3 * Copyright (c) 2000,2001 Fabrice Bellard
4 * Copyright (c) 2002-2004 Michael Niedermayer
6 * new motion estimation (X1/EPZS) by Michael Niedermayer <michaelni@gmx.at>
8 * This file is part of FFmpeg.
10 * FFmpeg is free software; you can redistribute it and/or
11 * modify it under the terms of the GNU Lesser General Public
12 * License as published by the Free Software Foundation; either
13 * version 2.1 of the License, or (at your option) any later version.
15 * FFmpeg is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 * Lesser General Public License for more details.
20 * You should have received a copy of the GNU Lesser General Public
21 * License along with FFmpeg; if not, write to the Free Software
22 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
37 #include "motion_est.h"
38 #include "mpegutils.h"
39 #include "mpegvideo.h"
43 #define P_TOPRIGHT P[3]
47 #define ME_MAP_SHIFT 3
48 #define ME_MAP_MV_BITS 11
50 static int sad_hpel_motion_search(MpegEncContext * s,
51 int *mx_ptr, int *my_ptr, int dmin,
52 int src_index, int ref_index,
55 static inline unsigned update_map_generation(MotionEstContext *c)
57 c->map_generation+= 1<<(ME_MAP_MV_BITS*2);
58 if(c->map_generation==0){
59 c->map_generation= 1<<(ME_MAP_MV_BITS*2);
60 memset(c->map, 0, sizeof(uint32_t)*ME_MAP_SIZE);
62 return c->map_generation;
65 /* shape adaptive search stuff */
66 typedef struct Minima{
72 static int minima_cmp(const void *a, const void *b){
73 const Minima *da = (const Minima *) a;
74 const Minima *db = (const Minima *) b;
76 return da->height - db->height;
79 #define FLAG_QPEL 1 //must be 1
83 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){
84 const int offset[3]= {
86 ((y*c->uvstride + x)>>1),
87 ((y*c->uvstride + x)>>1),
91 c->src[0][i]= src [i] + offset[i];
92 c->ref[0][i]= ref [i] + offset[i];
96 c->ref[ref_index][i]= ref2[i] + offset[i];
101 static int get_flags(MotionEstContext *c, int direct, int chroma){
102 return ((c->avctx->flags&AV_CODEC_FLAG_QPEL) ? FLAG_QPEL : 0)
103 + (direct ? FLAG_DIRECT : 0)
104 + (chroma ? FLAG_CHROMA : 0);
107 static av_always_inline int cmp_direct_inline(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
108 const int size, const int h, int ref_index, int src_index,
109 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, int qpel){
110 MotionEstContext * const c= &s->me;
111 const int stride= c->stride;
112 const int hx= subx + (x<<(1+qpel));
113 const int hy= suby + (y<<(1+qpel));
114 uint8_t * const * const ref= c->ref[ref_index];
115 uint8_t * const * const src= c->src[src_index];
117 //FIXME check chroma 4mv, (no crashes ...)
118 av_assert2(x >= c->xmin && hx <= c->xmax<<(qpel+1) && y >= c->ymin && hy <= c->ymax<<(qpel+1));
119 if(x >= c->xmin && hx <= c->xmax<<(qpel+1) && y >= c->ymin && hy <= c->ymax<<(qpel+1)){
120 const int time_pp= s->pp_time;
121 const int time_pb= s->pb_time;
122 const int mask= 2*qpel+1;
123 if(s->mv_type==MV_TYPE_8X8){
126 int fx = c->direct_basis_mv[i][0] + hx;
127 int fy = c->direct_basis_mv[i][1] + hy;
128 int bx = hx ? fx - c->co_located_mv[i][0] : c->co_located_mv[i][0]*(time_pb - time_pp)/time_pp + ((i &1)<<(qpel+4));
129 int by = hy ? fy - c->co_located_mv[i][1] : c->co_located_mv[i][1]*(time_pb - time_pp)/time_pp + ((i>>1)<<(qpel+4));
130 int fxy= (fx&mask) + ((fy&mask)<<(qpel+1));
131 int bxy= (bx&mask) + ((by&mask)<<(qpel+1));
133 uint8_t *dst= c->temp + 8*(i&1) + 8*stride*(i>>1);
135 c->qpel_put[1][fxy](dst, ref[0] + (fx>>2) + (fy>>2)*stride, stride);
136 c->qpel_avg[1][bxy](dst, ref[8] + (bx>>2) + (by>>2)*stride, stride);
138 c->hpel_put[1][fxy](dst, ref[0] + (fx>>1) + (fy>>1)*stride, stride, 8);
139 c->hpel_avg[1][bxy](dst, ref[8] + (bx>>1) + (by>>1)*stride, stride, 8);
143 int fx = c->direct_basis_mv[0][0] + hx;
144 int fy = c->direct_basis_mv[0][1] + hy;
145 int bx = hx ? fx - c->co_located_mv[0][0] : (c->co_located_mv[0][0]*(time_pb - time_pp)/time_pp);
146 int by = hy ? fy - c->co_located_mv[0][1] : (c->co_located_mv[0][1]*(time_pb - time_pp)/time_pp);
147 int fxy= (fx&mask) + ((fy&mask)<<(qpel+1));
148 int bxy= (bx&mask) + ((by&mask)<<(qpel+1));
151 c->qpel_put[1][fxy](c->temp , ref[0] + (fx>>2) + (fy>>2)*stride , stride);
152 c->qpel_put[1][fxy](c->temp + 8 , ref[0] + (fx>>2) + (fy>>2)*stride + 8 , stride);
153 c->qpel_put[1][fxy](c->temp + 8*stride, ref[0] + (fx>>2) + (fy>>2)*stride + 8*stride, stride);
154 c->qpel_put[1][fxy](c->temp + 8 + 8*stride, ref[0] + (fx>>2) + (fy>>2)*stride + 8 + 8*stride, stride);
155 c->qpel_avg[1][bxy](c->temp , ref[8] + (bx>>2) + (by>>2)*stride , stride);
156 c->qpel_avg[1][bxy](c->temp + 8 , ref[8] + (bx>>2) + (by>>2)*stride + 8 , stride);
157 c->qpel_avg[1][bxy](c->temp + 8*stride, ref[8] + (bx>>2) + (by>>2)*stride + 8*stride, stride);
158 c->qpel_avg[1][bxy](c->temp + 8 + 8*stride, ref[8] + (bx>>2) + (by>>2)*stride + 8 + 8*stride, stride);
160 av_assert2((fx>>1) + 16*s->mb_x >= -16);
161 av_assert2((fy>>1) + 16*s->mb_y >= -16);
162 av_assert2((fx>>1) + 16*s->mb_x <= s->width);
163 av_assert2((fy>>1) + 16*s->mb_y <= s->height);
164 av_assert2((bx>>1) + 16*s->mb_x >= -16);
165 av_assert2((by>>1) + 16*s->mb_y >= -16);
166 av_assert2((bx>>1) + 16*s->mb_x <= s->width);
167 av_assert2((by>>1) + 16*s->mb_y <= s->height);
169 c->hpel_put[0][fxy](c->temp, ref[0] + (fx>>1) + (fy>>1)*stride, stride, 16);
170 c->hpel_avg[0][bxy](c->temp, ref[8] + (bx>>1) + (by>>1)*stride, stride, 16);
173 d = cmp_func(s, c->temp, src[0], stride, 16);
179 static av_always_inline int cmp_inline(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
180 const int size, const int h, int ref_index, int src_index,
181 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, int qpel, int chroma){
182 MotionEstContext * const c= &s->me;
183 const int stride= c->stride;
184 const int uvstride= c->uvstride;
185 const int dxy= subx + (suby<<(1+qpel)); //FIXME log2_subpel?
186 const int hx= subx + x*(1<<(1+qpel));
187 const int hy= suby + y*(1<<(1+qpel));
188 uint8_t * const * const ref= c->ref[ref_index];
189 uint8_t * const * const src= c->src[src_index];
191 //FIXME check chroma 4mv, (no crashes ...)
