motion_est.c
Go to the documentation of this file.
1 /*
2  * Motion estimation
3  * Copyright (c) 2000,2001 Fabrice Bellard
4  * Copyright (c) 2002-2004 Michael Niedermayer
5  *
6  * new motion estimation (X1/EPZS) by Michael Niedermayer <michaelni@gmx.at>
7  *
8  * This file is part of Libav.
9  *
10  * Libav 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.
14  *
15  * Libav 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.
19  *
20  * You should have received a copy of the GNU Lesser General Public
21  * License along with Libav; if not, write to the Free Software
22  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
23  */
24 
30 #include <stdlib.h>
31 #include <stdio.h>
32 #include <limits.h>
33 
34 #include "avcodec.h"
35 #include "dsputil.h"
36 #include "mathops.h"
37 #include "mpegvideo.h"
38 
39 #undef NDEBUG
40 #include <assert.h>
41 
42 #define P_LEFT P[1]
43 #define P_TOP P[2]
44 #define P_TOPRIGHT P[3]
45 #define P_MEDIAN P[4]
46 #define P_MV1 P[9]
47 
49  int *mx_ptr, int *my_ptr, int dmin,
50  int src_index, int ref_index,
51  int size, int h);
52 
53 static inline unsigned update_map_generation(MotionEstContext *c)
54 {
55  c->map_generation+= 1<<(ME_MAP_MV_BITS*2);
56  if(c->map_generation==0){
57  c->map_generation= 1<<(ME_MAP_MV_BITS*2);
58  memset(c->map, 0, sizeof(uint32_t)*ME_MAP_SIZE);
59  }
60  return c->map_generation;
61 }
62 
63 /* shape adaptive search stuff */
64 typedef struct Minima{
65  int height;
66  int x, y;
67  int checked;
68 }Minima;
69 
70 static int minima_cmp(const void *a, const void *b){
71  const Minima *da = (const Minima *) a;
72  const Minima *db = (const Minima *) b;
73 
74  return da->height - db->height;
75 }
76 
77 #define FLAG_QPEL 1 //must be 1
78 #define FLAG_CHROMA 2
79 #define FLAG_DIRECT 4
80 
81 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){
82  const int offset[3]= {
83  y*c-> stride + x,
84  ((y*c->uvstride + x)>>1),
85  ((y*c->uvstride + x)>>1),
86  };
87  int i;
88  for(i=0; i<3; i++){
89  c->src[0][i]= src [i] + offset[i];
90  c->ref[0][i]= ref [i] + offset[i];
91  }
92  if(ref_index){
93  for(i=0; i<3; i++){
94  c->ref[ref_index][i]= ref2[i] + offset[i];
95  }
96  }
97 }
98 
99 static int get_flags(MotionEstContext *c, int direct, int chroma){
100  return ((c->avctx->flags&CODEC_FLAG_QPEL) ? FLAG_QPEL : 0)
101  + (direct ? FLAG_DIRECT : 0)
102  + (chroma ? FLAG_CHROMA : 0);
103 }
104 
105 static av_always_inline int cmp_direct_inline(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
106  const int size, const int h, int ref_index, int src_index,
107  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, int qpel){
108  MotionEstContext * const c= &s->me;
109  const int stride= c->stride;
110  const int hx= subx + (x<<(1+qpel));
111  const int hy= suby + (y<<(1+qpel));
112  uint8_t * const * const ref= c->ref[ref_index];
113  uint8_t * const * const src= c->src[src_index];
114  int d;
115  //FIXME check chroma 4mv, (no crashes ...)
116  assert(x >= c->xmin && hx <= c->xmax<<(qpel+1) && y >= c->ymin && hy <= c->ymax<<(qpel+1));
117  if(x >= c->xmin && hx <= c->xmax<<(qpel+1) && y >= c->ymin && hy <= c->ymax<<(qpel+1)){
118  const int time_pp= s->pp_time;
119  const int time_pb= s->pb_time;
120  const int mask= 2*qpel+1;
121  if(s->mv_type==MV_TYPE_8X8){
122  int i;
123  for(i=0; i<4; i++){
124  int fx = c->direct_basis_mv[i][0] + hx;
125  int fy = c->direct_basis_mv[i][1] + hy;
126  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));
127  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));
128  int fxy= (fx&mask) + ((fy&mask)<<(qpel+1));
129  int bxy= (bx&mask) + ((by&mask)<<(qpel+1));
130 
131  uint8_t *dst= c->temp + 8*(i&1) + 8*stride*(i>>1);
132  if(qpel){
133  c->qpel_put[1][fxy](dst, ref[0] + (fx>>2) + (fy>>2)*stride, stride);
134  c->qpel_avg[1][bxy](dst, ref[8] + (bx>>2) + (by>>2)*stride, stride);
135  }else{
136  c->hpel_put[1][fxy](dst, ref[0] + (fx>>1) + (fy>>1)*stride, stride, 8);
137  c->hpel_avg[1][bxy](dst, ref[8] + (bx>>1) + (by>>1)*stride, stride, 8);
138  }
139  }
140  }else{
141  int fx = c->direct_basis_mv[0][0] + hx;
142  int fy = c->direct_basis_mv[0][1] + hy;
143  int bx = hx ? fx - c->co_located_mv[0][0] : (c->co_located_mv[0][0]*(time_pb - time_pp)/time_pp);
144  int by = hy ? fy - c->co_located_mv[0][1] : (c->co_located_mv[0][1]*(time_pb - time_pp)/time_pp);
145  int fxy= (fx&mask) + ((fy&mask)<<(qpel+1));
146  int bxy= (bx&mask) + ((by&mask)<<(qpel+1));
147 
148  if(qpel){
149  c->qpel_put[1][fxy](c->temp , ref[0] + (fx>>2) + (fy>>2)*stride , stride);
150  c->qpel_put[1][fxy](c->temp + 8 , ref[0] + (fx>>2) + (fy>>2)*stride + 8 , stride);
151  c->qpel_put[1][fxy](c->temp + 8*stride, ref[0] + (fx>>2) + (fy>>2)*stride + 8*stride, stride);
152  c->qpel_put[1][fxy](c->temp + 8 + 8*stride, ref[0] + (fx>>2) + (fy>>2)*stride + 8 + 8*stride, stride);
153  c->qpel_avg[1][bxy](c->temp , ref[8] + (bx>>2) + (by>>2)*stride , stride);
154  c->qpel_avg[1][bxy](c->temp + 8 , ref[8] + (bx>>2) + (by>>2)*stride + 8 , stride);
155  c->qpel_avg[1][bxy](c->temp + 8*stride, ref[8] + (bx>>2) + (by>>2)*stride + 8*stride, stride);
156  c->qpel_avg[1][bxy](c->temp + 8 + 8*stride, ref[8] + (bx>>2) + (by>>2)*stride + 8 + 8*stride, stride);
157  }else{
158  assert((fx>>1) + 16*s->mb_x >= -16);
159  assert((fy>>1) + 16*s->mb_y >= -16);
160  assert((fx>>1) + 16*s->mb_x <= s->width);
161  assert((fy>>1) + 16*s->mb_y <= s->height);
162  assert((bx>>1) + 16*s->mb_x >= -16);
163  assert((by>>1) + 16*s->mb_y >= -16);
164  assert((bx>>1) + 16*s->mb_x <= s->width);
165  assert((by>>1) + 16*s->mb_y <= s->height);
166 
167  c->hpel_put[0][fxy](c->temp, ref[0] + (fx>>1) + (fy>>1)*stride, stride, 16);
168  c->hpel_avg[0][bxy](c->temp, ref[8] + (bx>>1) + (by>>1)*stride, stride, 16);
169  }
170  }
171  d = cmp_func(s, c->temp, src[0], stride, 16);
172  }else
173  d= 256*256*256*32;
174  return d;
175 }
176 
177 static av_always_inline int cmp_inline(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
178  const int size, const int h, int ref_index, int src_index,
179  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, int qpel, int chroma){
180  MotionEstContext * const c= &s->me;
181  const int stride= c->stride;
182  const int uvstride= c->uvstride;
183  const int dxy= subx + (suby<<(1+qpel)); //FIXME log2_subpel?
184  const int hx= subx + (x<<(1+qpel));
185  const int hy= suby + (y<<(1+qpel));
186  uint8_t * const * const ref= c->ref[ref_index];
187  uint8_t * const * const src= c->src[src_index];
188  int d;
189  //FIXME check chroma 4mv, (no crashes ...)
