Mercurial > mplayer.hg
changeset 12828:f9f8e49ff3d5
Cinepak and RoqA/V are now in ffmpeg
author | rtognimp |
---|---|
date | Thu, 15 Jul 2004 20:44:39 +0000 |
parents | f7c6d352c238 |
children | dcdee268bc3b |
files | libmpcodecs/native/cinepak.c libmpcodecs/native/roqav.c |
diffstat | 2 files changed, 0 insertions(+), 1563 deletions(-) [+] |
line wrap: on
line diff
--- a/libmpcodecs/native/cinepak.c Thu Jul 15 20:36:04 2004 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,886 +0,0 @@ -/* ------------------------------------------------------------------------ - * Radius Cinepak Video Decoder - * - * Dr. Tim Ferguson, 2001. - * For more details on the algorithm: - * http://www.csse.monash.edu.au/~timf/videocodec.html - * - * This is basically a vector quantiser with adaptive vector density. The - * frame is segmented into 4x4 pixel blocks, and each block is coded using - * either 1 or 4 vectors. - * - * There are still some issues with this code yet to be resolved. In - * particular with decoding in the strip boundaries. - * ------------------------------------------------------------------------ */ -#include <stdio.h> -#include <stdlib.h> -#include <string.h> -#include <sys/types.h> -#include <unistd.h> -#include <math.h> - -#include "config.h" -#include "mp_msg.h" -#include "bswap.h" - -#include "img_format.h" -#include "mp_image.h" - -#define DBUG 0 -#define MAX_STRIPS 32 - -/* ------------------------------------------------------------------------ */ -typedef struct -{ - unsigned char y0, y1, y2, y3; - char u, v; - - // These variables are for YV12 output: The v1 vars are for - // when the vector is doublesized and used by itself to paint a - // 4x4 block. - // This quad (y0 y0 y1 y1) is used on the 2 upper rows. - unsigned long yv12_v1_u; - // This quad (y2 y2 y3 y3) is used on the 2 lower rows. - unsigned long yv12_v1_l; - // The v4 vars are for when the vector is used as 1 of 4 vectors - // to paint a 4x4 block. - // Upper pair (y0 y1): - unsigned short yv12_v4_u; - // Lower pair (y2 y3): - unsigned short yv12_v4_l; - - // These longs are for YUY2 output: The v1 vars are for when the - // vector is doublesized and used by itself to paint a 4x4 block. - // The names stand for the upper-left, upper-right, - // lower-left, and lower-right YUY2 pixel pairs. - unsigned long yuy2_v1_ul, yuy2_v1_ur; - unsigned long yuy2_v1_ll, yuy2_v1_lr; - // The v4 vars are for when the vector is used as 1 of 4 vectors - // to paint a 4x4 block. The names stand for upper and lower - // YUY2 pixel pairs. - unsigned long yuy2_v4_u, yuy2_v4_l; - - // These longs are for BGR32 output - unsigned long rgb0, rgb1, rgb2, rgb3; - - // These char arrays are for BGR24 output - unsigned char r[4], g[4], b[4]; -} cvid_codebook; - -typedef struct { - cvid_codebook *v4_codebook[MAX_STRIPS]; - cvid_codebook *v1_codebook[MAX_STRIPS]; - unsigned long strip_num; -} cinepak_info; - - -/* ------------------------------------------------------------------------ */ -static unsigned char *in_buffer, uiclip[1024], *uiclp = NULL; - -#define SCALEBITS 16 -#define ONE_HALF ((long) 1 << (SCALEBITS-1)) -#define FIX(x) ((long) ((x) * (1L<<SCALEBITS) + 0.5)) -static long CU_Y_tab[256], CV_Y_tab[256], CU_Cb_tab[256], CV_Cb_tab[256], - CU_Cr_tab[256], CV_Cr_tab[256]; - -#define get_byte() *(in_buffer++) -#define skip_byte() in_buffer++ -#define get_word() ((unsigned short)(in_buffer += 2, \ - (in_buffer[-2] << 8 | in_buffer[-1]))) -#define get_long() ((unsigned long)(in_buffer += 4, \ - (in_buffer[-4] << 24 | in_buffer[-3] << 16 | in_buffer[-2] << 8 | in_buffer[-1]))) - - -/* ---------------------------------------------------------------------- */ - -// This PACKing macro packs the luminance bytes as y1-y1-y0-y0, which is -// stored on a little endian machine as y0-y0-y1-y1. Therefore, treat it as -// a little endian number and rearrange the bytes on big endian machines -// using the built-in byte order macros. -#define PACK_YV12_V1_Y(cb,y0,y1) le2me_32((((unsigned char)cb->y1)<<24)|(cb->y1<<16)|(((unsigned char)cb->y0)<<8)|(cb->y0)) -#define PACK_YV12_V4_Y(cb,y0,y1) le2me_16((((unsigned char)cb->y1)<<8)|(cb->y0)) - -static inline void read_codebook_yv12(cvid_codebook *c, int mode) -{ -unsigned char y0, y1, y2, y3, u, v; -int y_uv; - - if(mode) /* black and white */ - { - c->y0 = get_byte(); - c->y1 = get_byte(); - c->y2 = get_byte(); - c->y3 = get_byte(); - c->u = c->v = 128; - } - else /* colour */ - { - y0 = get_byte(); /* luma */ - y1 = get_byte(); - y2 = get_byte(); - y3 = get_byte(); - u = 128+get_byte(); /* chroma */ - v = 128+get_byte(); - - /* YUV * inv(CinYUV) - * | Y | | 1 -0.0655 0.0110 | | CY | - * | Cb | = | 0 1.1656 -0.0062 | | CU | - * | Cr | | 0 0.0467 1.4187 | | CV | - */ - y_uv = (int)((CU_Y_tab[u] + CV_Y_tab[v]) >> SCALEBITS); - c->y0 = uiclp[y0 + y_uv]; - c->y1 = uiclp[y1 + y_uv]; - c->y2 = uiclp[y2 + y_uv]; - c->y3 = uiclp[y3 + y_uv]; - c->u = uiclp[(int)((CU_Cb_tab[u] + CV_Cb_tab[v]) >> SCALEBITS)]; - c->v = uiclp[(int)((CU_Cr_tab[u] + CV_Cr_tab[v]) >> SCALEBITS)]; - - c->yv12_v1_u = PACK_YV12_V1_Y(c, y0, y1); - c->yv12_v1_l = PACK_YV12_V1_Y(c, y2, y3); - c->yv12_v4_u = PACK_YV12_V4_Y(c, y0, y1); - c->yv12_v4_l = PACK_YV12_V4_Y(c, y2, y3); - } -} - -/* ---------------------------------------------------------------------- */ - -inline void cvid_v1_yv12(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb) -{ -unsigned char *p; -int stride; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - // take care of the luminance - stride = mpi->stride[0]; - p = mpi->planes[0]+y*stride+x; - *((unsigned int*)p)=cb->yv12_v1_u; - *((unsigned int*)(p+stride))=cb->yv12_v1_u; - *((unsigned int*)(p+stride*2))=cb->yv12_v1_l; - *((unsigned int*)(p+stride*3))=cb->yv12_v1_l; - - // now for the chrominance - x/=2; y/=2; - - stride = mpi->stride[1]; - p = mpi->planes[1]+y*stride+x; - p[0]=p[1]=p[stride]=p[stride+1]=cb->u; - - stride = mpi->stride[2]; - p = mpi->planes[2]+y*stride+x; - p[0]=p[1]=p[stride]=p[stride+1]=cb->v; - -} - -/* ---------------------------------------------------------------------- */ -inline void cvid_v4_yv12(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb0, - cvid_codebook *cb1, cvid_codebook *cb2, cvid_codebook *cb3) -{ -unsigned char *p; -int stride; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - // take care of the luminance - stride = mpi->stride[0]; - p = mpi->planes[0]+y*stride+x; - ((unsigned short*)p)[0]=cb0->yv12_v4_u; - ((unsigned short*)p)[1]=cb1->yv12_v4_u; - ((unsigned short*)(p+stride))[0]=cb0->yv12_v4_l; - ((unsigned short*)(p+stride))[1]=cb1->yv12_v4_l; - ((unsigned short*)(p+stride*2))[0]=cb2->yv12_v4_u; - ((unsigned short*)(p+stride*2))[1]=cb3->yv12_v4_u; - ((unsigned short*)(p+stride*3))[0]=cb2->yv12_v4_l; - ((unsigned short*)(p+stride*3))[1]=cb3->yv12_v4_l; - - // now for the chrominance - x/=2; y/=2; - - stride = mpi->stride[1]; - p = mpi->planes[1]+y*stride+x; - p[0]=cb0->u; p[1]=cb1->u; - p[stride]=cb2->u; p[stride+1]=cb3->u; - - stride = mpi->stride[2]; - p = mpi->planes[2]+y*stride+x; - p[0]=cb0->v; p[1]=cb1->v; - p[stride]=cb2->v; p[stride+1]=cb3->v; - -} - -/* ---------------------------------------------------------------------- */ - -#define PACK_YUY2(cb,y0,y1,u,v) le2me_32(((((unsigned char)cb->v)<<24)|(cb->y1<<16)|(((unsigned char)cb->u)<<8)|(cb->y0))) - -static inline void read_codebook_yuy2(cvid_codebook *c, int mode) -{ -unsigned char y0, y1, y2, y3, u, v; -int y_uv; - - if(mode) /* black and white */ - { - c->y0 = get_byte(); - c->y1 = get_byte(); - c->y2 = get_byte(); - c->y3 = get_byte(); - c->u = c->v = 128; - } - else /* colour */ - { - y0 = get_byte(); /* luma */ - y1 = get_byte(); - y2 = get_byte(); - y3 = get_byte(); - u = 128+get_byte(); /* chroma */ - v = 128+get_byte(); - - /* YUV * inv(CinYUV) - * | Y | | 1 -0.0655 0.0110 | | CY | - * | Cb | = | 0 1.1656 -0.0062 | | CU | - * | Cr | | 0 0.0467 1.4187 | | CV | - */ - y_uv = (int)((CU_Y_tab[u] + CV_Y_tab[v]) >> SCALEBITS); - c->y0 = uiclp[y0 + y_uv]; - c->y1 = uiclp[y1 + y_uv]; - c->y2 = uiclp[y2 + y_uv]; - c->y3 = uiclp[y3 + y_uv]; - c->u = uiclp[(int)((CU_Cb_tab[u] + CV_Cb_tab[v]) >> SCALEBITS)]; - c->v = uiclp[(int)((CU_Cr_tab[u] + CV_Cr_tab[v]) >> SCALEBITS)]; - - c->yuy2_v4_u = PACK_YUY2(c, y0, y1, u, v); - c->yuy2_v4_l = PACK_YUY2(c, y2, y3, u, v); - c->yuy2_v1_ul = PACK_YUY2(c, y0, y0, u, v); - c->yuy2_v1_ur = PACK_YUY2(c, y1, y1, u, v); - c->yuy2_v1_ll = PACK_YUY2(c, y2, y2, u, v); - c->yuy2_v1_lr = PACK_YUY2(c, y3, y3, u, v); - } -} - -/* ------------------------------------------------------------------------ */ -inline void cvid_v1_yuy2(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb) -{ -int stride = mpi->stride[0] / 2; -unsigned long *vptr = (unsigned long *)mpi->planes[0]; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - vptr += (y * mpi->stride[0] + x) / 2; - - vptr[0] = cb->yuy2_v1_ul; - vptr[1] = cb->yuy2_v1_ur; - - vptr += stride; - vptr[0] = cb->yuy2_v1_ul; - vptr[1] = cb->yuy2_v1_ur; - - vptr += stride; - vptr[0] = cb->yuy2_v1_ll; - vptr[1] = cb->yuy2_v1_lr; - - vptr += stride; - vptr[0] = cb->yuy2_v1_ll; - vptr[1] = cb->yuy2_v1_lr; -} - - -/* ------------------------------------------------------------------------ */ -inline void cvid_v4_yuy2(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb0, - cvid_codebook *cb1, cvid_codebook *cb2, cvid_codebook *cb3) -{ -int stride = mpi->stride[0] / 2; -unsigned long *vptr = (unsigned long *)mpi->planes[0]; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - vptr += (y * mpi->stride[0] + x) / 2; - - vptr[0] = cb0->yuy2_v4_u; - vptr[1] = cb1->yuy2_v4_u; - - vptr += stride; - vptr[0] = cb0->yuy2_v4_l; - vptr[1] = cb1->yuy2_v4_l; - - vptr += stride; - vptr[0] = cb2->yuy2_v4_u; - vptr[1] = cb3->yuy2_v4_u; - - vptr += stride; - vptr[0] = cb2->yuy2_v4_l; - vptr[1] = cb3->yuy2_v4_l; -} - -/* ---------------------------------------------------------------------- */ -static inline void read_codebook_32(cvid_codebook *c, int mode) -{ -int uvr, uvg, uvb; - - if(mode) /* black and white */ - { - c->y0 = get_byte(); - c->y1 = get_byte(); - c->y2 = get_byte(); - c->y3 = get_byte(); - c->u = c->v = 0; - - c->rgb0 = (c->y0 << 16) | (c->y0 << 8) | c->y0; - c->rgb1 = (c->y1 << 16) | (c->y1 << 8) | c->y1; - c->rgb2 = (c->y2 << 16) | (c->y2 << 8) | c->y2; - c->rgb3 = (c->y3 << 16) | (c->y3 << 8) | c->y3; - } - else /* colour */ - { - c->y0 = get_byte(); /* luma */ - c->y1 = get_byte(); - c->y2 = get_byte(); - c->y3 = get_byte(); - c->u = get_byte(); /* chroma */ - c->v = get_byte(); - - uvr = c->v << 1; - uvg = -((c->u+1) >> 1) - c->v; - uvb = c->u << 1; - - c->rgb0 = le2me_32((uiclp[c->y0 + uvr] << 16) | (uiclp[c->y0 + uvg] << 8) | uiclp[c->y0 + uvb]); - c->rgb1 = le2me_32((uiclp[c->y1 + uvr] << 16) | (uiclp[c->y1 + uvg] << 8) | uiclp[c->y1 + uvb]); - c->rgb2 = le2me_32((uiclp[c->y2 + uvr] << 16) | (uiclp[c->y2 + uvg] << 8) | uiclp[c->y2 + uvb]); - c->rgb3 = le2me_32((uiclp[c->y3 + uvr] << 16) | (uiclp[c->y3 + uvg] << 8) | uiclp[c->y3 + uvb]); - } -} - - -/* ------------------------------------------------------------------------ */ -inline void cvid_v1_32(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb) -{ -int stride = mpi->stride[0]/4; -unsigned long *vptr = (unsigned long *)mpi->planes[0]; -unsigned long rgb; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - vptr += (y * stride + x); - - vptr[0] = rgb = cb->rgb0; vptr[1] = rgb; - vptr[2] = rgb = cb->rgb1; vptr[3] = rgb; - vptr += stride; - vptr[0] = rgb = cb->rgb0; vptr[1] = rgb; - vptr[2] = rgb = cb->rgb1; vptr[3] = rgb; - vptr += stride; - vptr[0] = rgb = cb->rgb2; vptr[1] = rgb; - vptr[2] = rgb = cb->rgb3; vptr[3] = rgb; - vptr += stride; - vptr[0] = rgb = cb->rgb2; vptr[1] = rgb; - vptr[2] = rgb = cb->rgb3; vptr[3] = rgb; -} - - -/* ------------------------------------------------------------------------ */ -inline void cvid_v4_32(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb0, - cvid_codebook *cb1, cvid_codebook *cb2, cvid_codebook *cb3) -{ -int stride = mpi->stride[0]/4; -unsigned long *vptr = (unsigned long *)mpi->planes[0]; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - vptr += (y * stride + x); - - vptr[0] = cb0->rgb0; - vptr[1] = cb0->rgb1; - vptr[2] = cb1->rgb0; - vptr[3] = cb1->rgb1; - vptr += stride; - vptr[0] = cb0->rgb2; - vptr[1] = cb0->rgb3; - vptr[2] = cb1->rgb2; - vptr[3] = cb1->rgb3; - vptr += stride; - vptr[0] = cb2->rgb0; - vptr[1] = cb2->rgb1; - vptr[2] = cb3->rgb0; - vptr[3] = cb3->rgb1; - vptr += stride; - vptr[0] = cb2->rgb2; - vptr[1] = cb2->rgb3; - vptr[2] = cb3->rgb2; - vptr[3] = cb3->rgb3; -} - - -/* ---------------------------------------------------------------------- */ -static inline void read_codebook_24(cvid_codebook *c, int mode) -{ -int uvr, uvg, uvb; - - if(mode) /* black and white */ - { - c->y0 = get_byte(); - c->y1 = get_byte(); - c->y2 = get_byte(); - c->y3 = get_byte(); - c->u = c->v = 0; - - c->r[0] = c->g[0] = c->b[0] = c->y0; - c->r[1] = c->g[1] = c->b[1] = c->y1; - c->r[2] = c->g[2] = c->b[2] = c->y2; - c->r[3] = c->g[3] = c->b[3] = c->y3; - } - else /* colour */ - { - c->y0 = get_byte(); /* luma */ - c->y1 = get_byte(); - c->y2 = get_byte(); - c->y3 = get_byte(); - c->u = get_byte(); /* chroma */ - c->v = get_byte(); - - uvr = c->v << 1; - uvg = -((c->u+1) >> 1) - c->v; - uvb = c->u << 1; - - c->r[0] = uiclp[c->y0 + uvr]; c->g[0] = uiclp[c->y0 + uvg]; c->b[0] = uiclp[c->y0 + uvb]; - c->r[1] = uiclp[c->y1 + uvr]; c->g[1] = uiclp[c->y1 + uvg]; c->b[1] = uiclp[c->y1 + uvb]; - c->r[2] = uiclp[c->y2 + uvr]; c->g[2] = uiclp[c->y2 + uvg]; c->b[2] = uiclp[c->y2 + uvb]; - c->r[3] = uiclp[c->y3 + uvr]; c->g[3] = uiclp[c->y3 + uvg]; c->b[3] = uiclp[c->y3 + uvb]; - } -} - - -/* ------------------------------------------------------------------------ */ -inline void cvid_v1_24(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb) -{ -unsigned char r, g, b; -int stride = (mpi->stride[0])-4*3; -unsigned char *vptr = mpi->planes[0] + (y * mpi->stride[0]) + x * 3; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - *vptr++ = b = cb->b[0]; *vptr++ = g = cb->g[0]; *vptr++ = r = cb->r[0]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - *vptr++ = b = cb->b[1]; *vptr++ = g = cb->g[1]; *vptr++ = r = cb->r[1]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - vptr += stride; - *vptr++ = b = cb->b[0]; *vptr++ = g = cb->g[0]; *vptr++ = r = cb->r[0]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - *vptr++ = b = cb->b[1]; *vptr++ = g = cb->g[1]; *vptr++ = r = cb->r[1]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - vptr += stride; - *vptr++ = b = cb->b[2]; *vptr++ = g = cb->g[2]; *vptr++ = r = cb->r[2]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - *vptr++ = b = cb->b[3]; *vptr++ = g = cb->g[3]; *vptr++ = r = cb->r[3]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - vptr += stride; - *vptr++ = b = cb->b[2]; *vptr++ = g = cb->g[2]; *vptr++ = r = cb->r[2]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; - *vptr++ = b = cb->b[3]; *vptr++ = g = cb->g[3]; *vptr++ = r = cb->r[3]; - *vptr++ = b; *vptr++ = g; *vptr++ = r; -} - - -/* ------------------------------------------------------------------------ */ -inline void cvid_v4_24(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb0, - cvid_codebook *cb1, cvid_codebook *cb2, cvid_codebook *cb3) -{ -int stride = (mpi->stride[0])-4*3; -unsigned char *vptr = mpi->planes[0] + (y * mpi->stride[0]) + x * 3; - - if(y+3>=(unsigned int)mpi->height) return; // avoid sig11 - - *vptr++ = cb0->b[0]; *vptr++ = cb0->g[0]; *vptr++ = cb0->r[0]; - *vptr++ = cb0->b[1]; *vptr++ = cb0->g[1]; *vptr++ = cb0->r[1]; - *vptr++ = cb1->b[0]; *vptr++ = cb1->g[0]; *vptr++ = cb1->r[0]; - *vptr++ = cb1->b[1]; *vptr++ = cb1->g[1]; *vptr++ = cb1->r[1]; - vptr += stride; - *vptr++ = cb0->b[2]; *vptr++ = cb0->g[2]; *vptr++ = cb0->r[2]; - *vptr++ = cb0->b[3]; *vptr++ = cb0->g[3]; *vptr++ = cb0->r[3]; - *vptr++ = cb1->b[2]; *vptr++ = cb1->g[2]; *vptr++ = cb1->r[2]; - *vptr++ = cb1->b[3]; *vptr++ = cb1->g[3]; *vptr++ = cb1->r[3]; - vptr += stride; - *vptr++ = cb2->b[0]; *vptr++ = cb2->g[0]; *vptr++ = cb2->r[0]; - *vptr++ = cb2->b[1]; *vptr++ = cb2->g[1]; *vptr++ = cb2->r[1]; - *vptr++ = cb3->b[0]; *vptr++ = cb3->g[0]; *vptr++ = cb3->r[0]; - *vptr++ = cb3->b[1]; *vptr++ = cb3->g[1]; *vptr++ = cb3->r[1]; - vptr += stride; - *vptr++ = cb2->b[2]; *vptr++ = cb2->g[2]; *vptr++ = cb2->r[2]; - *vptr++ = cb2->b[3]; *vptr++ = cb2->g[3]; *vptr++ = cb2->r[3]; - *vptr++ = cb3->b[2]; *vptr++ = cb3->g[2]; *vptr++ = cb3->r[2]; - *vptr++ = cb3->b[3]; *vptr++ = cb3->g[3]; *vptr++ = cb3->r[3]; -} - -/* ------------------------------------------------------------------------ - * Call this function once at the start of the sequence and save the - * returned context for calls to decode_cinepak(). - */ -void *decode_cinepak_init(void) -{ -cinepak_info *cvinfo; -int i, x; - - if((cvinfo = calloc(sizeof(cinepak_info), 1)) == NULL) return NULL; - cvinfo->strip_num = 0; - - if(uiclp == NULL) - { - uiclp = uiclip+512; - for(i = -512; i < 512; i++) - uiclp[i] = (i < 0 ? 0 : (i > 255 ? 255 : i)); - } - - for(i = 0, x = -128; i < 256; i++, x++) - { - CU_Y_tab[i] = (-FIX(0.0655)) * x; - CV_Y_tab[i] = (FIX(0.0110)) * x + ONE_HALF; - CU_Cb_tab[i] = (FIX(1.1656)) * x; - CV_Cb_tab[i] = (-FIX(0.0062)) * x + ONE_HALF + FIX(128); - CU_Cr_tab[i] = (FIX(0.0467)) * x; - CV_Cr_tab[i] = (FIX(1.4187)) * x + ONE_HALF + FIX(128); - } - - return (void *)cvinfo; -} - - -/* ------------------------------------------------------------------------ - * This function decodes a buffer containing a Cinepak encoded frame. - * - * context - the context created by decode_cinepak_init(). - * buf - the input buffer to be decoded - * size - the size of the input buffer - * frame - the output frame buffer - * width - the width of the output frame - * height - the height of the output frame - * bit_per_pixel - the number of bits per pixel allocated to the output - * frame; depths support: - * 32: BGR32 - * 24: BGR24 - * 16: YUY2 - * 12: YV12 - */ -void decode_cinepak(void *context, unsigned char *buf, int size, mp_image_t *mpi) -{ -cinepak_info *cvinfo = (cinepak_info *)context; -cvid_codebook *v4_codebook, *v1_codebook, *codebook = NULL; -unsigned long x, y, y_bottom, frame_flags, strips, cv_width, cv_height, cnum, - strip_id, chunk_id, x0, y0, x1, y1, ci, flag, mask; -long len, top_size, chunk_size; -unsigned int i, cur_strip, d0, d1, d2, d3; -int modulo; -void (*read_codebook)(cvid_codebook *c, int mode) = NULL; -void (*cvid_v1)(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb) = NULL; -void (*cvid_v4)(mp_image_t *mpi, unsigned int x, unsigned int y, cvid_codebook *cb0, - cvid_codebook *cb1, cvid_codebook *cb2, cvid_codebook *cb3) = NULL; - - x = y = 0; - y_bottom = 0; - in_buffer = buf; - - frame_flags = get_byte(); - len = get_byte() << 16; - len |= get_byte() << 8; - len |= get_byte(); - - switch(mpi->imgfmt) - { - case IMGFMT_I420: - case IMGFMT_IYUV: - case IMGFMT_YV12: // YV12 - read_codebook = read_codebook_yv12; - cvid_v1 = cvid_v1_yv12; - cvid_v4 = cvid_v4_yv12; - break; - case IMGFMT_YUY2: // YUY2 - read_codebook = read_codebook_yuy2; - cvid_v1 = cvid_v1_yuy2; - cvid_v4 = cvid_v4_yuy2; - break; - case IMGFMT_BGR24: // BGR24 - read_codebook = read_codebook_24; - cvid_v1 = cvid_v1_24; - cvid_v4 = cvid_v4_24; - break; - case IMGFMT_BGR32: // BGR32 - read_codebook = read_codebook_32; - cvid_v1 = cvid_v1_32; - cvid_v4 = cvid_v4_32; - break; - } - - if(len != size) - { - if(len & 0x01) len++; /* AVIs tend to have a size mismatch */ - if(len != size) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: corruption %d (QT/AVI) != %ld (CV)\n", size, len); - // return; - } - } - - cv_width = get_word(); - cv_height = get_word(); - strips = get_word(); - - if(strips > cvinfo->strip_num) - { - if(strips >= MAX_STRIPS) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: strip overflow (more than %d)\n", MAX_STRIPS); - return; - } - - for(i = cvinfo->strip_num; i < strips; i++) - { - if((cvinfo->v4_codebook[i] = (cvid_codebook *)calloc(sizeof(cvid_codebook), 260)) == NULL) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: codebook v4 alloc err\n"); - return; - } - - if((cvinfo->v1_codebook[i] = (cvid_codebook *)calloc(sizeof(cvid_codebook), 260)) == NULL) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: codebook v1 alloc err\n"); - return; - } - } - } - cvinfo->strip_num = strips; - -#if DBUG - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: <%ld,%ld> strips %ld\n", cv_width, cv_height, strips); -#endif - - for(cur_strip = 0; cur_strip < strips; cur_strip++) - { - v4_codebook = cvinfo->v4_codebook[cur_strip]; - v1_codebook = cvinfo->v1_codebook[cur_strip]; - - if((cur_strip > 0) && (!