comparison DOCS/man/en/mplayer.1 @ 13259:9f8b314406c2

MB, QP explanation, some more consistency

13258:7903766338f7

.IPs ec\ \ \ 
error concealment:
.RSss
1: Use strong deblock filter for damaged MBs.
.br
2: iterative MV search (slow)
.br
3: all (default)
.REss
.IPs er=<value>
error resilience:

.br
2: rate control
.br
4: bitstream
.br
8: MB type
.br
16: QP
.br
32: Motion vector
.br
0x0040: motion vector visualization (use \-noslices)
.br
0x0080: MB skip
.br
0x0100: startcode
.br
0x0200: PTS
.br

1: soft thresholding (better deringing, but blurrier)
.RE
.
.TP
.B qp=equation
QP change filter
.RSs
.IPs equation
some equation like "2+2*sin(PI*qp)"
.RE
.

.IPs mv\ \ \ 
Bits used for motion vectors
.IPs fCode
maximum length of motion vector in log2 scale
.IPs iCount
number of intra macro blocks / number of macro blocks
.IPs var\ \ 
spatial complexity
.IPs mcVar
temporal complexity
.IPs qComp

Larger values help against blockiness, if no deblocking filter is used for
decoding.
.br
.I HINT:
Crop any black borders completely as they will reduce the quality
of the macro blocks (also applies without scplx_mask).
.RSs
.IPs 0.0
disabled (default)
.IPs 0.0\-0.5
sane range

by the IDCT, which is obviously not the case) it simply finds a block for
the minimum of error and lambda*bits.
.PD 0
.RSs
.IPs lambda
QP dependant constant
.IPs bits\ 
amount of bits needed to encode the block
.IPs error
sum of squared errors of the quantization
.RE

.B mv0\ \ \ \ 
Try to encode each MB with MV=<0,0> and choose the better one.
This has no effect if mbd=0.
.
.TP
.B qprd\ \ \ 
rate distorted optimal QP for the given lambda of each macroblock, note this
needs mbd=2
.
.TP
.B last_pred=<0\-99>
Amount of motion predictors from the previous frame
.PD 0
.RSs
.IPs 0
(default)
.IPs a
will use 2a+1 x 2a+1 macro block square of motion vector predictors from the
previous frame
.RE
.PD 1
.
.TP

.B interlacing
Encode the fields of interlaced video material.
Turn this option on for interlaced content.
.br
.I NOTE:
Should you rescale the video, you would need a interlace-aware resizer,
which you can activate with -vf scale=<width>:<height>:1.
.
.TP
.B 4mv\ \ \ \ 
Use 4 motion vectors per macroblock.

.B qp_constant=<2-51>
This selects the quantizer to use.
20-40 is a useful range (default: 26).
Lower values result in better fidelity, but higher bitrates.
Note that quantization in H.264 works differently from MPEG[124].
H.264's QP is on a logarithmic scale.
As an example, the bitrate difference between QP=20 and QP=40
is about a factor of 10.
Useful quantizers in H.264 tend to be very large compared to MPEG[124].
.br
.I NOTE:

13259:9f8b314406c2

.IPs ec\ \ \ 
error concealment:
.RSss
1: Use strong deblock filter for damaged MBs.
.br
2: iterative motion vector (MV) search (slow)
.br
3: all (default)
.REss
.IPs er=<value>
error resilience:

.br
2: rate control
.br
4: bitstream
.br
8: macroblock (MB) type
.br
16: quantizer parameter (QP)
.br
32: motion vector
.br
0x0040: motion vector visualization (use \-noslices)
.br
0x0080: macroblock (MB) skip
.br
0x0100: startcode
.br
0x0200: PTS
.br

1: soft thresholding (better deringing, but blurrier)
.RE
.
.TP
.B qp=equation
quantizer parameter (QP) change filter
.RSs
.IPs equation
some equation like "2+2*sin(PI*qp)"
.RE
.

.IPs mv\ \ \ 
Bits used for motion vectors
.IPs fCode
maximum length of motion vector in log2 scale
.IPs iCount
number of intra macroblocks / number of macroblocks
.IPs var\ \ 
spatial complexity
.IPs mcVar
temporal complexity
.IPs qComp

Larger values help against blockiness, if no deblocking filter is used for
decoding.
.br
.I HINT:
Crop any black borders completely as they will reduce the quality
of the macroblocks (also applies without scplx_mask).
.RSs
.IPs 0.0
disabled (default)
.IPs 0.0\-0.5
sane range

by the IDCT, which is obviously not the case) it simply finds a block for
the minimum of error and lambda*bits.
.PD 0
.RSs
.IPs lambda
quantizer parameter (QP) dependant constant
.IPs bits\ 
amount of bits needed to encode the block
.IPs error
sum of squared errors of the quantization
.RE

.B mv0\ \ \ \ 
Try to encode each MB with MV=<0,0> and choose the better one.
This has no effect if mbd=0.
.
.TP
.B qprd (mbd=2 only)
rate distorted optimal quantizer parameter (QP) for the given
lambda of each macroblock
.
.TP
.B last_pred=<0\-99>
Amount of motion predictors from the previous frame
.PD 0
.RSs
.IPs 0
(default)
.IPs a
will use 2a+1 x 2a+1 macroblock square of motion vector predictors from the
previous frame
.RE
.PD 1
.
.TP

.B interlacing
Encode the fields of interlaced video material.
Turn this option on for interlaced content.
.br
.I NOTE:
Should you rescale the video, you would need an interlace-aware resizer,
which you can activate with -vf scale=<width>:<height>:1.
.
.TP
.B 4mv\ \ \ \ 
Use 4 motion vectors per macroblock.

.B qp_constant=<2-51>
This selects the quantizer to use.
20-40 is a useful range (default: 26).
Lower values result in better fidelity, but higher bitrates.
Note that quantization in H.264 works differently from MPEG[124].
H.264's quantizer parameter (QP) is on a logarithmic scale.
As an example, the bitrate difference between QP=20 and QP=40
is about a factor of 10.
Useful quantizers in H.264 tend to be very large compared to MPEG[124].
.br
.I NOTE: