *b5= buffer + mirror(y+4, height-1)*stride; |
2138 | 1290 |
1291 {START_TIMER | |
2998 | 1292 if(y+3<(unsigned)height) vertical_compose97iL1(b3, b4, b5, width); |
1293 if(y+2<(unsigned)height) vertical_compose97iH1(b2, b3, b4, width); | |
1294 if(y+1<(unsigned)height) vertical_compose97iL0(b1, b2, b3, width); | |
1295 if(y+0<(unsigned)height) vertical_compose97iH0(b0, b1, b2, width); | |
2138 | 1296 if(width>400){ |
1297 STOP_TIMER("vertical_compose97i")}} | |
1298 | |
1299 {START_TIMER | |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1300 if(y-1<(unsigned)height) ff_snow_horizontal_compose97i(b0, width); |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1301 if(y+0<(unsigned)height) ff_snow_horizontal_compose97i(b1, width); |
2138 | 1302 if(width>400 && b0 <= b2){ |
1303 STOP_TIMER("horizontal_compose97i")}} | |
2562 | 1304 |
1305 cs->b0=b2; | |
1306 cs->b1=b3; | |
1307 cs->b2=b4; | |
1308 cs->b3=b5; | |
1309 cs->y += 2; | |
1310 } | |
1311 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1312 static void spatial_compose97i(IDWTELEM *buffer, int width, int height, int stride){ |
2562 | 1313 dwt_compose_t cs; |
1314 spatial_compose97i_init(&cs, buffer, height, stride); | |
1315 while(cs.y <= height) | |
1316 spatial_compose97i_dy(&cs, buffer, width, height, stride); | |
1317 } | |
1318 | |
3075 | 1319 static void ff_spatial_idwt_buffered_init(dwt_compose_t *cs, slice_buffer * sb, int width, int height, int stride_line, int type, int decomposition_count){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1320 int level; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1321 for(level=decomposition_count-1; level>=0; level--){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1322 switch(type){ |
3326 | 1323 case DWT_97: spatial_compose97i_buffered_init(cs+level, sb, height>>level, stride_line<<level); break; |
1324 case DWT_53: spatial_compose53i_buffered_init(cs+level, sb, height>>level, stride_line<<level); break; | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1325 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1326 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1327 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1328 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1329 static void ff_spatial_idwt_init(dwt_compose_t *cs, IDWTELEM *buffer, int width, int height, int stride, int type, int decomposition_count){ |
2562 | 1330 int level; |
1331 for(level=decomposition_count-1; level>=0; level--){ | |
1332 switch(type){ | |
3326 | 1333 case DWT_97: spatial_compose97i_init(cs+level, buffer, height>>level, stride<<level); break; |
1334 case DWT_53: spatial_compose53i_init(cs+level, buffer, height>>level, stride<<level); break; | |
2562 | 1335 } |
1336 } | |
1337 } | |
1338 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1339 static void ff_spatial_idwt_slice(dwt_compose_t *cs, IDWTELEM *buffer, int width, int height, int stride, int type, int decomposition_count, int y){ |
2562 | 1340 const int support = type==1 ? 3 : 5; |
1341 int level; | |
1342 if(type==2) return; | |
1343 | |
1344 for(level=decomposition_count-1; level>=0; level--){ | |
1345 while(cs[level].y <= FFMIN((y>>level)+support, height>>level)){ | |
1346 switch(type){ | |
3326 | 1347 case DWT_97: spatial_compose97i_dy(cs+level, buffer, width>>level, height>>level, stride<<level); |
2562 | 1348 break; |
3326 | 1349 case DWT_53: spatial_compose53i_dy(cs+level, buffer, width>>level, height>>level, stride<<level); |
2562 | 1350 break; |
1351 } | |
1352 } | |
2138 | 1353 } |
1354 } | |
1355 | |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
1356 static void ff_spatial_idwt_buffered_slice(DSPContext *dsp, dwt_compose_t *cs, slice_buffer * slice_buf, int width, int height, int stride_line, int type, int decomposition_count, int y){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1357 const int support = type==1 ? 3 : 5; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1358 int level; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1359 if(type==2) return; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1360 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1361 for(level=decomposition_count-1; level>=0; level--){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1362 while(cs[level].y <= FFMIN((y>>level)+support, height>>level)){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1363 switch(type){ |
3326 | 1364 case DWT_97: spatial_compose97i_dy_buffered(dsp, cs+level, slice_buf, width>>level, height>>level, stride_line<<level); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1365 break; |
3326 | 1366 case DWT_53: spatial_compose53i_dy_buffered(cs+level, slice_buf, width>>level, height>>level, stride_line<<level); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1367 break; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1368 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1369 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1370 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1371 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1372 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1373 static void ff_spatial_idwt(IDWTELEM *buffer, int width, int height, int stride, int type, int decomposition_count){ |
2562 | 1374 dwt_compose_t cs[MAX_DECOMPOSITIONS]; |
1375 int y; | |
1376 ff_spatial_idwt_init(cs, buffer, width, height, stride, type, decomposition_count); | |
1377 for(y=0; y<height; y+=4) | |
1378 ff_spatial_idwt_slice(cs, buffer, width, height, stride, type, decomposition_count, y); | |
2138 | 1379 } |
1380 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1381 static int encode_subband_c0run(SnowContext *s, SubBand *b, IDWTELEM *src, IDWTELEM *parent, int stride, int orientation){ |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1382 const int w= b->width; |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1383 const int h= b->height; |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1384 int x, y; |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1385 |
2138 | 1386 if(1){ |
1387 int run=0; | |
2149 | 1388 int runs[w*h]; |
2138 | 1389 int run_index=0; |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1390 int max_index; |
2967 | 1391 |
2138 | 1392 for(y=0; y<h; y++){ |
1393 for(x=0; x<w; x++){ | |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1394 int v, p=0; |
2144 | 1395 int /*ll=0, */l=0, lt=0, t=0, rt=0; |
2149 | 1396 v= src[x + y*stride]; |
2138 | 1397 |
1398 if(y){ | |
2149 | 1399 t= src[x + (y-1)*stride]; |
2138 | 1400 if(x){ |
2149 | 1401 lt= src[x - 1 + (y-1)*stride]; |
2138 | 1402 } |
1403 if(x + 1 < w){ | |
2149 | 1404 rt= src[x + 1 + (y-1)*stride]; |
2138 | 1405 } |
1406 } | |
1407 if(x){ | |
2149 | 1408 l= src[x - 1 + y*stride]; |
2144 | 1409 /*if(x > 1){ |
1410 if(orientation==1) ll= src[y + (x-2)*stride]; | |
1411 else ll= src[x - 2 + y*stride]; | |
2138 | 1412 }*/ |
1413 } | |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1414 if(parent){ |
2149 | 1415 int px= x>>1; |
1416 int py= y>>1; | |
2967 | 1417 if(px<b->parent->width && py<b->parent->height) |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1418 p= parent[px + py*2*stride]; |
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1419 } |
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1420 if(!(/*ll|*/l|lt|t|rt|p)){ |
2138 | 1421 if(v){ |
1422 runs[run_index++]= run; | |
1423 run=0; | |
1424 }else{ | |
1425 run++; | |
1426 } | |
1427 } | |
1428 } | |
1429 } | |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1430 max_index= run_index; |
2138 | 1431 runs[run_index++]= run; |
1432 run_index=0; | |
1433 run= runs[run_index++]; | |
1434 | |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1435 put_symbol2(&s->c, b->state[30], max_index, 0); |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1436 if(run_index <= max_index) |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1437 put_symbol2(&s->c, b->state[1], run, 3); |
2967 | 1438 |
2138 | 1439 for(y=0; y<h; y++){ |
2435
c89ac0e70c66
10l patch by (matthieu castet <castet.matthieu free fr>)
michael
parents:
2422
diff
changeset
|
1440 if(s->c.bytestream_end - s->c.bytestream < w*40){ |
2422 | 1441 av_log(s->avctx, AV_LOG_ERROR, "encoded frame too large\n"); |
1442 return -1; | |
1443 } | |
2138 | 1444 for(x=0; x<w; x++){ |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1445 int v, p=0; |
2144 | 1446 int /*ll=0, */l=0, lt=0, t=0, rt=0; |
2149 | 1447 v= src[x + y*stride]; |
2138 | 1448 |
1449 if(y){ | |
2149 | 1450 t= src[x + (y-1)*stride]; |
2138 | 1451 if(x){ |
2149 | 1452 lt= src[x - 1 + (y-1)*stride]; |
2138 | 1453 } |
1454 if(x + 1 < w){ | |
2149 | 1455 rt= src[x + 1 + (y-1)*stride]; |
2138 | 1456 } |
1457 } | |
1458 if(x){ | |
2149 | 1459 l= src[x - 1 + y*stride]; |
2144 | 1460 /*if(x > 1){ |
1461 if(orientation==1) ll= src[y + (x-2)*stride]; | |
1462 else ll= src[x - 2 + y*stride]; | |
2138 | 1463 }*/ |
1464 } | |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1465 if(parent){ |
2149 | 1466 int px= x>>1; |
1467 int py= y>>1; | |
2967 | 1468 if(px<b->parent->width && py<b->parent->height) |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1469 p= parent[px + py*2*stride]; |
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1470 } |
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1471 if(/*ll|*/l|lt|t|rt|p){ |
4001 | 1472 int context= av_log2(/*FFABS(ll) + */3*FFABS(l) + FFABS(lt) + 2*FFABS(t) + FFABS(rt) + FFABS(p)); |
2144 | 1473 |
2335 | 1474 put_rac(&s->c, &b->state[0][context], !!v); |
2138 | 1475 }else{ |
1476 if(!run){ | |
1477 run= runs[run_index++]; | |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1478 |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1479 if(run_index <= max_index) |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1480 put_symbol2(&s->c, b->state[1], run, 3); |
2138 | 1481 assert(v); |
1482 }else{ | |
1483 run--; | |
1484 assert(!v); | |
1485 } | |
1486 } | |
1487 if(v){ | |
4001 | 1488 int context= av_log2(/*FFABS(ll) + */3*FFABS(l) + FFABS(lt) + 2*FFABS(t) + FFABS(rt) + FFABS(p)); |
1489 int l2= 2*FFABS(l) + (l<0); | |
1490 int t2= 2*FFABS(t) + (t<0); | |
1491 | |
1492 put_symbol2(&s->c, b->state[context + 2], FFABS(v)-1, context-4); | |
2596 | 1493 put_rac(&s->c, &b->state[0][16 + 1 + 3 + quant3bA[l2&0xFF] + 3*quant3bA[t2&0xFF]], v<0); |
2138 | 1494 } |
1495 } | |
1496 } | |
1497 } | |
2422 | 1498 return 0; |
2138 | 1499 } |
1500 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1501 static int encode_subband(SnowContext *s, SubBand *b, IDWTELEM *src, IDWTELEM *parent, int stride, int orientation){ |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1502 // encode_subband_qtree(s, b, src, parent, stride, orientation); |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1503 // encode_subband_z0run(s, b, src, parent, stride, orientation); |
2422 | 1504 return encode_subband_c0run(s, b, src, parent, stride, orientation); |
2155
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1505 // encode_subband_dzr(s, b, src, parent, stride, orientation); |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1506 } |
274a01d80f4a
various subband encoders (all either worse or complicated so they are commented out)
michael
parents:
2152
diff
changeset
|
1507 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1508 static inline void unpack_coeffs(SnowContext *s, SubBand *b, SubBand * parent, int orientation){ |
2138 | 1509 const int w= b->width; |
1510 const int h= b->height; | |
1511 int x,y; | |
2967 | 1512 |
2138 | 1513 if(1){ |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1514 int run, runs; |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1515 x_and_coeff *xc= b->x_coeff; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1516 x_and_coeff *prev_xc= NULL; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1517 x_and_coeff *prev2_xc= xc; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1518 x_and_coeff *parent_xc= parent ? parent->x_coeff : NULL; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1519 x_and_coeff *prev_parent_xc= parent_xc; |
2138 | 1520 |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1521 runs= get_symbol2(&s->c, b->state[30], 0); |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1522 if(runs-- > 0) run= get_symbol2(&s->c, b->state[1], 3); |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1523 else run= INT_MAX; |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1524 |
2138 | 1525 for(y=0; y<h; y++){ |
2193 | 1526 int v=0; |
1527 int lt=0, t=0, rt=0; | |
1528 | |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1529 if(y && prev_xc->x == 0){ |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1530 rt= prev_xc->coeff; |
2193 | 1531 } |
2138 | 1532 for(x=0; x<w; x++){ |
2193 | 1533 int p=0; |
1534 const int l= v; | |
2967 | 1535 |
2193 | 1536 lt= t; t= rt; |
1537 | |
2194 | 1538 if(y){ |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1539 if(prev_xc->x <= x) |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1540 prev_xc++; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1541 if(prev_xc->x == x + 1) |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1542 rt= prev_xc->coeff; |
2194 | 1543 else |
1544 rt=0; | |
1545 } | |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1546 if(parent_xc){ |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1547 if(x>>1 > parent_xc->x){ |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1548 parent_xc++; |
2192 | 1549 } |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1550 if(x>>1 == parent_xc->x){ |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1551 p= parent_xc->coeff; |
2194 | 1552 } |
2148
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1553 } |
678be5a8f282
use parent sample to predict significance & magnitude
michael
parents:
2146
diff
changeset
|
1554 if(/*ll|*/l|lt|t|rt|p){ |
4001 | 1555 int context= av_log2(/*FFABS(ll) + */3*(l>>1) + (lt>>1) + (t&~1) + (rt>>1) + (p>>1)); |
2144 | 1556 |
2335 | 1557 v=get_rac(&s->c, &b->state[0][context]); |
2605 | 1558 if(v){ |
1559 v= 2*(get_symbol2(&s->c, b->state[context + 2], context-4) + 1); | |
1560 v+=get_rac(&s->c, &b->state[0][16 + 1 + 3 + quant3bA[l&0xFF] + 3*quant3bA[t&0xFF]]); | |
2967 | 1561 |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1562 xc->x=x; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1563 (xc++)->coeff= v; |
2605 | 1564 } |
2138 | 1565 }else{ |
1566 if(!run){ | |
2609
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1567 if(runs-- > 0) run= get_symbol2(&s->c, b->state[1], 3); |
0f74a379a890
store the number of runs to avoid storing the last run value
michael
parents:
2608
diff
changeset
|
1568 else run= INT_MAX; |
2605 | 1569 v= 2*(get_symbol2(&s->c, b->state[0 + 2], 0-4) + 1); |
1570 v+=get_rac(&s->c, &b->state[0][16 + 1 + 3]); | |
2967 | 1571 |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1572 xc->x=x; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1573 (xc++)->coeff= v; |
2138 | 1574 }else{ |
2606 | 1575 int max_run; |
2138 | 1576 run--; |
1577 v=0; | |
2191 | 1578 |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1579 if(y) max_run= FFMIN(run, prev_xc->x - x - 2); |
2606 | 1580 else max_run= FFMIN(run, w-x-1); |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1581 if(parent_xc) |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1582 max_run= FFMIN(max_run, 2*parent_xc->x - x - 1); |
2606 | 1583 x+= max_run; |
1584 run-= max_run; | |
2138 | 1585 } |
1586 } | |
2192 | 1587 } |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1588 (xc++)->x= w+1; //end marker |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1589 prev_xc= prev2_xc; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1590 prev2_xc= xc; |
2967 | 1591 |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1592 if(parent_xc){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1593 if(y&1){ |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1594 while(parent_xc->x != parent->width+1) |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1595 parent_xc++; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1596 parent_xc++; |
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1597 prev_parent_xc= parent_xc; |
2192 | 1598 }else{ |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1599 parent_xc= prev_parent_xc; |
2138 | 1600 } |
1601 } | |
1602 } | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1603 |
2607
fde7b6fe2aaf
replace complicated pointer dereference + index stuff by pointers in unpack_coeffs()
michael
parents:
2606
diff
changeset
|
1604 (xc++)->x= w+1; //end marker |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1605 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1606 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1607 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1608 static inline void decode_subband_slice_buffered(SnowContext *s, SubBand *b, slice_buffer * sb, int start_y, int h, int save_state[1]){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1609 const int w= b->width; |
2893 | 1610 int y; |
4594 | 1611 const int qlog= av_clip(s->qlog + b->qlog, 0, QROOT*16); |
2600 | 1612 int qmul= qexp[qlog&(QROOT-1)]<<(qlog>>QSHIFT); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1613 int qadd= (s->qbias*qmul)>>QBIAS_SHIFT; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1614 int new_index = 0; |
2967 | 1615 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1616 START_TIMER |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1617 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1618 if(b->ibuf == s->spatial_idwt_buffer || s->qlog == LOSSLESS_QLOG){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1619 qadd= 0; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1620 qmul= 1<<QEXPSHIFT; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1621 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1622 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1623 /* If we are on the second or later slice, restore our index. */ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1624 if (start_y != 0) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1625 new_index = save_state[0]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1626 |
2967 | 1627 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1628 for(y=start_y; y<h; y++){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1629 int x = 0; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1630 int v; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1631 IDWTELEM * line = slice_buffer_get_line(sb, y * b->stride_line + b->buf_y_offset) + b->buf_x_offset; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
1632 memset(line, 0, b->width*sizeof(IDWTELEM)); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1633 v = b->x_coeff[new_index].coeff; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1634 x = b->x_coeff[new_index++].x; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1635 while(x < w) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1636 { |
2596 | 1637 register int t= ( (v>>1)*qmul + qadd)>>QEXPSHIFT; |
1638 register int u= -(v&1); | |
1639 line[x] = (t^u) - u; | |
1640 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1641 v = b->x_coeff[new_index].coeff; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1642 x = b->x_coeff[new_index++].x; |
2138 | 1643 } |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1644 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1645 if(w > 200 && start_y != 0/*level+1 == s->spatial_decomposition_count*/){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1646 STOP_TIMER("decode_subband") |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1647 } |
2967 | 1648 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1649 /* Save our variables for the next slice. */ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1650 save_state[0] = new_index; |
2967 | 1651 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
1652 return; |
2138 | 1653 } |
1654 | |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
1655 static void reset_contexts(SnowContext *s){ //FIXME better initial contexts |
2138 | 1656 int plane_index, level, orientation; |
1657 | |
2199 | 1658 for(plane_index=0; plane_index<3; plane_index++){ |
2138 | 1659 for(level=0; level<s->spatial_decomposition_count; level++){ |
1660 for(orientation=level ? 1:0; orientation<4; orientation++){ | |
2335 | 1661 memset(s->plane[plane_index].band[level][orientation].state, MID_STATE, sizeof(s->plane[plane_index].band[level][orientation].state)); |
2138 | 1662 } |
1663 } | |
1664 } | |
2335 | 1665 memset(s->header_state, MID_STATE, sizeof(s->header_state)); |
1666 memset(s->block_state, MID_STATE, sizeof(s->block_state)); | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1667 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1668 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1669 static int alloc_blocks(SnowContext *s){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1670 int w= -((-s->avctx->width )>>LOG2_MB_SIZE); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1671 int h= -((-s->avctx->height)>>LOG2_MB_SIZE); |
2967 | 1672 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1673 s->b_width = w; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1674 s->b_height= h; |
2967 | 1675 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1676 s->block= av_mallocz(w * h * sizeof(BlockNode) << (s->block_max_depth*2)); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1677 return 0; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1678 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1679 |
2335 | 1680 static inline void copy_rac_state(RangeCoder *d, RangeCoder *s){ |
1681 uint8_t *bytestream= d->bytestream; | |
1682 uint8_t *bytestream_start= d->bytestream_start; | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1683 *d= *s; |
2335 | 1684 d->bytestream= bytestream; |
1685 d->bytestream_start= bytestream_start; | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1686 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1687 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1688 //near copy & paste from dsputil, FIXME |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1689 static int pix_sum(uint8_t * pix, int line_size, int w) |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1690 { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1691 int s, i, j; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1692 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1693 s = 0; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1694 for (i = 0; i < w; i++) { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1695 for (j = 0; j < w; j++) { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1696 s += pix[0]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1697 pix ++; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1698 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1699 pix += line_size - w; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1700 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1701 return s; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1702 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1703 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1704 //near copy & paste from dsputil, FIXME |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1705 static int pix_norm1(uint8_t * pix, int line_size, int w) |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1706 { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1707 int s, i, j; |
4179 | 1708 uint32_t *sq = ff_squareTbl + 256; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1709 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1710 s = 0; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1711 for (i = 0; i < w; i++) { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1712 for (j = 0; j < w; j ++) { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1713 s += sq[pix[0]]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1714 pix ++; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1715 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1716 pix += line_size - w; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1717 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1718 return s; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1719 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1720 |
3314 | 1721 static inline void set_blocks(SnowContext *s, int level, int x, int y, int l, int cb, int cr, int mx, int my, int ref, int type){ |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1722 const int w= s->b_width << s->block_max_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1723 const int rem_depth= s->block_max_depth - level; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1724 const int index= (x + y*w) << rem_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1725 const int block_w= 1<<rem_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1726 BlockNode block; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1727 int i,j; |
2967 | 1728 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1729 block.color[0]= l; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1730 block.color[1]= cb; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1731 block.color[2]= cr; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1732 block.mx= mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1733 block.my= my; |
3314 | 1734 block.ref= ref; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1735 block.type= type; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1736 block.level= level; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1737 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1738 for(j=0; j<block_w; j++){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1739 for(i=0; i<block_w; i++){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1740 s->block[index + i + j*w]= block; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1741 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1742 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1743 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1744 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1745 static inline void init_ref(MotionEstContext *c, uint8_t *src[3], uint8_t *ref[3], uint8_t *ref2[3], int x, int y, int ref_index){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1746 const int offset[3]= { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1747 y*c-> stride + x, |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1748 ((y*c->uvstride + x)>>1), |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1749 ((y*c->uvstride + x)>>1), |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1750 }; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1751 int i; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1752 for(i=0; i<3; i++){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1753 c->src[0][i]= src [i]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1754 c->ref[0][i]= ref [i] + offset[i]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1755 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1756 assert(!ref_index); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1757 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1758 |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1759 static inline void pred_mv(SnowContext *s, int *mx, int *my, int ref, |
4408
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1760 const BlockNode *left, const BlockNode *top, const BlockNode *tr){ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1761 if(s->ref_frames == 1){ |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1762 *mx = mid_pred(left->mx, top->mx, tr->mx); |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1763 *my = mid_pred(left->my, top->my, tr->my); |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1764 }else{ |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1765 const int *scale = scale_mv_ref[ref]; |
4407
15688bdfe1b5
Brings down the number of snow.c warnings from 33 to 27 by putting parentheses
takis
parents:
4360
diff
changeset
|
1766 *mx = mid_pred((left->mx * scale[left->ref] + 128) >>8, |
15688bdfe1b5
Brings down the number of snow.c warnings from 33 to 27 by putting parentheses
takis
parents:
4360
diff
changeset
|
1767 (top ->mx * scale[top ->ref] + 128) >>8, |
15688bdfe1b5
Brings down the number of snow.c warnings from 33 to 27 by putting parentheses
takis
parents:
4360
diff
changeset
|
1768 (tr ->mx * scale[tr ->ref] + 128) >>8); |
15688bdfe1b5
Brings down the number of snow.c warnings from 33 to 27 by putting parentheses
takis
parents:
4360
diff
changeset
|
1769 *my = mid_pred((left->my * scale[left->ref] + 128) >>8, |
15688bdfe1b5
Brings down the number of snow.c warnings from 33 to 27 by putting parentheses
takis
parents:
4360
diff
changeset
|
1770 (top ->my * scale[top ->ref] + 128) >>8, |
15688bdfe1b5
Brings down the number of snow.c warnings from 33 to 27 by putting parentheses
takis
parents:
4360
diff
changeset
|
1771 (tr ->my * scale[tr ->ref] + 128) >>8); |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1772 } |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1773 } |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1774 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1775 //FIXME copy&paste |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1776 #define P_LEFT P[1] |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1777 #define P_TOP P[2] |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1778 #define P_TOPRIGHT P[3] |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1779 #define P_MEDIAN P[4] |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1780 #define P_MV1 P[9] |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1781 #define FLAG_QPEL 1 //must be 1 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1782 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1783 static int encode_q_branch(SnowContext *s, int level, int x, int y){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1784 uint8_t p_buffer[1024]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1785 uint8_t i_buffer[1024]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1786 uint8_t p_state[sizeof(s->block_state)]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1787 uint8_t i_state[sizeof(s->block_state)]; |
2335 | 1788 RangeCoder pc, ic; |
1789 uint8_t *pbbak= s->c.bytestream; | |
1790 uint8_t *pbbak_start= s->c.bytestream_start; | |
5082 | 1791 int score, score2, iscore, i_len, p_len, block_s, sum, base_bits; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1792 const int w= s->b_width << s->block_max_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1793 const int h= s->b_height << s->block_max_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1794 const int rem_depth= s->block_max_depth - level; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1795 const int index= (x + y*w) << rem_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1796 const int block_w= 1<<(LOG2_MB_SIZE - level); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1797 int trx= (x+1)<<rem_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1798 int try= (y+1)<<rem_depth; |
4408
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1799 const BlockNode *left = x ? &s->block[index-1] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1800 const BlockNode *top = y ? &s->block[index-w] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1801 const BlockNode *right = trx<w ? &s->block[index+1] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1802 const BlockNode *bottom= try<h ? &s->block[index+w] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1803 const BlockNode *tl = y && x ? &s->block[index-w-1] : left; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
1804 const BlockNode *tr = y && trx<w && ((x&1)==0 || level==0) ? &s->block[index-w+(1<<rem_depth)] : tl; //FIXME use lt |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1805 int pl = left->color[0]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1806 int pcb= left->color[1]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1807 int pcr= left->color[2]; |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1808 int pmx, pmy; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1809 int mx=0, my=0; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1810 int l,cr,cb; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1811 const int stride= s->current_picture.linesize[0]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1812 const int uvstride= s->current_picture.linesize[1]; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1813 uint8_t *current_data[3]= { s->input_picture.data[0] + (x + y* stride)*block_w, |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1814 s->input_picture.data[1] + (x + y*uvstride)*block_w/2, |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1815 s->input_picture.data[2] + (x + y*uvstride)*block_w/2}; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1816 int P[10][2]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1817 int16_t last_mv[3][2]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1818 int qpel= !!(s->avctx->flags & CODEC_FLAG_QPEL); //unused |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1819 const int shift= 1+qpel; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1820 MotionEstContext *c= &s->m.me; |
3314 | 1821 int ref_context= av_log2(2*left->ref) + av_log2(2*top->ref); |
4001 | 1822 int mx_context= av_log2(2*FFABS(left->mx - top->mx)); |
1823 int my_context= av_log2(2*FFABS(left->my - top->my)); | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1824 int s_context= 2*left->level + 2*top->level + tl->level + tr->level; |
3314 | 1825 int ref, best_ref, ref_score, ref_mx, ref_my; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1826 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1827 assert(sizeof(s->block_state) >= 256); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1828 if(s->keyframe){ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1829 set_blocks(s, level, x, y, pl, pcb, pcr, 0, 0, 0, BLOCK_INTRA); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1830 return 0; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1831 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1832 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1833 // clip predictors / edge ? |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1834 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1835 P_LEFT[0]= left->mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1836 P_LEFT[1]= left->my; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1837 P_TOP [0]= top->mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1838 P_TOP [1]= top->my; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1839 P_TOPRIGHT[0]= tr->mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1840 P_TOPRIGHT[1]= tr->my; |
2967 | 1841 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1842 last_mv[0][0]= s->block[index].mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1843 last_mv[0][1]= s->block[index].my; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1844 last_mv[1][0]= right->mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1845 last_mv[1][1]= right->my; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1846 last_mv[2][0]= bottom->mx; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1847 last_mv[2][1]= bottom->my; |
2967 | 1848 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1849 s->m.mb_stride=2; |
2967 | 1850 s->m.mb_x= |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1851 s->m.mb_y= 0; |
4360
5f887a3c7281
Change the Snow encoder to always use the available MotionEstContext pointer.