190  int uvdxy; /* no, it might not be used uninitialized */
191  if(dxy){
192  if(qpel){
193  c->qpel_put[size][dxy](c->temp, ref[0] + x + y*stride, stride); //FIXME prototype (add h)
194  if(chroma){
195  int cx= hx/2;
196  int cy= hy/2;
197  cx= (cx>>1)|(cx&1);
198  cy= (cy>>1)|(cy&1);
199  uvdxy= (cx&1) + 2*(cy&1);
200  //FIXME x/y wrong, but mpeg4 qpel is sick anyway, we should drop as much of it as possible in favor for h264
201  }
202  }else{
203  c->hpel_put[size][dxy](c->temp, ref[0] + x + y*stride, stride, h);
204  if(chroma)
205  uvdxy= dxy | (x&1) | (2*(y&1));
206  }
207  d = cmp_func(s, c->temp, src[0], stride, h);
208  }else{
209  d = cmp_func(s, src[0], ref[0] + x + y*stride, stride, h);
210  if(chroma)
211  uvdxy= (x&1) + 2*(y&1);
212  }
213  if(chroma){
214  uint8_t * const uvtemp= c->temp + 16*stride;
215  c->hpel_put[size+1][uvdxy](uvtemp , ref[1] + (x>>1) + (y>>1)*uvstride, uvstride, h>>1);
216  c->hpel_put[size+1][uvdxy](uvtemp+8, ref[2] + (x>>1) + (y>>1)*uvstride, uvstride, h>>1);
217  d += chroma_cmp_func(s, uvtemp , src[1], uvstride, h>>1);
218  d += chroma_cmp_func(s, uvtemp+8, src[2], uvstride, h>>1);
219  }
220  return d;
221 }
222 
223 static int cmp_simple(MpegEncContext *s, const int x, const int y,
224  int ref_index, int src_index,
225  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func){
226  return cmp_inline(s,x,y,0,0,0,16,ref_index,src_index, cmp_func, chroma_cmp_func, 0, 0);
227 }
228 
229 static int cmp_fpel_internal(MpegEncContext *s, const int x, const int y,
230  const int size, const int h, int ref_index, int src_index,
231  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
232  if(flags&FLAG_DIRECT){
233  return cmp_direct_inline(s,x,y,0,0,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags&FLAG_QPEL);
234  }else{
235  return cmp_inline(s,x,y,0,0,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 0, flags&FLAG_CHROMA);
236  }
237 }
238 
239 static int cmp_internal(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
240  const int size, const int h, int ref_index, int src_index,
241  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
242  if(flags&FLAG_DIRECT){
243  return cmp_direct_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags&FLAG_QPEL);
244  }else{
245  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);
246  }
247 }
248 
252 static av_always_inline int cmp(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
253  const int size, const int h, int ref_index, int src_index,
254  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
257  && flags==0 && h==16 && size==0 && subx==0 && suby==0){
258  return cmp_simple(s,x,y,ref_index,src_index, cmp_func, chroma_cmp_func);
259  }else if(av_builtin_constant_p(subx) && av_builtin_constant_p(suby)
260  && subx==0 && suby==0){
261  return cmp_fpel_internal(s,x,y,size,h,ref_index,src_index, cmp_func, chroma_cmp_func,flags);
262  }else{
263  return cmp_internal(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, flags);
264  }
265 }
266 
267 static int cmp_hpel(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
268  const int size, const int h, int ref_index, int src_index,
269  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
270  if(flags&FLAG_DIRECT){
271  return cmp_direct_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 0);
272  }else{
273  return cmp_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 0, flags&FLAG_CHROMA);
274  }
275 }
276 
277 static int cmp_qpel(MpegEncContext *s, const int x, const int y, const int subx, const int suby,