192 int uvdxy; /* no, it might not be used uninitialized */
195 if (h << size == 16) {
196 c->qpel_put[size][dxy](c->temp, ref[0] + x + y*stride, stride); //FIXME prototype (add h)
197 } else if (size == 0 && h == 8) {
198 c->qpel_put[1][dxy](c->temp , ref[0] + x + y*stride , stride);
199 c->qpel_put[1][dxy](c->temp + 8, ref[0] + x + y*stride + 8, stride);
207 uvdxy= (cx&1) + 2*(cy&1);
208 // FIXME x/y wrong, but MPEG-4 qpel is sick anyway, we should drop as much of it as possible in favor for H.264
211 c->hpel_put[size][dxy](c->temp, ref[0] + x + y*stride, stride, h);
213 uvdxy= dxy | (x&1) | (2*(y&1));
215 d = cmp_func(s, c->temp, src[0], stride, h);
217 d = cmp_func(s, src[0], ref[0] + x + y*stride, stride, h);
219 uvdxy= (x&1) + 2*(y&1);
222 uint8_t * const uvtemp= c->temp + 16*stride;
223 c->hpel_put[size+1][uvdxy](uvtemp , ref[1] + (x>>1) + (y>>1)*uvstride, uvstride, h>>1);
224 c->hpel_put[size+1][uvdxy](uvtemp+8, ref[2] + (x>>1) + (y>>1)*uvstride, uvstride, h>>1);
225 d += chroma_cmp_func(s, uvtemp , src[1], uvstride, h>>1);
226 d += chroma_cmp_func(s, uvtemp+8, src[2], uvstride, h>>1);
231 static int cmp_simple(MpegEncContext *s, const int x, const int y,
232 int ref_index, int src_index,
233 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func){
234 return cmp_inline(s,x,y,0,0,0,16,ref_index,src_index, cmp_func, chroma_cmp_func, 0, 0);
237 static int cmp_fpel_internal(MpegEncContext *s, const int x, const int y,
238 const int size, const int h, int ref_index, int src_index,
239 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
240 if(flags&FLAG_DIRECT){
241 return cmp_direct_inline(s,x,y,0,0,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags&FLAG_QPEL);
243 return cmp_inline(s,x,y,0,0,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 0, flags&FLAG_CHROMA);
247 static int cmp_internal(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
248 const int size, const int h, int ref_index, int src_index,
249 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
250 if(flags&FLAG_DIRECT){
251 return cmp_direct_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags&FLAG_QPEL);
253 return cmp_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags&FLAG_QPEL, flags&FLAG_CHROMA);
257 /** @brief compares a block (either a full macroblock or a partition thereof)
258 against a proposed motion-compensated prediction of that block
260 static av_always_inline int cmp(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
261 const int size, const int h, int ref_index, int src_index,
262 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
263 if(av_builtin_constant_p(flags) && av_builtin_constant_p(h) && av_builtin_constant_p(size)
264 && av_builtin_constant_p(subx) && av_builtin_constant_p(suby)
265 && flags==0 && h==16 && size==0 && subx==0 && suby==0){
266 return cmp_simple(s,x,y,ref_index,src_index, cmp_func, chroma_cmp_func);
267 }else if(av_builtin_constant_p(subx) && av_builtin_constant_p(suby)
268 && subx==0 && suby==0){
269 return cmp_fpel_internal(s,x,y,size,h,ref_index,src_index, cmp_func, chroma_cmp_func,flags);
271 return cmp_internal(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags);
275 static int cmp_hpel(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
276 const int size, const int h, int ref_index, int src_index,
277 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
278 if(flags&FLAG_DIRECT){
279 return cmp_direct_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 0);
281 return cmp_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 0, flags&FLAG_CHROMA);
285 static int cmp_qpel(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
286 const int size, const int h, int ref_index, int src_index,
287 me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
288 if(flags&FLAG_DIRECT){
289 return cmp_direct_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 1);
291 return cmp_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 1, flags&FLAG_CHROMA);
295 #include "motion_est_template.c"
297 static int zero_cmp(MpegEncContext *s, uint8_t *a, uint8_t *b,
298 ptrdiff_t stride, int h)
303 static void zero_hpel(uint8_t *a, const uint8_t *b, ptrdiff_t stride, int h){
306 int ff_init_me(MpegEncContext *s){
307 MotionEstContext * const c= &s->me;
308 int cache_size= FFMIN(ME_MAP_SIZE>>ME_MAP_SHIFT, 1<<ME_MAP_SHIFT);
309 int dia_size= FFMAX(FFABS(s->avctx->dia_size)&255, FFABS(s->avctx->pre_dia_size)&255);
311 if(FFMIN(s->avctx->dia_size, s->avctx->pre_dia_size) < -FFMIN(ME_MAP_SIZE, MAX_SAB_SIZE)){
312 av_log(s->avctx, AV_LOG_ERROR, "ME_MAP size is too small for SAB diamond\n");
318 if(s->codec_id == AV_CODEC_ID_H261)
319 c->avctx->me_sub_cmp = c->avctx->me_cmp;
321 if(cache_size < 2*dia_size && !c->stride){
322 av_log(s->avctx, AV_LOG_INFO, "ME_MAP size may be a little small for the selected diamond size\n");
325 ff_set_cmp(&s->mecc, s->mecc.me_pre_cmp, c->avctx->me_pre_cmp);
326 ff_set_cmp(&s->mecc, s->mecc.me_cmp, c->avctx->me_cmp);
327 ff_set_cmp(&s->mecc, s->mecc.me_sub_cmp, c->avctx->me_sub_cmp);
328 ff_set_cmp(&s->mecc, s->mecc.mb_cmp, c->avctx->mb_cmp);
330 c->flags = get_flags(c, 0, c->avctx->me_cmp &FF_CMP_CHROMA);
331 c->sub_flags= get_flags(c, 0, c->avctx->me_sub_cmp&FF_CMP_CHROMA);
332 c->mb_flags = get_flags(c, 0, c->avctx->mb_cmp &FF_CMP_CHROMA);
334 /*FIXME s->no_rounding b_type*/
335 if (s->avctx->flags & AV_CODEC_FLAG_QPEL) {
336 c->sub_motion_search= qpel_motion_search;
337 c->qpel_avg = s->qdsp.avg_qpel_pixels_tab;
339 c->qpel_put = s->qdsp.put_no_rnd_qpel_pixels_tab;
341 c->qpel_put = s->qdsp.put_qpel_pixels_tab;
343 if(c->avctx->me_sub_cmp&FF_CMP_CHROMA)
344 c->sub_motion_search= hpel_motion_search;
345 else if( c->avctx->me_sub_cmp == FF_CMP_SAD
346 && c->avctx-> me_cmp == FF_CMP_SAD
347 && c->avctx-> mb_cmp == FF_CMP_SAD)
348 c->sub_motion_search= sad_hpel_motion_search; // 2050 vs. 2450 cycles
350 c->sub_motion_search= hpel_motion_search;
352 c->hpel_avg = s->hdsp.avg_pixels_tab;
354 c->hpel_put = s->hdsp.put_no_rnd_pixels_tab;
356 c->hpel_put = s->hdsp.put_pixels_tab;
359 c->stride = s->linesize;
360 c->uvstride= s->uvlinesize;
362 c->stride = 16*s->mb_width + 32;
363 c->uvstride= 8*s->mb_width + 16;
366 /* 8x8 fullpel search would need a 4x4 chroma compare, which we do
367 * not have yet, and even if we had, the motion estimation code
368 * does not expect it. */
369 if (s->codec_id != AV_CODEC_ID_SNOW) {
370 if ((c->avctx->me_cmp & FF_CMP_CHROMA) /* && !s->mecc.me_cmp[2] */)
371 s->mecc.me_cmp[2] = zero_cmp;
372 if ((c->avctx->me_sub_cmp & FF_CMP_CHROMA) && !s->mecc.me_sub_cmp[2])
373 s->mecc.me_sub_cmp[2] = zero_cmp;
374 c->hpel_put[2][0]= c->hpel_put[2][1]=
375 c->hpel_put[2][2]= c->hpel_put[2][3]= zero_hpel;
378 if(s->codec_id == AV_CODEC_ID_H261){
379 c->sub_motion_search= no_sub_motion_search;
385 #define CHECK_SAD_HALF_MV(suffix, x, y) \
387 d = s->mecc.pix_abs[size][(x ? 1 : 0) + (y ? 2 : 0)](NULL, pix, ptr + ((x) >> 1), stride, h); \
388 d += (mv_penalty[pen_x + x] + mv_penalty[pen_y + y])*penalty_factor;\
389 COPY3_IF_LT(dminh, d, dx, x, dy, y)\
392 static int sad_hpel_motion_search(MpegEncContext * s,
393 int *mx_ptr, int *my_ptr, int dmin,
394 int src_index, int ref_index,
397 MotionEstContext * const c= &s->me;
398 const int penalty_factor= c->sub_penalty_factor;
401 int stride= c->stride;
404 av_assert2(c->sub_flags == 0);
412 pix = c->src[src_index][0];
416 ptr = c->ref[ref_index][0] + (my * stride) + mx;
420 if (mx > xmin && mx < xmax &&
421 my > ymin && my < ymax) {
424 const int index= my*(1<<ME_MAP_SHIFT) + mx;
425 const int t= score_map[(index-(1<<ME_MAP_SHIFT))&(ME_MAP_SIZE-1)];
426 const int l= score_map[(index- 1 )&(ME_MAP_SIZE-1)];
427 const int r= score_map[(index+ 1 )&(ME_MAP_SIZE-1)];
428 const int b= score_map[(index+(1<<ME_MAP_SHIFT))&(ME_MAP_SIZE-1)];
438 CHECK_SAD_HALF_MV(y2 , 0, -1)
440 CHECK_SAD_HALF_MV(xy2, -1, -1)
442 CHECK_SAD_HALF_MV(xy2, +1, -1)
446 CHECK_SAD_HALF_MV(xy2, -1, +1)
448 CHECK_SAD_HALF_MV(x2 , -1, 0)
450 CHECK_SAD_HALF_MV(xy2, +1, -1)
452 CHECK_SAD_HALF_MV(xy2, -1, -1)
456 CHECK_SAD_HALF_MV(xy2, +1, +1)
458 CHECK_SAD_HALF_MV(x2 , +1, 0)
463 CHECK_SAD_HALF_MV(xy2, -1, -1)
467 CHECK_SAD_HALF_MV(xy2, +1, +1)
469 CHECK_SAD_HALF_MV(x2 , -1, 0)
470 CHECK_SAD_HALF_MV(xy2, -1, +1)
473 CHECK_SAD_HALF_MV(xy2, +1, -1)
477 CHECK_SAD_HALF_MV(xy2, -1, +1)
479 CHECK_SAD_HALF_MV(x2 , +1, 0)
480 CHECK_SAD_HALF_MV(xy2, +1, +1)
482 CHECK_SAD_HALF_MV(y2 , 0, +1)
497 static inline void set_p_mv_tables(MpegEncContext * s, int mx, int my, int mv4)
499 const int xy= s->mb_x + s->mb_y*s->mb_stride;
501 s->p_mv_table[xy][0] = mx;
502 s->p_mv_table[xy][1] = my;
504 /* has already been set to the 4 MV if 4MV is done */
506 int mot_xy= s->block_index[0];
508 s->current_picture.motion_val[0][mot_xy ][0] = mx;
509 s->current_picture.motion_val[0][mot_xy ][1] = my;
510 s->current_picture.motion_val[0][mot_xy + 1][0] = mx;
511 s->current_picture.motion_val[0][mot_xy + 1][1] = my;
513 mot_xy += s->b8_stride;
514 s->current_picture.motion_val[0][mot_xy ][0] = mx;
515 s->current_picture.motion_val[0][mot_xy ][1] = my;
516 s->current_picture.motion_val[0][mot_xy + 1][0] = mx;
517 s->current_picture.motion_val[0][mot_xy + 1][1] = my;
522 * get fullpel ME search limits.
524 static inline void get_limits(MpegEncContext *s, int x, int y)
526 MotionEstContext * const c= &s->me;
527 int range= c->avctx->me_range >> (1 + !!(c->flags&FLAG_QPEL));
528 int max_range = MAX_MV >> (1 + !!(c->flags&FLAG_QPEL));
530 if(c->avctx->me_range) c->range= c->avctx->me_range >> 1;
533 if (s->unrestricted_mv) {
536 c->xmax = - x + s->width;
537 c->ymax = - y + s->height;
538 } else if (s->out_format == FMT_H261){
539 // Search range of H.261 is different from other codec standards
540 c->xmin = (x > 15) ? - 15 : 0;
541 c->ymin = (y > 15) ? - 15 : 0;
542 c->xmax = (x < s->mb_width * 16 - 16) ? 15 : 0;
543 c->ymax = (y < s->mb_height * 16 - 16) ? 15 : 0;
547 c->xmax = - x + s->mb_width *16 - 16;
548 c->ymax = - y + s->mb_height*16 - 16;
550 if(!range || range > max_range)
553 c->xmin = FFMAX(c->xmin,-range);
554 c->xmax = FFMIN(c->xmax, range);
555 c->ymin = FFMAX(c->ymin,-range);
556 c->ymax = FFMIN(c->ymax, range);
560 static inline void init_mv4_ref(MotionEstContext *c){
561 const int stride= c->stride;
563 c->ref[1][0] = c->ref[0][0] + 8;
564 c->ref[2][0] = c->ref[0][0] + 8*stride;
565 c->ref[3][0] = c->ref[2][0] + 8;
566 c->src[1][0] = c->src[0][0] + 8;
567 c->src[2][0] = c->src[0][0] + 8*stride;
568 c->src[3][0] = c->src[2][0] + 8;
571 static inline int h263_mv4_search(MpegEncContext *s, int mx, int my, int shift)
573 MotionEstContext * const c= &s->me;
578 int dmin_sum=0, mx4_sum=0, my4_sum=0, i;
580 const int stride= c->stride;
581 uint8_t *mv_penalty= c->current_mv_penalty;
582 int safety_clipping= s->unrestricted_mv && (s->width&15) && (s->height&15);
586 for(block=0; block<4; block++){
588 int pred_x4, pred_y4;
590 static const int off[4]= {2, 1, 1, -1};
591 const int mot_stride = s->b8_stride;
592 const int mot_xy = s->block_index[block];
595 c->xmax = - 16*s->mb_x + s->width - 8*(block &1);
596 c->ymax = - 16*s->mb_y + s->height - 8*(block>>1);
599 P_LEFT[0] = s->current_picture.motion_val[0][mot_xy - 1][0];
600 P_LEFT[1] = s->current_picture.motion_val[0][mot_xy - 1][1];
602 if(P_LEFT[0] > (c->xmax<<shift)) P_LEFT[0] = (c->xmax<<shift);
604 /* special case for first line */
605 if (s->first_slice_line && block<2) {
606 c->pred_x= pred_x4= P_LEFT[0];
607 c->pred_y= pred_y4= P_LEFT[1];
609 P_TOP[0] = s->current_picture.motion_val[0][mot_xy - mot_stride ][0];
610 P_TOP[1] = s->current_picture.motion_val[0][mot_xy - mot_stride ][1];
611 P_TOPRIGHT[0] = s->current_picture.motion_val[0][mot_xy - mot_stride + off[block]][0];
612 P_TOPRIGHT[1] = s->current_picture.motion_val[0][mot_xy - mot_stride + off[block]][1];
613 if(P_TOP[1] > (c->ymax<<shift)) P_TOP[1] = (c->ymax<<shift);
614 if(P_TOPRIGHT[0] < (c->xmin<<shift)) P_TOPRIGHT[0]= (c->xmin<<shift);
615 if(P_TOPRIGHT[0] > (c->xmax<<shift)) P_TOPRIGHT[0]= (c->xmax<<shift);
616 if(P_TOPRIGHT[1] > (c->ymax<<shift)) P_TOPRIGHT[1]= (c->ymax<<shift);
618 P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
619 P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
621 c->pred_x= pred_x4 = P_MEDIAN[0];
622 c->pred_y= pred_y4 = P_MEDIAN[1];
628 if (s->first_slice_line && block<2 && i>1 && i<9)
632 if(P[i][0] > (c->xmax<<shift)) P[i][0]= (c->xmax<<shift);
633 if(P[i][1] > (c->ymax<<shift)) P[i][1]= (c->ymax<<shift);
636 dmin4 = epzs_motion_search4(s, &mx4, &my4, P, block, block, s->p_mv_table, (1<<16)>>shift);
638 dmin4= c->sub_motion_search(s, &mx4, &my4, dmin4, block, block, size, h);
640 if (s->mecc.me_sub_cmp[0] != s->mecc.mb_cmp[0]) {
642 const int offset= ((block&1) + (block>>1)*stride)*8;
643 uint8_t *dest_y = c->scratchpad + offset;
644 if(s->quarter_sample){
645 uint8_t *ref= c->ref[block][0] + (mx4>>2) + (my4>>2)*stride;
646 dxy = ((my4 & 3) << 2) | (mx4 & 3);
649 s->qdsp.put_no_rnd_qpel_pixels_tab[1][dxy](dest_y, ref, stride);
651 s->qdsp.put_qpel_pixels_tab[1][dxy](dest_y, ref, stride);
653 uint8_t *ref= c->ref[block][0] + (mx4>>1) + (my4>>1)*stride;
654 dxy = ((my4 & 1) << 1) | (mx4 & 1);
657 s->hdsp.put_no_rnd_pixels_tab[1][dxy](dest_y , ref , stride, h);
659 s->hdsp.put_pixels_tab [1][dxy](dest_y , ref , stride, h);
661 dmin_sum+= (mv_penalty[mx4-pred_x4] + mv_penalty[my4-pred_y4])*c->mb_penalty_factor;
665 if(s->quarter_sample){
673 s->current_picture.motion_val[0][s->block_index[block]][0] = mx4;
674 s->current_picture.motion_val[0][s->block_index[block]][1] = my4;
676 if(mx4 != mx || my4 != my) same=0;
682 if (s->mecc.me_sub_cmp[0] != s->mecc.mb_cmp[0]) {
683 dmin_sum += s->mecc.mb_cmp[0](s,
684 s->new_picture.f->data[0] +
685 s->mb_x * 16 + s->mb_y * 16 * stride,
686 c->scratchpad, stride, 16);
689 if(c->avctx->mb_cmp&FF_CMP_CHROMA){
694 mx= ff_h263_round_chroma(mx4_sum);
695 my= ff_h263_round_chroma(my4_sum);
696 dxy = ((my & 1) << 1) | (mx & 1);
698 offset= (s->mb_x*8 + (mx>>1)) + (s->mb_y*8 + (my>>1))*s->uvlinesize;
701 s->hdsp.put_no_rnd_pixels_tab[1][dxy](c->scratchpad , s->last_picture.f->data[1] + offset, s->uvlinesize, 8);
702 s->hdsp.put_no_rnd_pixels_tab[1][dxy](c->scratchpad + 8, s->last_picture.f->data[2] + offset, s->uvlinesize, 8);
704 s->hdsp.put_pixels_tab [1][dxy](c->scratchpad , s->last_picture.f->data[1] + offset, s->uvlinesize, 8);
705 s->hdsp.put_pixels_tab [1][dxy](c->scratchpad + 8, s->last_picture.f->data[2] + offset, s->uvlinesize, 8);
708 dmin_sum += s->mecc.mb_cmp[1](s, s->new_picture.f->data[1] + s->mb_x * 8 + s->mb_y * 8 * s->uvlinesize, c->scratchpad, s->uvlinesize, 8);
709 dmin_sum += s->mecc.mb_cmp[1](s, s->new_picture.f->data[2] + s->mb_x * 8 + s->mb_y * 8 * s->uvlinesize, c->scratchpad + 8, s->uvlinesize, 8);
715 switch(c->avctx->mb_cmp&0xFF){
717 return dmin_sum+ 32*s->qscale*s->qscale;*/
721 return dmin_sum+ 11*c->mb_penalty_factor;
725 static inline void init_interlaced_ref(MpegEncContext *s, int ref_index){
726 MotionEstContext * const c= &s->me;
728 c->ref[1+ref_index][0] = c->ref[0+ref_index][0] + s->linesize;
729 c->src[1][0] = c->src[0][0] + s->linesize;
730 if(c->flags & FLAG_CHROMA){
731 c->ref[1+ref_index][1] = c->ref[0+ref_index][1] + s->uvlinesize;
732 c->ref[1+ref_index][2] = c->ref[0+ref_index][2] + s->uvlinesize;
733 c->src[1][1] = c->src[0][1] + s->uvlinesize;
734 c->src[1][2] = c->src[0][2] + s->uvlinesize;
738 static int interlaced_search(MpegEncContext *s, int ref_index,
739 int16_t (*mv_tables[2][2])[2], uint8_t *field_select_tables[2], int mx, int my, int user_field_select)