(frame_flags & 0x01))) - { - memcpy(cvinfo->v4_codebook[cur_strip], cvinfo->v4_codebook[cur_strip-1], 260 * sizeof(cvid_codebook)); - memcpy(cvinfo->v1_codebook[cur_strip], cvinfo->v1_codebook[cur_strip-1], 260 * sizeof(cvid_codebook)); - } - - strip_id = get_word(); /* 1000 = key strip, 1100 = iter strip */ - top_size = get_word(); - y0 = get_word(); /* FIXME: most of these are ignored at the moment */ - x0 = get_word(); - y1 = get_word(); - x1 = get_word(); - - y_bottom += y1; - top_size -= 12; - x = 0; -// if(x1 != (unsigned int)mpi->width) -// mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: Warning x1 (%ld) != width (%d)\n", x1, mpi->width); - -//x1 = mpi->width; -#if DBUG - mp_msg(MSGT_DECVIDEO, MSGL_WARN, " %d) %04lx %04ld <%ld,%ld> <%ld,%ld> yt %ld %d\n", - cur_strip, strip_id, top_size, x0, y0, x1, y1, y_bottom); -#endif - - while(top_size > 0) - { - chunk_id = get_word(); - chunk_size = get_word(); - -#if DBUG - mp_msg(MSGT_DECVIDEO, MSGL_WARN, " %04lx %04ld\n", chunk_id, chunk_size); -#endif - top_size -= chunk_size; - chunk_size -= 4; - - switch(chunk_id) - { - /* -------------------- Codebook Entries -------------------- */ - case 0x2000: - case 0x2200: - modulo = chunk_size % 6; - codebook = (chunk_id == 0x2200 ? v1_codebook : v4_codebook); - cnum = (chunk_size - modulo) / 6; - for(i = 0; i < cnum; i++) read_codebook(codebook+i, 0); - while (modulo--) - in_buffer++; - break; - - case 0x2400: - case 0x2600: /* 8 bit per pixel */ - codebook = (chunk_id == 0x2600 ? v1_codebook : v4_codebook); - cnum = chunk_size/4; - for(i = 0; i < cnum; i++) read_codebook(codebook+i, 1); - break; - - case 0x2100: - case 0x2300: - codebook = (chunk_id == 0x2300 ? v1_codebook : v4_codebook); - - ci = 0; - while(chunk_size > 3) - { - flag = get_long(); - chunk_size -= 4; - - for(i = 0; i < 32; i++) - { - if(flag & 0x80000000) - { - chunk_size -= 6; - read_codebook(codebook+ci, 0); - } - - ci++; - flag <<= 1; - } - } - while(chunk_size > 0) { skip_byte(); chunk_size--; } - break; - - case 0x2500: - case 0x2700: /* 8 bit per pixel */ - codebook = (chunk_id == 0x2700 ? v1_codebook : v4_codebook); - - ci = 0; - while(chunk_size > 0) - { - flag = get_long(); - chunk_size -= 4; - - for(i = 0; i < 32; i++) - { - if(flag & 0x80000000) - { - chunk_size -= 4; - read_codebook(codebook+ci, 1); - } - - ci++; - flag <<= 1; - } - } - while(chunk_size > 0) { skip_byte(); chunk_size--; } - break; - - /* -------------------- Frame -------------------- */ - case 0x3000: - while((chunk_size > 0) && (y < y_bottom)) - { - flag = get_long(); - chunk_size -= 4; - - for(i = 0; i < 32; i++) - { - if(y >= y_bottom) break; - if(flag & 0x80000000) /* 4 bytes per block */ - { - d0 = get_byte(); - d1 = get_byte(); - d2 = get_byte(); - d3 = get_byte(); - chunk_size -= 4; - cvid_v4(mpi, x, y, v4_codebook+d0, v4_codebook+d1, v4_codebook+d2, v4_codebook+d3); - } - else /* 1 byte per block */ - { - cvid_v1(mpi, x, y, v1_codebook + get_byte()); - chunk_size--; - } - - x += 4; - if(x >= (unsigned int)x1) - { - x = 0; - y += 4; - } - flag <<= 1; - } - } - while(chunk_size > 0) { skip_byte(); chunk_size--; } - break; - - case 0x3100: - while((chunk_size > 0) && (y < y_bottom)) - { - /* ---- flag bits: 0 = SKIP, 10 = V1, 11 = V4 ---- */ - flag = (unsigned long)get_long(); - chunk_size -= 4; - mask = 0x80000000; - - while((mask) && (y < y_bottom)) - { - if(flag & mask) - { - if(mask == 1) - { - if(chunk_size < 0) break; - flag = (unsigned long)get_long(); - chunk_size -= 4; - mask = 0x80000000; - } - else mask >>= 1; - - if(flag & mask) /* V4 */ - { - d0 = get_byte(); - d1 = get_byte(); - d2 = get_byte(); - d3 = get_byte(); - chunk_size -= 4; - cvid_v4(mpi, x, y, v4_codebook+d0, v4_codebook+d1, v4_codebook+d2, v4_codebook+d3); - } - else /* V1 */ - { - chunk_size--; - cvid_v1(mpi, x, y, v1_codebook + get_byte()); - } - } /* else SKIP */ - - mask >>= 1; - x += 4; - if(x >= (unsigned int)x1) - { - x = 0; - y += 4; - } - } - } - - while(chunk_size > 0) { skip_byte(); chunk_size--; } - break; - - case 0x3200: /* each byte is a V1 codebook */ - while((chunk_size > 0) && (y < y_bottom)) - { - cvid_v1(mpi, x, y, v1_codebook + get_byte()); - chunk_size--; - x += 4; - if(x >= (unsigned int)x1) - { - x = 0; - y += 4; - } - } - while(chunk_size > 0) { skip_byte(); chunk_size--; } - break; - - default: - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: unknown chunk_id %08lx\n", chunk_id); - while(chunk_size > 0) { skip_byte(); chunk_size--; } - break; - } - } - } - - if(len != size) - { - if(len & 0x01) len++; /* AVIs tend to have a size mismatch */ - if(len != size) - { - long xlen; - skip_byte(); - xlen = get_byte() << 16; - xlen |= get_byte() << 8; - xlen |= get_byte(); /* Read Len */ - mp_msg(MSGT_DECVIDEO, MSGL_WARN, "CVID: END INFO chunk size %d cvid size1 %ld cvid size2 %ld\n", size, len, xlen); - } - } -} -
--- a/libmpcodecs/native/roqav.c Thu Jul 15 20:36:04 2004 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,677 +0,0 @@ -/* - RoQ A/V decoder for the MPlayer program - by Mike Melanson - based on Dr. Tim Ferguson's RoQ document and accompanying source - code found at: - http://www.csse.monash.edu.au/~timf/videocodec.html -*/ - -#include <stdio.h> -#include <stdlib.h> -#include <string.h> -#include "config.h" -#include "bswap.h" -#include "mp_msg.h" -#include "mp_image.h" - -#define LE_16(x) (le2me_16(*(unsigned short *)(x))) -#define LE_32(x) (le2me_32(*(unsigned int *)(x))) - -#define CLAMP_S16(x) if (x < -32768) x = -32768; \ - else if (x > 32767) x = 32767; -#define SE_16BIT(x) if (x & 0x8000) x -= 0x10000; - -// RoQ chunk types -#define RoQ_INFO 0x1001 -#define RoQ_QUAD_CODEBOOK 0x1002 -#define RoQ_QUAD_VQ 0x1011 -#define RoQ_SOUND_MONO 0x1020 -#define RoQ_SOUND_STEREO 0x1021 - -#define MAX_ROQ_CODEBOOK_SIZE 256 - -// codebook entry for 2x2 vector -typedef struct -{ - // upper and lower luminance value pairs of 2x2 vector: [y0 y1], [y2 y3] - unsigned short v2_y_u; - unsigned short v2_y_l; - - // chrominance components - unsigned char u, v; - - // these variables are for rendering