278  const int size, const int h, int ref_index, int src_index,
279  me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags){
280  if(flags&FLAG_DIRECT){
281  return cmp_direct_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 1);
282  }else{
283  return cmp_inline(s,x,y,subx,suby,size,h,ref_index,src_index, cmp_func, chroma_cmp_func, 1, flags&FLAG_CHROMA);
284  }
285 }
286 
287 #include "motion_est_template.c"
288 
289 static int zero_cmp(void *s, uint8_t *a, uint8_t *b, int stride, int h){
290  return 0;
291 }
292 
293 static void zero_hpel(uint8_t *a, const uint8_t *b, int stride, int h){
294 }
295 
297  MotionEstContext * const c= &s->me;
298  int cache_size= FFMIN(ME_MAP_SIZE>>ME_MAP_SHIFT, 1<<ME_MAP_SHIFT);
299  int dia_size= FFMAX(FFABS(s->avctx->dia_size)&255, FFABS(s->avctx->pre_dia_size)&255);
300 
302  av_log(s->avctx, AV_LOG_ERROR, "ME_MAP size is too small for SAB diamond\n");
303  return -1;
304  }
305  //special case of snow is needed because snow uses its own iterative ME code
307  av_log(s->avctx, AV_LOG_ERROR, "me_method is only allowed to be set to zero and epzs; for hex,umh,full and others see dia_size\n");
308  return -1;
309  }
310 
311  c->avctx= s->avctx;
312 
313  if(cache_size < 2*dia_size && !c->stride){
314  av_log(s->avctx, AV_LOG_INFO, "ME_MAP size may be a little small for the selected diamond size\n");
315  }
316 
318  ff_set_cmp(&s->dsp, s->dsp.me_cmp, c->avctx->me_cmp);
320  ff_set_cmp(&s->dsp, s->dsp.mb_cmp, c->avctx->mb_cmp);
321 
322  c->flags = get_flags(c, 0, c->avctx->me_cmp &FF_CMP_CHROMA);
324  c->mb_flags = get_flags(c, 0, c->avctx->mb_cmp &FF_CMP_CHROMA);
325 
326 /*FIXME s->no_rounding b_type*/
327  if(s->flags&CODEC_FLAG_QPEL){
331  else c->qpel_put= s->dsp.put_qpel_pixels_tab;
332  }else{
335  else if( c->avctx->me_sub_cmp == FF_CMP_SAD
336  && c->avctx-> me_cmp == FF_CMP_SAD
337  && c->avctx-> mb_cmp == FF_CMP_SAD)
338  c->sub_motion_search= sad_hpel_motion_search; // 2050 vs. 2450 cycles
339  else
341  }
342  c->hpel_avg= s->dsp.avg_pixels_tab;
344  else c->hpel_put= s->dsp.put_pixels_tab;
345 
346  if(s->linesize){
347  c->stride = s->linesize;
348  c->uvstride= s->uvlinesize;
349  }else{
350  c->stride = 16*s->mb_width + 32;
351  c->uvstride= 8*s->mb_width + 16;
352  }
353 
354  /* 8x8 fullpel search would need a 4x4 chroma compare, which we do
355  * not have yet, and even if we had, the motion estimation code
356  * does not expect it. */
357  if(s->codec_id != AV_CODEC_ID_SNOW){
358  if((c->avctx->me_cmp&FF_CMP_CHROMA)/* && !s->dsp.me_cmp[2]*/){
359  s->dsp.me_cmp[2]= zero_cmp;
360  }
361  if((c->avctx->me_sub_cmp&FF_CMP_CHROMA) && !s->dsp.me_sub_cmp[2]){
362  s->dsp.me_sub_cmp[2]= zero_cmp;
363  }
364  c->hpel_put[2][0]= c->hpel_put[2][1]=
365  c->hpel_put[2][2]= c->hpel_put[2][3]= zero_hpel;
366  }
367 
368  if(s->codec_id == AV_CODEC_ID_H261){
370  }
371 
372  return 0;
373 }
374 
375 #define CHECK_SAD_HALF_MV(suffix, x, y) \
376 {\
377  d= s->dsp.pix_abs[size][(x?1:0)+(y?2:0)](NULL, pix, ptr+((x)>>1), stride, h);\
378  d += (mv_penalty[pen_x + x] + mv_penalty[pen_y + y])*penalty_factor;\
379  COPY3_IF_LT(dminh, d, dx, x, dy, y)\
380 }
381 
383  int *mx_ptr, int *my_ptr, int dmin,
384  int src_index, int ref_index,
385  int size, int h)
386 {
387  MotionEstContext * const c= &s->me;
388  const int penalty_factor= c->sub_penalty_factor;
389  int mx, my, dminh;
390  uint8_t *pix, *ptr;
391  int stride= c->stride;
392  const int flags= c->sub_flags;
394 
395  assert(flags == 0);
396 
397  if(c->skip){
398  *mx_ptr = 0;
399  *my_ptr = 0;
400  return dmin;