takis
parents:
4332
diff
changeset
|
1852 c->skip= 0; |
5f887a3c7281
Change the Snow encoder to always use the available MotionEstContext pointer.
takis
parents:
4332
diff
changeset
|
1853 |
5f887a3c7281
Change the Snow encoder to always use the available MotionEstContext pointer.
takis
parents:
4332
diff
changeset
|
1854 assert(c-> stride == stride); |
5f887a3c7281
Change the Snow encoder to always use the available MotionEstContext pointer.
takis
parents:
4332
diff
changeset
|
1855 assert(c->uvstride == uvstride); |
2967 | 1856 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1857 c->penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_cmp); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1858 c->sub_penalty_factor= get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_sub_cmp); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1859 c->mb_penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->mb_cmp); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1860 c->current_mv_penalty= c->mv_penalty[s->m.f_code=1] + MAX_MV; |
2967 | 1861 |
2206 | 1862 c->xmin = - x*block_w - 16+2; |
1863 c->ymin = - y*block_w - 16+2; | |
1864 c->xmax = - (x+1)*block_w + (w<<(LOG2_MB_SIZE - s->block_max_depth)) + 16-2; | |
1865 c->ymax = - (y+1)*block_w + (h<<(LOG2_MB_SIZE - s->block_max_depth)) + 16-2; | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1866 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1867 if(P_LEFT[0] > (c->xmax<<shift)) P_LEFT[0] = (c->xmax<<shift); |
2967 | 1868 if(P_LEFT[1] > (c->ymax<<shift)) P_LEFT[1] = (c->ymax<<shift); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1869 if(P_TOP[0] > (c->xmax<<shift)) P_TOP[0] = (c->xmax<<shift); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1870 if(P_TOP[1] > (c->ymax<<shift)) P_TOP[1] = (c->ymax<<shift); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1871 if(P_TOPRIGHT[0] < (c->xmin<<shift)) P_TOPRIGHT[0]= (c->xmin<<shift); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1872 if(P_TOPRIGHT[0] > (c->xmax<<shift)) P_TOPRIGHT[0]= (c->xmax<<shift); //due to pmx no clip |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1873 if(P_TOPRIGHT[1] > (c->ymax<<shift)) P_TOPRIGHT[1]= (c->ymax<<shift); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1874 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1875 P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1876 P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1877 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1878 if (!y) { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1879 c->pred_x= P_LEFT[0]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1880 c->pred_y= P_LEFT[1]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1881 } else { |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1882 c->pred_x = P_MEDIAN[0]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1883 c->pred_y = P_MEDIAN[1]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1884 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1885 |
3314 | 1886 score= INT_MAX; |
1887 best_ref= 0; | |
1888 for(ref=0; ref<s->ref_frames; ref++){ | |
1889 init_ref(c, current_data, s->last_picture[ref].data, NULL, block_w*x, block_w*y, 0); | |
1890 | |
1891 ref_score= ff_epzs_motion_search(&s->m, &ref_mx, &ref_my, P, 0, /*ref_index*/ 0, last_mv, | |
1892 (1<<16)>>shift, level-LOG2_MB_SIZE+4, block_w); | |
1893 | |
1894 assert(ref_mx >= c->xmin); | |
1895 assert(ref_mx <= c->xmax); | |
1896 assert(ref_my >= c->ymin); | |
1897 assert(ref_my <= c->ymax); | |
1898 | |
4360
5f887a3c7281
Change the Snow encoder to always use the available MotionEstContext pointer.
takis
parents:
4332
diff
changeset
|
1899 ref_score= c->sub_motion_search(&s->m, &ref_mx, &ref_my, ref_score, 0, 0, level-LOG2_MB_SIZE+4, block_w); |
3314 | 1900 ref_score= ff_get_mb_score(&s->m, ref_mx, ref_my, 0, 0, level-LOG2_MB_SIZE+4, block_w, 0); |
1901 ref_score+= 2*av_log2(2*ref)*c->penalty_factor; | |
1902 if(s->ref_mvs[ref]){ | |
1903 s->ref_mvs[ref][index][0]= ref_mx; | |
1904 s->ref_mvs[ref][index][1]= ref_my; | |
1905 s->ref_scores[ref][index]= ref_score; | |
1906 } | |
1907 if(score > ref_score){ | |
1908 score= ref_score; | |
1909 best_ref= ref; | |
1910 mx= ref_mx; | |
1911 my= ref_my; | |
1912 } | |
1913 } | |
5127 | 1914 //FIXME if mb_cmp != SSE then intra cannot be compared currently and mb_penalty vs. lambda2 |
2967 | 1915 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1916 // subpel search |
5085 | 1917 base_bits= get_rac_count(&s->c) - 8*(s->c.bytestream - s->c.bytestream_start); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1918 pc= s->c; |
2335 | 1919 pc.bytestream_start= |
1920 pc.bytestream= p_buffer; //FIXME end/start? and at the other stoo | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1921 memcpy(p_state, s->block_state, sizeof(s->block_state)); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1922 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1923 if(level!=s->block_max_depth) |
2335 | 1924 put_rac(&pc, &p_state[4 + s_context], 1); |
1925 put_rac(&pc, &p_state[1 + left->type + top->type], 0); | |
3314 | 1926 if(s->ref_frames > 1) |
1927 put_symbol(&pc, &p_state[128 + 1024 + 32*ref_context], best_ref, 0); | |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1928 pred_mv(s, &pmx, &pmy, best_ref, left, top, tr); |
3314 | 1929 put_symbol(&pc, &p_state[128 + 32*(mx_context + 16*!!best_ref)], mx - pmx, 1); |
1930 put_symbol(&pc, &p_state[128 + 32*(my_context + 16*!!best_ref)], my - pmy, 1); | |
2335 | 1931 p_len= pc.bytestream - pc.bytestream_start; |
5082 | 1932 score += (s->lambda2*(get_rac_count(&pc)-base_bits))>>FF_LAMBDA_SHIFT; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1933 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1934 block_s= block_w*block_w; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1935 sum = pix_sum(current_data[0], stride, block_w); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1936 l= (sum + block_s/2)/block_s; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1937 iscore = pix_norm1(current_data[0], stride, block_w) - 2*l*sum + l*l*block_s; |
2967 | 1938 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1939 block_s= block_w*block_w>>2; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1940 sum = pix_sum(current_data[1], uvstride, block_w>>1); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1941 cb= (sum + block_s/2)/block_s; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1942 // iscore += pix_norm1(¤t_mb[1][0], uvstride, block_w>>1) - 2*cb*sum + cb*cb*block_s; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
1943 sum = pix_sum(current_data[2], uvstride, block_w>>1); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1944 cr= (sum + block_s/2)/block_s; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1945 // iscore += pix_norm1(¤t_mb[2][0], uvstride, block_w>>1) - 2*cr*sum + cr*cr*block_s; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1946 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1947 ic= s->c; |
2335 | 1948 ic.bytestream_start= |
1949 ic.bytestream= i_buffer; //FIXME end/start? and at the other stoo | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1950 memcpy(i_state, s->block_state, sizeof(s->block_state)); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1951 if(level!=s->block_max_depth) |
2335 | 1952 put_rac(&ic, &i_state[4 + s_context], 1); |
1953 put_rac(&ic, &i_state[1 + left->type + top->type], 1); | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1954 put_symbol(&ic, &i_state[32], l-pl , 1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1955 put_symbol(&ic, &i_state[64], cb-pcb, 1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1956 put_symbol(&ic, &i_state[96], cr-pcr, 1); |
2335 | 1957 i_len= ic.bytestream - ic.bytestream_start; |
5082 | 1958 iscore += (s->lambda2*(get_rac_count(&ic)-base_bits))>>FF_LAMBDA_SHIFT; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1959 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1960 // assert(score==256*256*256*64-1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1961 assert(iscore < 255*255*256 + s->lambda2*10); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1962 assert(iscore >= 0); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1963 assert(l>=0 && l<=255); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1964 assert(pl>=0 && pl<=255); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1965 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1966 if(level==0){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1967 int varc= iscore >> 8; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1968 int vard= score >> 8; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1969 if (vard <= 64 || vard < varc) |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1970 c->scene_change_score+= ff_sqrt(vard) - ff_sqrt(varc); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1971 else |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1972 c->scene_change_score+= s->m.qscale; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1973 } |
2967 | 1974 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1975 if(level!=s->block_max_depth){ |
2335 | 1976 put_rac(&s->c, &s->block_state[4 + s_context], 0); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1977 score2 = encode_q_branch(s, level+1, 2*x+0, 2*y+0); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1978 score2+= encode_q_branch(s, level+1, 2*x+1, 2*y+0); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1979 score2+= encode_q_branch(s, level+1, 2*x+0, 2*y+1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1980 score2+= encode_q_branch(s, level+1, 2*x+1, 2*y+1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1981 score2+= s->lambda2>>FF_LAMBDA_SHIFT; //FIXME exact split overhead |
2967 | 1982 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1983 if(score2 < score && score2 < iscore) |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1984 return score2; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1985 } |
2967 | 1986 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1987 if(iscore < score){ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
1988 pred_mv(s, &pmx, &pmy, 0, left, top, tr); |
2335 | 1989 memcpy(pbbak, i_buffer, i_len); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1990 s->c= ic; |
2335 | 1991 s->c.bytestream_start= pbbak_start; |
1992 s->c.bytestream= pbbak + i_len; | |
3314 | 1993 set_blocks(s, level, x, y, l, cb, cr, pmx, pmy, 0, BLOCK_INTRA); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1994 memcpy(s->block_state, i_state, sizeof(s->block_state)); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1995 return iscore; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1996 }else{ |
2335 | 1997 memcpy(pbbak, p_buffer, p_len); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
1998 s->c= pc; |
2335 | 1999 s->c.bytestream_start= pbbak_start; |
2000 s->c.bytestream= pbbak + p_len; | |
3314 | 2001 set_blocks(s, level, x, y, pl, pcb, pcr, mx, my, best_ref, 0); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2002 memcpy(s->block_state, p_state, sizeof(s->block_state)); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2003 return score; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2004 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2005 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2006 |
4283
d6f83e2f8804
rename always_inline to av_always_inline and move to common.h
mru
parents:
4197
diff
changeset
|
2007 static av_always_inline int same_block(BlockNode *a, BlockNode *b){ |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2008 if((a->type&BLOCK_INTRA) && (b->type&BLOCK_INTRA)){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2009 return !((a->color[0] - b->color[0]) | (a->color[1] - b->color[1]) | (a->color[2] - b->color[2])); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2010 }else{ |
3314 | 2011 return !((a->mx - b->mx) | (a->my - b->my) | (a->ref - b->ref) | ((a->type ^ b->type)&BLOCK_INTRA)); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2012 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2013 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2014 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2015 static void encode_q_branch2(SnowContext *s, int level, int x, int y){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2016 const int w= s->b_width << s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2017 const int rem_depth= s->block_max_depth - level; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2018 const int index= (x + y*w) << rem_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2019 int trx= (x+1)<<rem_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2020 BlockNode *b= &s->block[index]; |
4408
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2021 const BlockNode *left = x ? &s->block[index-1] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2022 const BlockNode *top = y ? &s->block[index-w] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2023 const BlockNode *tl = y && x ? &s->block[index-w-1] : left; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2024 const BlockNode *tr = y && trx<w && ((x&1)==0 || level==0) ? &s->block[index-w+(1<<rem_depth)] : tl; //FIXME use lt |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2025 int pl = left->color[0]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2026 int pcb= left->color[1]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2027 int pcr= left->color[2]; |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2028 int pmx, pmy; |
3314 | 2029 int ref_context= av_log2(2*left->ref) + av_log2(2*top->ref); |
4001 | 2030 int mx_context= av_log2(2*FFABS(left->mx - top->mx)) + 16*!!b->ref; |
2031 int my_context= av_log2(2*FFABS(left->my - top->my)) + 16*!!b->ref; | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2032 int s_context= 2*left->level + 2*top->level + tl->level + tr->level; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2033 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2034 if(s->keyframe){ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2035 set_blocks(s, level, x, y, pl, pcb, pcr, 0, 0, 0, BLOCK_INTRA); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2036 return; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2037 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2038 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2039 if(level!=s->block_max_depth){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2040 if(same_block(b,b+1) && same_block(b,b+w) && same_block(b,b+w+1)){ |
2995
dfc271b90fe6
4mv + iter ME 10l fix (still not bugfree but better)
michael
parents:
2994
diff
changeset
|
2041 put_rac(&s->c, &s->block_state[4 + s_context], 1); |
dfc271b90fe6
4mv + iter ME 10l fix (still not bugfree but better)
michael
parents:
2994
diff
changeset
|
2042 }else{ |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2043 put_rac(&s->c, &s->block_state[4 + s_context], 0); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2044 encode_q_branch2(s, level+1, 2*x+0, 2*y+0); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2045 encode_q_branch2(s, level+1, 2*x+1, 2*y+0); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2046 encode_q_branch2(s, level+1, 2*x+0, 2*y+1); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2047 encode_q_branch2(s, level+1, 2*x+1, 2*y+1); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2048 return; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2049 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2050 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2051 if(b->type & BLOCK_INTRA){ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2052 pred_mv(s, &pmx, &pmy, 0, left, top, tr); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2053 put_rac(&s->c, &s->block_state[1 + (left->type&1) + (top->type&1)], 1); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2054 put_symbol(&s->c, &s->block_state[32], b->color[0]-pl , 1); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2055 put_symbol(&s->c, &s->block_state[64], b->color[1]-pcb, 1); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2056 put_symbol(&s->c, &s->block_state[96], b->color[2]-pcr, 1); |
3314 | 2057 set_blocks(s, level, x, y, b->color[0], b->color[1], b->color[2], pmx, pmy, 0, BLOCK_INTRA); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2058 }else{ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2059 pred_mv(s, &pmx, &pmy, b->ref, left, top, tr); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2060 put_rac(&s->c, &s->block_state[1 + (left->type&1) + (top->type&1)], 0); |
3314 | 2061 if(s->ref_frames > 1) |
2062 put_symbol(&s->c, &s->block_state[128 + 1024 + 32*ref_context], b->ref, 0); | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2063 put_symbol(&s->c, &s->block_state[128 + 32*mx_context], b->mx - pmx, 1); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2064 put_symbol(&s->c, &s->block_state[128 + 32*my_context], b->my - pmy, 1); |
3314 | 2065 set_blocks(s, level, x, y, pl, pcb, pcr, b->mx, b->my, b->ref, 0); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2066 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2067 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2068 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2069 static void decode_q_branch(SnowContext *s, int level, int x, int y){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2070 const int w= s->b_width << s->block_max_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2071 const int rem_depth= s->block_max_depth - level; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2072 const int index= (x + y*w) << rem_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2073 int trx= (x+1)<<rem_depth; |
4408
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2074 const BlockNode *left = x ? &s->block[index-1] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2075 const BlockNode *top = y ? &s->block[index-w] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2076 const BlockNode *tl = y && x ? &s->block[index-w-1] : left; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2077 const BlockNode *tr = y && trx<w && ((x&1)==0 || level==0) ? &s->block[index-w+(1<<rem_depth)] : tl; //FIXME use lt |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2078 int s_context= 2*left->level + 2*top->level + tl->level + tr->level; |
2967 | 2079 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2080 if(s->keyframe){ |
3314 | 2081 set_blocks(s, level, x, y, null_block.color[0], null_block.color[1], null_block.color[2], null_block.mx, null_block.my, null_block.ref, BLOCK_INTRA); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2082 return; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2083 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2084 |
2335 | 2085 if(level==s->block_max_depth || get_rac(&s->c, &s->block_state[4 + s_context])){ |
4332 | 2086 int type, mx, my; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2087 int l = left->color[0]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2088 int cb= left->color[1]; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2089 int cr= left->color[2]; |
3314 | 2090 int ref = 0; |
2091 int ref_context= av_log2(2*left->ref) + av_log2(2*top->ref); | |
4001 | 2092 int mx_context= av_log2(2*FFABS(left->mx - top->mx)) + 0*av_log2(2*FFABS(tr->mx - top->mx)); |
2093 int my_context= av_log2(2*FFABS(left->my - top->my)) + 0*av_log2(2*FFABS(tr->my - top->my)); | |
2967 | 2094 |
2335 | 2095 type= get_rac(&s->c, &s->block_state[1 + left->type + top->type]) ? BLOCK_INTRA : 0; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2096 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2097 if(type){ |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2098 pred_mv(s, &mx, &my, 0, left, top, tr); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2099 l += get_symbol(&s->c, &s->block_state[32], 1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2100 cb+= get_symbol(&s->c, &s->block_state[64], 1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2101 cr+= get_symbol(&s->c, &s->block_state[96], 1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2102 }else{ |
3314 | 2103 if(s->ref_frames > 1) |
2104 ref= get_symbol(&s->c, &s->block_state[128 + 1024 + 32*ref_context], 0); | |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2105 pred_mv(s, &mx, &my, ref, left, top, tr); |
3314 | 2106 mx+= get_symbol(&s->c, &s->block_state[128 + 32*(mx_context + 16*!!ref)], 1); |
2107 my+= get_symbol(&s->c, &s->block_state[128 + 32*(my_context + 16*!!ref)], 1); | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2108 } |
3314 | 2109 set_blocks(s, level, x, y, l, cb, cr, mx, my, ref, type); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2110 }else{ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2111 decode_q_branch(s, level+1, 2*x+0, 2*y+0); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2112 decode_q_branch(s, level+1, 2*x+1, 2*y+0); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2113 decode_q_branch(s, level+1, 2*x+0, 2*y+1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2114 decode_q_branch(s, level+1, 2*x+1, 2*y+1); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2115 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2116 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2117 |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
2118 static void encode_blocks(SnowContext *s, int search){ |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2119 int x, y; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2120 int w= s->b_width; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2121 int h= s->b_height; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2122 |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
2123 if(s->avctx->me_method == ME_ITER && !s->keyframe && search) |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2124 iterative_me(s); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2125 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2126 for(y=0; y<h; y++){ |
2435
c89ac0e70c66
10l patch by (matthieu castet <castet.matthieu free fr>)
michael
parents:
2422
diff
changeset
|
2127 if(s->c.bytestream_end - s->c.bytestream < w*MB_SIZE*MB_SIZE*3){ //FIXME nicer limit |
2422 | 2128 av_log(s->avctx, AV_LOG_ERROR, "encoded frame too large\n"); |
2129 return; | |
2130 } | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2131 for(x=0; x<w; x++){ |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
2132 if(s->avctx->me_method == ME_ITER || !search) |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2133 encode_q_branch2(s, 0, x, y); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2134 else |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2135 encode_q_branch (s, 0, x, y); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2136 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2137 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2138 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2139 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2140 static void decode_blocks(SnowContext *s){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2141 int x, y; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2142 int w= s->b_width; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2143 int h= s->b_height; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2144 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2145 for(y=0; y<h; y++){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2146 for(x=0; x<w; x++){ |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2147 decode_q_branch(s, 0, x, y); |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2148 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2149 } |
2138 | 2150 } |
2151 | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2152 static void mc_block(Plane *p, uint8_t *dst, const uint8_t *src, uint8_t *tmp, int stride, int b_w, int b_h, int dx, int dy){ |
5648 | 2153 const static uint8_t weight[64]={ |
2154 8,7,6,5,4,3,2,1, | |
2155 7,7,0,0,0,0,0,1, | |
2156 6,0,6,0,0,0,2,0, | |
2157 5,0,0,5,0,3,0,0, | |
2158 4,0,0,0,4,0,0,0, | |
2159 3,0,0,5,0,3,0,0, | |
2160 2,0,6,0,0,0,2,0, | |
2161 1,7,0,0,0,0,0,1, | |
2162 }; | |
2163 | |
2164 const static uint8_t brane[256]={ | |
2165 0x00,0x01,0x01,0x01,0x01,0x01,0x01,0x01,0x11,0x12,0x12,0x12,0x12,0x12,0x12,0x12, | |
2166 0x04,0x05,0xcc,0xcc,0xcc,0xcc,0xcc,0x41,0x15,0x16,0xcc,0xcc,0xcc,0xcc,0xcc,0x52, | |
2167 0x04,0xcc,0x05,0xcc,0xcc,0xcc,0x41,0xcc,0x15,0xcc,0x16,0xcc,0xcc,0xcc,0x52,0xcc, | |