741 MotionEstContext * const c= &s->me;
746 uint8_t * const mv_penalty= c->current_mv_penalty;
748 const int stride= 2*s->linesize;
750 const int mot_stride= s->mb_stride;
751 const int xy= s->mb_x + s->mb_y*mot_stride;
757 init_interlaced_ref(s, ref_index);
759 for(block=0; block<2; block++){
761 int best_dmin= INT_MAX;
764 for(field_select=0; field_select<2; field_select++){
765 int dmin, mx_i, my_i;
766 int16_t (*mv_table)[2]= mv_tables[block][field_select];
768 if(user_field_select){
769 av_assert1(field_select==0 || field_select==1);
770 av_assert1(field_select_tables[block][xy]==0 || field_select_tables[block][xy]==1);
771 if(field_select_tables[block][xy] != field_select)
775 P_LEFT[0] = mv_table[xy - 1][0];
776 P_LEFT[1] = mv_table[xy - 1][1];
777 if(P_LEFT[0] > (c->xmax<<1)) P_LEFT[0] = (c->xmax<<1);
779 c->pred_x= P_LEFT[0];
780 c->pred_y= P_LEFT[1];
782 if(!s->first_slice_line){
783 P_TOP[0] = mv_table[xy - mot_stride][0];
784 P_TOP[1] = mv_table[xy - mot_stride][1];
785 P_TOPRIGHT[0] = mv_table[xy - mot_stride + 1][0];
786 P_TOPRIGHT[1] = mv_table[xy - mot_stride + 1][1];
787 if(P_TOP[1] > (c->ymax<<1)) P_TOP[1] = (c->ymax<<1);
788 if(P_TOPRIGHT[0] < (c->xmin<<1)) P_TOPRIGHT[0]= (c->xmin<<1);
789 if(P_TOPRIGHT[0] > (c->xmax<<1)) P_TOPRIGHT[0]= (c->xmax<<1);
790 if(P_TOPRIGHT[1] > (c->ymax<<1)) P_TOPRIGHT[1]= (c->ymax<<1);
792 P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
793 P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
795 P_MV1[0]= mx; //FIXME not correct if block != field_select
798 dmin = epzs_motion_search2(s, &mx_i, &my_i, P, block, field_select+ref_index, mv_table, (1<<16)>>1);
800 dmin= c->sub_motion_search(s, &mx_i, &my_i, dmin, block, field_select+ref_index, size, h);
802 mv_table[xy][0]= mx_i;
803 mv_table[xy][1]= my_i;
805 if (s->mecc.me_sub_cmp[0] != s->mecc.mb_cmp[0]) {
809 uint8_t *ref= c->ref[field_select+ref_index][0] + (mx_i>>1) + (my_i>>1)*stride;
810 dxy = ((my_i & 1) << 1) | (mx_i & 1);
813 s->hdsp.put_no_rnd_pixels_tab[size][dxy](c->scratchpad, ref , stride, h);
815 s->hdsp.put_pixels_tab [size][dxy](c->scratchpad, ref , stride, h);
817 dmin = s->mecc.mb_cmp[size](s, c->src[block][0], c->scratchpad, stride, h);
818 dmin+= (mv_penalty[mx_i-c->pred_x] + mv_penalty[my_i-c->pred_y] + 1)*c->mb_penalty_factor;
820 dmin+= c->mb_penalty_factor; //field_select bits
822 dmin += field_select != block; //slightly prefer same field
824 if(dmin < best_dmin){
826 best_field= field_select;
830 int16_t (*mv_table)[2]= mv_tables[block][best_field];
832 if(mv_table[xy][0] != mx) same=0; //FIXME check if these checks work and are any good at all
833 if(mv_table[xy][1]&1) same=0;
834 if(mv_table[xy][1]*2 != my) same=0;
835 if(best_field != block) same=0;
838 field_select_tables[block][xy]= best_field;
839 dmin_sum += best_dmin;
850 switch(c->avctx->mb_cmp&0xFF){
852 return dmin_sum+ 32*s->qscale*s->qscale;*/
856 return dmin_sum+ 11*c->mb_penalty_factor;
860 static inline int get_penalty_factor(int lambda, int lambda2, int type){
864 return lambda>>FF_LAMBDA_SHIFT;
866 return (3*lambda)>>(FF_LAMBDA_SHIFT+1);
868 return (4*lambda)>>(FF_LAMBDA_SHIFT);
870 return (2*lambda)>>(FF_LAMBDA_SHIFT);
873 return (2*lambda)>>FF_LAMBDA_SHIFT;
878 return lambda2>>FF_LAMBDA_SHIFT;
880 case FF_CMP_MEDIAN_SAD:
885 void ff_estimate_p_frame_motion(MpegEncContext * s,
888 MotionEstContext * const c= &s->me;
890 int sum, mx = 0, my = 0, dmin = 0;
891 int varc; ///< the variance of the block (sum of squared (p[y][x]-average))
892 int vard; ///< sum of squared differences with the estimated motion vector
894 const int shift= 1+s->quarter_sample;
896 Picture * const pic= &s->current_picture;
898 init_ref(c, s->new_picture.f->data, s->last_picture.f->data, NULL, 16*mb_x, 16*mb_y, 0);
900 av_assert0(s->quarter_sample==0 || s->quarter_sample==1);
901 av_assert0(s->linesize == c->stride);
902 av_assert0(s->uvlinesize == c->uvstride);
904 c->penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_cmp);
905 c->sub_penalty_factor= get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_sub_cmp);
906 c->mb_penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->mb_cmp);
907 c->current_mv_penalty= c->mv_penalty[s->f_code] + MAX_DMV;
909 get_limits(s, 16*mb_x, 16*mb_y);
912 /* intra / predictive decision */
914 sum = s->mpvencdsp.pix_sum(pix, s->linesize);
915 varc = s->mpvencdsp.pix_norm1(pix, s->linesize) -
916 (((unsigned) sum * sum) >> 8) + 500;
918 pic->mb_mean[s->mb_stride * mb_y + mb_x] = (sum+128)>>8;
919 pic->mb_var [s->mb_stride * mb_y + mb_x] = (varc+128)>>8;
920 c->mb_var_sum_temp += (varc+128)>>8;
922 if (s->motion_est != FF_ME_ZERO) {
923 const int mot_stride = s->b8_stride;
924 const int mot_xy = s->block_index[0];
926 P_LEFT[0] = s->current_picture.motion_val[0][mot_xy - 1][0];
927 P_LEFT[1] = s->current_picture.motion_val[0][mot_xy - 1][1];
929 if (P_LEFT[0] > (c->xmax << shift))
930 P_LEFT[0] = c->xmax << shift;
932 if (!s->first_slice_line) {
933 P_TOP[0] = s->current_picture.motion_val[0][mot_xy - mot_stride ][0];
934 P_TOP[1] = s->current_picture.motion_val[0][mot_xy - mot_stride ][1];
935 P_TOPRIGHT[0] = s->current_picture.motion_val[0][mot_xy - mot_stride + 2][0];
936 P_TOPRIGHT[1] = s->current_picture.motion_val[0][mot_xy - mot_stride + 2][1];
937 if (P_TOP[1] > (c->ymax << shift))
938 P_TOP[1] = c->ymax << shift;
939 if (P_TOPRIGHT[0] < (c->xmin * (1 << shift)))
940 P_TOPRIGHT[0] = c->xmin * (1 << shift);
941 if (P_TOPRIGHT[1] > (c->ymax * (1 << shift)))
942 P_TOPRIGHT[1] = c->ymax * (1 << shift);
944 P_MEDIAN[0] = mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
945 P_MEDIAN[1] = mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
947 if (s->out_format == FMT_H263) {
948 c->pred_x = P_MEDIAN[0];
949 c->pred_y = P_MEDIAN[1];
950 } else { /* MPEG-1 at least */
951 c->pred_x = P_LEFT[0];
952 c->pred_y = P_LEFT[1];
955 c->pred_x = P_LEFT[0];
956 c->pred_y = P_LEFT[1];
958 dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, 0, s->p_mv_table, (1<<16)>>shift, 0, 16);
961 /* At this point (mx,my) are full-pell and the relative displacement */
962 ppix = c->ref[0][0] + (my * s->linesize) + mx;
964 vard = s->mecc.sse[0](NULL, pix, ppix, s->linesize, 16);
966 pic->mc_mb_var[s->mb_stride * mb_y + mb_x] = (vard+128)>>8;
967 c->mc_mb_var_sum_temp += (vard+128)>>8;