a doublesized 8x8 block; e.g.: - // v2_y_rows12 = [y0 y0 y1 y1] - // v2_y_rows34 = [y2 y2 y3 y3] - unsigned long v2d_y_rows_12; - unsigned long v2d_y_rows_34; - // ex: v2_u_row1 = [u u] - // v2_v_row2 = [v v] - unsigned short v2d_u_rows_12; - unsigned short v2d_v_rows_12; - - // maintain separate bytes for the luminance values as well - unsigned char y0, y1, y2, y3; -} roq_v2_codebook; - -// codebook entry for 4x4 vector -typedef struct -{ - unsigned char v2_index[4]; -} roq_v4_codebook; - -typedef struct -{ - roq_v2_codebook v2[MAX_ROQ_CODEBOOK_SIZE]; - roq_v4_codebook v4[MAX_ROQ_CODEBOOK_SIZE]; - mp_image_t *prev_frame; - uint32_t numframe; -} roqvideo_info; - - -// This function fills in the missing information for a v2 vector after -// loading the Y, U and V values. -inline void prep_v2(roq_v2_codebook *v2) -{ - v2->v2_y_u = be2me_16((v2->y0 << 8) | v2->y1); - v2->v2_y_l = be2me_16((v2->y2 << 8) | v2->y3); - - v2->v2d_y_rows_12 = be2me_32((v2->y0 << 24) | (v2->y0 << 16) | - (v2->y1 << 8) | v2->y1); - v2->v2d_y_rows_34 = be2me_32((v2->y2 << 24) | (v2->y2 << 16) | - (v2->y3 << 8) | v2->y3); - - // no reason to swap these for endianness since they're the same bytes - v2->v2d_u_rows_12 = (v2->u << 8) | v2->u; - v2->v2d_v_rows_12 = (v2->v << 8) | v2->v; -} - -inline void paint_v2double_block( - unsigned char *y_plane, - unsigned char *u_plane, - unsigned char *v_plane, - roq_v2_codebook *v2, - unsigned int y_stride, - unsigned int u_stride, - unsigned int v_stride) -{ - // render the luminance components - *(unsigned int *)y_plane = v2->v2d_y_rows_12; - y_plane += y_stride; - *(unsigned int *)y_plane = v2->v2d_y_rows_12; - y_plane += y_stride; - *(unsigned int *)y_plane = v2->v2d_y_rows_34; - y_plane += y_stride; - *(unsigned int *)y_plane = v2->v2d_y_rows_34; - - // render the color planes - *(unsigned short *)u_plane = v2->v2d_u_rows_12; - u_plane += u_stride; - *(unsigned short *)u_plane = v2->v2d_u_rows_12; - - *(unsigned short *)v_plane = v2->v2d_v_rows_12; - v_plane += v_stride; - *(unsigned short *)v_plane = v2->v2d_v_rows_12; -} - -inline void paint_v4_block( - unsigned char *y_plane, - unsigned char *u_plane, - unsigned char *v_plane, - unsigned int y_stride, - unsigned int u_stride, - unsigned int v_stride, - roq_v2_codebook *v2_a, - roq_v2_codebook *v2_b, - roq_v2_codebook *v2_c, - roq_v2_codebook *v2_d) -{ - // render luminance components - ((unsigned short *)y_plane)[0] = v2_a->v2_y_u; - ((unsigned short *)y_plane)[1] = v2_b->v2_y_u; - - y_plane += y_stride; - ((unsigned short *)y_plane)[0] = v2_a->v2_y_l; - ((unsigned short *)y_plane)[1] = v2_b->v2_y_l; - - y_plane += y_stride; - ((unsigned short *)y_plane)[0] = v2_c->v2_y_u; - ((unsigned short *)y_plane)[1] = v2_d->v2_y_u; - - y_plane += y_stride; - ((unsigned short *)y_plane)[0] = v2_c->v2_y_l; - ((unsigned short *)y_plane)[1] = v2_d->v2_y_l; - - // render the color planes - u_plane[0] = v2_a->u; - u_plane[1] = v2_b->u; - u_plane += u_stride; - u_plane[0] = v2_c->u; - u_plane[1] = v2_d->u; - - v_plane[0] = v2_a->v; - v_plane[1] = v2_b->v; - v_plane += v_stride; - v_plane[0] = v2_c->v; - v_plane[1] = v2_d->v; -} - -// This function copies the 4x4 block from the prev_*_planes to the -// current *_planes. -inline void copy_4x4_block( - unsigned char *y_plane, - unsigned char *u_plane, - unsigned char *v_plane, - unsigned char *prev_y_plane, - unsigned char *prev_u_plane, - unsigned char *prev_v_plane, - unsigned int y_stride, - unsigned int u_stride, - unsigned int v_stride) -{ - int i; - - // copy over the luminance components (4 rows, 1 uint each) - for (i = 0; i < 4; i++) - { - *(unsigned int *)y_plane = *(unsigned int *)prev_y_plane; - y_plane += y_stride; - prev_y_plane += y_stride; - } - - // copy the chrominance values - for (i = 0; i < 2; i++) - { - *(unsigned short*)u_plane = *(unsigned short*)prev_u_plane; - u_plane += u_stride; - prev_u_plane += u_stride; - *(unsigned short*)v_plane = *(unsigned short*)prev_v_plane; - v_plane += v_stride; - prev_v_plane += v_stride; - } -} - -// This function copies the 8x8 block from the prev_*_planes to the -// current *_planes. -inline void copy_8x8_block( - unsigned char *y_plane, - unsigned char *u_plane, - unsigned char *v_plane, - unsigned char *prev_y_plane, - unsigned char *prev_u_plane, - unsigned char *prev_v_plane, - unsigned int y_stride, - unsigned int u_stride, - unsigned int v_stride) -{ - int i; - - // copy over the luminance components (8 rows, 2 uints each) - for (i = 0; i < 8; i++) - { - ((unsigned int *)y_plane)[0] = ((unsigned int *)prev_y_plane)[0]; - ((unsigned int *)y_plane)[1] = ((unsigned int *)prev_y_plane)[1]; - y_plane += y_stride; - prev_y_plane += y_stride; - } - - // copy the chrominance values - for (i = 0; i < 4; i++) - { - *(unsigned int*)u_plane = *(unsigned int*)prev_u_plane; - u_plane += u_stride; - prev_u_plane += u_stride; - *(unsigned int*)v_plane = *(unsigned int*)prev_v_plane; - v_plane += v_stride; - prev_v_plane += v_stride; - } -} - -// This function creates storage space for the vector codebooks. -void *roq_decode_video_init(void) -{ - roqvideo_info *info = - (roqvideo_info *)calloc(sizeof(roqvideo_info), 1); - - info->prev_frame = NULL; - info->numframe=0; - - return info; -} - -#define EMPTY_ROQ_CODEWORD 0xFFFF0000 - -#define FETCH_NEXT_CODE() \ - if (current_roq_codeword == EMPTY_ROQ_CODEWORD) \ - { \ - if (stream_ptr + 2 > encoded_size) \ - { \ - mp_msg(MSGT_DECVIDEO, MSGL_WARN, \ - "RoQ video: stream pointer just went out of bounds (1)\n"); \ - return; \ - } \ - current_roq_codeword = (0x0000FFFF) | \ - (encoded[stream_ptr + 0] << 16) | \ - (encoded[stream_ptr + 1] << 24); \ - stream_ptr += 2; \ - } \ - roq_code = ((current_roq_codeword >> 30) & 0x03); \ - current_roq_codeword <<= 2; - -#define FETCH_NEXT_ARGUMENT() \ - if (stream_ptr + 1 > encoded_size) \ - { \ - mp_msg(MSGT_DECVIDEO, MSGL_WARN, \ - "RoQ video: stream pointer just went out of bounds (2)\n"); \ - return; \ - } \ - argument = encoded[stream_ptr++]; - -#define CHECK_PREV_FRAME() \ - if (!info->prev_frame) \ - { \ - mp_msg(MSGT_DECVIDEO, MSGL_WARN, \ - "RoQ video: can't handle motion vector when there's no previous frame\n"); \ - return; \ - } - -void roq_decode_video(void *context, unsigned char *encoded, - int encoded_size, mp_image_t *mpi) -{ - roqvideo_info *info = (roqvideo_info *)context; - - int stream_ptr = 0; - int i, j; - int chunk_length; - int v2_count; - int v4_count; - - int roq_code; - unsigned int current_roq_codeword = EMPTY_ROQ_CODEWORD; - unsigned char argument = 0; - char mean_motion_x; - char mean_motion_y; - int mx, my; // for calculating the motion vector - - int mblock_x = 0; - int mblock_y = 0; - int quad8_x, quad8_y; // for pointing to 8x8 blocks in a macroblock - int quad4_x, quad4_y; // for pointing to 4x4 blocks in an 8x8 block - - unsigned char *y_plane; - unsigned char *u_plane; - unsigned char *v_plane; - unsigned char *prev_y_plane; - unsigned char *prev_u_plane; - unsigned char *prev_v_plane; - unsigned int y_stride = mpi->stride[0]; - unsigned int u_stride = mpi->stride[1]; - unsigned int v_stride = mpi->stride[2]; - - roq_v4_codebook v4; - - // make sure the encoded chunk is of minimal acceptable length - if (encoded_size < 8) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video: chunk isn't even 8 bytes long (minimum acceptable length)\n"); - return; - } - - // make sure the resolution checks out - if ((mpi->width % 16 != 0) || (mpi->height % 16 != 0)) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video resolution: %d x %d; expected dimensions divisible by 16\n", - mpi->width, mpi->height); - return; - } - - if (LE_16(&encoded[stream_ptr]) == RoQ_QUAD_CODEBOOK) - { - stream_ptr += 2; - chunk_length = LE_32(&encoded[stream_ptr]); - stream_ptr += 4; - v4_count = encoded[stream_ptr++]; - v2_count = encoded[stream_ptr++]; - if (v2_count == 0) - v2_count = 256; - if ((v4_count == 0) && (v2_count * 6 < chunk_length)) - v4_count = 256; - - // make sure the lengths agree with each other - if (((v2_count * 6) + (v4_count * 4)) != chunk_length) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video: encountered quad codebook chunk with weird lengths (1)\n"); - return; - } - if ((v2_count * 6) > (encoded_size - stream_ptr)) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video: encountered quad codebook chunk with weird lengths (2)\n"); - return; - } - - // load the 2x2 vectors - for (i = 0; i < v2_count; i++) - { - info->v2[i].y0 = encoded[stream_ptr++]; - info->v2[i].y1 = encoded[stream_ptr++]; - info->v2[i].y2 = encoded[stream_ptr++]; - info->v2[i].y3 = encoded[stream_ptr++]; - info->v2[i].u = encoded[stream_ptr++]; - info->v2[i].v = encoded[stream_ptr++]; - prep_v2(&info->v2[i]); - } - - if ((v4_count * 4) > (encoded_size - stream_ptr)) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video: encountered quad codebook chunk with weird lengths (3)\n"); - return; - } - - // load the 4x4 vectors - for (i = 0; i < v4_count; i++) - { - info->v4[i].v2_index[0] = encoded[stream_ptr++]; - info->v4[i].v2_index[1] = encoded[stream_ptr++]; - info->v4[i].v2_index[2] = encoded[stream_ptr++]; - info->v4[i].v2_index[3] = encoded[stream_ptr++]; - } - } - - if (LE_16(&encoded[stream_ptr]) == RoQ_QUAD_VQ) - { - stream_ptr += 2; - chunk_length = LE_32(&encoded[stream_ptr]); - stream_ptr += 4; - mean_motion_y = encoded[stream_ptr++]; - mean_motion_x = encoded[stream_ptr++]; - - //RoQ reuses its buffers so a transparent block keeps content - //from 2 frames ago. The only exception is 2'd frame (#1) - if(info->numframe==1) - { - CHECK_PREV_FRAME(); - memcpy(mpi->planes[0], info->prev_frame->planes[0], - mpi->width * mpi->height); - memcpy(mpi->planes[1], info->prev_frame->planes[1], - (mpi->width * mpi->height) / 4); - memcpy(mpi->planes[2], info->prev_frame->planes[2], - (mpi->width * mpi->height) / 4); - } - - // iterate through the 16x16 macroblocks - for (mblock_y = 0; mblock_y < mpi->height; mblock_y += 16) - { - for (mblock_x = 0; mblock_x < mpi->width; mblock_x += 16) - { - // iterate through the 4 quadrants of the macroblock - for (i = 0; i < 4; i++) - { - quad8_x = mblock_x; - quad8_y = mblock_y; - if (i & 0x01) quad8_x += 8; - if (i & 0x02) quad8_y += 8; - - // set up the planes - y_plane = mpi->planes[0] + quad8_y * y_stride + quad8_x; - u_plane = mpi->planes[1] + (quad8_y / 2) * u_stride + (quad8_x / 2); - v_plane = mpi->planes[2] + (quad8_y / 2) * v_stride + (quad8_x / 2); - - // decide how to handle this 8x8 quad - FETCH_NEXT_CODE(); - switch(roq_code) - { - // 8x8 block is the same as in the previous frame; - // skip it - case 0: - break; - - // 8x8 block is painted with an 8x8 block from the last frame - // (i.e., motion compensation) - case 1: - CHECK_PREV_FRAME(); - - // prepare the pointers to the planes in the previous frame - FETCH_NEXT_ARGUMENT(); // argument contains motion vectors - - // figure out the motion vectors - mx = quad8_x + 8 - (argument >> 4) - mean_motion_x; - my = quad8_y + 8 - (argument & 0x0F) - mean_motion_y; - - prev_y_plane = info->prev_frame->planes[0] + - my * y_stride + mx; - prev_u_plane = info->prev_frame->planes[1] + - (my / 2) * u_stride + (mx + 1) / 2; - prev_v_plane = info->prev_frame->planes[2] + - (my / 2) * v_stride + (mx + 1) / 2; - -// sanity check before rendering - copy_8x8_block( - y_plane, - u_plane, - v_plane, - prev_y_plane, - prev_u_plane, - prev_v_plane, - y_stride, - u_stride, - v_stride - ); - break; - - // 8x8 block is painted with a doublesized 4x4 vector - case 2: - FETCH_NEXT_ARGUMENT(); - v4 = info->v4[argument]; - -// sanity check before rendering - // iterate through 4 4x4 blocks - for (j = 0; j < 4; j++) - { - quad4_x = quad8_x; - quad4_y = quad8_y; - if (j & 0x01) quad4_x += 4; - if (j & 0x02) quad4_y += 4; - - // set up the planes - y_plane = mpi->planes[0] + quad4_y * y_stride + quad4_x; - u_plane = mpi->planes[1] + - (quad4_y / 2) * u_stride + (quad4_x / 2); - v_plane = mpi->planes[2] + - (quad4_y / 2) * v_stride + (quad4_x / 2); - - paint_v2double_block( - y_plane, - u_plane, - v_plane, - &info->v2[v4.v2_index[j]], - y_stride, - u_stride, - v_stride - ); - } - break; - - // 8x8 block is broken down into 4 4x4 blocks and painted using - // 4 different codes. - case 3: - // iterate through 4 4x4 blocks - for (j = 0; j < 4; j++) - { - quad4_x = quad8_x; - quad4_y = quad8_y; - if (j & 0x01) quad4_x += 4; - if (j & 0x02) quad4_y += 4; - - // set up the planes - y_plane = mpi->planes[0] + quad4_y * y_stride + quad4_x; - u_plane = mpi->planes[1] + - (quad4_y / 2) * u_stride + (quad4_x / 2); - v_plane = mpi->planes[2] + - (quad4_y / 2) * v_stride + (quad4_x / 2); - - // decide how to handle this 4x4 quad - FETCH_NEXT_CODE(); - switch(roq_code) - { - // 4x4 block is the same as in the previous frame; - // skip it - case 0: - break; - - // 4x4 block is motion compensated from the previous frame - case 1: - CHECK_PREV_FRAME(); - // prepare the pointers to the planes in the previous frame - FETCH_NEXT_ARGUMENT(); // argument contains motion vectors - - // figure out the motion vectors - mx = quad4_x + 8 - (argument >> 4) - mean_motion_x; - my = quad4_y + 8 - (argument & 0x0F) - mean_motion_y; - - prev_y_plane = info->prev_frame->planes[0] + - my * y_stride + mx; - prev_u_plane = info->prev_frame->planes[1] + - (my / 2) * u_stride + (mx + 1) / 2; - prev_v_plane = info->prev_frame->planes[2] + - (my / 2) * u_stride + (mx + 1) / 2; - -// sanity check before rendering - copy_4x4_block( - y_plane, - u_plane, - v_plane, - prev_y_plane, - prev_u_plane, - prev_v_plane, - y_stride, - u_stride, - v_stride - ); - break; - - // 4x4 block is copied directly from v4 vector table - case 2: - FETCH_NEXT_ARGUMENT(); - v4 = info->v4[argument]; - - paint_v4_block( - y_plane, - u_plane, - v_plane, - y_stride, - u_stride, - v_stride, - &info->v2[v4.v2_index[0]], - &info->v2[v4.v2_index[1]], - &info->v2[v4.v2_index[2]], - &info->v2[v4.v2_index[3]]); - break; - - // 4x4 block is built from 4 2x2 vectors - case 3: - if (stream_ptr + 4 > encoded_size) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video: stream pointer just went out of bounds (2)\n"); - return; - } - paint_v4_block( - y_plane, - u_plane, - v_plane, - y_stride, - u_stride, - v_stride, - &info->v2[encoded[stream_ptr + 0]], - &info->v2[encoded[stream_ptr + 1]], - &info->v2[encoded[stream_ptr + 2]], - &info->v2[encoded[stream_ptr + 3]]); - stream_ptr += 4; - break; - } - } - break; - } - } - } - } - } - - // one last sanity check on the way out - // (apparently, it's not unusual to have 2 bytes left over after decode) - if (stream_ptr < encoded_size - 2) - { - mp_msg(MSGT_DECVIDEO, MSGL_WARN, - "RoQ video: completed frame decode with bytes left over (%d < %d)\n", - stream_ptr, encoded_size); - } - - // save the current frame as the previous frame for the next iteration - info->numframe++; - info->prev_frame = mpi; -} - -// Initialize the RoQ audio decoder, which is to say, initialize the table -// of squares. -void *roq_decode_audio_init(void) -{ - short *square_array; - short square; - int i; - - square_array = (short *)malloc(256 * sizeof(short)); - if (!square_array) - return NULL; - - for (i = 0; i < 128; i++) - { - square = i * i; - square_array[i] = square; - square_array[i + 128] = -square; - } - - return square_array; -} - -int roq_decode_audio( - unsigned short *output, - unsigned char *input, - int encoded_size, - int channels, - void *context) -{ - short *square_array = (short *)context; - int i; - int predictor[2]; - int channel_number = 0; - - // prepare the initial predictors - if (channels == 1) - predictor[0] = LE_16(&input[0]); - else - { - predictor[0] = input[1] << 8; - predictor[1] = input[0] << 8; - } - SE_16BIT(predictor[0]); - SE_16BIT(predictor[1]); - - // decode the samples - for (i = 2; i < encoded_size; i++) - { - predictor[channel_number] += square_array[input[i]]; - CLAMP_S16(predictor[channel_number]); - output[i - 2] = predictor[channel_number]; - - // toggle channel - channel_number ^= channels - 1; - } - - // return the number of samples decoded - return (encoded_size - 2); -}