2168 0x04,0xcc,0xcc,0x05,0xcc,0x41,0xcc,0xcc,0x15,0xcc,0xcc,0x16,0xcc,0x52,0xcc,0xcc, | |
2169 0x04,0xcc,0xcc,0xcc,0x41,0xcc,0xcc,0xcc,0x15,0xcc,0xcc,0xcc,0x16,0xcc,0xcc,0xcc, | |
2170 0x04,0xcc,0xcc,0x41,0xcc,0x05,0xcc,0xcc,0x15,0xcc,0xcc,0x52,0xcc,0x16,0xcc,0xcc, | |
2171 0x04,0xcc,0x41,0xcc,0xcc,0xcc,0x05,0xcc,0x15,0xcc,0x52,0xcc,0xcc,0xcc,0x16,0xcc, | |
2172 0x04,0x41,0xcc,0xcc,0xcc,0xcc,0xcc,0x05,0x15,0x52,0xcc,0xcc,0xcc,0xcc,0xcc,0x16, | |
2173 0x44,0x45,0x45,0x45,0x45,0x45,0x45,0x45,0x55,0x56,0x56,0x56,0x56,0x56,0x56,0x56, | |
2174 0x48,0x49,0xcc,0xcc,0xcc,0xcc,0xcc,0x85,0x59,0x5A,0xcc,0xcc,0xcc,0xcc,0xcc,0x96, | |
2175 0x48,0xcc,0x49,0xcc,0xcc,0xcc,0x85,0xcc,0x59,0xcc,0x5A,0xcc,0xcc,0xcc,0x96,0xcc, | |
2176 0x48,0xcc,0xcc,0x49,0xcc,0x85,0xcc,0xcc,0x59,0xcc,0xcc,0x5A,0xcc,0x96,0xcc,0xcc, | |
2177 0x48,0xcc,0xcc,0xcc,0x49,0xcc,0xcc,0xcc,0x59,0xcc,0xcc,0xcc,0x96,0xcc,0xcc,0xcc, | |
2178 0x48,0xcc,0xcc,0x85,0xcc,0x49,0xcc,0xcc,0x59,0xcc,0xcc,0x96,0xcc,0x5A,0xcc,0xcc, | |
2179 0x48,0xcc,0x85,0xcc,0xcc,0xcc,0x49,0xcc,0x59,0xcc,0x96,0xcc,0xcc,0xcc,0x5A,0xcc, | |
2180 0x48,0x85,0xcc,0xcc,0xcc,0xcc,0xcc,0x49,0x59,0x96,0xcc,0xcc,0xcc,0xcc,0xcc,0x5A, | |
2181 }; | |
2182 | |
2183 const static uint8_t needs[16]={ | |
2184 0,1,0,0, | |
2185 2,4,2,0, | |
2186 0,1,0,0, | |
2187 15 | |
2188 }; | |
2189 | |
2190 int x, y, b, r, l; | |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2191 int16_t tmpIt [64*(32+HTAPS_MAX)]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2192 uint8_t tmp2t[3][stride*(32+HTAPS_MAX)]; |
5648 | 2193 int16_t *tmpI= tmpIt; |
2194 uint8_t *tmp2= tmp2t[0]; | |
2195 uint8_t *hpel[11]; | |
2221 | 2196 START_TIMER |
5648 | 2197 assert(dx<16 && dy<16); |
2198 r= brane[dx + 16*dy]&15; | |
2199 l= brane[dx + 16*dy]>>4; | |
2200 | |
2201 b= needs[l] | needs[r]; | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2202 if(p && !p->diag_mc) |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2203 b= 15; |
5648 | 2204 |
2205 if(b&5){ | |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2206 for(y=0; y < b_h+HTAPS_MAX-1; y++){ |
5649 | 2207 for(x=0; x < b_w; x++){ |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2208 int a_1=src[x + HTAPS_MAX/2-4]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2209 int a0= src[x + HTAPS_MAX/2-3]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2210 int a1= src[x + HTAPS_MAX/2-2]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2211 int a2= src[x + HTAPS_MAX/2-1]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2212 int a3= src[x + HTAPS_MAX/2+0]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2213 int a4= src[x + HTAPS_MAX/2+1]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2214 int a5= src[x + HTAPS_MAX/2+2]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2215 int a6= src[x + HTAPS_MAX/2+3]; |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2216 int am=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2217 if(!p || p->fast_mc){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2218 am= 20*(a2+a3) - 5*(a1+a4) + (a0+a5); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2219 tmpI[x]= am; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2220 am= (am+16)>>5; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2221 }else{ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2222 am= p->hcoeff[0]*(a2+a3) + p->hcoeff[1]*(a1+a4) + p->hcoeff[2]*(a0+a5) + p->hcoeff[3]*(a_1+a6); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2223 tmpI[x]= am; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2224 am= (am+32)>>6; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2225 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2226 |
5649 | 2227 if(am&(~255)) am= ~(am>>31); |
2228 tmp2[x]= am; | |
2229 } | |
2230 tmpI+= 64; | |
2231 tmp2+= stride; | |
2232 src += stride; | |
2138 | 2233 } |
5649 | 2234 src -= stride*y; |
5648 | 2235 } |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2236 src += HTAPS_MAX/2 - 1; |
5648 | 2237 tmp2= tmp2t[1]; |
2238 | |
2239 if(b&2){ | |
5649 | 2240 for(y=0; y < b_h; y++){ |
2241 for(x=0; x < b_w+1; x++){ | |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2242 int a_1=src[x + (HTAPS_MAX/2-4)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2243 int a0= src[x + (HTAPS_MAX/2-3)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2244 int a1= src[x + (HTAPS_MAX/2-2)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2245 int a2= src[x + (HTAPS_MAX/2-1)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2246 int a3= src[x + (HTAPS_MAX/2+0)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2247 int a4= src[x + (HTAPS_MAX/2+1)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2248 int a5= src[x + (HTAPS_MAX/2+2)*stride]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2249 int a6= src[x + (HTAPS_MAX/2+3)*stride]; |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2250 int am=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2251 if(!p || p->fast_mc) |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2252 am= (20*(a2+a3) - 5*(a1+a4) + (a0+a5) + 16)>>5; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2253 else |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2254 am= (p->hcoeff[0]*(a2+a3) + p->hcoeff[1]*(a1+a4) + p->hcoeff[2]*(a0+a5) + p->hcoeff[3]*(a_1+a6) + 32)>>6; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2255 |
5649 | 2256 if(am&(~255)) am= ~(am>>31); |
2257 tmp2[x]= am; | |
2258 } | |
2259 src += stride; | |
2260 tmp2+= stride; | |
5648 | 2261 } |
5649 | 2262 src -= stride*y; |
5648 | 2263 } |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2264 src += stride*(HTAPS_MAX/2 - 1); |
5648 | 2265 tmp2= tmp2t[2]; |
2266 tmpI= tmpIt; | |
2267 if(b&4){ | |
2268 for(y=0; y < b_h; y++){ | |
2269 for(x=0; x < b_w; x++){ | |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2270 int a_1=tmpI[x + (HTAPS_MAX/2-4)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2271 int a0= tmpI[x + (HTAPS_MAX/2-3)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2272 int a1= tmpI[x + (HTAPS_MAX/2-2)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2273 int a2= tmpI[x + (HTAPS_MAX/2-1)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2274 int a3= tmpI[x + (HTAPS_MAX/2+0)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2275 int a4= tmpI[x + (HTAPS_MAX/2+1)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2276 int a5= tmpI[x + (HTAPS_MAX/2+2)*64]; |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2277 int a6= tmpI[x + (HTAPS_MAX/2+3)*64]; |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2278 int am=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2279 if(!p || p->fast_mc) |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2280 am= (20*(a2+a3) - 5*(a1+a4) + (a0+a5) + 512)>>10; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2281 else |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2282 am= (p->hcoeff[0]*(a2+a3) + p->hcoeff[1]*(a1+a4) + p->hcoeff[2]*(a0+a5) + p->hcoeff[3]*(a_1+a6) + 2048)>>12; |
5648 | 2283 if(am&(~255)) am= ~(am>>31); |
2284 tmp2[x]= am; | |
2285 } | |
2286 tmpI+= 64; | |
2287 tmp2+= stride; | |
2138 | 2288 } |
5648 | 2289 } |
2290 | |
2291 hpel[ 0]= src; | |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2292 hpel[ 1]= tmp2t[0] + stride*(HTAPS_MAX/2-1); |
5648 | 2293 hpel[ 2]= src + 1; |
2294 | |
2295 hpel[ 4]= tmp2t[1]; | |
2296 hpel[ 5]= tmp2t[2]; | |
2297 hpel[ 6]= tmp2t[1] + 1; | |
2298 | |
2299 hpel[ 8]= src + stride; | |
2300 hpel[ 9]= hpel[1] + stride; | |
2301 hpel[10]= hpel[8] + 1; | |
2302 | |
2303 if(b==15){ | |
2304 uint8_t *src1= hpel[dx/8 + dy/8*4 ]; | |
2305 uint8_t *src2= hpel[dx/8 + dy/8*4+1]; | |
2306 uint8_t *src3= hpel[dx/8 + dy/8*4+4]; | |
2307 uint8_t *src4= hpel[dx/8 + dy/8*4+5]; | |
2308 dx&=7; | |
2309 dy&=7; | |
2310 for(y=0; y < b_h; y++){ | |
2311 for(x=0; x < b_w; x++){ | |
2312 dst[x]= ((8-dx)*(8-dy)*src1[x] + dx*(8-dy)*src2[x]+ | |
2313 (8-dx)* dy *src3[x] + dx* dy *src4[x]+32)>>6; | |
2314 } | |
2315 src1+=stride; | |
2316 src2+=stride; | |
2317 src3+=stride; | |
2318 src4+=stride; | |
2319 dst +=stride; | |
2320 } | |
2321 }else{ | |
2322 uint8_t *src1= hpel[l]; | |
2323 uint8_t *src2= hpel[r]; | |
2324 int a= weight[((dx&7) + (8*(dy&7)))]; | |
2325 int b= 8-a; | |
2326 for(y=0; y < b_h; y++){ | |
2327 for(x=0; x < b_w; x++){ | |
2328 dst[x]= (a*src1[x] + b*src2[x] + 4)>>3; | |
2329 } | |
2330 src1+=stride; | |
2331 src2+=stride; | |
2332 dst +=stride; | |
2333 } | |
2138 | 2334 } |
2221 | 2335 STOP_TIMER("mc_block") |
2138 | 2336 } |
2337 | |
2338 #define mca(dx,dy,b_w)\ | |
5254 | 2339 static void mc_block_hpel ## dx ## dy ## b_w(uint8_t *dst, const uint8_t *src, int stride, int h){\ |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2340 uint8_t tmp[stride*(b_w+HTAPS_MAX-1)];\ |
2138 | 2341 assert(h==b_w);\ |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2342 mc_block(NULL, dst, src-(HTAPS_MAX/2-1)-(HTAPS_MAX/2-1)*stride, tmp, stride, b_w, b_w, dx, dy);\ |
2138 | 2343 } |
2344 | |
2345 mca( 0, 0,16) | |
2346 mca( 8, 0,16) | |
2347 mca( 0, 8,16) | |
2348 mca( 8, 8,16) | |
2224
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
2349 mca( 0, 0,8) |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
2350 mca( 8, 0,8) |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
2351 mca( 0, 8,8) |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
2352 mca( 8, 8,8) |
2138 | 2353 |
3314 | 2354 static void pred_block(SnowContext *s, uint8_t *dst, uint8_t *tmp, int stride, int sx, int sy, int b_w, int b_h, BlockNode *block, int plane_index, int w, int h){ |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2355 if(block->type & BLOCK_INTRA){ |
2206 | 2356 int x, y; |
3018 | 2357 const int color = block->color[plane_index]; |
2358 const int color4= color*0x01010101; | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2359 if(b_w==32){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2360 for(y=0; y < b_h; y++){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2361 *(uint32_t*)&dst[0 + y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2362 *(uint32_t*)&dst[4 + y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2363 *(uint32_t*)&dst[8 + y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2364 *(uint32_t*)&dst[12+ y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2365 *(uint32_t*)&dst[16+ y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2366 *(uint32_t*)&dst[20+ y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2367 *(uint32_t*)&dst[24+ y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2368 *(uint32_t*)&dst[28+ y*stride]= color4; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2369 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2370 }else if(b_w==16){ |
3018 | 2371 for(y=0; y < b_h; y++){ |
2372 *(uint32_t*)&dst[0 + y*stride]= color4; | |
2373 *(uint32_t*)&dst[4 + y*stride]= color4; | |
2374 *(uint32_t*)&dst[8 + y*stride]= color4; | |
2375 *(uint32_t*)&dst[12+ y*stride]= color4; | |
2376 } | |
2377 }else if(b_w==8){ | |
2378 for(y=0; y < b_h; y++){ | |
2379 *(uint32_t*)&dst[0 + y*stride]= color4; | |
2380 *(uint32_t*)&dst[4 + y*stride]= color4; | |
2381 } | |
2382 }else if(b_w==4){ | |
2383 for(y=0; y < b_h; y++){ | |
2384 *(uint32_t*)&dst[0 + y*stride]= color4; | |
2385 } | |
2386 }else{ | |
2387 for(y=0; y < b_h; y++){ | |
2388 for(x=0; x < b_w; x++){ | |
2389 dst[x + y*stride]= color; | |
2390 } | |
2138 | 2391 } |
2392 } | |
2393 }else{ | |
3314 | 2394 uint8_t *src= s->last_picture[block->ref].data[plane_index]; |
2206 | 2395 const int scale= plane_index ? s->mv_scale : 2*s->mv_scale; |
2396 int mx= block->mx*scale; | |
2397 int my= block->my*scale; | |
2223 | 2398 const int dx= mx&15; |
2399 const int dy= my&15; | |
3020
c75fb0747e74
use h264 MC functions for 2xX Xx2 blocks in snow too
michael
parents:
3018
diff
changeset
|
2400 const int tab_index= 3 - (b_w>>2) + (b_w>>4); |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2401 sx += (mx>>4) - (HTAPS_MAX/2-1); |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2402 sy += (my>>4) - (HTAPS_MAX/2-1); |
2206 | 2403 src += sx + sy*stride; |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2404 if( (unsigned)sx >= w - b_w - (HTAPS_MAX-2) |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2405 || (unsigned)sy >= h - b_h - (HTAPS_MAX-2)){ |
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2406 ff_emulated_edge_mc(tmp + MB_SIZE, src, stride, b_w+HTAPS_MAX-1, b_h+HTAPS_MAX-1, sx, sy, w, h); |
2206 | 2407 src= tmp + MB_SIZE; |
2138 | 2408 } |
3189 | 2409 // assert(b_w == b_h || 2*b_w == b_h || b_w == 2*b_h); |
2410 // assert(!(b_w&(b_w-1))); | |
3018 | 2411 assert(b_w>1 && b_h>1); |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2412 assert(tab_index>=0 && tab_index<4 || b_w==32); |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2413 if((dx&3) || (dy&3) || !(b_w == b_h || 2*b_w == b_h || b_w == 2*b_h) || (b_w&(b_w-1)) || !s->plane[plane_index].fast_mc ) |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2414 mc_block(&s->plane[plane_index], dst, src, tmp, stride, b_w, b_h, dx, dy); |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2415 else if(b_w==32){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2416 int y; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2417 for(y=0; y<b_h; y+=16){ |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2418 s->dsp.put_h264_qpel_pixels_tab[0][dy+(dx>>2)](dst + y*stride, src + 3 + (y+3)*stride,stride); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2419 s->dsp.put_h264_qpel_pixels_tab[0][dy+(dx>>2)](dst + 16 + y*stride, src + 19 + (y+3)*stride,stride); |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2420 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2421 }else if(b_w==b_h) |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2422 s->dsp.put_h264_qpel_pixels_tab[tab_index ][dy+(dx>>2)](dst,src + 3 + 3*stride,stride); |
3018 | 2423 else if(b_w==2*b_h){ |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2424 s->dsp.put_h264_qpel_pixels_tab[tab_index+1][dy+(dx>>2)](dst ,src + 3 + 3*stride,stride); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2425 s->dsp.put_h264_qpel_pixels_tab[tab_index+1][dy+(dx>>2)](dst+b_h,src + 3 + b_h + 3*stride,stride); |
3018 | 2426 }else{ |
2427 assert(2*b_w==b_h); | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2428 s->dsp.put_h264_qpel_pixels_tab[tab_index ][dy+(dx>>2)](dst ,src + 3 + 3*stride ,stride); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
2429 s->dsp.put_h264_qpel_pixels_tab[tab_index ][dy+(dx>>2)](dst+b_w*stride,src + 3 + 3*stride+b_w*stride,stride); |
3018 | 2430 } |
2138 | 2431 } |
2432 } | |
2433 | |
4436
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2434 void ff_snow_inner_add_yblock(const uint8_t *obmc, const int obmc_stride, uint8_t * * block, int b_w, int b_h, |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2435 int src_x, int src_y, int src_stride, slice_buffer * sb, int add, uint8_t * dst8){ |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2436 int y, x; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2437 IDWTELEM * dst; |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2438 for(y=0; y<b_h; y++){ |
5409 | 2439 //FIXME ugly misuse of obmc_stride |
4436
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2440 const uint8_t *obmc1= obmc + y*obmc_stride; |
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2441 const uint8_t *obmc2= obmc1+ (obmc_stride>>1); |
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2442 const uint8_t *obmc3= obmc1+ obmc_stride*(obmc_stride>>1); |
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2443 const uint8_t *obmc4= obmc3+ (obmc_stride>>1); |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2444 dst = slice_buffer_get_line(sb, src_y + y); |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2445 for(x=0; x<b_w; x++){ |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2446 int v= obmc1[x] * block[3][x + y*src_stride] |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2447 +obmc2[x] * block[2][x + y*src_stride] |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2448 +obmc3[x] * block[1][x + y*src_stride] |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2449 +obmc4[x] * block[0][x + y*src_stride]; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2450 |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2451 v <<= 8 - LOG2_OBMC_MAX; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2452 if(FRAC_BITS != 8){ |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2453 v >>= 8 - FRAC_BITS; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2454 } |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2455 if(add){ |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2456 v += dst[x + src_x]; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2457 v = (v + (1<<(FRAC_BITS-1))) >> FRAC_BITS; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2458 if(v&(~255)) v= ~(v>>31); |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2459 dst8[x + y*src_stride] = v; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2460 }else{ |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2461 dst[x + src_x] -= v; |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2462 } |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2463 } |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2464 } |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2465 } |
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
2466 |
2206 | 2467 //FIXME name clenup (b_w, block_w, b_width stuff) |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2468 static av_always_inline void add_yblock(SnowContext *s, int sliced, slice_buffer *sb, IDWTELEM *dst, uint8_t *dst8, const uint8_t *obmc, int src_x, int src_y, int b_w, int b_h, int w, int h, int dst_stride, int src_stride, int obmc_stride, int b_x, int b_y, int add, int offset_dst, int plane_index){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2469 const int b_width = s->b_width << s->block_max_depth; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2470 const int b_height= s->b_height << s->block_max_depth; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2471 const int b_stride= b_width; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2472 BlockNode *lt= &s->block[b_x + b_y*b_stride]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2473 BlockNode *rt= lt+1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2474 BlockNode *lb= lt+b_stride; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2475 BlockNode *rb= lb+1; |
2967 | 2476 uint8_t *block[4]; |
2842 | 2477 int tmp_step= src_stride >= 7*MB_SIZE ? MB_SIZE : MB_SIZE*src_stride; |
2478 uint8_t tmp[src_stride*7*MB_SIZE]; //FIXME align | |
2479 uint8_t *ptmp; | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2480 int x,y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2481 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2482 if(b_x<0){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2483 lt= rt; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2484 lb= rb; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2485 }else if(b_x + 1 >= b_width){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2486 rt= lt; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2487 rb= lb; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2488 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2489 if(b_y<0){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2490 lt= lb; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2491 rt= rb; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2492 }else if(b_y + 1 >= b_height){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2493 lb= lt; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2494 rb= rt; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2495 } |
2967 | 2496 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2497 if(src_x<0){ //FIXME merge with prev & always round internal width upto *16 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2498 obmc -= src_x; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2499 b_w += src_x; |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2500 if(!sliced && !offset_dst) |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2501 dst -= src_x; |
2206 | 2502 src_x=0; |
2503 }else if(src_x + b_w > w){ | |
2504 b_w = w - src_x; | |
2505 } | |
2506 if(src_y<0){ | |
2507 obmc -= src_y*obmc_stride; | |
2508 b_h += src_y; | |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2509 if(!sliced && !offset_dst) |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2510 dst -= src_y*dst_stride; |
2206 | 2511 src_y=0; |
2512 }else if(src_y + b_h> h){ | |
2513 b_h = h - src_y; | |
2514 } | |
2967 | 2515 |
2206 | 2516 if(b_w<=0 || b_h<=0) return; |
2517 | |
2842 | 2518 assert(src_stride > 2*MB_SIZE + 5); |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2519 if(!sliced && offset_dst) |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2520 dst += src_x + src_y*dst_stride; |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2521 dst8+= src_x + src_y*src_stride; |
2206 | 2522 // src += src_x + src_y*src_stride; |
2523 | |
2842 | 2524 ptmp= tmp + 3*tmp_step; |
2525 block[0]= ptmp; | |
2526 ptmp+=tmp_step; | |
3314 | 2527 pred_block(s, block[0], tmp, src_stride, src_x, src_y, b_w, b_h, lt, plane_index, w, h); |
2206 | 2528 |
2529 if(same_block(lt, rt)){ | |
2530 block[1]= block[0]; | |
2531 }else{ | |
2842 | 2532 block[1]= ptmp; |
2533 ptmp+=tmp_step; | |
3314 | 2534 pred_block(s, block[1], tmp, src_stride, src_x, src_y, b_w, b_h, rt, plane_index, w, h); |
2206 | 2535 } |
2967 | 2536 |
2206 | 2537 if(same_block(lt, lb)){ |
2538 block[2]= block[0]; | |
2539 }else if(same_block(rt, lb)){ | |
2540 block[2]= block[1]; | |
2541 }else{ | |
2842 | 2542 block[2]= ptmp; |
2543 ptmp+=tmp_step; | |
3314 | 2544 pred_block(s, block[2], tmp, src_stride, src_x, src_y, b_w, b_h, lb, plane_index, w, h); |
2206 | 2545 } |
2546 | |
2547 if(same_block(lt, rb) ){ | |
2548 block[3]= block[0]; | |
2549 }else if(same_block(rt, rb)){ | |
2550 block[3]= block[1]; | |
2551 }else if(same_block(lb, rb)){ | |
2552 block[3]= block[2]; | |
2553 }else{ | |
2842 | 2554 block[3]= ptmp; |
3314 | 2555 pred_block(s, block[3], tmp, src_stride, src_x, src_y, b_w, b_h, rb, plane_index, w, h); |
2206 | 2556 } |
2557 #if 0 | |
2558 for(y=0; y<b_h; y++){ | |
2559 for(x=0; x<b_w; x++){ | |
2560 int v= obmc [x + y*obmc_stride] * block[3][x + y*src_stride] * (256/OBMC_MAX); | |
2561 if(add) dst[x + y*dst_stride] += v; | |
2562 else dst[x + y*dst_stride] -= v; | |
2563 } | |
2564 } | |
2565 for(y=0; y<b_h; y++){ | |
2566 uint8_t *obmc2= obmc + (obmc_stride>>1); | |
2567 for(x=0; x<b_w; x++){ | |
2568 int v= obmc2[x + y*obmc_stride] * block[2][x + y*src_stride] * (256/OBMC_MAX); | |
2569 if(add) dst[x + y*dst_stride] += v; | |
2570 else dst[x + y*dst_stride] -= v; | |
2571 } | |
2572 } | |
2573 for(y=0; y<b_h; y++){ | |
2574 uint8_t *obmc3= obmc + obmc_stride*(obmc_stride>>1); | |
2575 for(x=0; x<b_w; x++){ | |
2576 int v= obmc3[x + y*obmc_stride] * block[1][x + y*src_stride] * (256/OBMC_MAX); | |
2577 if(add) dst[x + y*dst_stride] += v; | |
2578 else dst[x + y*dst_stride] -= v; | |
2579 } | |
2580 } | |
2581 for(y=0; y<b_h; y++){ | |
2582 uint8_t *obmc3= obmc + obmc_stride*(obmc_stride>>1); | |
2583 uint8_t *obmc4= obmc3+ (obmc_stride>>1); | |
2584 for(x=0; x<b_w; x++){ | |
2585 int v= obmc4[x + y*obmc_stride] * block[0][x + y*src_stride] * (256/OBMC_MAX); | |
2586 if(add) dst[x + y*dst_stride] += v; | |
2587 else dst[x + y*dst_stride] -= v; | |
2588 } | |
2589 } | |
2590 #else | |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2591 if(sliced){ |
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2592 START_TIMER |
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2593 |
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2594 s->dsp.inner_add_yblock(obmc, obmc_stride, block, b_w, b_h, src_x,src_y, src_stride, sb, add, dst8); |
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2595 STOP_TIMER("inner_add_yblock") |
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2596 }else |
2206 | 2597 for(y=0; y<b_h; y++){ |
5409 | 2598 //FIXME ugly misuse of obmc_stride |
4436
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2599 const uint8_t *obmc1= obmc + y*obmc_stride; |
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2600 const uint8_t *obmc2= obmc1+ (obmc_stride>>1); |
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2601 const uint8_t *obmc3= obmc1+ obmc_stride*(obmc_stride>>1); |
d3e389536b0a
Add the const specifier as needed to reduce the number of warnings.