969 if (c->avctx->mb_decision > FF_MB_DECISION_SIMPLE) {
970 int p_score= FFMIN(vard, varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*100);
971 int i_score= varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*20;
972 c->scene_change_score+= ff_sqrt(p_score) - ff_sqrt(i_score);
974 if (vard*2 + 200*256 > varc)
975 mb_type|= CANDIDATE_MB_TYPE_INTRA;
976 if (varc*2 + 200*256 > vard || s->qscale > 24){
977 // if (varc*2 + 200*256 + 50*(s->lambda2>>FF_LAMBDA_SHIFT) > vard){
978 mb_type|= CANDIDATE_MB_TYPE_INTER;
979 c->sub_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
980 if (s->mpv_flags & FF_MPV_FLAG_MV0)
982 mb_type |= CANDIDATE_MB_TYPE_SKIPPED; //FIXME check difference
987 if ((s->avctx->flags & AV_CODEC_FLAG_4MV)
988 && !c->skip && varc>50<<8 && vard>10<<8){
989 if(h263_mv4_search(s, mx, my, shift) < INT_MAX)
990 mb_type|=CANDIDATE_MB_TYPE_INTER4V;
992 set_p_mv_tables(s, mx, my, 0);
994 set_p_mv_tables(s, mx, my, 1);
995 if ((s->avctx->flags & AV_CODEC_FLAG_INTERLACED_ME)
996 && !c->skip){ //FIXME varc/d checks
997 if(interlaced_search(s, 0, s->p_field_mv_table, s->p_field_select_table, mx, my, 0) < INT_MAX)
998 mb_type |= CANDIDATE_MB_TYPE_INTER_I;
1002 mb_type= CANDIDATE_MB_TYPE_INTER;
1004 dmin= c->sub_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1005 if(c->avctx->me_sub_cmp != c->avctx->mb_cmp && !c->skip)
1006 dmin= get_mb_score(s, mx, my, 0, 0, 0, 16, 1);
1008 if ((s->avctx->flags & AV_CODEC_FLAG_4MV)
1009 && !c->skip && varc>50<<8 && vard>10<<8){
1010 int dmin4= h263_mv4_search(s, mx, my, shift);
1012 mb_type= CANDIDATE_MB_TYPE_INTER4V;
1016 if ((s->avctx->flags & AV_CODEC_FLAG_INTERLACED_ME)
1017 && !c->skip){ //FIXME varc/d checks
1018 int dmin_i= interlaced_search(s, 0, s->p_field_mv_table, s->p_field_select_table, mx, my, 0);
1020 mb_type = CANDIDATE_MB_TYPE_INTER_I;
1025 set_p_mv_tables(s, mx, my, mb_type!=CANDIDATE_MB_TYPE_INTER4V);
1027 /* get intra luma score */
1028 if((c->avctx->mb_cmp&0xFF)==FF_CMP_SSE){
1029 intra_score= varc - 500;
1031 unsigned mean = (sum+128)>>8;
1034 for(i=0; i<16; i++){
1035 *(uint32_t*)(&c->scratchpad[i*s->linesize+ 0]) = mean;
1036 *(uint32_t*)(&c->scratchpad[i*s->linesize+ 4]) = mean;
1037 *(uint32_t*)(&c->scratchpad[i*s->linesize+ 8]) = mean;
1038 *(uint32_t*)(&c->scratchpad[i*s->linesize+12]) = mean;
1041 intra_score= s->mecc.mb_cmp[0](s, c->scratchpad, pix, s->linesize, 16);
1043 intra_score += c->mb_penalty_factor*16;
1045 if(intra_score < dmin){
1046 mb_type= CANDIDATE_MB_TYPE_INTRA;
1047 s->current_picture.mb_type[mb_y*s->mb_stride + mb_x] = CANDIDATE_MB_TYPE_INTRA; //FIXME cleanup
1049 s->current_picture.mb_type[mb_y*s->mb_stride + mb_x] = 0;
1052 int p_score= FFMIN(vard, varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*100);
1053 int i_score= varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*20;
1054 c->scene_change_score+= ff_sqrt(p_score) - ff_sqrt(i_score);
1058 s->mb_type[mb_y*s->mb_stride + mb_x]= mb_type;
1061 int ff_pre_estimate_p_frame_motion(MpegEncContext * s,
1064 MotionEstContext * const c= &s->me;
1067 const int shift= 1+s->quarter_sample;
1068 const int xy= mb_x + mb_y*s->mb_stride;
1069 init_ref(c, s->new_picture.f->data, s->last_picture.f->data, NULL, 16*mb_x, 16*mb_y, 0);
1071 av_assert0(s->quarter_sample==0 || s->quarter_sample==1);
1073 c->pre_penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_pre_cmp);
1074 c->current_mv_penalty= c->mv_penalty[s->f_code] + MAX_DMV;
1076 get_limits(s, 16*mb_x, 16*mb_y);
1079 P_LEFT[0] = s->p_mv_table[xy + 1][0];
1080 P_LEFT[1] = s->p_mv_table[xy + 1][1];
1082 if(P_LEFT[0] < (c->xmin<<shift)) P_LEFT[0] = (c->xmin<<shift);
1084 /* special case for first line */
1085 if (s->first_slice_line) {
1086 c->pred_x= P_LEFT[0];
1087 c->pred_y= P_LEFT[1];
1088 P_TOP[0]= P_TOPRIGHT[0]= P_MEDIAN[0]=
1089 P_TOP[1]= P_TOPRIGHT[1]= P_MEDIAN[1]= 0; //FIXME
1091 P_TOP[0] = s->p_mv_table[xy + s->mb_stride ][0];
1092 P_TOP[1] = s->p_mv_table[xy + s->mb_stride ][1];
1093 P_TOPRIGHT[0] = s->p_mv_table[xy + s->mb_stride - 1][0];
1094 P_TOPRIGHT[1] = s->p_mv_table[xy + s->mb_stride - 1][1];
1095 if(P_TOP[1] < (c->ymin<<shift)) P_TOP[1] = (c->ymin<<shift);
1096 if(P_TOPRIGHT[0] > (c->xmax<<shift)) P_TOPRIGHT[0]= (c->xmax<<shift);
1097 if(P_TOPRIGHT[1] < (c->ymin<<shift)) P_TOPRIGHT[1]= (c->ymin<<shift);
1099 P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1100 P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1102 c->pred_x = P_MEDIAN[0];
1103 c->pred_y = P_MEDIAN[1];
1106 dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, 0, s->p_mv_table, (1<<16)>>shift, 0, 16);
1108 s->p_mv_table[xy][0] = mx<<shift;
1109 s->p_mv_table[xy][1] = my<<shift;
1114 static int estimate_motion_b(MpegEncContext *s, int mb_x, int mb_y,
1115 int16_t (*mv_table)[2], int ref_index, int f_code)
1117 MotionEstContext * const c= &s->me;
1118 int mx = 0, my = 0, dmin = 0;
1120 const int shift= 1+s->quarter_sample;
1121 const int mot_stride = s->mb_stride;
1122 const int mot_xy = mb_y*mot_stride + mb_x;
1123 uint8_t * const mv_penalty= c->mv_penalty[f_code] + MAX_DMV;
1126 c->penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_cmp);
1127 c->sub_penalty_factor= get_penalty_factor(s->lambda, s->lambda2, c->avctx->me_sub_cmp);
1128 c->mb_penalty_factor = get_penalty_factor(s->lambda, s->lambda2, c->avctx->mb_cmp);
1129 c->current_mv_penalty= mv_penalty;
1131 get_limits(s, 16*mb_x, 16*mb_y);
1133 if (s->motion_est != FF_ME_ZERO) {
1134 P_LEFT[0] = mv_table[mot_xy - 1][0];
1135 P_LEFT[1] = mv_table[mot_xy - 1][1];
1137 if (P_LEFT[0] > (c->xmax << shift)) P_LEFT[0] = (c->xmax << shift);
1139 /* special case for first line */
1140 if (!s->first_slice_line) {
1141 P_TOP[0] = mv_table[mot_xy - mot_stride ][0];
1142 P_TOP[1] = mv_table[mot_xy - mot_stride ][1];
1143 P_TOPRIGHT[0] = mv_table[mot_xy - mot_stride + 1][0];
1144 P_TOPRIGHT[1] = mv_table[mot_xy - mot_stride + 1][1];
1145 if (P_TOP[1] > (c->ymax << shift)) P_TOP[1] = (c->ymax << shift);
1146 if (P_TOPRIGHT[0] < (c->xmin << shift)) P_TOPRIGHT[0] = (c->xmin << shift);
1147 if (P_TOPRIGHT[1] > (c->ymax << shift)) P_TOPRIGHT[1] = (c->ymax << shift);
1149 P_MEDIAN[0] = mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1150 P_MEDIAN[1] = mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1152 c->pred_x = P_LEFT[0];