takis
parents:
4409
diff
changeset
|
2602 const uint8_t *obmc4= obmc3+ (obmc_stride>>1); |
2206 | 2603 for(x=0; x<b_w; x++){ |
2604 int v= obmc1[x] * block[3][x + y*src_stride] | |
2605 +obmc2[x] * block[2][x + y*src_stride] | |
2606 +obmc3[x] * block[1][x + y*src_stride] | |
2607 +obmc4[x] * block[0][x + y*src_stride]; | |
2967 | 2608 |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2609 v <<= 8 - LOG2_OBMC_MAX; |
2246 | 2610 if(FRAC_BITS != 8){ |
2611 v >>= 8 - FRAC_BITS; | |
2612 } | |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2613 if(add){ |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2614 v += dst[x + y*dst_stride]; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2615 v = (v + (1<<(FRAC_BITS-1))) >> FRAC_BITS; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2616 if(v&(~255)) v= ~(v>>31); |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2617 dst8[x + y*src_stride] = v; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2618 }else{ |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2619 dst[x + y*dst_stride] -= v; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2620 } |
2206 | 2621 } |
2622 } | |
2623 #endif | |
2624 } | |
2625 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2626 static av_always_inline void predict_slice_buffered(SnowContext *s, slice_buffer * sb, IDWTELEM * old_buffer, int plane_index, int add, int mb_y){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2627 Plane *p= &s->plane[plane_index]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2628 const int mb_w= s->b_width << s->block_max_depth; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2629 const int mb_h= s->b_height << s->block_max_depth; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2630 int x, y, mb_x; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2631 int block_size = MB_SIZE >> s->block_max_depth; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2632 int block_w = plane_index ? block_size/2 : block_size; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2633 const uint8_t *obmc = plane_index ? obmc_tab[s->block_max_depth+1] : obmc_tab[s->block_max_depth]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2634 int obmc_stride= plane_index ? block_size : 2*block_size; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2635 int ref_stride= s->current_picture.linesize[plane_index]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2636 uint8_t *dst8= s->current_picture.data[plane_index]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2637 int w= p->width; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2638 int h= p->height; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2639 START_TIMER |
2967 | 2640 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2641 if(s->keyframe || (s->avctx->debug&512)){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2642 if(mb_y==mb_h) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2643 return; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2644 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2645 if(add){ |
2604
b7e6c3d31c65
Snow segfault bug in revision 1.43 patch by (Yartrebo /yartrebo earthlink net/)
michael
parents:
2602
diff
changeset
|
2646 for(y=block_w*mb_y; y<FFMIN(h,block_w*(mb_y+1)); y++) |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2647 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2648 // DWTELEM * line = slice_buffer_get_line(sb, y); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2649 IDWTELEM * line = sb->line[y]; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2650 for(x=0; x<w; x++) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2651 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2652 // int v= buf[x + y*w] + (128<<FRAC_BITS) + (1<<(FRAC_BITS-1)); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2653 int v= line[x] + (128<<FRAC_BITS) + (1<<(FRAC_BITS-1)); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2654 v >>= FRAC_BITS; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2655 if(v&(~255)) v= ~(v>>31); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2656 dst8[x + y*ref_stride]= v; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2657 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2658 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2659 }else{ |
2604
b7e6c3d31c65
Snow segfault bug in revision 1.43 patch by (Yartrebo /yartrebo earthlink net/)
michael
parents:
2602
diff
changeset
|
2660 for(y=block_w*mb_y; y<FFMIN(h,block_w*(mb_y+1)); y++) |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2661 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2662 // DWTELEM * line = slice_buffer_get_line(sb, y); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2663 IDWTELEM * line = sb->line[y]; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2664 for(x=0; x<w; x++) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2665 { |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2666 line[x] -= 128 << FRAC_BITS; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2667 // buf[x + y*w]-= 128<<FRAC_BITS; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2668 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2669 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2670 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2671 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2672 return; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2673 } |
2967 | 2674 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2675 for(mb_x=0; mb_x<=mb_w; mb_x++){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2676 START_TIMER |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2677 |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2678 add_yblock(s, 1, sb, old_buffer, dst8, obmc, |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2679 block_w*mb_x - block_w/2, |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2680 block_w*mb_y - block_w/2, |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2681 block_w, block_w, |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2682 w, h, |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2683 w, ref_stride, obmc_stride, |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2684 mb_x - 1, mb_y - 1, |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2685 add, 0, plane_index); |
2967 | 2686 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2687 STOP_TIMER("add_yblock") |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2688 } |
2967 | 2689 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2690 STOP_TIMER("predict_slice") |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2691 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
2692 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2693 static av_always_inline void predict_slice(SnowContext *s, IDWTELEM *buf, int plane_index, int add, int mb_y){ |
2138 | 2694 Plane *p= &s->plane[plane_index]; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2695 const int mb_w= s->b_width << s->block_max_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2696 const int mb_h= s->b_height << s->block_max_depth; |
2562 | 2697 int x, y, mb_x; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2698 int block_size = MB_SIZE >> s->block_max_depth; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
2699 int block_w = plane_index ? block_size/2 : block_size; |
2206 | 2700 const uint8_t *obmc = plane_index ? obmc_tab[s->block_max_depth+1] : obmc_tab[s->block_max_depth]; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2701 const int obmc_stride= plane_index ? block_size : 2*block_size; |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2702 int ref_stride= s->current_picture.linesize[plane_index]; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2703 uint8_t *dst8= s->current_picture.data[plane_index]; |
2138 | 2704 int w= p->width; |
2705 int h= p->height; | |
2197 | 2706 START_TIMER |
2967 | 2707 |
2206 | 2708 if(s->keyframe || (s->avctx->debug&512)){ |
2562 | 2709 if(mb_y==mb_h) |
2710 return; | |
2711 | |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2712 if(add){ |
2604
b7e6c3d31c65
Snow segfault bug in revision 1.43 patch by (Yartrebo /yartrebo earthlink net/)
michael
parents:
2602
diff
changeset
|
2713 for(y=block_w*mb_y; y<FFMIN(h,block_w*(mb_y+1)); y++){ |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2714 for(x=0; x<w; x++){ |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2715 int v= buf[x + y*w] + (128<<FRAC_BITS) + (1<<(FRAC_BITS-1)); |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2716 v >>= FRAC_BITS; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2717 if(v&(~255)) v= ~(v>>31); |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2718 dst8[x + y*ref_stride]= v; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2719 } |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2720 } |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2721 }else{ |
2604
b7e6c3d31c65
Snow segfault bug in revision 1.43 patch by (Yartrebo /yartrebo earthlink net/)
michael
parents:
2602
diff
changeset
|
2722 for(y=block_w*mb_y; y<FFMIN(h,block_w*(mb_y+1)); y++){ |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2723 for(x=0; x<w; x++){ |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2724 buf[x + y*w]-= 128<<FRAC_BITS; |
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
2725 } |
2206 | 2726 } |
2138 | 2727 } |
2206 | 2728 |
2729 return; | |
2138 | 2730 } |
2967 | 2731 |
2206 | 2732 for(mb_x=0; mb_x<=mb_w; mb_x++){ |
2197 | 2733 START_TIMER |
2206 | 2734 |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2735 add_yblock(s, 0, NULL, buf, dst8, obmc, |
2206 | 2736 block_w*mb_x - block_w/2, |
2138 | 2737 block_w*mb_y - block_w/2, |
2206 | 2738 block_w, block_w, |
2138 | 2739 w, h, |
2206 | 2740 w, ref_stride, obmc_stride, |
2741 mb_x - 1, mb_y - 1, | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2742 add, 1, plane_index); |
2967 | 2743 |
2206 | 2744 STOP_TIMER("add_yblock") |
2138 | 2745 } |
2967 | 2746 |
2562 | 2747 STOP_TIMER("predict_slice") |
2748 } | |
2749 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2750 static av_always_inline void predict_plane(SnowContext *s, IDWTELEM *buf, int plane_index, int add){ |
2562 | 2751 const int mb_h= s->b_height << s->block_max_depth; |
2752 int mb_y; | |
2753 for(mb_y=0; mb_y<=mb_h; mb_y++) | |
2754 predict_slice(s, buf, plane_index, add, mb_y); | |
2138 | 2755 } |
2756 | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2757 static int get_dc(SnowContext *s, int mb_x, int mb_y, int plane_index){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2758 int i, x2, y2; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2759 Plane *p= &s->plane[plane_index]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2760 const int block_size = MB_SIZE >> s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2761 const int block_w = plane_index ? block_size/2 : block_size; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2762 const uint8_t *obmc = plane_index ? obmc_tab[s->block_max_depth+1] : obmc_tab[s->block_max_depth]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2763 const int obmc_stride= plane_index ? block_size : 2*block_size; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2764 const int ref_stride= s->current_picture.linesize[plane_index]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2765 uint8_t *src= s-> input_picture.data[plane_index]; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2766 IDWTELEM *dst= (IDWTELEM*)s->m.obmc_scratchpad + plane_index*block_size*block_size*4; //FIXME change to unsigned |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2767 const int b_stride = s->b_width << s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2768 const int w= p->width; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2769 const int h= p->height; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2770 int index= mb_x + mb_y*b_stride; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2771 BlockNode *b= &s->block[index]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2772 BlockNode backup= *b; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2773 int ab=0; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2774 int aa=0; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2775 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2776 b->type|= BLOCK_INTRA; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2777 b->color[plane_index]= 0; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2778 memset(dst, 0, obmc_stride*obmc_stride*sizeof(IDWTELEM)); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2779 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2780 for(i=0; i<4; i++){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2781 int mb_x2= mb_x + (i &1) - 1; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2782 int mb_y2= mb_y + (i>>1) - 1; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2783 int x= block_w*mb_x2 + block_w/2; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2784 int y= block_w*mb_y2 + block_w/2; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2785 |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2786 add_yblock(s, 0, NULL, dst + ((i&1)+(i>>1)*obmc_stride)*block_w, NULL, obmc, |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2787 x, y, block_w, block_w, w, h, obmc_stride, ref_stride, obmc_stride, mb_x2, mb_y2, 0, 0, plane_index); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2788 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2789 for(y2= FFMAX(y, 0); y2<FFMIN(h, y+block_w); y2++){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2790 for(x2= FFMAX(x, 0); x2<FFMIN(w, x+block_w); x2++){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2791 int index= x2-(block_w*mb_x - block_w/2) + (y2-(block_w*mb_y - block_w/2))*obmc_stride; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2792 int obmc_v= obmc[index]; |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2793 int d; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2794 if(y<0) obmc_v += obmc[index + block_w*obmc_stride]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2795 if(x<0) obmc_v += obmc[index + block_w]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2796 if(y+block_w>h) obmc_v += obmc[index - block_w*obmc_stride]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2797 if(x+block_w>w) obmc_v += obmc[index - block_w]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2798 //FIXME precalc this or simplify it somehow else |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2799 |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2800 d = -dst[index] + (1<<(FRAC_BITS-1)); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2801 dst[index] = d; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2802 ab += (src[x2 + y2*ref_stride] - (d>>FRAC_BITS)) * obmc_v; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2803 aa += obmc_v * obmc_v; //FIXME precalclate this |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2804 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2805 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2806 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2807 *b= backup; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2808 |
5127 | 2809 return av_clip(((ab<<LOG2_OBMC_MAX) + aa/2)/aa, 0, 255); //FIXME we should not need clipping |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2810 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2811 |
3051 | 2812 static inline int get_block_bits(SnowContext *s, int x, int y, int w){ |
2813 const int b_stride = s->b_width << s->block_max_depth; | |
2814 const int b_height = s->b_height<< s->block_max_depth; | |
2815 int index= x + y*b_stride; | |
4408
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2816 const BlockNode *b = &s->block[index]; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2817 const BlockNode *left = x ? &s->block[index-1] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2818 const BlockNode *top = y ? &s->block[index-b_stride] : &null_block; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2819 const BlockNode *tl = y && x ? &s->block[index-b_stride-1] : left; |
c3be13eac2d6
Bring down the number of snow.c warnings from 27 to 17 by using the const
takis
parents:
4407
diff
changeset
|
2820 const BlockNode *tr = y && x+w<b_stride ? &s->block[index-b_stride+w] : tl; |
3051 | 2821 int dmx, dmy; |
4001 | 2822 // int mx_context= av_log2(2*FFABS(left->mx - top->mx)); |
2823 // int my_context= av_log2(2*FFABS(left->my - top->my)); | |
3051 | 2824 |
2825 if(x<0 || x>=b_stride || y>=b_height) | |
2826 return 0; | |
2827 /* | |
2828 1 0 0 | |
2829 01X 1-2 1 | |
2830 001XX 3-6 2-3 | |
2831 0001XXX 7-14 4-7 | |
2832 00001XXXX 15-30 8-15 | |
2833 */ | |
2834 //FIXME try accurate rate | |
2835 //FIXME intra and inter predictors if surrounding blocks arent the same type | |
2836 if(b->type & BLOCK_INTRA){ | |
4001 | 2837 return 3+2*( av_log2(2*FFABS(left->color[0] - b->color[0])) |
2838 + av_log2(2*FFABS(left->color[1] - b->color[1])) | |
2839 + av_log2(2*FFABS(left->color[2] - b->color[2]))); | |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2840 }else{ |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2841 pred_mv(s, &dmx, &dmy, b->ref, left, top, tr); |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2842 dmx-= b->mx; |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2843 dmy-= b->my; |
4001 | 2844 return 2*(1 + av_log2(2*FFABS(dmx)) //FIXME kill the 2* can be merged in lambda |
2845 + av_log2(2*FFABS(dmy)) | |
3314 | 2846 + av_log2(2*b->ref)); |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
2847 } |
3051 | 2848 } |
2849 | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2850 static int get_block_rd(SnowContext *s, int mb_x, int mb_y, int plane_index, const uint8_t *obmc_edged){ |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2851 Plane *p= &s->plane[plane_index]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2852 const int block_size = MB_SIZE >> s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2853 const int block_w = plane_index ? block_size/2 : block_size; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2854 const int obmc_stride= plane_index ? block_size : 2*block_size; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2855 const int ref_stride= s->current_picture.linesize[plane_index]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2856 uint8_t *dst= s->current_picture.data[plane_index]; |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2857 uint8_t *src= s-> input_picture.data[plane_index]; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2858 IDWTELEM *pred= (IDWTELEM*)s->m.obmc_scratchpad + plane_index*block_size*block_size*4; |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2859 uint8_t cur[ref_stride*2*MB_SIZE]; //FIXME alignment |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
2860 uint8_t tmp[ref_stride*(2*MB_SIZE+HTAPS_MAX-1)]; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2861 const int b_stride = s->b_width << s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2862 const int b_height = s->b_height<< s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2863 const int w= p->width; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2864 const int h= p->height; |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2865 int distortion; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2866 int rate= 0; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2867 const int penalty_factor= get_penalty_factor(s->lambda, s->lambda2, s->avctx->me_cmp); |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2868 int sx= block_w*mb_x - block_w/2; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2869 int sy= block_w*mb_y - block_w/2; |
3206 | 2870 int x0= FFMAX(0,-sx); |
2871 int y0= FFMAX(0,-sy); | |
2872 int x1= FFMIN(block_w*2, w-sx); | |
2873 int y1= FFMIN(block_w*2, h-sy); | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2874 int i,x,y; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2875 |
3314 | 2876 pred_block(s, cur, tmp, ref_stride, sx, sy, block_w*2, block_w*2, &s->block[mb_x + mb_y*b_stride], plane_index, w, h); |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2877 |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2878 for(y=y0; y<y1; y++){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2879 const uint8_t *obmc1= obmc_edged + y*obmc_stride; |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2880 const IDWTELEM *pred1 = pred + y*obmc_stride; |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2881 uint8_t *cur1 = cur + y*ref_stride; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2882 uint8_t *dst1 = dst + sx + (sy+y)*ref_stride; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2883 for(x=x0; x<x1; x++){ |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2884 #if FRAC_BITS >= LOG2_OBMC_MAX |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2885 int v = (cur1[x] * obmc1[x]) << (FRAC_BITS - LOG2_OBMC_MAX); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2886 #else |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2887 int v = (cur1[x] * obmc1[x] + (1<<(LOG2_OBMC_MAX - FRAC_BITS-1))) >> (LOG2_OBMC_MAX - FRAC_BITS); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2888 #endif |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2889 v = (v + pred1[x]) >> FRAC_BITS; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2890 if(v&(~255)) v= ~(v>>31); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2891 dst1[x] = v; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2892 } |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2893 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2894 |
3206 | 2895 /* copy the regions where obmc[] = (uint8_t)256 */ |
2896 if(LOG2_OBMC_MAX == 8 | |
2897 && (mb_x == 0 || mb_x == b_stride-1) | |
2898 && (mb_y == 0 || mb_y == b_height-1)){ | |
2899 if(mb_x == 0) | |
2900 x1 = block_w; | |
2901 else | |
2902 x0 = block_w; | |
2903 if(mb_y == 0) | |
2904 y1 = block_w; | |
2905 else | |
2906 y0 = block_w; | |
2907 for(y=y0; y<y1; y++) | |
2908 memcpy(dst + sx+x0 + (sy+y)*ref_stride, cur + x0 + y*ref_stride, x1-x0); | |
2909 } | |
2910 | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2911 if(block_w==16){ |
3323
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2912 /* FIXME rearrange dsputil to fit 32x32 cmp functions */ |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2913 /* FIXME check alignment of the cmp wavelet vs the encoding wavelet */ |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2914 /* FIXME cmps overlap but don't cover the wavelet's whole support, |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2915 * so improving the score of one block is not strictly guaranteed to |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2916 * improve the score of the whole frame, so iterative motion est |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2917 * doesn't always converge. */ |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2918 if(s->avctx->me_cmp == FF_CMP_W97) |
4197 | 2919 distortion = w97_32_c(&s->m, src + sx + sy*ref_stride, dst + sx + sy*ref_stride, ref_stride, 32); |
3323
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2920 else if(s->avctx->me_cmp == FF_CMP_W53) |
4197 | 2921 distortion = w53_32_c(&s->m, src + sx + sy*ref_stride, dst + sx + sy*ref_stride, ref_stride, 32); |
3323
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2922 else{ |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2923 distortion = 0; |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2924 for(i=0; i<4; i++){ |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2925 int off = sx+16*(i&1) + (sy+16*(i>>1))*ref_stride; |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2926 distortion += s->dsp.me_cmp[0](&s->m, src + off, dst + off, ref_stride, 16); |
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
2927 } |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2928 } |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2929 }else{ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2930 assert(block_w==8); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
2931 distortion = s->dsp.me_cmp[0](&s->m, src + sx + sy*ref_stride, dst + sx + sy*ref_stride, ref_stride, block_w*2); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2932 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2933 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2934 if(plane_index==0){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2935 for(i=0; i<4; i++){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2936 /* ..RRr |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2937 * .RXx. |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2938 * rxx.. |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2939 */ |
3051 | 2940 rate += get_block_bits(s, mb_x + (i&1) - (i>>1), mb_y + (i>>1), 1); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2941 } |
3057 | 2942 if(mb_x == b_stride-2) |
2943 rate += get_block_bits(s, mb_x + 1, mb_y + 1, 1); | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
2944 } |
3051 | 2945 return distortion + rate*penalty_factor; |
2946 } | |
2947 | |
2948 static int get_4block_rd(SnowContext *s, int mb_x, int mb_y, int plane_index){ | |
2949 int i, y2; | |
2950 Plane *p= &s->plane[plane_index]; | |
2951 const int block_size = MB_SIZE >> s->block_max_depth; | |
2952 const int block_w = plane_index ? block_size/2 : block_size; | |
2953 const uint8_t *obmc = plane_index ? obmc_tab[s->block_max_depth+1] : obmc_tab[s->block_max_depth]; | |
2954 const int obmc_stride= plane_index ? block_size : 2*block_size; | |
2955 const int ref_stride= s->current_picture.linesize[plane_index]; | |
2956 uint8_t *dst= s->current_picture.data[plane_index]; | |
2957 uint8_t *src= s-> input_picture.data[plane_index]; | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
2958 static const IDWTELEM zero_dst[4096]; //FIXME |
3051 | 2959 const int b_stride = s->b_width << s->block_max_depth; |
2960 const int w= p->width; | |
2961 const int h= p->height; | |
2962 int distortion= 0; | |
2963 int rate= 0; | |
2964 const int penalty_factor= get_penalty_factor(s->lambda, s->lambda2, s->avctx->me_cmp); | |
2965 | |
2966 for(i=0; i<9; i++){ | |
2967 int mb_x2= mb_x + (i%3) - 1; | |
2968 int mb_y2= mb_y + (i/3) - 1; | |
2969 int x= block_w*mb_x2 + block_w/2; | |
2970 int y= block_w*mb_y2 + block_w/2; | |
2971 | |
3662
fc714e9a5419
snow cosmetics: merge the sliced and non-sliced versions of add_yblock
lorenm
parents:
3661
diff
changeset
|
2972 add_yblock(s, 0, NULL, zero_dst, dst, obmc, |
3051 | 2973 x, y, block_w, block_w, w, h, /*dst_stride*/0, ref_stride, obmc_stride, mb_x2, mb_y2, 1, 1, plane_index); |
2974 | |
2975 //FIXME find a cleaner/simpler way to skip the outside stuff | |
2976 for(y2= y; y2<0; y2++) | |
2977 memcpy(dst + x + y2*ref_stride, src + x + y2*ref_stride, block_w); | |
2978 for(y2= h; y2<y+block_w; y2++) | |
2979 memcpy(dst + x + y2*ref_stride, src + x + y2*ref_stride, block_w); | |
2980 if(x<0){ | |
2981 for(y2= y; y2<y+block_w; y2++) | |
2982 memcpy(dst + x + y2*ref_stride, src + x + y2*ref_stride, -x); | |
2983 } | |
2984 if(x+block_w > w){ | |
2985 for(y2= y; y2<y+block_w; y2++) | |
2986 memcpy(dst + w + y2*ref_stride, src + w + y2*ref_stride, x+block_w - w); | |
2987 } | |
2988 | |
2989 assert(block_w== 8 || block_w==16); | |
2990 distortion += s->dsp.me_cmp[block_w==8](&s->m, src + x + y*ref_stride, dst + x + y*ref_stride, ref_stride, block_w); | |
2991 } | |
2992 | |
2993 if(plane_index==0){ | |
2994 BlockNode *b= &s->block[mb_x+mb_y*b_stride]; | |
2995 int merged= same_block(b,b+1) && same_block(b,b+b_stride) && same_block(b,b+b_stride+1); | |
2996 | |
2997 /* ..RRRr | |
2998 * .RXXx. | |
2999 * .RXXx. | |
3000 * rxxx. | |
3001 */ | |
3002 if(merged) | |
3003 rate = get_block_bits(s, mb_x, mb_y, 2); | |
3004 for(i=merged?4:0; i<9; i++){ | |
3005 static const int dxy[9][2] = {{0,0},{1,0},{0,1},{1,1},{2,0},{2,1},{-1,2},{0,2},{1,2}}; | |
3006 rate += get_block_bits(s, mb_x + dxy[i][0], mb_y + dxy[i][1], 1); | |
3007 } | |
3008 } | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3009 return distortion + rate*penalty_factor; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3010 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3011 |
4283
d6f83e2f8804
rename always_inline to av_always_inline and move to common.h
mru
parents:
4197
diff
changeset
|
3012 static av_always_inline int check_block(SnowContext *s, int mb_x, int mb_y, int p[3], int intra, const uint8_t *obmc_edged, int *best_rd){ |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3013 const int b_stride= s->b_width << s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3014 BlockNode *block= &s->block[mb_x + mb_y * b_stride]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3015 BlockNode backup= *block; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3016 int rd, index, value; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3017 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3018 assert(mb_x>=0 && mb_y>=0); |
2994 | 3019 assert(mb_x<b_stride); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3020 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3021 if(intra){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3022 block->color[0] = p[0]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3023 block->color[1] = p[1]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3024 block->color[2] = p[2]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3025 block->type |= BLOCK_INTRA; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3026 }else{ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3027 index= (p[0] + 31*p[1]) & (ME_CACHE_SIZE-1); |
3314 | 3028 value= s->me_cache_generation + (p[0]>>10) + (p[1]<<6) + (block->ref<<12); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3029 if(s->me_cache[index] == value) |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3030 return 0; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3031 s->me_cache[index]= value; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3032 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3033 block->mx= p[0]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3034 block->my= p[1]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3035 block->type &= ~BLOCK_INTRA; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3036 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3037 |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3038 rd= get_block_rd(s, mb_x, mb_y, 0, obmc_edged); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3039 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3040 //FIXME chroma |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3041 if(rd < *best_rd){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3042 *best_rd= rd; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3043 return 1; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3044 }else{ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3045 *block= backup; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3046 return 0; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3047 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3048 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3049 |
2999
bc83e7a080a5
this fixes compilation with gcc 2.95 which seems to not like array casting... special case inline of check_block which makes the array and passes it
mmu_man
parents:
2998
diff
changeset
|
3050 /* special case for int[2] args we discard afterward, fixes compilation prob with gcc 2.95 */ |
4283
d6f83e2f8804
rename always_inline to av_always_inline and move to common.h
mru
parents:
4197
diff
changeset
|
3051 static av_always_inline int check_block_inter(SnowContext *s, int mb_x, int mb_y, int p0, int p1, const uint8_t *obmc_edged, int *best_rd){ |
2999
bc83e7a080a5
this fixes compilation with gcc 2.95 which seems to not like array casting... special case inline of check_block which makes the array and passes it
mmu_man
parents:
2998
diff
changeset
|
3052 int p[2] = {p0, p1}; |
3197 | 3053 return check_block(s, mb_x, mb_y, p, 0, obmc_edged, best_rd); |
2999
bc83e7a080a5
this fixes compilation with gcc 2.95 which seems to not like array casting... special case inline of check_block which makes the array and passes it
mmu_man
parents:
2998
diff
changeset
|
3054 } |
bc83e7a080a5
this fixes compilation with gcc 2.95 which seems to not like array casting... special case inline of check_block which makes the array and passes it
mmu_man
parents:
2998
diff
changeset
|
3055 |
4283
d6f83e2f8804
rename always_inline to av_always_inline and move to common.h
mru
parents:
4197
diff
changeset
|
3056 static av_always_inline int check_4block_inter(SnowContext *s, int mb_x, int mb_y, int p0, int p1, int ref, int *best_rd){ |
3051 | 3057 const int b_stride= s->b_width << s->block_max_depth; |
3058 BlockNode *block= &s->block[mb_x + mb_y * b_stride]; | |
3059 BlockNode backup[4]= {block[0], block[1], block[b_stride], block[b_stride+1]}; | |
3060 int rd, index, value; | |
3061 | |
3062 assert(mb_x>=0 && mb_y>=0); | |
3063 assert(mb_x<b_stride); | |
3064 assert(((mb_x|mb_y)&1) == 0); | |
3065 | |
3066 index= (p0 + 31*p1) & (ME_CACHE_SIZE-1); | |
3314 | 3067 value= s->me_cache_generation + (p0>>10) + (p1<<6) + (block->ref<<12); |
3051 | 3068 if(s->me_cache[index] == value) |
3069 return 0; | |
3070 s->me_cache[index]= value; | |
3071 | |
3072 block->mx= p0; | |
3073 block->my= p1; | |
3314 | 3074 block->ref= ref; |
3051 | 3075 block->type &= ~BLOCK_INTRA; |
3076 block[1]= block[b_stride]= block[b_stride+1]= *block; | |
3077 | |
3078 rd= get_4block_rd(s, mb_x, mb_y, 0); | |
3079 | |
3080 //FIXME chroma | |
3081 if(rd < *best_rd){ | |
3082 *best_rd= rd; | |
3083 return 1; | |
3084 }else{ | |
3085 block[0]= backup[0]; | |
3086 block[1]= backup[1]; | |
3087 block[b_stride]= backup[2]; | |
3088 block[b_stride+1]= backup[3]; | |
3089 return 0; | |
3090 } | |
3091 } | |
3092 | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3093 static void iterative_me(SnowContext *s){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3094 int pass, mb_x, mb_y; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3095 const int b_width = s->b_width << s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3096 const int b_height= s->b_height << s->block_max_depth; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3097 const int b_stride= b_width; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3098 int color[3]; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3099 |
3194
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3100 { |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3101 RangeCoder r = s->c; |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3102 uint8_t state[sizeof(s->block_state)]; |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3103 memcpy(state, s->block_state, sizeof(s->block_state)); |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3104 for(mb_y= 0; mb_y<s->b_height; mb_y++) |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3105 for(mb_x= 0; mb_x<s->b_width; mb_x++) |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3106 encode_q_branch(s, 0, mb_x, mb_y); |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3107 s->c = r; |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3108 memcpy(s->block_state, state, sizeof(s->block_state)); |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3109 } |
c30e9bcbb716
seed iterative_me with mvs from conventional search.
lorenm
parents:
3193
diff
changeset
|
3110 |
3323
87c54a3f8d19
Snow: fix subband weighting in wavelet cmp functions. use 32x32 cmp in iterative motion estimation.
lorenm
parents:
3322
diff
changeset
|
3111 for(pass=0; pass<25; pass++){ |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3112 int change= 0; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3113 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3114 for(mb_y= 0; mb_y<b_height; mb_y++){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3115 for(mb_x= 0; mb_x<b_width; mb_x++){ |
3314 | 3116 int dia_change, i, j, ref; |
3117 int best_rd= INT_MAX, ref_rd; | |
3118 BlockNode backup, ref_b; | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3119 const int index= mb_x + mb_y * b_stride; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3120 BlockNode *block= &s->block[index]; |
3324 | 3121 BlockNode *tb = mb_y ? &s->block[index-b_stride ] : NULL; |
3122 BlockNode *lb = mb_x ? &s->block[index -1] : NULL; | |
3123 BlockNode *rb = mb_x+1<b_width ? &s->block[index +1] : NULL; | |
3124 BlockNode *bb = mb_y+1<b_height ? &s->block[index+b_stride ] : NULL; | |
3125 BlockNode *tlb= mb_x && mb_y ? &s->block[index-b_stride-1] : NULL; | |
3126 BlockNode *trb= mb_x+1<b_width && mb_y ? &s->block[index-b_stride+1] : NULL; | |
3127 BlockNode *blb= mb_x && mb_y+1<b_height ? &s->block[index+b_stride-1] : NULL; | |
3128 BlockNode *brb= mb_x+1<b_width && mb_y+1<b_height ? &s->block[index+b_stride+1] : NULL; | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3129 const int b_w= (MB_SIZE >> s->block_max_depth); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3130 uint8_t obmc_edged[b_w*2][b_w*2]; |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3131 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3132 if(pass && (block->type & BLOCK_OPT)) |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3133 continue; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3134 block->type |= BLOCK_OPT; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3135 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3136 backup= *block; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3137 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3138 if(!s->me_cache_generation) |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3139 memset(s->me_cache, 0, sizeof(s->me_cache)); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3140 s->me_cache_generation += 1<<22; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3141 |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3142 //FIXME precalc |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3143 { |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3144 int x, y; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3145 memcpy(obmc_edged, obmc_tab[s->block_max_depth], b_w*b_w*4); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3146 if(mb_x==0) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3147 for(y=0; y<b_w*2; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3148 memset(obmc_edged[y], obmc_edged[y][0] + obmc_edged[y][b_w-1], b_w); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3149 if(mb_x==b_stride-1) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3150 for(y=0; y<b_w*2; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3151 memset(obmc_edged[y]+b_w, obmc_edged[y][b_w] + obmc_edged[y][b_w*2-1], b_w); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3152 if(mb_y==0){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3153 for(x=0; x<b_w*2; x++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3154 obmc_edged[0][x] += obmc_edged[b_w-1][x]; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3155 for(y=1; y<b_w; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3156 memcpy(obmc_edged[y], obmc_edged[0], b_w*2); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3157 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3158 if(mb_y==b_height-1){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3159 for(x=0; x<b_w*2; x++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3160 obmc_edged[b_w*2-1][x] += obmc_edged[b_w][x]; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3161 for(y=b_w; y<b_w*2-1; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3162 memcpy(obmc_edged[y], obmc_edged[b_w*2-1], b_w*2); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3163 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3164 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3165 |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3166 //skip stuff outside the picture |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3167 if(mb_x==0 || mb_y==0 || mb_x==b_width-1 || mb_y==b_height-1) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3168 { |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3169 uint8_t *src= s-> input_picture.data[0]; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3170 uint8_t *dst= s->current_picture.data[0]; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3171 const int stride= s->current_picture.linesize[0]; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3172 const int block_w= MB_SIZE >> s->block_max_depth; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3173 const int sx= block_w*mb_x - block_w/2; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3174 const int sy= block_w*mb_y - block_w/2; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3175 const int w= s->plane[0].width; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3176 const int h= s->plane[0].height; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3177 int y; |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3178 |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3179 for(y=sy; y<0; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3180 memcpy(dst + sx + y*stride, src + sx + y*stride, block_w*2); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3181 for(y=h; y<sy+block_w*2; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3182 memcpy(dst + sx + y*stride, src + sx + y*stride, block_w*2); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3183 if(sx<0){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3184 for(y=sy; y<sy+block_w*2; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3185 memcpy(dst + sx + y*stride, src + sx + y*stride, -sx); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3186 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3187 if(sx+block_w*2 > w){ |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3188 for(y=sy; y<sy+block_w*2; y++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3189 memcpy(dst + w + y*stride, src + w + y*stride, sx+block_w*2 - w); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3190 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3191 } |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3192 |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3193 // intra(black) = neighbors' contribution to the current block |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3194 for(i=0; i<3; i++) |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3195 color[i]= get_dc(s, mb_x, mb_y, i); |
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3196 |
5127 | 3197 // get previous score (cannot be cached due to OBMC) |
3057 | 3198 if(pass > 0 && (block->type&BLOCK_INTRA)){ |
3199 int color0[3]= {block->color[0], block->color[1], block->color[2]}; | |
3200 check_block(s, mb_x, mb_y, color0, 1, *obmc_edged, &best_rd); | |
3201 }else | |
3197 | 3202 check_block_inter(s, mb_x, mb_y, block->mx, block->my, *obmc_edged, &best_rd); |
3203 | |
3314 | 3204 ref_b= *block; |
3205 ref_rd= best_rd; | |
3206 for(ref=0; ref < s->ref_frames; ref++){ | |
3207 int16_t (*mvr)[2]= &s->ref_mvs[ref][index]; | |
3208 if(s->ref_scores[ref][index] > s->ref_scores[ref_b.ref][index]*3/2) //FIXME tune threshold | |
3209 continue; | |
3210 block->ref= ref; | |
3211 best_rd= INT_MAX; | |
3212 | |
3213 check_block_inter(s, mb_x, mb_y, mvr[0][0], mvr[0][1], *obmc_edged, &best_rd); | |
3214 check_block_inter(s, mb_x, mb_y, 0, 0, *obmc_edged, &best_rd); | |
3324 | 3215 if(tb) |
3314 | 3216 check_block_inter(s, mb_x, mb_y, mvr[-b_stride][0], mvr[-b_stride][1], *obmc_edged, &best_rd); |
3324 | 3217 if(lb) |
3314 | 3218 check_block_inter(s, mb_x, mb_y, mvr[-1][0], mvr[-1][1], *obmc_edged, &best_rd); |
3324 | 3219 if(rb) |
3314 | 3220 check_block_inter(s, mb_x, mb_y, mvr[1][0], mvr[1][1], *obmc_edged, &best_rd); |
3324 | 3221 if(bb) |
3314 | 3222 check_block_inter(s, mb_x, mb_y, mvr[b_stride][0], mvr[b_stride][1], *obmc_edged, &best_rd); |
3223 | |
3224 /* fullpel ME */ | |
3225 //FIXME avoid subpel interpol / round to nearest integer | |
3226 do{ | |
3227 dia_change=0; | |
3228 for(i=0; i<FFMAX(s->avctx->dia_size, 1); i++){ | |
3229 for(j=0; j<i; j++){ | |
3230 dia_change |= check_block_inter(s, mb_x, mb_y, block->mx+4*(i-j), block->my+(4*j), *obmc_edged, &best_rd); | |
3231 dia_change |= check_block_inter(s, mb_x, mb_y, block->mx-4*(i-j), block->my-(4*j), *obmc_edged, &best_rd); | |
3232 dia_change |= check_block_inter(s, mb_x, mb_y, block->mx+4*(i-j), block->my-(4*j), *obmc_edged, &best_rd); | |
3233 dia_change |= check_block_inter(s, mb_x, mb_y, block->mx-4*(i-j), block->my+(4*j), *obmc_edged, &best_rd); | |
3234 } | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3235 } |
3314 | 3236 }while(dia_change); |
3237 /* subpel ME */ | |
3238 do{ | |
3239 static const int square[8][2]= {{+1, 0},{-1, 0},{ 0,+1},{ 0,-1},{+1,+1},{-1,-1},{+1,-1},{-1,+1},}; | |
3240 dia_change=0; | |
3241 for(i=0; i<8; i++) | |
3242 dia_change |= check_block_inter(s, mb_x, mb_y, block->mx+square[i][0], block->my+square[i][1], *obmc_edged, &best_rd); | |
3243 }while(dia_change); | |
3244 //FIXME or try the standard 2 pass qpel or similar | |
3245 | |
3246 mvr[0][0]= block->mx; | |
3247 mvr[0][1]= block->my; | |
3248 if(ref_rd > best_rd){ | |
3249 ref_rd= best_rd; | |
3250 ref_b= *block; | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3251 } |
3314 | 3252 } |
3253 best_rd= ref_rd; | |
3254 *block= ref_b; | |
2998 | 3255 #if 1 |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3256 check_block(s, mb_x, mb_y, color, 1, *obmc_edged, &best_rd); |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3257 //FIXME RD style color selection |
2998 | 3258 #endif |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3259 if(!same_block(block, &backup)){ |
3324 | 3260 if(tb ) tb ->type &= ~BLOCK_OPT; |
3261 if(lb ) lb ->type &= ~BLOCK_OPT; | |
3262 if(rb ) rb ->type &= ~BLOCK_OPT; | |
3263 if(bb ) bb ->type &= ~BLOCK_OPT; | |
3264 if(tlb) tlb->type &= ~BLOCK_OPT; | |
3265 if(trb) trb->type &= ~BLOCK_OPT; | |
3266 if(blb) blb->type &= ~BLOCK_OPT; | |
3267 if(brb) brb->type &= ~BLOCK_OPT; | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3268 change ++; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3269 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3270 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3271 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3272 av_log(NULL, AV_LOG_ERROR, "pass:%d changed:%d\n", pass, change); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3273 if(!change) |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3274 break; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3275 } |
3051 | 3276 |
3277 if(s->block_max_depth == 1){ | |
3278 int change= 0; | |
3279 for(mb_y= 0; mb_y<b_height; mb_y+=2){ | |
3280 for(mb_x= 0; mb_x<b_width; mb_x+=2){ | |
3324 | 3281 int i; |
3051 | 3282 int best_rd, init_rd; |
3283 const int index= mb_x + mb_y * b_stride; | |
3284 BlockNode *b[4]; | |
3285 | |
3286 b[0]= &s->block[index]; | |
3287 b[1]= b[0]+1; | |
3288 b[2]= b[0]+b_stride; | |
3289 b[3]= b[2]+1; | |
3290 if(same_block(b[0], b[1]) && | |
3291 same_block(b[0], b[2]) && | |
3292 same_block(b[0], b[3])) | |
3293 continue; | |
3294 | |
3295 if(!s->me_cache_generation) | |
3296 memset(s->me_cache, 0, sizeof(s->me_cache)); | |
3297 s->me_cache_generation += 1<<22; | |
3298 | |
3299 init_rd= best_rd= get_4block_rd(s, mb_x, mb_y, 0); | |
3300 | |
3314 | 3301 //FIXME more multiref search? |
3051 | 3302 check_4block_inter(s, mb_x, mb_y, |
3303 (b[0]->mx + b[1]->mx + b[2]->mx + b[3]->mx + 2) >> 2, | |
3314 | 3304 (b[0]->my + b[1]->my + b[2]->my + b[3]->my + 2) >> 2, 0, &best_rd); |
3051 | 3305 |
3306 for(i=0; i<4; i++) | |
3307 if(!(b[i]->type&BLOCK_INTRA)) | |
3314 | 3308 check_4block_inter(s, mb_x, mb_y, b[i]->mx, b[i]->my, b[i]->ref, &best_rd); |
3051 | 3309 |
3310 if(init_rd != best_rd) | |
3311 change++; | |
3312 } | |
3313 } | |
3314 av_log(NULL, AV_LOG_ERROR, "pass:4mv changed:%d\n", change*4); | |
3315 } | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3316 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3317 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3318 static void quantize(SnowContext *s, SubBand *b, IDWTELEM *dst, DWTELEM *src, int stride, int bias){ |
2138 | 3319 const int level= b->level; |
3320 const int w= b->width; | |
3321 const int h= b->height; | |
4594 | 3322 const int qlog= av_clip(s->qlog + b->qlog, 0, QROOT*16); |
5575 | 3323 const int qmul= qexp[qlog&(QROOT-1)]<<((qlog>>QSHIFT) + ENCODER_EXTRA_BITS); |
2150 | 3324 int x,y, thres1, thres2; |
2893 | 3325 // START_TIMER |
2138 | 3326 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3327 if(s->qlog == LOSSLESS_QLOG){ |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3328 for(y=0; y<h; y++) |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3329 for(x=0; x<w; x++) |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3330 dst[x + y*stride]= src[x + y*stride]; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3331 return; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3332 } |
2967 | 3333 |
2138 | 3334 bias= bias ? 0 : (3*qmul)>>3; |
2150 | 3335 thres1= ((qmul - bias)>>QEXPSHIFT) - 1; |
3336 thres2= 2*thres1; | |
2967 | 3337 |
2138 | 3338 if(!bias){ |
3339 for(y=0; y<h; y++){ | |
3340 for(x=0; x<w; x++){ | |
2150 | 3341 int i= src[x + y*stride]; |
2967 | 3342 |
2150 | 3343 if((unsigned)(i+thres1) > thres2){ |
3344 if(i>=0){ | |
3345 i<<= QEXPSHIFT; | |
3346 i/= qmul; //FIXME optimize | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3347 dst[x + y*stride]= i; |
2150 | 3348 }else{ |
3349 i= -i; | |
3350 i<<= QEXPSHIFT; | |
3351 i/= qmul; //FIXME optimize | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3352 dst[x + y*stride]= -i; |
2150 | 3353 } |
3354 }else | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3355 dst[x + y*stride]= 0; |
2138 | 3356 } |
3357 } | |
3358 }else{ | |
3359 for(y=0; y<h; y++){ | |
3360 for(x=0; x<w; x++){ | |
2967 | 3361 int i= src[x + y*stride]; |
3362 | |
2150 | 3363 if((unsigned)(i+thres1) > thres2){ |
3364 if(i>=0){ | |
3365 i<<= QEXPSHIFT; | |
3366 i= (i + bias) / qmul; //FIXME optimize | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3367 dst[x + y*stride]= i; |
2150 | 3368 }else{ |
3369 i= -i; | |
3370 i<<= QEXPSHIFT; | |
3371 i= (i + bias) / qmul; //FIXME optimize | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3372 dst[x + y*stride]= -i; |
2150 | 3373 } |
3374 }else | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3375 dst[x + y*stride]= 0; |
2138 | 3376 } |
3377 } | |
3378 } | |
2150 | 3379 if(level+1 == s->spatial_decomposition_count){ |
3380 // STOP_TIMER("quantize") | |
3381 } | |
2138 | 3382 } |
3383 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3384 static void dequantize_slice_buffered(SnowContext *s, slice_buffer * sb, SubBand *b, IDWTELEM *src, int stride, int start_y, int end_y){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3385 const int w= b->width; |
4594 | 3386 const int qlog= av_clip(s->qlog + b->qlog, 0, QROOT*16); |
2600 | 3387 const int qmul= qexp[qlog&(QROOT-1)]<<(qlog>>QSHIFT); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3388 const int qadd= (s->qbias*qmul)>>QBIAS_SHIFT; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3389 int x,y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3390 START_TIMER |
2967 | 3391 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3392 if(s->qlog == LOSSLESS_QLOG) return; |
2967 | 3393 |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
3394 for(y=start_y; y<end_y; y++){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3395 // DWTELEM * line = slice_buffer_get_line_from_address(sb, src + (y * stride)); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3396 IDWTELEM * line = slice_buffer_get_line(sb, (y * b->stride_line) + b->buf_y_offset) + b->buf_x_offset; |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3397 for(x=0; x<w; x++){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3398 int i= line[x]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3399 if(i<0){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3400 line[x]= -((-i*qmul + qadd)>>(QEXPSHIFT)); //FIXME try different bias |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3401 }else if(i>0){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3402 line[x]= (( i*qmul + qadd)>>(QEXPSHIFT)); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3403 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3404 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3405 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3406 if(w > 200 /*level+1 == s->spatial_decomposition_count*/){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3407 STOP_TIMER("dquant") |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3408 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3409 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3410 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3411 static void dequantize(SnowContext *s, SubBand *b, IDWTELEM *src, int stride){ |
2138 | 3412 const int w= b->width; |
3413 const int h= b->height; | |
4594 | 3414 const int qlog= av_clip(s->qlog + b->qlog, 0, QROOT*16); |
2600 | 3415 const int qmul= qexp[qlog&(QROOT-1)]<<(qlog>>QSHIFT); |
2138 | 3416 const int qadd= (s->qbias*qmul)>>QBIAS_SHIFT; |
3417 int x,y; | |
2195 | 3418 START_TIMER |
2967 | 3419 |
2161 | 3420 if(s->qlog == LOSSLESS_QLOG) return; |
2967 | 3421 |
2138 | 3422 for(y=0; y<h; y++){ |
3423 for(x=0; x<w; x++){ | |
3424 int i= src[x + y*stride]; | |
3425 if(i<0){ | |
3426 src[x + y*stride]= -((-i*qmul + qadd)>>(QEXPSHIFT)); //FIXME try different bias | |
3427 }else if(i>0){ | |
3428 src[x + y*stride]= (( i*qmul + qadd)>>(QEXPSHIFT)); | |
3429 } | |
3430 } | |
3431 } | |
2195 | 3432 if(w > 200 /*level+1 == s->spatial_decomposition_count*/){ |
3433 STOP_TIMER("dquant") | |
3434 } | |
2138 | 3435 } |
3436 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3437 static void decorrelate(SnowContext *s, SubBand *b, IDWTELEM *src, int stride, int inverse, int use_median){ |
2138 | 3438 const int w= b->width; |
3439 const int h= b->height; | |
3440 int x,y; | |
2967 | 3441 |
2138 | 3442 for(y=h-1; y>=0; y--){ |
3443 for(x=w-1; x>=0; x--){ | |
3444 int i= x + y*stride; | |
2967 | 3445 |
2138 | 3446 if(x){ |
3447 if(use_median){ | |
3448 if(y && x+1<w) src[i] -= mid_pred(src[i - 1], src[i - stride], src[i - stride + 1]); | |
3449 else src[i] -= src[i - 1]; | |
3450 }else{ | |
3451 if(y) src[i] -= mid_pred(src[i - 1], src[i - stride], src[i - 1] + src[i - stride] - src[i - 1 - stride]); | |
3452 else src[i] -= src[i - 1]; | |
3453 } | |
3454 }else{ | |
3455 if(y) src[i] -= src[i - stride]; | |
3456 } | |
3457 } | |
3458 } | |
3459 } | |
3460 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3461 static void correlate_slice_buffered(SnowContext *s, slice_buffer * sb, SubBand *b, IDWTELEM *src, int stride, int inverse, int use_median, int start_y, int end_y){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3462 const int w= b->width; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3463 int x,y; |
2967 | 3464 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3465 // START_TIMER |
2967 | 3466 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3467 IDWTELEM * line=0; // silence silly "could be used without having been initialized" warning |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3468 IDWTELEM * prev; |
2967 | 3469 |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
3470 if (start_y != 0) |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
3471 line = slice_buffer_get_line(sb, ((start_y - 1) * b->stride_line) + b->buf_y_offset) + b->buf_x_offset; |
2967 | 3472 |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
3473 for(y=start_y; y<end_y; y++){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3474 prev = line; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3475 // line = slice_buffer_get_line_from_address(sb, src + (y * stride)); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3476 line = slice_buffer_get_line(sb, (y * b->stride_line) + b->buf_y_offset) + b->buf_x_offset; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3477 for(x=0; x<w; x++){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3478 if(x){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3479 if(use_median){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3480 if(y && x+1<w) line[x] += mid_pred(line[x - 1], prev[x], prev[x + 1]); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3481 else line[x] += line[x - 1]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3482 }else{ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3483 if(y) line[x] += mid_pred(line[x - 1], prev[x], line[x - 1] + prev[x] - prev[x - 1]); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3484 else line[x] += line[x - 1]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3485 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3486 }else{ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3487 if(y) line[x] += prev[x]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3488 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3489 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3490 } |
2967 | 3491 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3492 // STOP_TIMER("correlate") |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3493 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3494 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3495 static void correlate(SnowContext *s, SubBand *b, IDWTELEM *src, int stride, int inverse, int use_median){ |
2138 | 3496 const int w= b->width; |
3497 const int h= b->height; | |
3498 int x,y; | |
2967 | 3499 |
2138 | 3500 for(y=0; y<h; y++){ |
3501 for(x=0; x<w; x++){ | |
3502 int i= x + y*stride; | |
2967 | 3503 |
2138 | 3504 if(x){ |
3505 if(use_median){ | |
3506 if(y && x+1<w) src[i] += mid_pred(src[i - 1], src[i - stride], src[i - stride + 1]); | |
3507 else src[i] += src[i - 1]; | |
3508 }else{ | |
3509 if(y) src[i] += mid_pred(src[i - 1], src[i - stride], src[i - 1] + src[i - stride] - src[i - 1 - stride]); | |
3510 else src[i] += src[i - 1]; | |
3511 } | |
3512 }else{ | |
3513 if(y) src[i] += src[i - stride]; | |
3514 } | |
3515 } | |
3516 } | |
3517 } | |
3518 | |
3519 static void encode_header(SnowContext *s){ | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3520 int plane_index, level, orientation, i; |
2967 | 3521 uint8_t kstate[32]; |
3522 | |
3523 memset(kstate, MID_STATE, sizeof(kstate)); | |
2138 | 3524 |
2335 | 3525 put_rac(&s->c, kstate, s->keyframe); |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3526 if(s->keyframe || s->always_reset){ |
2199 | 3527 reset_contexts(s); |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3528 s->last_spatial_decomposition_type= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3529 s->last_qlog= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3530 s->last_qbias= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3531 s->last_mv_scale= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3532 s->last_block_max_depth= 0; |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3533 for(plane_index=0; plane_index<2; plane_index++){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3534 Plane *p= &s->plane[plane_index]; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3535 p->last_htaps=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3536 p->last_diag_mc=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3537 memset(p->last_hcoeff, 0, sizeof(p->last_hcoeff)); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3538 } |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3539 } |
2138 | 3540 if(s->keyframe){ |
3541 put_symbol(&s->c, s->header_state, s->version, 0); | |
2335 | 3542 put_rac(&s->c, s->header_state, s->always_reset); |
2138 | 3543 put_symbol(&s->c, s->header_state, s->temporal_decomposition_type, 0); |
3544 put_symbol(&s->c, s->header_state, s->temporal_decomposition_count, 0); | |
3545 put_symbol(&s->c, s->header_state, s->spatial_decomposition_count, 0); | |
3546 put_symbol(&s->c, s->header_state, s->colorspace_type, 0); | |
3547 put_symbol(&s->c, s->header_state, s->chroma_h_shift, 0); | |
3548 put_symbol(&s->c, s->header_state, s->chroma_v_shift, 0); | |
2335 | 3549 put_rac(&s->c, s->header_state, s->spatial_scalability); |
3550 // put_rac(&s->c, s->header_state, s->rate_scalability); | |
3314 | 3551 put_symbol(&s->c, s->header_state, s->max_ref_frames-1, 0); |
2138 | 3552 |
3553 for(plane_index=0; plane_index<2; plane_index++){ | |
3554 for(level=0; level<s->spatial_decomposition_count; level++){ | |
3555 for(orientation=level ? 1:0; orientation<4; orientation++){ | |
3556 if(orientation==2) continue; | |
3557 put_symbol(&s->c, s->header_state, s->plane[plane_index].band[level][orientation].qlog, 1); | |
3558 } | |
3559 } | |
3560 } | |
3561 } | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3562 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3563 if(!s->keyframe){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3564 int update_mc=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3565 for(plane_index=0; plane_index<2; plane_index++){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3566 Plane *p= &s->plane[plane_index]; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3567 update_mc |= p->last_htaps != p->htaps; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3568 update_mc |= p->last_diag_mc != p->diag_mc; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3569 update_mc |= !!memcmp(p->last_hcoeff, p->hcoeff, sizeof(p->hcoeff)); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3570 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3571 if(!s->always_reset) |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3572 put_rac(&s->c, s->header_state, update_mc); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3573 if(update_mc){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3574 for(plane_index=0; plane_index<2; plane_index++){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3575 Plane *p= &s->plane[plane_index]; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3576 put_rac(&s->c, s->header_state, p->diag_mc); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3577 put_symbol(&s->c, s->header_state, p->htaps/2-1, 0); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3578 for(i= p->htaps/2; i; i--) |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3579 put_symbol(&s->c, s->header_state, FFABS(p->hcoeff[i]), 0); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3580 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3581 p->last_diag_mc= p->diag_mc; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3582 p->last_htaps= p->htaps; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3583 memcpy(p->last_hcoeff, p->hcoeff, sizeof(p->hcoeff)); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3584 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3585 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3586 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3587 |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3588 put_symbol(&s->c, s->header_state, s->spatial_decomposition_type - s->last_spatial_decomposition_type, 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3589 put_symbol(&s->c, s->header_state, s->qlog - s->last_qlog , 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3590 put_symbol(&s->c, s->header_state, s->mv_scale - s->last_mv_scale, 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3591 put_symbol(&s->c, s->header_state, s->qbias - s->last_qbias , 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3592 put_symbol(&s->c, s->header_state, s->block_max_depth - s->last_block_max_depth, 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3593 |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3594 s->last_spatial_decomposition_type= s->spatial_decomposition_type; |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3595 s->last_qlog = s->qlog; |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3596 s->last_qbias = s->qbias; |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3597 s->last_mv_scale = s->mv_scale; |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3598 s->last_block_max_depth = s->block_max_depth; |
2138 | 3599 } |
3600 | |
3601 static int decode_header(SnowContext *s){ | |
3602 int plane_index, level, orientation; | |
2335 | 3603 uint8_t kstate[32]; |
2138 | 3604 |
2967 | 3605 memset(kstate, MID_STATE, sizeof(kstate)); |
2335 | 3606 |
3607 s->keyframe= get_rac(&s->c, kstate); | |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3608 if(s->keyframe || s->always_reset){ |
2199 | 3609 reset_contexts(s); |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3610 s->spatial_decomposition_type= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3611 s->qlog= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3612 s->qbias= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3613 s->mv_scale= |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3614 s->block_max_depth= 0; |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3615 } |
2138 | 3616 if(s->keyframe){ |
3617 s->version= get_symbol(&s->c, s->header_state, 0); | |
3618 if(s->version>0){ | |
3619 av_log(s->avctx, AV_LOG_ERROR, "version %d not supported", s->version); | |
3620 return -1; | |
3621 } | |
2335 | 3622 s->always_reset= get_rac(&s->c, s->header_state); |
2138 | 3623 s->temporal_decomposition_type= get_symbol(&s->c, s->header_state, 0); |
3624 s->temporal_decomposition_count= get_symbol(&s->c, s->header_state, 0); | |
3625 s->spatial_decomposition_count= get_symbol(&s->c, s->header_state, 0); | |
3626 s->colorspace_type= get_symbol(&s->c, s->header_state, 0); | |
3627 s->chroma_h_shift= get_symbol(&s->c, s->header_state, 0); | |
3628 s->chroma_v_shift= get_symbol(&s->c, s->header_state, 0); | |
2335 | 3629 s->spatial_scalability= get_rac(&s->c, s->header_state); |
3630 // s->rate_scalability= get_rac(&s->c, s->header_state); | |
3314 | 3631 s->max_ref_frames= get_symbol(&s->c, s->header_state, 0)+1; |
2138 | 3632 |
3633 for(plane_index=0; plane_index<3; plane_index++){ | |
3634 for(level=0; level<s->spatial_decomposition_count; level++){ | |
3635 for(orientation=level ? 1:0; orientation<4; orientation++){ | |
3636 int q; | |
3637 if (plane_index==2) q= s->plane[1].band[level][orientation].qlog; | |
3638 else if(orientation==2) q= s->plane[plane_index].band[level][1].qlog; | |
3639 else q= get_symbol(&s->c, s->header_state, 1); | |
3640 s->plane[plane_index].band[level][orientation].qlog= q; | |
3641 } | |
3642 } | |
3643 } | |
3644 } | |
2967 | 3645 |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3646 if(!s->keyframe){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3647 if(s->always_reset || get_rac(&s->c, s->header_state)){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3648 for(plane_index=0; plane_index<2; plane_index++){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3649 int htaps, i, sum=0, absum=0; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3650 Plane *p= &s->plane[plane_index]; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3651 p->diag_mc= get_rac(&s->c, s->header_state); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3652 htaps= get_symbol(&s->c, s->header_state, 0)*2 + 2; |
5652
941e5deeb2a4
rename HTAPS -> HTAPS_MAX (later is correct after the previous change)
michael
parents:
5651
diff
changeset
|
3653 if((unsigned)htaps > HTAPS_MAX || htaps==0) |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3654 return -1; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3655 p->htaps= htaps; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3656 for(i= htaps/2; i; i--){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3657 p->hcoeff[i]= get_symbol(&s->c, s->header_state, 0) * (1-2*(i&1)); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3658 sum += p->hcoeff[i]; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3659 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3660 p->hcoeff[0]= 32-sum; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3661 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3662 s->plane[2].diag_mc= s->plane[1].diag_mc; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3663 s->plane[2].htaps = s->plane[1].htaps; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3664 memcpy(s->plane[2].hcoeff, s->plane[1].hcoeff, sizeof(s->plane[1].hcoeff)); |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3665 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3666 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3667 |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3668 s->spatial_decomposition_type+= get_symbol(&s->c, s->header_state, 1); |
5588
effa59ca89b3
we only have 2 wavelets, the 3rd was just for experimentation ...
michael
parents:
5587
diff
changeset
|
3669 if(s->spatial_decomposition_type > 1){ |
2138 | 3670 av_log(s->avctx, AV_LOG_ERROR, "spatial_decomposition_type %d not supported", s->spatial_decomposition_type); |
3671 return -1; | |
3672 } | |
2967 | 3673 |
4331
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3674 s->qlog += get_symbol(&s->c, s->header_state, 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3675 s->mv_scale += get_symbol(&s->c, s->header_state, 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3676 s->qbias += get_symbol(&s->c, s->header_state, 1); |
e571dfe677be
store a few values in the header as difference to the last
michael
parents:
4283
diff
changeset
|
3677 s->block_max_depth+= get_symbol(&s->c, s->header_state, 1); |
3303
68721b62a528
sanity checks, some might have been exploitable ...
michael
parents:
3206
diff
changeset
|
3678 if(s->block_max_depth > 1 || s->block_max_depth < 0){ |
2952 | 3679 av_log(s->avctx, AV_LOG_ERROR, "block_max_depth= %d is too large", s->block_max_depth); |
3680 s->block_max_depth= 0; | |
3681 return -1; | |
3682 } | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3683 |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3684 return 0; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3685 } |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3686 |
3075 | 3687 static void init_qexp(void){ |
2600 | 3688 int i; |
3689 double v=128; | |
3690 | |
3691 for(i=0; i<QROOT; i++){ | |
3692 qexp[i]= lrintf(v); | |
2967 | 3693 v *= pow(2, 1.0 / QROOT); |
2600 | 3694 } |
3695 } | |
3696 | |
2138 | 3697 static int common_init(AVCodecContext *avctx){ |
3698 SnowContext *s = avctx->priv_data; | |
3699 int width, height; | |
5663 | 3700 int level, orientation, plane_index; |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
3701 int i, j; |
2138 | 3702 |
3703 s->avctx= avctx; | |
2967 | 3704 |
2138 | 3705 dsputil_init(&s->dsp, avctx); |
3706 | |
3707 #define mcf(dx,dy)\ | |
3708 s->dsp.put_qpel_pixels_tab [0][dy+dx/4]=\ | |
3709 s->dsp.put_no_rnd_qpel_pixels_tab[0][dy+dx/4]=\ | |
2224
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3710 s->dsp.put_h264_qpel_pixels_tab[0][dy+dx/4];\ |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3711 s->dsp.put_qpel_pixels_tab [1][dy+dx/4]=\ |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3712 s->dsp.put_no_rnd_qpel_pixels_tab[1][dy+dx/4]=\ |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3713 s->dsp.put_h264_qpel_pixels_tab[1][dy+dx/4]; |
2138 | 3714 |
3715 mcf( 0, 0) | |
3716 mcf( 4, 0) | |
3717 mcf( 8, 0) | |
3718 mcf(12, 0) | |
3719 mcf( 0, 4) | |
3720 mcf( 4, 4) | |
3721 mcf( 8, 4) | |
3722 mcf(12, 4) | |
3723 mcf( 0, 8) | |
3724 mcf( 4, 8) | |
3725 mcf( 8, 8) | |
3726 mcf(12, 8) | |
3727 mcf( 0,12) | |
3728 mcf( 4,12) | |
3729 mcf( 8,12) | |
3730 mcf(12,12) | |
3731 | |
3732 #define mcfh(dx,dy)\ | |
3733 s->dsp.put_pixels_tab [0][dy/4+dx/8]=\ | |
3734 s->dsp.put_no_rnd_pixels_tab[0][dy/4+dx/8]=\ | |
2224
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3735 mc_block_hpel ## dx ## dy ## 16;\ |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3736 s->dsp.put_pixels_tab [1][dy/4+dx/8]=\ |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3737 s->dsp.put_no_rnd_pixels_tab[1][dy/4+dx/8]=\ |
11d54cb7ac4e
100l (forgot to set the 8x8 mc functions, so ME did use the mpeg ones ...)
michael
parents:
2223
diff
changeset
|
3738 mc_block_hpel ## dx ## dy ## 8; |
2138 | 3739 |
3740 mcfh(0, 0) | |
3741 mcfh(8, 0) | |
3742 mcfh(0, 8) | |
3743 mcfh(8, 8) | |
2600 | 3744 |
3745 if(!qexp[0]) | |
3746 init_qexp(); | |
3747 | |
5663 | 3748 s->spatial_decomposition_count= 5; |
2138 | 3749 s->spatial_decomposition_type= avctx->prediction_method; //FIXME add decorrelator type r transform_type |
2967 | 3750 |
2138 | 3751 s->chroma_h_shift= 1; //FIXME XXX |
3752 s->chroma_v_shift= 1; | |
2967 | 3753 |
2138 | 3754 // dec += FFMAX(s->chroma_h_shift, s->chroma_v_shift); |
2967 | 3755 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3756 width= s->avctx->width; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3757 height= s->avctx->height; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3758 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3759 s->spatial_idwt_buffer= av_mallocz(width*height*sizeof(IDWTELEM)); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3760 s->spatial_dwt_buffer= av_mallocz(width*height*sizeof(DWTELEM)); //FIXME this doesnt belong here |
2967 | 3761 |
2138 | 3762 s->mv_scale= (s->avctx->flags & CODEC_FLAG_QPEL) ? 2 : 4; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3763 s->block_max_depth= (s->avctx->flags & CODEC_FLAG_4MV) ? 1 : 0; |
2967 | 3764 |
3765 for(plane_index=0; plane_index<3; plane_index++){ | |
2138 | 3766 int w= s->avctx->width; |
3767 int h= s->avctx->height; | |
3768 | |
3769 if(plane_index){ | |
3770 w>>= s->chroma_h_shift; | |
3771 h>>= s->chroma_v_shift; | |
3772 } | |
3773 s->plane[plane_index].width = w; | |
3774 s->plane[plane_index].height= h; | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3775 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3776 s->plane[plane_index].diag_mc= 1; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3777 s->plane[plane_index].htaps= 6; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3778 s->plane[plane_index].hcoeff[0]= 40; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3779 s->plane[plane_index].hcoeff[1]= -10; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3780 s->plane[plane_index].hcoeff[2]= 2; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3781 s->plane[plane_index].fast_mc= 1; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
3782 |
2160 | 3783 //av_log(NULL, AV_LOG_DEBUG, "%d %d\n", w, h); |
2138 | 3784 for(level=s->spatial_decomposition_count-1; level>=0; level--){ |
3785 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
3786 SubBand *b= &s->plane[plane_index].band[level][orientation]; | |
2967 | 3787 |
2138 | 3788 b->buf= s->spatial_dwt_buffer; |
3789 b->level= level; | |
3790 b->stride= s->plane[plane_index].width << (s->spatial_decomposition_count - level); | |
3791 b->width = (w + !(orientation&1))>>1; | |
3792 b->height= (h + !(orientation>1))>>1; | |
2967 | 3793 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3794 b->stride_line = 1 << (s->spatial_decomposition_count - level); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3795 b->buf_x_offset = 0; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3796 b->buf_y_offset = 0; |
2967 | 3797 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3798 if(orientation&1){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3799 b->buf += (w+1)>>1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3800 b->buf_x_offset = (w+1)>>1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3801 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3802 if(orientation>1){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3803 b->buf += b->stride>>1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3804 b->buf_y_offset = b->stride_line >> 1; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3805 } |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3806 b->ibuf= s->spatial_idwt_buffer + (b->buf - s->spatial_dwt_buffer); |
2967 | 3807 |
2138 | 3808 if(level) |
3809 b->parent= &s->plane[plane_index].band[level-1][orientation]; | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
3810 b->x_coeff=av_mallocz(((b->width+1) * b->height+1)*sizeof(x_and_coeff)); |
2138 | 3811 } |
3812 w= (w+1)>>1; | |
3813 h= (h+1)>>1; | |
3814 } | |
3815 } | |
2967 | 3816 |
3325
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
3817 for(i=0; i<MAX_REF_FRAMES; i++) |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
3818 for(j=0; j<MAX_REF_FRAMES; j++) |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
3819 scale_mv_ref[i][j] = 256*(i+1)/(j+1); |
c2a017de6bea
Snow: scale predicted mv based on which reference frame the neighbors used.
lorenm
parents:
3324
diff
changeset
|
3820 |
2967 | 3821 /* |
2138 | 3822 width= s->width= avctx->width; |
3823 height= s->height= avctx->height; | |
2967 | 3824 |
2138 | 3825 assert(width && height); |
3826 */ | |
3827 s->avctx->get_buffer(s->avctx, &s->mconly_picture); | |
2967 | 3828 |
2138 | 3829 return 0; |
3830 } | |
3831 | |
3322
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
3832 static int qscale2qlog(int qscale){ |
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
3833 return rint(QROOT*log(qscale / (float)FF_QP2LAMBDA)/log(2)) |
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
3834 + 61*QROOT/8; //<64 >60 |
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
3835 } |
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
3836 |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
3837 static int ratecontrol_1pass(SnowContext *s, AVFrame *pict) |
3313 | 3838 { |
3839 /* estimate the frame's complexity as a sum of weighted dwt coefs. | |
3840 * FIXME we know exact mv bits at this point, | |
3841 * but ratecontrol isn't set up to include them. */ | |
3842 uint32_t coef_sum= 0; | |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
3843 int level, orientation, delta_qlog; |
3313 | 3844 |
3845 for(level=0; level<s->spatial_decomposition_count; level++){ | |
3846 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
3847 SubBand *b= &s->plane[0].band[level][orientation]; | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3848 IDWTELEM *buf= b->ibuf; |
3313 | 3849 const int w= b->width; |
3850 const int h= b->height; | |
3851 const int stride= b->stride; | |
4594 | 3852 const int qlog= av_clip(2*QROOT + b->qlog, 0, QROOT*16); |
3313 | 3853 const int qmul= qexp[qlog&(QROOT-1)]<<(qlog>>QSHIFT); |
3854 const int qdiv= (1<<16)/qmul; | |
3855 int x, y; | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3856 //FIXME this is ugly |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3857 for(y=0; y<h; y++) |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3858 for(x=0; x<w; x++) |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3859 buf[x+y*stride]= b->buf[x+y*stride]; |
3313 | 3860 if(orientation==0) |
3861 decorrelate(s, b, buf, stride, 1, 0); | |
3862 for(y=0; y<h; y++) | |
3863 for(x=0; x<w; x++) | |
3864 coef_sum+= abs(buf[x+y*stride]) * qdiv >> 16; | |
3865 } | |
3866 } | |
3867 | |
3868 /* ugly, ratecontrol just takes a sqrt again */ | |
3869 coef_sum = (uint64_t)coef_sum * coef_sum >> 16; | |
3870 assert(coef_sum < INT_MAX); | |
3871 | |
3872 if(pict->pict_type == I_TYPE){ | |
3873 s->m.current_picture.mb_var_sum= coef_sum; | |
3874 s->m.current_picture.mc_mb_var_sum= 0; | |
3875 }else{ | |
3876 s->m.current_picture.mc_mb_var_sum= coef_sum; | |
3877 s->m.current_picture.mb_var_sum= 0; | |
3878 } | |
3879 | |
3880 pict->quality= ff_rate_estimate_qscale(&s->m, 1); | |
3766 | 3881 if (pict->quality < 0) |
4011
5bce97c30a69
-1 is a valid return value in ratecontrol_1pass() -> 100l for takis
michael
parents:
4001
diff
changeset
|
3882 return INT_MIN; |
3313 | 3883 s->lambda= pict->quality * 3/2; |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
3884 delta_qlog= qscale2qlog(pict->quality) - s->qlog; |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
3885 s->qlog+= delta_qlog; |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
3886 return delta_qlog; |
3313 | 3887 } |
2138 | 3888 |
3889 static void calculate_vissual_weight(SnowContext *s, Plane *p){ | |
3890 int width = p->width; | |
3891 int height= p->height; | |
2198 | 3892 int level, orientation, x, y; |
2138 | 3893 |
3894 for(level=0; level<s->spatial_decomposition_count; level++){ | |
3895 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
3896 SubBand *b= &p->band[level][orientation]; | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3897 IDWTELEM *ibuf= b->ibuf; |
2138 | 3898 int64_t error=0; |
2967 | 3899 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3900 memset(s->spatial_idwt_buffer, 0, sizeof(*s->spatial_idwt_buffer)*width*height); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3901 ibuf[b->width/2 + b->height/2*b->stride]= 256*16; |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3902 ff_spatial_idwt(s->spatial_idwt_buffer, width, height, width, s->spatial_decomposition_type, s->spatial_decomposition_count); |
2138 | 3903 for(y=0; y<height; y++){ |
3904 for(x=0; x<width; x++){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
3905 int64_t d= s->spatial_idwt_buffer[x + y*width]*16; |
2138 | 3906 error += d*d; |
3907 } | |
3908 } | |
3909 | |
3910 b->qlog= (int)(log(352256.0/sqrt(error)) / log(pow(2.0, 1.0/QROOT))+0.5); | |
2164 | 3911 // av_log(NULL, AV_LOG_DEBUG, "%d %d %d\n", level, orientation, b->qlog/*, sqrt(error)*/); |
2138 | 3912 } |
3913 } | |
3914 } | |
3915 | |
3916 static int encode_init(AVCodecContext *avctx) | |
3917 { | |
3918 SnowContext *s = avctx->priv_data; | |
2198 | 3919 int plane_index; |
2138 | 3920 |
2658
d1609cfeb1d0
#defines for strict_std_compliance and split between inofficial extensions and non standarized things
michael
parents:
2635
diff
changeset
|
3921 if(avctx->strict_std_compliance > FF_COMPLIANCE_EXPERIMENTAL){ |
2628
511e3afc43e1
Ministry of English Composition, reporting for duty (and the word is "skipped", not "skiped"; "skiped" would rhyme with "hyped")
melanson
parents:
2610
diff
changeset
|
3922 av_log(avctx, AV_LOG_ERROR, "this codec is under development, files encoded with it may not be decodable with future versions!!!\n" |
2658
d1609cfeb1d0
#defines for strict_std_compliance and split between inofficial extensions and non standarized things
michael
parents:
2635
diff
changeset
|
3923 "use vstrict=-2 / -strict -2 to use it anyway\n"); |
2151 | 3924 return -1; |
3925 } | |
2967 | 3926 |
3327
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3927 if(avctx->prediction_method == DWT_97 |
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3928 && (avctx->flags & CODEC_FLAG_QSCALE) |
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3929 && avctx->global_quality == 0){ |
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3930 av_log(avctx, AV_LOG_ERROR, "the 9/7 wavelet is incompatible with lossless mode\n"); |
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3931 return -1; |
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3932 } |
955096780e7c
Snow: don't try to encode lossless with 9/7 wavelet, because it isn't lossless.