1153 c->pred_y = P_LEFT[1];
1155 if(mv_table == s->b_forw_mv_table){
1156 mv_scale= (s->pb_time<<16) / (s->pp_time<<shift);
1158 mv_scale= ((s->pb_time - s->pp_time)<<16) / (s->pp_time<<shift);
1161 dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, ref_index, s->p_mv_table, mv_scale, 0, 16);
1164 dmin= c->sub_motion_search(s, &mx, &my, dmin, 0, ref_index, 0, 16);
1166 if(c->avctx->me_sub_cmp != c->avctx->mb_cmp && !c->skip)
1167 dmin= get_mb_score(s, mx, my, 0, ref_index, 0, 16, 1);
1169 // s->mb_type[mb_y*s->mb_width + mb_x]= mb_type;
1170 mv_table[mot_xy][0]= mx;
1171 mv_table[mot_xy][1]= my;
1176 static inline int check_bidir_mv(MpegEncContext * s,
1177 int motion_fx, int motion_fy,
1178 int motion_bx, int motion_by,
1179 int pred_fx, int pred_fy,
1180 int pred_bx, int pred_by,
1184 //FIXME better f_code prediction (max mv & distance)
1186 MotionEstContext * const c= &s->me;
1187 uint8_t * const mv_penalty_f= c->mv_penalty[s->f_code] + MAX_DMV; // f_code of the prev frame
1188 uint8_t * const mv_penalty_b= c->mv_penalty[s->b_code] + MAX_DMV; // f_code of the prev frame
1189 int stride= c->stride;
1190 uint8_t *dest_y = c->scratchpad;
1195 uint8_t **src_data= c->src[0];
1196 uint8_t **ref_data= c->ref[0];
1197 uint8_t **ref2_data= c->ref[2];
1199 if(s->quarter_sample){
1200 dxy = ((motion_fy & 3) << 2) | (motion_fx & 3);
1201 src_x = motion_fx >> 2;
1202 src_y = motion_fy >> 2;
1204 ptr = ref_data[0] + (src_y * stride) + src_x;
1205 s->qdsp.put_qpel_pixels_tab[0][dxy](dest_y, ptr, stride);
1207 dxy = ((motion_by & 3) << 2) | (motion_bx & 3);
1208 src_x = motion_bx >> 2;
1209 src_y = motion_by >> 2;
1211 ptr = ref2_data[0] + (src_y * stride) + src_x;
1212 s->qdsp.avg_qpel_pixels_tab[size][dxy](dest_y, ptr, stride);
1214 dxy = ((motion_fy & 1) << 1) | (motion_fx & 1);
1215 src_x = motion_fx >> 1;
1216 src_y = motion_fy >> 1;
1218 ptr = ref_data[0] + (src_y * stride) + src_x;
1219 s->hdsp.put_pixels_tab[size][dxy](dest_y , ptr , stride, h);
1221 dxy = ((motion_by & 1) << 1) | (motion_bx & 1);
1222 src_x = motion_bx >> 1;
1223 src_y = motion_by >> 1;
1225 ptr = ref2_data[0] + (src_y * stride) + src_x;
1226 s->hdsp.avg_pixels_tab[size][dxy](dest_y , ptr , stride, h);
1229 fbmin = (mv_penalty_f[motion_fx-pred_fx] + mv_penalty_f[motion_fy-pred_fy])*c->mb_penalty_factor
1230 +(mv_penalty_b[motion_bx-pred_bx] + mv_penalty_b[motion_by-pred_by])*c->mb_penalty_factor
1231 + s->mecc.mb_cmp[size](s, src_data[0], dest_y, stride, h); // FIXME new_pic
1233 if(c->avctx->mb_cmp&FF_CMP_CHROMA){
1240 /* refine the bidir vectors in hq mode and return the score in both lq & hq mode*/
1241 static inline int bidir_refine(MpegEncContext * s, int mb_x, int mb_y)
1243 MotionEstContext * const c= &s->me;
1244 const int mot_stride = s->mb_stride;
1245 const int xy = mb_y *mot_stride + mb_x;
1247 int pred_fx= s->b_bidir_forw_mv_table[xy-1][0];
1248 int pred_fy= s->b_bidir_forw_mv_table[xy-1][1];
1249 int pred_bx= s->b_bidir_back_mv_table[xy-1][0];
1250 int pred_by= s->b_bidir_back_mv_table[xy-1][1];
1251 int motion_fx= s->b_bidir_forw_mv_table[xy][0]= s->b_forw_mv_table[xy][0];
1252 int motion_fy= s->b_bidir_forw_mv_table[xy][1]= s->b_forw_mv_table[xy][1];
1253 int motion_bx= s->b_bidir_back_mv_table[xy][0]= s->b_back_mv_table[xy][0];
1254 int motion_by= s->b_bidir_back_mv_table[xy][1]= s->b_back_mv_table[xy][1];
1255 const int flags= c->sub_flags;
1256 const int qpel= flags&FLAG_QPEL;
1257 const int shift= 1+qpel;
1258 const int xmin= c->xmin<<shift;
1259 const int ymin= c->ymin<<shift;
1260 const int xmax= c->xmax<<shift;
1261 const int ymax= c->ymax<<shift;
1262 #define HASH(fx,fy,bx,by) ((fx)+17*(fy)+63*(bx)+117*(by))
1263 #define HASH8(fx,fy,bx,by) ((uint8_t)HASH(fx,fy,bx,by))
1264 int hashidx= HASH(motion_fx,motion_fy, motion_bx, motion_by);
1265 uint8_t map[256] = { 0 };
1267 map[hashidx&255] = 1;
1269 fbmin= check_bidir_mv(s, motion_fx, motion_fy,
1270 motion_bx, motion_by,
1275 if(s->avctx->bidir_refine){
1277 static const uint8_t limittab[5]={0,8,32,64,80};
1278 const int limit= limittab[s->avctx->bidir_refine];
1279 static const int8_t vect[][4]={
1280 { 0, 0, 0, 1}, { 0, 0, 0,-1}, { 0, 0, 1, 0}, { 0, 0,-1, 0}, { 0, 1, 0, 0}, { 0,-1, 0, 0}, { 1, 0, 0, 0}, {-1, 0, 0, 0},
1282 { 0, 0, 1, 1}, { 0, 0,-1,-1}, { 0, 1, 1, 0}, { 0,-1,-1, 0}, { 1, 1, 0, 0}, {-1,-1, 0, 0}, { 1, 0, 0, 1}, {-1, 0, 0,-1},
1283 { 0, 1, 0, 1}, { 0,-1, 0,-1}, { 1, 0, 1, 0}, {-1, 0,-1, 0},
1284 { 0, 0,-1, 1}, { 0, 0, 1,-1}, { 0,-1, 1, 0}, { 0, 1,-1, 0}, {-1, 1, 0, 0}, { 1,-1, 0, 0}, { 1, 0, 0,-1}, {-1, 0, 0, 1},
1285 { 0,-1, 0, 1}, { 0, 1, 0,-1}, {-1, 0, 1, 0}, { 1, 0,-1, 0},
1287 { 0, 1, 1, 1}, { 0,-1,-1,-1}, { 1, 1, 1, 0}, {-1,-1,-1, 0}, { 1, 1, 0, 1}, {-1,-1, 0,-1}, { 1, 0, 1, 1}, {-1, 0,-1,-1},
1288 { 0,-1, 1, 1}, { 0, 1,-1,-1}, {-1, 1, 1, 0}, { 1,-1,-1, 0}, { 1, 1, 0,-1}, {-1,-1, 0, 1}, { 1, 0,-1, 1}, {-1, 0, 1,-1},
1289 { 0, 1,-1, 1}, { 0,-1, 1,-1}, { 1,-1, 1, 0}, {-1, 1,-1, 0}, {-1, 1, 0, 1}, { 1,-1, 0,-1}, { 1, 0, 1,-1}, {-1, 0,-1, 1},
1290 { 0, 1, 1,-1}, { 0,-1,-1, 1}, { 1, 1,-1, 0}, {-1,-1, 1, 0}, { 1,-1, 0, 1}, {-1, 1, 0,-1}, {-1, 0, 1, 1}, { 1, 0,-1,-1},
1292 { 1, 1, 1, 1}, {-1,-1,-1,-1},
1293 { 1, 1, 1,-1}, {-1,-1,-1, 1}, { 1, 1,-1, 1}, {-1,-1, 1,-1}, { 1,-1, 1, 1}, {-1, 1,-1,-1}, {-1, 1, 1, 1}, { 1,-1,-1,-1},
1294 { 1, 1,-1,-1}, {-1,-1, 1, 1}, { 1,-1,-1, 1}, {-1, 1, 1,-1}, { 1,-1, 1,-1}, {-1, 1,-1, 1},
1296 static const uint8_t hash[]={
1297 HASH8( 0, 0, 0, 1), HASH8( 0, 0, 0,-1), HASH8( 0, 0, 1, 0), HASH8( 0, 0,-1, 0), HASH8( 0, 1, 0, 0), HASH8( 0,-1, 0, 0), HASH8( 1, 0, 0, 0), HASH8(-1, 0, 0, 0),
1299 HASH8( 0, 0, 1, 1), HASH8( 0, 0,-1,-1), HASH8( 0, 1, 1, 0), HASH8( 0,-1,-1, 0), HASH8( 1, 1, 0, 0), HASH8(-1,-1, 0, 0), HASH8( 1, 0, 0, 1), HASH8(-1, 0, 0,-1),
1300 HASH8( 0, 1, 0, 1), HASH8( 0,-1, 0,-1), HASH8( 1, 0, 1, 0), HASH8(-1, 0,-1, 0),
1301 HASH8( 0, 0,-1, 1), HASH8( 0, 0, 1,-1), HASH8( 0,-1, 1, 0), HASH8( 0, 1,-1, 0), HASH8(-1, 1, 0, 0), HASH8( 1,-1, 0, 0), HASH8( 1, 0, 0,-1), HASH8(-1, 0, 0, 1),
1302 HASH8( 0,-1, 0, 1), HASH8( 0, 1, 0,-1), HASH8(-1, 0, 1, 0), HASH8( 1, 0,-1, 0),
1304 HASH8( 0, 1, 1, 1), HASH8( 0,-1,-1,-1), HASH8( 1, 1, 1, 0), HASH8(-1,-1,-1, 0), HASH8( 1, 1, 0, 1), HASH8(-1,-1, 0,-1), HASH8( 1, 0, 1, 1), HASH8(-1, 0,-1,-1),
1305 HASH8( 0,-1, 1, 1), HASH8( 0, 1,-1,-1), HASH8(-1, 1, 1, 0), HASH8( 1,-1,-1, 0), HASH8( 1, 1, 0,-1), HASH8(-1,-1, 0, 1), HASH8( 1, 0,-1, 1), HASH8(-1, 0, 1,-1),
1306 HASH8( 0, 1,-1, 1), HASH8( 0,-1, 1,-1), HASH8( 1,-1, 1, 0), HASH8(-1, 1,-1, 0), HASH8(-1, 1, 0, 1), HASH8( 1,-1, 0,-1), HASH8( 1, 0, 1,-1), HASH8(-1, 0,-1, 1),
1307 HASH8( 0, 1, 1,-1), HASH8( 0,-1,-1, 1), HASH8( 1, 1,-1, 0), HASH8(-1,-1, 1, 0), HASH8( 1,-1, 0, 1), HASH8(-1, 1, 0,-1), HASH8(-1, 0, 1, 1), HASH8( 1, 0,-1,-1),
1309 HASH8( 1, 1, 1, 1), HASH8(-1,-1,-1,-1),
1310 HASH8( 1, 1, 1,-1), HASH8(-1,-1,-1, 1), HASH8( 1, 1,-1, 1), HASH8(-1,-1, 1,-1), HASH8( 1,-1, 1, 1), HASH8(-1, 1,-1,-1), HASH8(-1, 1, 1, 1), HASH8( 1,-1,-1,-1),
1311 HASH8( 1, 1,-1,-1), HASH8(-1,-1, 1, 1), HASH8( 1,-1,-1, 1), HASH8(-1, 1, 1,-1), HASH8( 1,-1, 1,-1), HASH8(-1, 1,-1, 1),
1314 #define CHECK_BIDIR(fx,fy,bx,by)\
1315 if( !map[(hashidx+HASH(fx,fy,bx,by))&255]\
1316 &&(fx<=0 || motion_fx+fx<=xmax) && (fy<=0 || motion_fy+fy<=ymax) && (bx<=0 || motion_bx+bx<=xmax) && (by<=0 || motion_by+by<=ymax)\
1317 &&(fx>=0 || motion_fx+fx>=xmin) && (fy>=0 || motion_fy+fy>=ymin) && (bx>=0 || motion_bx+bx>=xmin) && (by>=0 || motion_by+by>=ymin)){\
1319 map[(hashidx+HASH(fx,fy,bx,by))&255] = 1;\
1320 score= check_bidir_mv(s, motion_fx+fx, motion_fy+fy, motion_bx+bx, motion_by+by, pred_fx, pred_fy, pred_bx, pred_by, 0, 16);\
1322 hashidx += HASH(fx,fy,bx,by);\
1331 #define CHECK_BIDIR2(a,b,c,d)\
1332 CHECK_BIDIR(a,b,c,d)\
1333 CHECK_BIDIR(-(a),-(b),-(c),-(d))
1340 CHECK_BIDIR2(0,0,0,1)
1341 CHECK_BIDIR2(0,0,1,0)
1342 CHECK_BIDIR2(0,1,0,0)
1343 CHECK_BIDIR2(1,0,0,0)
1345 for(i=8; i<limit; i++){
1346 int fx= motion_fx+vect[i][0];
1347 int fy= motion_fy+vect[i][1];
1348 int bx= motion_bx+vect[i][2];
1349 int by= motion_by+vect[i][3];
1351 int a= (xmax - FFMAX(fx,bx))|(FFMIN(fx,bx) - xmin);
1352 int b= (ymax - FFMAX(fy,by))|(FFMIN(fy,by) - ymin);
1354 map[(hashidx+hash[i])&255] = 1;
1356 if(!map[(hashidx+hash[i])&255]){
1358 map[(hashidx+hash[i])&255] = 1;
1359 score= check_bidir_mv(s, fx, fy, bx, by, pred_fx, pred_fy, pred_bx, pred_by, 0, 16);
1370 int a= FFMIN(xmax - FFMAX(fx,bx), FFMIN(fx,bx) - xmin);
1371 int b= FFMIN(ymax - FFMAX(fy,by), FFMIN(fy,by) - ymin);
1372 borderdist= FFMIN(a,b);
1380 s->b_bidir_forw_mv_table[xy][0]= motion_fx;
1381 s->b_bidir_forw_mv_table[xy][1]= motion_fy;
1382 s->b_bidir_back_mv_table[xy][0]= motion_bx;
1383 s->b_bidir_back_mv_table[xy][1]= motion_by;
1388 static inline int direct_search(MpegEncContext * s, int mb_x, int mb_y)
1390 MotionEstContext * const c= &s->me;
1392 const int mot_stride = s->mb_stride;
1393 const int mot_xy = mb_y*mot_stride + mb_x;
1394 const int shift= 1+s->quarter_sample;
1396 const int time_pp= s->pp_time;
1397 const int time_pb= s->pb_time;
1398 int mx, my, xmin, xmax, ymin, ymax;
1399 int16_t (*mv_table)[2]= s->b_direct_mv_table;
1401 c->current_mv_penalty= c->mv_penalty[1] + MAX_DMV;
1402 ymin= xmin=(-32)>>shift;
1403 ymax= xmax= 31>>shift;
1405 if (IS_8X8(s->next_picture.mb_type[mot_xy])) {
1406 s->mv_type= MV_TYPE_8X8;
1408 s->mv_type= MV_TYPE_16X16;
1412 int index= s->block_index[i];
1415 c->co_located_mv[i][0] = s->next_picture.motion_val[0][index][0];
1416 c->co_located_mv[i][1] = s->next_picture.motion_val[0][index][1];
1417 c->direct_basis_mv[i][0]= c->co_located_mv[i][0]*time_pb/time_pp + ((i& 1)<<(shift+3));
1418 c->direct_basis_mv[i][1]= c->co_located_mv[i][1]*time_pb/time_pp + ((i>>1)<<(shift+3));
1419 // c->direct_basis_mv[1][i][0]= c->co_located_mv[i][0]*(time_pb - time_pp)/time_pp + ((i &1)<<(shift+3);
1420 // c->direct_basis_mv[1][i][1]= c->co_located_mv[i][1]*(time_pb - time_pp)/time_pp + ((i>>1)<<(shift+3);
1422 max= FFMAX(c->direct_basis_mv[i][0], c->direct_basis_mv[i][0] - c->co_located_mv[i][0])>>shift;
1423 min= FFMIN(c->direct_basis_mv[i][0], c->direct_basis_mv[i][0] - c->co_located_mv[i][0])>>shift;
1424 max+= 16*mb_x + 1; // +-1 is for the simpler rounding
1426 xmax= FFMIN(xmax, s->width - max);
1427 xmin= FFMAX(xmin, - 16 - min);
1429 max= FFMAX(c->direct_basis_mv[i][1], c->direct_basis_mv[i][1] - c->co_located_mv[i][1])>>shift;
1430 min= FFMIN(c->direct_basis_mv[i][1], c->direct_basis_mv[i][1] - c->co_located_mv[i][1])>>shift;
1431 max+= 16*mb_y + 1; // +-1 is for the simpler rounding
1433 ymax= FFMIN(ymax, s->height - max);
1434 ymin= FFMAX(ymin, - 16 - min);
1436 if(s->mv_type == MV_TYPE_16X16) break;
1439 av_assert2(xmax <= 15 && ymax <= 15 && xmin >= -16 && ymin >= -16);
1441 if(xmax < 0 || xmin >0 || ymax < 0 || ymin > 0){
1442 s->b_direct_mv_table[mot_xy][0]= 0;
1443 s->b_direct_mv_table[mot_xy][1]= 0;
1445 return 256*256*256*64;
1452 c->flags |= FLAG_DIRECT;
1453 c->sub_flags |= FLAG_DIRECT;
1457 P_LEFT[0] = av_clip(mv_table[mot_xy - 1][0], xmin<<shift, xmax<<shift);
1458 P_LEFT[1] = av_clip(mv_table[mot_xy - 1][1], ymin<<shift, ymax<<shift);
1460 /* special case for first line */
1461 if (!s->first_slice_line) { //FIXME maybe allow this over thread boundary as it is clipped
1462 P_TOP[0] = av_clip(mv_table[mot_xy - mot_stride ][0], xmin<<shift, xmax<<shift);
1463 P_TOP[1] = av_clip(mv_table[mot_xy - mot_stride ][1], ymin<<shift, ymax<<shift);
1464 P_TOPRIGHT[0] = av_clip(mv_table[mot_xy - mot_stride + 1 ][0], xmin<<shift, xmax<<shift);
1465 P_TOPRIGHT[1] = av_clip(mv_table[mot_xy - mot_stride + 1 ][1], ymin<<shift, ymax<<shift);
1467 P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1468 P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1471 dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, 0, mv_table, 1<<(16-shift), 0, 16);
1472 if(c->sub_flags&FLAG_QPEL)
1473 dmin = qpel_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1475 dmin = hpel_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1477 if(c->avctx->me_sub_cmp != c->avctx->mb_cmp && !c->skip)
1478 dmin= get_mb_score(s, mx, my, 0, 0, 0, 16, 1);
1480 get_limits(s, 16*mb_x, 16*mb_y); //restore c->?min/max, maybe not needed
1482 mv_table[mot_xy][0]= mx;
1483 mv_table[mot_xy][1]= my;
1484 c->flags &= ~FLAG_DIRECT;
1485 c->sub_flags &= ~FLAG_DIRECT;
1490 void ff_estimate_b_frame_motion(MpegEncContext * s,
1493 MotionEstContext * const c= &s->me;
1494 const int penalty_factor= c->mb_penalty_factor;
1495 int fmin, bmin, dmin, fbmin, bimin, fimin;
1497 const int xy = mb_y*s->mb_stride + mb_x;
1498 init_ref(c, s->new_picture.f->data, s->last_picture.f->data,
1499 s->next_picture.f->data, 16 * mb_x, 16 * mb_y, 2);