lorenm
parents:
3326
diff
changeset
|
3933 |
2138 | 3934 common_init(avctx); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
3935 alloc_blocks(s); |
2967 | 3936 |
2138 | 3937 s->version=0; |
2967 | 3938 |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3939 s->m.avctx = avctx; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3940 s->m.flags = avctx->flags; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3941 s->m.bit_rate= avctx->bit_rate; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3942 |
2138 | 3943 s->m.me.scratchpad= av_mallocz((avctx->width+64)*2*16*2*sizeof(uint8_t)); |
3944 s->m.me.map = av_mallocz(ME_MAP_SIZE*sizeof(uint32_t)); | |
3945 s->m.me.score_map = av_mallocz(ME_MAP_SIZE*sizeof(uint32_t)); | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
3946 s->m.obmc_scratchpad= av_mallocz(MB_SIZE*MB_SIZE*12*sizeof(uint32_t)); |
2138 | 3947 h263_encode_init(&s->m); //mv_penalty |
3948 | |
3314 | 3949 s->max_ref_frames = FFMAX(FFMIN(avctx->refs, MAX_REF_FRAMES), 1); |
3950 | |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3951 if(avctx->flags&CODEC_FLAG_PASS1){ |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3952 if(!avctx->stats_out) |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3953 avctx->stats_out = av_mallocz(256); |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3954 } |
3322
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
3955 if((avctx->flags&CODEC_FLAG_PASS2) || !(avctx->flags&CODEC_FLAG_QSCALE)){ |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3956 if(ff_rate_control_init(&s->m) < 0) |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3957 return -1; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3958 } |
3313 | 3959 s->pass1_rc= !(avctx->flags & (CODEC_FLAG_QSCALE|CODEC_FLAG_PASS2)); |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
3960 |
2138 | 3961 for(plane_index=0; plane_index<3; plane_index++){ |
3962 calculate_vissual_weight(s, &s->plane[plane_index]); | |
3963 } | |
2967 | 3964 |
3965 | |
2138 | 3966 avctx->coded_frame= &s->current_picture; |
3967 switch(avctx->pix_fmt){ | |
3968 // case PIX_FMT_YUV444P: | |
3969 // case PIX_FMT_YUV422P: | |
3970 case PIX_FMT_YUV420P: | |
3971 case PIX_FMT_GRAY8: | |
3972 // case PIX_FMT_YUV411P: | |
3973 // case PIX_FMT_YUV410P: | |
3974 s->colorspace_type= 0; | |
3975 break; | |
4494
ce643a22f049
Replace deprecated PIX_FMT names by the newer variants.
diego
parents:
4436
diff
changeset
|
3976 /* case PIX_FMT_RGB32: |
2138 | 3977 s->colorspace= 1; |
3978 break;*/ | |
3979 default: | |
3980 av_log(avctx, AV_LOG_ERROR, "format not supported\n"); | |
3981 return -1; | |
3982 } | |
3983 // avcodec_get_chroma_sub_sample(avctx->pix_fmt, &s->chroma_h_shift, &s->chroma_v_shift); | |
3984 s->chroma_h_shift= 1; | |
3985 s->chroma_v_shift= 1; | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3986 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3987 ff_set_cmp(&s->dsp, s->dsp.me_cmp, s->avctx->me_cmp); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3988 ff_set_cmp(&s->dsp, s->dsp.me_sub_cmp, s->avctx->me_sub_cmp); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3989 |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3990 s->avctx->get_buffer(s->avctx, &s->input_picture); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
3991 |
3314 | 3992 if(s->avctx->me_method == ME_ITER){ |
3993 int i; | |
3994 int size= s->b_width * s->b_height << 2*s->block_max_depth; | |
3995 for(i=0; i<s->max_ref_frames; i++){ | |
3996 s->ref_mvs[i]= av_mallocz(size*sizeof(int16_t[2])); | |
3997 s->ref_scores[i]= av_mallocz(size*sizeof(uint32_t)); | |
3998 } | |
3999 } | |
4000 | |
2138 | 4001 return 0; |
4002 } | |
4003 | |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4004 static void halfpel_interpol(SnowContext *s, uint8_t *halfpel[4][4], AVFrame *frame){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4005 int p,x,y; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4006 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4007 assert(!(s->avctx->flags & CODEC_FLAG_EMU_EDGE)); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4008 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4009 for(p=0; p<3; p++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4010 int is_chroma= !!p; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4011 int w= s->avctx->width >>is_chroma; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4012 int h= s->avctx->height >>is_chroma; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4013 int ls= frame->linesize[p]; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4014 uint8_t *src= frame->data[p]; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4015 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4016 halfpel[1][p]= (uint8_t*)av_malloc(ls * (h+2*EDGE_WIDTH)) + EDGE_WIDTH*(1+ls); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4017 halfpel[2][p]= (uint8_t*)av_malloc(ls * (h+2*EDGE_WIDTH)) + EDGE_WIDTH*(1+ls); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4018 halfpel[3][p]= (uint8_t*)av_malloc(ls * (h+2*EDGE_WIDTH)) + EDGE_WIDTH*(1+ls); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4019 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4020 halfpel[0][p]= src; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4021 for(y=0; y<h; y++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4022 for(x=0; x<w; x++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4023 int i= y*ls + x; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4024 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4025 halfpel[1][p][i]= (20*(src[i] + src[i+1]) - 5*(src[i-1] + src[i+2]) + (src[i-2] + src[i+3]) + 16 )>>5; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4026 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4027 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4028 for(y=0; y<h; y++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4029 for(x=0; x<w; x++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4030 int i= y*ls + x; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4031 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4032 halfpel[2][p][i]= (20*(src[i] + src[i+ls]) - 5*(src[i-ls] + src[i+2*ls]) + (src[i-2*ls] + src[i+3*ls]) + 16 )>>5; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4033 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4034 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4035 src= halfpel[1][p]; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4036 for(y=0; y<h; y++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4037 for(x=0; x<w; x++){ |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4038 int i= y*ls + x; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4039 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4040 halfpel[3][p][i]= (20*(src[i] + src[i+ls]) - 5*(src[i-ls] + src[i+2*ls]) + (src[i-2*ls] + src[i+3*ls]) + 16 )>>5; |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4041 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4042 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4043 |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4044 //FIXME border! |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4045 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4046 } |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4047 |
2138 | 4048 static int frame_start(SnowContext *s){ |
4049 AVFrame tmp; | |
2187 | 4050 int w= s->avctx->width; //FIXME round up to x16 ? |
4051 int h= s->avctx->height; | |
2138 | 4052 |
2187 | 4053 if(s->current_picture.data[0]){ |
4054 draw_edges(s->current_picture.data[0], s->current_picture.linesize[0], w , h , EDGE_WIDTH ); | |
4055 draw_edges(s->current_picture.data[1], s->current_picture.linesize[1], w>>1, h>>1, EDGE_WIDTH/2); | |
4056 draw_edges(s->current_picture.data[2], s->current_picture.linesize[2], w>>1, h>>1, EDGE_WIDTH/2); | |
4057 } | |
4058 | |
3314 | 4059 tmp= s->last_picture[s->max_ref_frames-1]; |
4060 memmove(s->last_picture+1, s->last_picture, (s->max_ref_frames-1)*sizeof(AVFrame)); | |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4061 memmove(s->halfpel_plane+1, s->halfpel_plane, (s->max_ref_frames-1)*sizeof(void*)*4*4); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4062 #ifdef USE_HALFPEL_PLANE |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4063 if(s->current_picture.data[0]) |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4064 halfpel_interpol(s, s->halfpel_plane[0], &s->current_picture); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4065 #endif |
3314 | 4066 s->last_picture[0]= s->current_picture; |
2138 | 4067 s->current_picture= tmp; |
2967 | 4068 |
3314 | 4069 if(s->keyframe){ |
4070 s->ref_frames= 0; | |
4071 }else{ | |
4072 int i; | |
4073 for(i=0; i<s->max_ref_frames && s->last_picture[i].data[0]; i++) | |
4074 if(i && s->last_picture[i-1].key_frame) | |
4075 break; | |
4076 s->ref_frames= i; | |
4077 } | |
4078 | |
2138 | 4079 s->current_picture.reference= 1; |
4080 if(s->avctx->get_buffer(s->avctx, &s->current_picture) < 0){ | |
4081 av_log(s->avctx, AV_LOG_ERROR, "get_buffer() failed\n"); | |
4082 return -1; | |
4083 } | |
2967 | 4084 |
3314 | 4085 s->current_picture.key_frame= s->keyframe; |
4086 | |
2138 | 4087 return 0; |
4088 } | |
4089 | |
4090 static int encode_frame(AVCodecContext *avctx, unsigned char *buf, int buf_size, void *data){ | |
4091 SnowContext *s = avctx->priv_data; | |
2335 | 4092 RangeCoder * const c= &s->c; |
2138 | 4093 AVFrame *pict = data; |
4094 const int width= s->avctx->width; | |
4095 const int height= s->avctx->height; | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4096 int level, orientation, plane_index, i, y; |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4097 uint8_t rc_header_bak[sizeof(s->header_state)]; |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4098 uint8_t rc_block_bak[sizeof(s->block_state)]; |
2138 | 4099 |
2335 | 4100 ff_init_range_encoder(c, buf, buf_size); |
4101 ff_build_rac_states(c, 0.05*(1LL<<32), 256-8); | |
2967 | 4102 |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4103 for(i=0; i<3; i++){ |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4104 int shift= !!i; |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4105 for(y=0; y<(height>>shift); y++) |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4106 memcpy(&s->input_picture.data[i][y * s->input_picture.linesize[i]], |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4107 &pict->data[i][y * pict->linesize[i]], |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4108 width>>shift); |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4109 } |
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4110 s->new_picture = *pict; |
2138 | 4111 |
3313 | 4112 s->m.picture_number= avctx->frame_number; |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4113 if(avctx->flags&CODEC_FLAG_PASS2){ |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4114 s->m.pict_type = |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4115 pict->pict_type= s->m.rc_context.entry[avctx->frame_number].new_pict_type; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4116 s->keyframe= pict->pict_type==FF_I_TYPE; |
3766 | 4117 if(!(avctx->flags&CODEC_FLAG_QSCALE)) { |
3193 | 4118 pict->quality= ff_rate_estimate_qscale(&s->m, 0); |
3766 | 4119 if (pict->quality < 0) |
4120 return -1; | |
4121 } | |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4122 }else{ |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4123 s->keyframe= avctx->gop_size==0 || avctx->frame_number % avctx->gop_size == 0; |
3313 | 4124 s->m.pict_type= |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4125 pict->pict_type= s->keyframe ? FF_I_TYPE : FF_P_TYPE; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4126 } |
2967 | 4127 |
3313 | 4128 if(s->pass1_rc && avctx->frame_number == 0) |
4129 pict->quality= 2*FF_QP2LAMBDA; | |
2161 | 4130 if(pict->quality){ |
3322
0b4f548dfb44
Snow: 10l. 1pass ratecontrol failed to set snow's internal quant.
lorenm
parents:
3314
diff
changeset
|
4131 s->qlog= qscale2qlog(pict->quality); |
3313 | 4132 s->lambda = pict->quality * 3/2; |
4133 } | |
4134 if(s->qlog < 0 || (!pict->quality && (avctx->flags & CODEC_FLAG_QSCALE))){ | |
2161 | 4135 s->qlog= LOSSLESS_QLOG; |
3313 | 4136 s->lambda = 0; |
4137 }//else keep previous frame's qlog until after motion est | |
2138 | 4138 |
4139 frame_start(s); | |
4140 | |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4141 s->m.current_picture_ptr= &s->m.current_picture; |
2138 | 4142 if(pict->pict_type == P_TYPE){ |
4143 int block_width = (width +15)>>4; | |
4144 int block_height= (height+15)>>4; | |
4145 int stride= s->current_picture.linesize[0]; | |
2967 | 4146 |
2138 | 4147 assert(s->current_picture.data[0]); |
3314 | 4148 assert(s->last_picture[0].data[0]); |
2967 | 4149 |
2138 | 4150 s->m.avctx= s->avctx; |
4151 s->m.current_picture.data[0]= s->current_picture.data[0]; | |
3314 | 4152 s->m. last_picture.data[0]= s->last_picture[0].data[0]; |
2138 | 4153 s->m. new_picture.data[0]= s-> input_picture.data[0]; |
4154 s->m. last_picture_ptr= &s->m. last_picture; | |
4155 s->m.linesize= | |
4156 s->m. last_picture.linesize[0]= | |
4157 s->m. new_picture.linesize[0]= | |
4158 s->m.current_picture.linesize[0]= stride; | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4159 s->m.uvlinesize= s->current_picture.linesize[1]; |
2138 | 4160 s->m.width = width; |
4161 s->m.height= height; | |
4162 s->m.mb_width = block_width; | |
4163 s->m.mb_height= block_height; | |
4164 s->m.mb_stride= s->m.mb_width+1; | |
4165 s->m.b8_stride= 2*s->m.mb_width+1; | |
4166 s->m.f_code=1; | |
4167 s->m.pict_type= pict->pict_type; | |
4168 s->m.me_method= s->avctx->me_method; | |
4169 s->m.me.scene_change_score=0; | |
4170 s->m.flags= s->avctx->flags; | |
4171 s->m.quarter_sample= (s->avctx->flags & CODEC_FLAG_QPEL)!=0; | |
4172 s->m.out_format= FMT_H263; | |
4173 s->m.unrestricted_mv= 1; | |
4174 | |
3313 | 4175 s->m.lambda = s->lambda; |
2138 | 4176 s->m.qscale= (s->m.lambda*139 + FF_LAMBDA_SCALE*64) >> (FF_LAMBDA_SHIFT + 7); |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4177 s->lambda2= s->m.lambda2= (s->m.lambda*s->m.lambda + FF_LAMBDA_SCALE/2) >> FF_LAMBDA_SHIFT; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4178 |
2138 | 4179 s->m.dsp= s->dsp; //move |
4180 ff_init_me(&s->m); | |
2993
cb0e26759cca
iterative overlapped block based motion estimation for snow
michael
parents:
2979
diff
changeset
|
4181 s->dsp= s->m.dsp; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4182 } |
2967 | 4183 |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4184 if(s->pass1_rc){ |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4185 memcpy(rc_header_bak, s->header_state, sizeof(s->header_state)); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4186 memcpy(rc_block_bak, s->block_state, sizeof(s->block_state)); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4187 } |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4188 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4189 redo_frame: |
2967 | 4190 |
3313 | 4191 s->m.pict_type = pict->pict_type; |
2138 | 4192 s->qbias= pict->pict_type == P_TYPE ? 2 : 0; |
4193 | |
4194 encode_header(s); | |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4195 s->m.misc_bits = 8*(s->c.bytestream - s->c.bytestream_start); |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4196 encode_blocks(s, 1); |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4197 s->m.mv_bits = 8*(s->c.bytestream - s->c.bytestream_start) - s->m.misc_bits; |
2967 | 4198 |
2138 | 4199 for(plane_index=0; plane_index<3; plane_index++){ |
4200 Plane *p= &s->plane[plane_index]; | |
4201 int w= p->width; | |
4202 int h= p->height; | |
4203 int x, y; | |
2198 | 4204 // int bits= put_bits_count(&s->c.pb); |
2138 | 4205 |
3338
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4206 if(!(avctx->flags2 & CODEC_FLAG2_MEMC_ONLY)){ |
2138 | 4207 //FIXME optimize |
4208 if(pict->data[plane_index]) //FIXME gray hack | |
4209 for(y=0; y<h; y++){ | |
4210 for(x=0; x<w; x++){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4211 s->spatial_idwt_buffer[y*w + x]= pict->data[plane_index][y*pict->linesize[plane_index] + x]<<FRAC_BITS; |
2138 | 4212 } |
4213 } | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4214 predict_plane(s, s->spatial_idwt_buffer, plane_index, 0); |
2967 | 4215 |
4216 if( plane_index==0 | |
4217 && pict->pict_type == P_TYPE | |
3313 | 4218 && !(avctx->flags&CODEC_FLAG_PASS2) |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4219 && s->m.me.scene_change_score > s->avctx->scenechange_threshold){ |
2335 | 4220 ff_init_range_encoder(c, buf, buf_size); |
4221 ff_build_rac_states(c, 0.05*(1LL<<32), 256-8); | |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4222 pict->pict_type= FF_I_TYPE; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4223 s->keyframe=1; |
3314 | 4224 s->current_picture.key_frame=1; |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4225 goto redo_frame; |
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4226 } |
2967 | 4227 |
2161 | 4228 if(s->qlog == LOSSLESS_QLOG){ |
4229 for(y=0; y<h; y++){ | |
4230 for(x=0; x<w; x++){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4231 s->spatial_dwt_buffer[y*w + x]= (s->spatial_idwt_buffer[y*w + x] + (1<<(FRAC_BITS-1))-1)>>FRAC_BITS; |
2161 | 4232 } |
4233 } | |
5575 | 4234 }else{ |
4235 for(y=0; y<h; y++){ | |
4236 for(x=0; x<w; x++){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4237 s->spatial_dwt_buffer[y*w + x]=s->spatial_idwt_buffer[y*w + x]<<ENCODER_EXTRA_BITS; |
5575 | 4238 } |
4239 } | |
2161 | 4240 } |
2967 | 4241 |
2164 | 4242 ff_spatial_dwt(s->spatial_dwt_buffer, w, h, w, s->spatial_decomposition_type, s->spatial_decomposition_count); |
2161 | 4243 |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4244 if(s->pass1_rc && plane_index==0){ |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4245 int delta_qlog = ratecontrol_1pass(s, pict); |
4011
5bce97c30a69
-1 is a valid return value in ratecontrol_1pass() -> 100l for takis
michael
parents:
4001
diff
changeset
|
4246 if (delta_qlog <= INT_MIN) |
3766 | 4247 return -1; |
3661
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4248 if(delta_qlog){ |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4249 //reordering qlog in the bitstream would eliminate this reset |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4250 ff_init_range_encoder(c, buf, buf_size); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4251 memcpy(s->header_state, rc_header_bak, sizeof(s->header_state)); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4252 memcpy(s->block_state, rc_block_bak, sizeof(s->block_state)); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4253 encode_header(s); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4254 encode_blocks(s, 0); |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4255 } |
b4425339894b
fix snow 1pass rc: the qlog used and the one written to the bitstream weren't always the same.
lorenm
parents:
3556
diff
changeset
|
4256 } |
3313 | 4257 |
2138 | 4258 for(level=0; level<s->spatial_decomposition_count; level++){ |
4259 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
4260 SubBand *b= &p->band[level][orientation]; | |
2967 | 4261 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4262 quantize(s, b, b->ibuf, b->buf, b->stride, s->qbias); |
2138 | 4263 if(orientation==0) |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4264 decorrelate(s, b, b->ibuf, b->stride, pict->pict_type == P_TYPE, 0); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4265 encode_subband(s, b, b->ibuf, b->parent ? b->parent->ibuf : NULL, b->stride, orientation); |
2138 | 4266 assert(b->parent==NULL || b->parent->stride == b->stride*2); |
4267 if(orientation==0) | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4268 correlate(s, b, b->ibuf, b->stride, 1, 0); |
2138 | 4269 } |
4270 } | |
4271 // av_log(NULL, AV_LOG_DEBUG, "plane:%d bits:%d\n", plane_index, put_bits_count(&s->c.pb) - bits); | |
4272 | |
4273 for(level=0; level<s->spatial_decomposition_count; level++){ | |
4274 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
4275 SubBand *b= &p->band[level][orientation]; | |
4276 | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4277 dequantize(s, b, b->ibuf, b->stride); |
2138 | 4278 } |
4279 } | |
2161 | 4280 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4281 ff_spatial_idwt(s->spatial_idwt_buffer, w, h, w, s->spatial_decomposition_type, s->spatial_decomposition_count); |
2161 | 4282 if(s->qlog == LOSSLESS_QLOG){ |
4283 for(y=0; y<h; y++){ | |
4284 for(x=0; x<w; x++){ | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4285 s->spatial_idwt_buffer[y*w + x]<<=FRAC_BITS; |
2161 | 4286 } |
4287 } | |
4288 } | |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
4289 {START_TIMER |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4290 predict_plane(s, s->spatial_idwt_buffer, plane_index, 1); |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
4291 STOP_TIMER("pred-conv")} |
3338
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4292 }else{ |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4293 //ME/MC only |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4294 if(pict->pict_type == I_TYPE){ |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4295 for(y=0; y<h; y++){ |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4296 for(x=0; x<w; x++){ |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4297 s->current_picture.data[plane_index][y*s->current_picture.linesize[plane_index] + x]= |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4298 pict->data[plane_index][y*pict->linesize[plane_index] + x]; |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4299 } |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4300 } |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4301 }else{ |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4302 memset(s->spatial_idwt_buffer, 0, sizeof(IDWTELEM)*w*h); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4303 predict_plane(s, s->spatial_idwt_buffer, plane_index, 1); |
3338
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4304 } |
937f14bb0f23
support doing motion estimation and compensation without any residual transform or coding
michael
parents:
3327
diff
changeset
|
4305 } |
2138 | 4306 if(s->avctx->flags&CODEC_FLAG_PSNR){ |
4307 int64_t error= 0; | |
2967 | 4308 |
2138 | 4309 if(pict->data[plane_index]) //FIXME gray hack |
4310 for(y=0; y<h; y++){ | |
4311 for(x=0; x<w; x++){ | |
2161 | 4312 int d= s->current_picture.data[plane_index][y*s->current_picture.linesize[plane_index] + x] - pict->data[plane_index][y*pict->linesize[plane_index] + x]; |
2138 | 4313 error += d*d; |
4314 } | |
4315 } | |
4316 s->avctx->error[plane_index] += error; | |
2232 | 4317 s->current_picture.error[plane_index] = error; |
2138 | 4318 } |
4319 } | |
4320 | |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4321 if(s->last_picture[s->max_ref_frames-1].data[0]){ |
3314 | 4322 avctx->release_buffer(avctx, &s->last_picture[s->max_ref_frames-1]); |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4323 for(i=0; i<9; i++) |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4324 if(s->halfpel_plane[s->max_ref_frames-1][1+i/3][i%3]) |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4325 av_free(s->halfpel_plane[s->max_ref_frames-1][1+i/3][i%3] - EDGE_WIDTH*(1+s->current_picture.linesize[i%3])); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4326 } |
2138 | 4327 |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4328 s->current_picture.coded_picture_number = avctx->frame_number; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4329 s->current_picture.pict_type = pict->pict_type; |
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4330 s->current_picture.quality = pict->quality; |
3313 | 4331 s->m.frame_bits = 8*(s->c.bytestream - s->c.bytestream_start); |
4332 s->m.p_tex_bits = s->m.frame_bits - s->m.misc_bits - s->m.mv_bits; | |
4333 s->m.current_picture.display_picture_number = | |
4334 s->m.current_picture.coded_picture_number = avctx->frame_number; | |
4335 s->m.current_picture.quality = pict->quality; | |
4336 s->m.total_bits += 8*(s->c.bytestream - s->c.bytestream_start); | |
4337 if(s->pass1_rc) | |
3766 | 4338 if (ff_rate_estimate_qscale(&s->m, 0) < 0) |
4339 return -1; | |
3313 | 4340 if(avctx->flags&CODEC_FLAG_PASS1) |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4341 ff_write_pass1_stats(&s->m); |
3313 | 4342 s->m.last_pict_type = s->m.pict_type; |
4123
5c86acb39889
outputs bit spent on various encoding functions (motion vectors, overhead, etc)
gpoirier
parents:
4122
diff
changeset
|
4343 avctx->frame_bits = s->m.frame_bits; |
5c86acb39889
outputs bit spent on various encoding functions (motion vectors, overhead, etc)
gpoirier
parents:
4122
diff
changeset
|
4344 avctx->mv_bits = s->m.mv_bits; |
5c86acb39889
outputs bit spent on various encoding functions (motion vectors, overhead, etc)
gpoirier
parents:
4122
diff
changeset
|
4345 avctx->misc_bits = s->m.misc_bits; |
5c86acb39889
outputs bit spent on various encoding functions (motion vectors, overhead, etc)
gpoirier
parents:
4122
diff
changeset
|
4346 avctx->p_tex_bits = s->m.p_tex_bits; |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4347 |
2138 | 4348 emms_c(); |
2967 | 4349 |
2335 | 4350 return ff_rac_terminate(c); |
2138 | 4351 } |
4352 | |
4353 static void common_end(SnowContext *s){ | |
3314 | 4354 int plane_index, level, orientation, i; |
2192 | 4355 |
2138 | 4356 av_freep(&s->spatial_dwt_buffer); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4357 av_freep(&s->spatial_idwt_buffer); |
2138 | 4358 |
2967 | 4359 av_freep(&s->m.me.scratchpad); |
2138 | 4360 av_freep(&s->m.me.map); |
4361 av_freep(&s->m.me.score_map); | |
3033
e8599ab02b38
faster iterative_me: avoid duplicate mc of neighboring blocks.