1501 get_limits(s, 16*mb_x, 16*mb_y);
1505 if (s->codec_id == AV_CODEC_ID_MPEG4 && s->next_picture.mbskip_table[xy]) {
1506 int score= direct_search(s, mb_x, mb_y); //FIXME just check 0,0
1508 score= ((unsigned)(score*score + 128*256))>>16;
1509 c->mc_mb_var_sum_temp += score;
1510 s->current_picture.mc_mb_var[mb_y*s->mb_stride + mb_x] = score; //FIXME use SSE
1511 s->mb_type[mb_y*s->mb_stride + mb_x]= CANDIDATE_MB_TYPE_DIRECT0;
1516 if (s->codec_id == AV_CODEC_ID_MPEG4)
1517 dmin= direct_search(s, mb_x, mb_y);
1520 // FIXME penalty stuff for non-MPEG-4
1522 fmin = estimate_motion_b(s, mb_x, mb_y, s->b_forw_mv_table, 0, s->f_code) +
1526 bmin = estimate_motion_b(s, mb_x, mb_y, s->b_back_mv_table, 2, s->b_code) +
1528 ff_dlog(s, " %d %d ", s->b_forw_mv_table[xy][0], s->b_forw_mv_table[xy][1]);
1531 fbmin= bidir_refine(s, mb_x, mb_y) + penalty_factor;
1532 ff_dlog(s, "%d %d %d %d\n", dmin, fmin, bmin, fbmin);
1534 if (s->avctx->flags & AV_CODEC_FLAG_INTERLACED_ME) {
1535 //FIXME mb type penalty
1537 c->current_mv_penalty= c->mv_penalty[s->f_code] + MAX_DMV;
1538 fimin= interlaced_search(s, 0,
1539 s->b_field_mv_table[0], s->b_field_select_table[0],
1540 s->b_forw_mv_table[xy][0], s->b_forw_mv_table[xy][1], 0);
1541 c->current_mv_penalty= c->mv_penalty[s->b_code] + MAX_DMV;
1542 bimin= interlaced_search(s, 2,
1543 s->b_field_mv_table[1], s->b_field_select_table[1],
1544 s->b_back_mv_table[xy][0], s->b_back_mv_table[xy][1], 0);
1546 fimin= bimin= INT_MAX;
1550 type = CANDIDATE_MB_TYPE_FORWARD;
1554 type = CANDIDATE_MB_TYPE_DIRECT;
1558 type= CANDIDATE_MB_TYPE_BACKWARD;
1562 type= CANDIDATE_MB_TYPE_BIDIR;
1566 type= CANDIDATE_MB_TYPE_FORWARD_I;
1570 type= CANDIDATE_MB_TYPE_BACKWARD_I;
1573 score= ((unsigned)(score*score + 128*256))>>16;
1574 c->mc_mb_var_sum_temp += score;
1575 s->current_picture.mc_mb_var[mb_y*s->mb_stride + mb_x] = score; //FIXME use SSE
1578 if(c->avctx->mb_decision > FF_MB_DECISION_SIMPLE){
1579 type= CANDIDATE_MB_TYPE_FORWARD | CANDIDATE_MB_TYPE_BACKWARD | CANDIDATE_MB_TYPE_BIDIR | CANDIDATE_MB_TYPE_DIRECT;
1581 type |= CANDIDATE_MB_TYPE_FORWARD_I;
1583 type |= CANDIDATE_MB_TYPE_BACKWARD_I;
1584 if(fimin < INT_MAX && bimin < INT_MAX){
1585 type |= CANDIDATE_MB_TYPE_BIDIR_I;
1587 //FIXME something smarter
1588 if(dmin>256*256*16) type&= ~CANDIDATE_MB_TYPE_DIRECT; //do not try direct mode if it is invalid for this MB
1589 if (s->codec_id == AV_CODEC_ID_MPEG4 && type&CANDIDATE_MB_TYPE_DIRECT &&
1590 s->mpv_flags & FF_MPV_FLAG_MV0 && *(uint32_t*)s->b_direct_mv_table[xy])
1591 type |= CANDIDATE_MB_TYPE_DIRECT0;
1594 s->mb_type[mb_y*s->mb_stride + mb_x]= type;
1597 /* find best f_code for ME which do unlimited searches */
1598 int ff_get_best_fcode(MpegEncContext * s, int16_t (*mv_table)[2], int type)
1600 if (s->motion_est != FF_ME_ZERO) {
1602 int i, y, range= s->avctx->me_range ? s->avctx->me_range : (INT_MAX/2);
1603 uint8_t * fcode_tab= s->fcode_tab;
1605 int best_score=-10000000;
1607 if(s->msmpeg4_version)
1608 range= FFMIN(range, 16);
1609 else if(s->codec_id == AV_CODEC_ID_MPEG2VIDEO && s->avctx->strict_std_compliance >= FF_COMPLIANCE_NORMAL)
1610 range= FFMIN(range, 256);
1612 for(i=0; i<8; i++) score[i]= s->mb_num*(8-i);
1614 for(y=0; y<s->mb_height; y++){
1616 int xy= y*s->mb_stride;
1617 for(x=0; x<s->mb_width; x++){
1618 if(s->mb_type[xy] & type){
1619 int mx= mv_table[xy][0];
1620 int my= mv_table[xy][1];
1621 int fcode= FFMAX(fcode_tab[mx + MAX_MV],
1622 fcode_tab[my + MAX_MV]);
1625 if(mx >= range || mx < -range ||
1626 my >= range || my < -range)
1629 for(j=0; j<fcode && j<8; j++){
1630 if(s->pict_type==AV_PICTURE_TYPE_B || s->current_picture.mc_mb_var[xy] < s->current_picture.mb_var[xy])
1639 if(score[i] > best_score){
1640 best_score= score[i];
1651 void ff_fix_long_p_mvs(MpegEncContext * s)
1653 MotionEstContext * const c= &s->me;
1654 const int f_code= s->f_code;
1656 av_assert0(s->pict_type==AV_PICTURE_TYPE_P);
1658 range = (((s->out_format == FMT_MPEG1 || s->msmpeg4_version) ? 8 : 16) << f_code);
1660 av_assert0(range <= 16 || !s->msmpeg4_version);
1661 av_assert0(range <=256 || !(s->codec_id == AV_CODEC_ID_MPEG2VIDEO && s->avctx->strict_std_compliance >= FF_COMPLIANCE_NORMAL));
1663 if(c->avctx->me_range && range > c->avctx->me_range) range= c->avctx->me_range;
1665 if (s->avctx->flags & AV_CODEC_FLAG_4MV) {
1666 const int wrap= s->b8_stride;
1668 /* clip / convert to intra 8x8 type MVs */
1669 for(y=0; y<s->mb_height; y++){
1671 int i= y*s->mb_stride;
1674 for(x=0; x<s->mb_width; x++){
1675 if(s->mb_type[i]&CANDIDATE_MB_TYPE_INTER4V){
1677 for(block=0; block<4; block++){
1678 int off= (block& 1) + (block>>1)*wrap;
1679 int mx = s->current_picture.motion_val[0][ xy + off ][0];
1680 int my = s->current_picture.motion_val[0][ xy + off ][1];
1682 if( mx >=range || mx <-range
1683 || my >=range || my <-range){
1684 s->mb_type[i] &= ~CANDIDATE_MB_TYPE_INTER4V;
1685 s->mb_type[i] |= CANDIDATE_MB_TYPE_INTRA;
1686 s->current_picture.mb_type[i] = CANDIDATE_MB_TYPE_INTRA;
1698 * @param truncate 1 for truncation, 0 for using intra
1700 void ff_fix_long_mvs(MpegEncContext * s, uint8_t *field_select_table, int field_select,
1701 int16_t (*mv_table)[2], int f_code, int type, int truncate)
1703 MotionEstContext * const c= &s->me;
1704 int y, h_range, v_range;
1706 // RAL: 8 in MPEG-1, 16 in MPEG-4
1707 int range = (((s->out_format == FMT_MPEG1 || s->msmpeg4_version) ? 8 : 16) << f_code);
1709 if(c->avctx->me_range && range > c->avctx->me_range) range= c->avctx->me_range;
1712 v_range= field_select_table ? range>>1 : range;
1714 /* clip / convert to intra 16x16 type MVs */
1715 for(y=0; y<s->mb_height; y++){
1717 int xy= y*s->mb_stride;
1718 for(x=0; x<s->mb_width; x++){
1719 if (s->mb_type[xy] & type){ // RAL: "type" test added...
1720 if (!field_select_table || field_select_table[xy] == field_select) {
1721 if( mv_table[xy][0] >=h_range || mv_table[xy][0] <-h_range
1722 || mv_table[xy][1] >=v_range || mv_table[xy][1] <-v_range){
1725 if (mv_table[xy][0] > h_range-1) mv_table[xy][0]= h_range-1;
1726 else if(mv_table[xy][0] < -h_range ) mv_table[xy][0]= -h_range;
1727 if (mv_table[xy][1] > v_range-1) mv_table[xy][1]= v_range-1;
1728 else if(mv_table[xy][1] < -v_range ) mv_table[xy][1]= -v_range;
1730 s->mb_type[xy] &= ~type;
1731 s->mb_type[xy] |= CANDIDATE_MB_TYPE_INTRA;