lorenm
parents:
3020
diff
changeset
|
4362 av_freep(&s->m.obmc_scratchpad); |
2967 | 4363 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4364 av_freep(&s->block); |
2192 | 4365 |
3314 | 4366 for(i=0; i<MAX_REF_FRAMES; i++){ |
4367 av_freep(&s->ref_mvs[i]); | |
4368 av_freep(&s->ref_scores[i]); | |
4369 if(s->last_picture[i].data[0]) | |
4370 s->avctx->release_buffer(s->avctx, &s->last_picture[i]); | |
4371 } | |
4372 | |
2967 | 4373 for(plane_index=0; plane_index<3; plane_index++){ |
2192 | 4374 for(level=s->spatial_decomposition_count-1; level>=0; level--){ |
4375 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
4376 SubBand *b= &s->plane[plane_index].band[level][orientation]; | |
2967 | 4377 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4378 av_freep(&b->x_coeff); |
2192 | 4379 } |
4380 } | |
4381 } | |
2138 | 4382 } |
4383 | |
4384 static int encode_end(AVCodecContext *avctx) | |
4385 { | |
4386 SnowContext *s = avctx->priv_data; | |
4387 | |
4388 common_end(s); | |
2608
4fb7fa34050b
allow 2pass ratecontrol. also fixes psnr displayed by mencoder.
lorenm
parents:
2607
diff
changeset
|
4389 av_free(avctx->stats_out); |
2138 | 4390 |
4391 return 0; | |
4392 } | |
4393 | |
4394 static int decode_init(AVCodecContext *avctx) | |
4395 { | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4396 SnowContext *s = avctx->priv_data; |
2967 | 4397 |
2635 | 4398 avctx->pix_fmt= PIX_FMT_YUV420P; |
2138 | 4399 |
4400 common_init(avctx); | |
2967 | 4401 |
2138 | 4402 return 0; |
4403 } | |
4404 | |
4405 static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8_t *buf, int buf_size){ | |
4406 SnowContext *s = avctx->priv_data; | |
2335 | 4407 RangeCoder * const c= &s->c; |
2138 | 4408 int bytes_read; |
4409 AVFrame *picture = data; | |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4410 int level, orientation, plane_index, i; |
2138 | 4411 |
2335 | 4412 ff_init_range_decoder(c, buf, buf_size); |
4413 ff_build_rac_states(c, 0.05*(1LL<<32), 256-8); | |
2138 | 4414 |
4415 s->current_picture.pict_type= FF_I_TYPE; //FIXME I vs. P | |
4416 decode_header(s); | |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4417 |
5662
77297fe0311c
allocate slice buffer after parsing the header containing needed parameters
michael
parents:
5661
diff
changeset
|
4418 // realloc slice buffer for the case that spatial_decomposition_count changed |
77297fe0311c
allocate slice buffer after parsing the header containing needed parameters
michael
parents:
5661
diff
changeset
|
4419 slice_buffer_destroy(&s->sb); |
77297fe0311c
allocate slice buffer after parsing the header containing needed parameters
michael
parents:
5661
diff
changeset
|
4420 slice_buffer_init(&s->sb, s->plane[0].height, (MB_SIZE >> s->block_max_depth) + s->spatial_decomposition_count * 8 + 1, s->plane[0].width, s->spatial_idwt_buffer); |
77297fe0311c
allocate slice buffer after parsing the header containing needed parameters
michael
parents:
5661
diff
changeset
|
4421 |
5651
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4422 for(plane_index=0; plane_index<3; plane_index++){ |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4423 Plane *p= &s->plane[plane_index]; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4424 p->fast_mc= p->diag_mc && p->htaps==6 && p->hcoeff[0]==40 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4425 && p->hcoeff[1]==-10 |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4426 && p->hcoeff[2]==2; |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4427 } |
ab023c9f03d0
store halfpel filter coefficients in the header as well as the
michael
parents:
5650
diff
changeset
|
4428 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4429 if(!s->block) alloc_blocks(s); |
2138 | 4430 |
4431 frame_start(s); | |
4432 //keyframe flag dupliaction mess FIXME | |
4433 if(avctx->debug&FF_DEBUG_PICT_INFO) | |
4434 av_log(avctx, AV_LOG_ERROR, "keyframe:%d qlog:%d\n", s->keyframe, s->qlog); | |
2967 | 4435 |
2189
70b27300a496
quad tree based motion compensation (currently only 16x16 & 8x8 OBMC blocks, but can be extended to other block sizes easily)
michael
parents:
2187
diff
changeset
|
4436 decode_blocks(s); |
2138 | 4437 |
4438 for(plane_index=0; plane_index<3; plane_index++){ | |
4439 Plane *p= &s->plane[plane_index]; | |
4440 int w= p->width; | |
4441 int h= p->height; | |
4442 int x, y; | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4443 int decode_state[MAX_DECOMPOSITIONS][4][1]; /* Stored state info for unpack_coeffs. 1 variable per instance. */ |
2967 | 4444 |
2138 | 4445 if(s->avctx->debug&2048){ |
4446 memset(s->spatial_dwt_buffer, 0, sizeof(DWTELEM)*w*h); | |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4447 predict_plane(s, s->spatial_idwt_buffer, plane_index, 1); |
2138 | 4448 |
4449 for(y=0; y<h; y++){ | |
4450 for(x=0; x<w; x++){ | |
2249
2b1a5e1fd449
merge predict_plane() with DWTELEM->8bit conversation (21% faster)
michael
parents:
2246
diff
changeset
|
4451 int v= s->current_picture.data[plane_index][y*s->current_picture.linesize[plane_index] + x]; |
2138 | 4452 s->mconly_picture.data[plane_index][y*s->mconly_picture.linesize[plane_index] + x]= v; |
4453 } | |
4454 } | |
4455 } | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4456 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4457 { START_TIMER |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4458 for(level=0; level<s->spatial_decomposition_count; level++){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4459 for(orientation=level ? 1 : 0; orientation<4; orientation++){ |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4460 SubBand *b= &p->band[level][orientation]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4461 unpack_coeffs(s, b, b->parent, orientation); |
2138 | 4462 } |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4463 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4464 STOP_TIMER("unpack coeffs"); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4465 } |
2138 | 4466 |
2562 | 4467 {START_TIMER |
4468 const int mb_h= s->b_height << s->block_max_depth; | |
4469 const int block_size = MB_SIZE >> s->block_max_depth; | |
4470 const int block_w = plane_index ? block_size/2 : block_size; | |
4471 int mb_y; | |
4472 dwt_compose_t cs[MAX_DECOMPOSITIONS]; | |
4473 int yd=0, yq=0; | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4474 int y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4475 int end_y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4476 |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4477 ff_spatial_idwt_buffered_init(cs, &s->sb, w, h, 1, s->spatial_decomposition_type, s->spatial_decomposition_count); |
2562 | 4478 for(mb_y=0; mb_y<=mb_h; mb_y++){ |
2967 | 4479 |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4480 int slice_starty = block_w*mb_y; |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4481 int slice_h = block_w*(mb_y+1); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4482 if (!(s->keyframe || s->avctx->debug&512)){ |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4483 slice_starty = FFMAX(0, slice_starty - (block_w >> 1)); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4484 slice_h -= (block_w >> 1); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4485 } |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4486 |
2967 | 4487 { |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4488 START_TIMER |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4489 for(level=0; level<s->spatial_decomposition_count; level++){ |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4490 for(orientation=level ? 1 : 0; orientation<4; orientation++){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4491 SubBand *b= &p->band[level][orientation]; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4492 int start_y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4493 int end_y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4494 int our_mb_start = mb_y; |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4495 int our_mb_end = (mb_y + 1); |
3012 | 4496 const int extra= 3; |
4497 start_y = (mb_y ? ((block_w * our_mb_start) >> (s->spatial_decomposition_count - level)) + s->spatial_decomposition_count - level + extra: 0); | |
4498 end_y = (((block_w * our_mb_end) >> (s->spatial_decomposition_count - level)) + s->spatial_decomposition_count - level + extra); | |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4499 if (!(s->keyframe || s->avctx->debug&512)){ |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4500 start_y = FFMAX(0, start_y - (block_w >> (1+s->spatial_decomposition_count - level))); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4501 end_y = FFMAX(0, end_y - (block_w >> (1+s->spatial_decomposition_count - level))); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4502 } |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4503 start_y = FFMIN(b->height, start_y); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4504 end_y = FFMIN(b->height, end_y); |
2967 | 4505 |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4506 if (start_y != end_y){ |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4507 if (orientation == 0){ |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4508 SubBand * correlate_band = &p->band[0][0]; |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4509 int correlate_end_y = FFMIN(b->height, end_y + 1); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4510 int correlate_start_y = FFMIN(b->height, (start_y ? start_y + 1 : 0)); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4511 decode_subband_slice_buffered(s, correlate_band, &s->sb, correlate_start_y, correlate_end_y, decode_state[0][0]); |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4512 correlate_slice_buffered(s, &s->sb, correlate_band, correlate_band->ibuf, correlate_band->stride, 1, 0, correlate_start_y, correlate_end_y); |
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4513 dequantize_slice_buffered(s, &s->sb, correlate_band, correlate_band->ibuf, correlate_band->stride, start_y, end_y); |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4514 } |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4515 else |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4516 decode_subband_slice_buffered(s, b, &s->sb, start_y, end_y, decode_state[level][orientation]); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4517 } |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4518 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4519 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4520 STOP_TIMER("decode_subband_slice"); |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4521 } |
2967 | 4522 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4523 { START_TIMER |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4524 for(; yd<slice_h; yd+=4){ |
3198
6b9f0c4fbdbe
First part of a series of speed-enchancing patches.
gpoirier
parents:
3197
diff
changeset
|
4525 ff_spatial_idwt_buffered_slice(&s->dsp, cs, &s->sb, w, h, 1, s->spatial_decomposition_type, s->spatial_decomposition_count, yd); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4526 } |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4527 STOP_TIMER("idwt slice");} |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4528 |
2967 | 4529 |
2161 | 4530 if(s->qlog == LOSSLESS_QLOG){ |
2562 | 4531 for(; yq<slice_h && yq<h; yq++){ |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4532 IDWTELEM * line = slice_buffer_get_line(&s->sb, yq); |
2161 | 4533 for(x=0; x<w; x++){ |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4534 line[x] <<= FRAC_BITS; |
2161 | 4535 } |
4536 } | |
4537 } | |
2562 | 4538 |
5587
3ae03eacbe9f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
michael
parents:
5575
diff
changeset
|
4539 predict_slice_buffered(s, &s->sb, s->spatial_idwt_buffer, plane_index, 1, mb_y); |
2967 | 4540 |
2634
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4541 y = FFMIN(p->height, slice_starty); |
337217ecbb3e
tighter snow slicing patch by (Yartrebo || yartrebo earthlink net)
michael
parents:
2628
diff
changeset
|
4542 end_y = FFMIN(p->height, slice_h); |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4543 while(y < end_y) |
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4544 slice_buffer_release(&s->sb, y++); |
2562 | 4545 } |
2967 | 4546 |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4547 slice_buffer_flush(&s->sb); |
2967 | 4548 |
2562 | 4549 STOP_TIMER("idwt + predict_slices")} |
2138 | 4550 } |
2967 | 4551 |
2138 | 4552 emms_c(); |
4553 | |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4554 if(s->last_picture[s->max_ref_frames-1].data[0]){ |
3314 | 4555 avctx->release_buffer(avctx, &s->last_picture[s->max_ref_frames-1]); |
5633
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4556 for(i=0; i<9; i++) |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4557 if(s->halfpel_plane[s->max_ref_frames-1][1+i/3][i%3]) |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4558 av_free(s->halfpel_plane[s->max_ref_frames-1][1+i/3][i%3] - EDGE_WIDTH*(1+s->current_picture.linesize[i%3])); |
873ea64637d9
code to do halfpel interpolation per frame (unfinished and under ifdef but it
michael
parents:
5627
diff
changeset
|
4559 } |
2138 | 4560 |
2967 | 4561 if(!(s->avctx->debug&2048)) |
2138 | 4562 *picture= s->current_picture; |
4563 else | |
4564 *picture= s->mconly_picture; | |
2967 | 4565 |
2138 | 4566 *data_size = sizeof(AVFrame); |
2967 | 4567 |
2335 | 4568 bytes_read= c->bytestream - c->bytestream_start; |
4569 if(bytes_read ==0) av_log(s->avctx, AV_LOG_ERROR, "error at end of frame\n"); //FIXME | |
2138 | 4570 |
4571 return bytes_read; | |
4572 } | |
4573 | |
4574 static int decode_end(AVCodecContext *avctx) | |
4575 { | |
4576 SnowContext *s = avctx->priv_data; | |
4577 | |
2589
a5a62827f195
Snow Slicing patch by (Yartrebo) yartrebo earthlink net
michael
parents:
2562
diff
changeset
|
4578 slice_buffer_destroy(&s->sb); |
2967 | 4579 |
2138 | 4580 common_end(s); |
4581 | |
4582 return 0; | |
4583 } | |
4584 | |
4585 AVCodec snow_decoder = { | |
4586 "snow", | |
4587 CODEC_TYPE_VIDEO, | |
4588 CODEC_ID_SNOW, | |
4589 sizeof(SnowContext), | |
4590 decode_init, | |
4591 NULL, | |
4592 decode_end, | |
4593 decode_frame, | |
4594 0 /*CODEC_CAP_DR1*/ /*| CODEC_CAP_DRAW_HORIZ_BAND*/, | |
4595 NULL | |
4596 }; | |
4597 | |
5224 | 4598 #ifdef CONFIG_SNOW_ENCODER |
2138 | 4599 AVCodec snow_encoder = { |
4600 "snow", | |
4601 CODEC_TYPE_VIDEO, | |
4602 CODEC_ID_SNOW, | |
4603 sizeof(SnowContext), | |
4604 encode_init, | |
4605 encode_frame, | |
4606 encode_end, | |
4607 }; | |
2408
a6e4da1c28ee
Disable encoders patch by (Gianluigi Tiesi <mplayer netfarm it>)
michael
parents:
2368
diff
changeset
|
4608 #endif |
2138 | 4609 |
4610 | |
4611 #if 0 | |
4612 #undef malloc | |
4613 #undef free | |
4614 #undef printf | |
5380
389366aa3458
Fix the self tests which are contained in some codecs and are using random().
takis
parents:
5254
diff
changeset
|
4615 #undef random |
2138 | 4616 |
4617 int main(){ | |
4618 int width=256; | |
4619 int height=256; | |
4620 int buffer[2][width*height]; | |
4621 SnowContext s; | |
4622 int i; | |
4623 s.spatial_decomposition_count=6; | |
4624 s.spatial_decomposition_type=1; | |
2967 | 4625 |
2138 | 4626 printf("testing 5/3 DWT\n"); |
4627 for(i=0; i<width*height; i++) | |
4628 buffer[0][i]= buffer[1][i]= random()%54321 - 12345; | |
2967 | 4629 |
2951 | 4630 ff_spatial_dwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); |
4631 ff_spatial_idwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); | |
2967 | 4632 |
2138 | 4633 for(i=0; i<width*height; i++) |
4634 if(buffer[0][i]!= buffer[1][i]) printf("fsck: %d %d %d\n",i, buffer[0][i], buffer[1][i]); | |
4635 | |
4636 printf("testing 9/7 DWT\n"); | |
4637 s.spatial_decomposition_type=0; | |
4638 for(i=0; i<width*height; i++) | |
4639 buffer[0][i]= buffer[1][i]= random()%54321 - 12345; | |
2967 | 4640 |
2951 | 4641 ff_spatial_dwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); |
4642 ff_spatial_idwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); | |
2967 | 4643 |
2138 | 4644 for(i=0; i<width*height; i++) |
4001 | 4645 if(FFABS(buffer[0][i] - buffer[1][i])>20) printf("fsck: %d %d %d\n",i, buffer[0][i], buffer[1][i]); |
2967 | 4646 |
2951 | 4647 #if 0 |
2138 | 4648 printf("testing AC coder\n"); |
4649 memset(s.header_state, 0, sizeof(s.header_state)); | |
2335 | 4650 ff_init_range_encoder(&s.c, buffer[0], 256*256); |
2138 | 4651 ff_init_cabac_states(&s.c, ff_h264_lps_range, ff_h264_mps_state, ff_h264_lps_state, 64); |
2967 | 4652 |
2138 | 4653 for(i=-256; i<256; i++){ |
4654 START_TIMER | |
4001 | 4655 put_symbol(&s.c, s.header_state, i*i*i/3*FFABS(i), 1); |
2138 | 4656 STOP_TIMER("put_symbol") |
4657 } | |
2335 | 4658 ff_rac_terminate(&s.c); |
2138 | 4659 |
4660 memset(s.header_state, 0, sizeof(s.header_state)); | |
2335 | 4661 ff_init_range_decoder(&s.c, buffer[0], 256*256); |
2138 | 4662 ff_init_cabac_states(&s.c, ff_h264_lps_range, ff_h264_mps_state, ff_h264_lps_state, 64); |
2967 | 4663 |
2138 | 4664 for(i=-256; i<256; i++){ |
4665 int j; | |
4666 START_TIMER | |
4667 j= get_symbol(&s.c, s.header_state, 1); | |
4668 STOP_TIMER("get_symbol") | |
4001 | 4669 if(j!=i*i*i/3*FFABS(i)) printf("fsck: %d != %d\n", i, j); |
2138 | 4670 } |
2951 | 4671 #endif |
2138 | 4672 { |
4673 int level, orientation, x, y; | |
4674 int64_t errors[8][4]; | |
4675 int64_t g=0; | |
4676 | |
4677 memset(errors, 0, sizeof(errors)); | |
4678 s.spatial_decomposition_count=3; | |
4679 s.spatial_decomposition_type=0; | |
4680 for(level=0; level<s.spatial_decomposition_count; level++){ | |
4681 for(orientation=level ? 1 : 0; orientation<4; orientation++){ | |
4682 int w= width >> (s.spatial_decomposition_count-level); | |
4683 int h= height >> (s.spatial_decomposition_count-level); | |
4684 int stride= width << (s.spatial_decomposition_count-level); | |
4685 DWTELEM *buf= buffer[0]; | |
4686 int64_t error=0; | |
4687 | |
4688 if(orientation&1) buf+=w; | |
4689 if(orientation>1) buf+=stride>>1; | |
2967 | 4690 |
2138 | 4691 memset(buffer[0], 0, sizeof(int)*width*height); |
4692 buf[w/2 + h/2*stride]= 256*256; | |
2951 | 4693 ff_spatial_idwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); |
2138 | 4694 for(y=0; y<height; y++){ |
4695 for(x=0; x<width; x++){ | |
4696 int64_t d= buffer[0][x + y*width]; | |
4697 error += d*d; | |
4122
daae66c03857
Replace most of the %lld and %llx by their (cleaner) PRI*64 counterparts.
diego
parents:
4011
diff
changeset
|
4698 if(FFABS(width/2-x)<9 && FFABS(height/2-y)<9 && level==2) printf("%8"PRId64" ", d); |
2138 | 4699 } |
4001 | 4700 if(FFABS(height/2-y)<9 && level==2) printf("\n"); |
2138 | 4701 } |
4702 error= (int)(sqrt(error)+0.5); | |
4703 errors[level][orientation]= error; | |
4704 if(g) g=ff_gcd(g, error); | |
4705 else g= error; | |
4706 } | |
4707 } | |
4708 printf("static int const visual_weight[][4]={\n"); | |
4709 for(level=0; level<s.spatial_decomposition_count; level++){ | |
4710 printf(" {"); | |
4711 for(orientation=0; orientation<4; orientation++){ | |
4122
daae66c03857
Replace most of the %lld and %llx by their (cleaner) PRI*64 counterparts.
diego
parents:
4011
diff
changeset
|
4712 printf("%8"PRId64",", errors[level][orientation]/g); |
2138 | 4713 } |
4714 printf("},\n"); | |
4715 } | |
4716 printf("};\n"); | |
4717 { | |
4718 int level=2; | |
4719 int orientation=3; | |
4720 int w= width >> (s.spatial_decomposition_count-level); | |
4721 int h= height >> (s.spatial_decomposition_count-level); | |
4722 int stride= width << (s.spatial_decomposition_count-level); | |
4723 DWTELEM *buf= buffer[0]; | |
4724 int64_t error=0; | |
4725 | |
4726 buf+=w; | |
4727 buf+=stride>>1; | |
2967 | 4728 |
2138 | 4729 memset(buffer[0], 0, sizeof(int)*width*height); |
4730 #if 1 | |
4731 for(y=0; y<height; y++){ | |
4732 for(x=0; x<width; x++){ | |
4733 int tab[4]={0,2,3,1}; | |
4734 buffer[0][x+width*y]= 256*256*tab[(x&1) + 2*(y&1)]; | |
4735 } | |
4736 } | |
2951 | 4737 ff_spatial_dwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); |
2138 | 4738 #else |
4739 for(y=0; y<h; y++){ | |
4740 for(x=0; x<w; x++){ | |
4741 buf[x + y*stride ]=169; | |
4742 buf[x + y*stride-w]=64; | |
4743 } | |
4744 } | |
2951 | 4745 ff_spatial_idwt(buffer[0], width, height, width, s.spatial_decomposition_type, s.spatial_decomposition_count); |
2138 | 4746 #endif |
4747 for(y=0; y<height; y++){ | |
4748 for(x=0; x<width; x++){ | |
4749 int64_t d= buffer[0][x + y*width]; | |
4750 error += d*d; | |
4122
daae66c03857
Replace most of the %lld and %llx by their (cleaner) PRI*64 counterparts.
diego
parents:
4011
diff
changeset
|
4751 if(FFABS(width/2-x)<9 && FFABS(height/2-y)<9) printf("%8"PRId64" ", d); |
2138 | 4752 } |
4001 | 4753 if(FFABS(height/2-y)<9) printf("\n"); |
2138 | 4754 } |
4755 } | |
4756 | |
4757 } | |
4758 return 0; | |
4759 } | |
4760 #endif | |
4761 |