401  }
402 
403  pix = c->src[src_index][0];
404 
405  mx = *mx_ptr;
406  my = *my_ptr;
407  ptr = c->ref[ref_index][0] + (my * stride) + mx;
408 
409  dminh = dmin;
410 
411  if (mx > xmin && mx < xmax &&
412  my > ymin && my < ymax) {
413  int dx=0, dy=0;
414  int d, pen_x, pen_y;
415  const int index= (my<<ME_MAP_SHIFT) + mx;
416  const int t= score_map[(index-(1<<ME_MAP_SHIFT))&(ME_MAP_SIZE-1)];
417  const int l= score_map[(index- 1 )&(ME_MAP_SIZE-1)];
418  const int r= score_map[(index+ 1 )&(ME_MAP_SIZE-1)];
419  const int b= score_map[(index+(1<<ME_MAP_SHIFT))&(ME_MAP_SIZE-1)];
420  mx<<=1;
421  my<<=1;
422 
423 
424  pen_x= pred_x + mx;
425  pen_y= pred_y + my;
426 
427  ptr-= stride;
428  if(t<=b){
429  CHECK_SAD_HALF_MV(y2 , 0, -1)
430  if(l<=r){
431  CHECK_SAD_HALF_MV(xy2, -1, -1)
432  if(t+r<=b+l){
433  CHECK_SAD_HALF_MV(xy2, +1, -1)
434  ptr+= stride;
435  }else{
436  ptr+= stride;
437  CHECK_SAD_HALF_MV(xy2, -1, +1)
438  }
439  CHECK_SAD_HALF_MV(x2 , -1, 0)
440  }else{
441  CHECK_SAD_HALF_MV(xy2, +1, -1)
442  if(t+l<=b+r){
443  CHECK_SAD_HALF_MV(xy2, -1, -1)
444  ptr+= stride;
445  }else{
446  ptr+= stride;
447  CHECK_SAD_HALF_MV(xy2, +1, +1)
448  }
449  CHECK_SAD_HALF_MV(x2 , +1, 0)
450  }
451  }else{
452  if(l<=r){
453  if(t+l<=b+r){
454  CHECK_SAD_HALF_MV(xy2, -1, -1)
455  ptr+= stride;
456  }else{
457  ptr+= stride;
458  CHECK_SAD_HALF_MV(xy2, +1, +1)
459  }
460  CHECK_SAD_HALF_MV(x2 , -1, 0)
461  CHECK_SAD_HALF_MV(xy2, -1, +1)
462  }else{
463  if(t+r<=b+l){
464  CHECK_SAD_HALF_MV(xy2, +1, -1)
465  ptr+= stride;
466  }else{
467  ptr+= stride;
468  CHECK_SAD_HALF_MV(xy2, -1, +1)
469  }
470  CHECK_SAD_HALF_MV(x2 , +1, 0)
471  CHECK_SAD_HALF_MV(xy2, +1, +1)
472  }
473  CHECK_SAD_HALF_MV(y2 , 0, +1)
474  }
475  mx+=dx;
476  my+=dy;
477 
478  }else{
479  mx<<=1;
480  my<<=1;
481  }
482 
483  *mx_ptr = mx;
484  *my_ptr = my;
485  return dminh;
486 }
487 
488 static inline void set_p_mv_tables(MpegEncContext * s, int mx, int my, int mv4)
489 {
490  const int xy= s->mb_x + s->mb_y*s->mb_stride;
491 
492  s->p_mv_table[xy][0] = mx;
493  s->p_mv_table[xy][1] = my;
494 
495  /* has already been set to the 4 MV if 4MV is done */
496  if(mv4){
497  int mot_xy= s->block_index[0];
498 
499  s->current_picture.f.motion_val[0][mot_xy ][0] = mx;
500  s->current_picture.f.motion_val[0][mot_xy ][1] = my;
501  s->current_picture.f.motion_val[0][mot_xy + 1][0] = mx;
502  s->current_picture.f.motion_val[0][mot_xy + 1][1] = my;
503 
504  mot_xy += s->b8_stride;
505  s->current_picture.f.motion_val[0][mot_xy ][0] = mx;
506  s->current_picture.f.motion_val[0][mot_xy ][1] = my;
507  s->current_picture.f.motion_val[0][mot_xy + 1][0] = mx;
508  s->current_picture.f.motion_val[0][mot_xy + 1][1] = my;
509  }
510 }
511 
515 static inline void get_limits(MpegEncContext *s, int x, int y)
516 {
517  MotionEstContext * const c= &s->me;
518  int range= c->avctx->me_range >> (1 + !!(c->flags&FLAG_QPEL));
519 /*
520  if(c->avctx->me_range) c->range= c->avctx->me_range >> 1;
521  else c->range= 16;
522 */
523  if (s->unrestricted_mv) {
524  c->xmin = - x - 16;
525  c->ymin = - y - 16;
526  c->xmax = - x + s->mb_width *16;
527  c->ymax = - y + s->mb_height*16;
528  } else if (s->out_format == FMT_H261){
529  // Search range of H261 is different from other codec standards
530  c->xmin = (x > 15) ? - 15 : 0;
531  c->ymin = (y > 15) ? - 15 : 0;
532  c->xmax = (x < s->mb_width * 16 - 16) ? 15 : 0;
533  c->ymax = (y < s->mb_height * 16 - 16) ? 15 : 0;
534  } else {
535  c->xmin = - x;
536  c->ymin = - y;
537  c->xmax = - x + s->mb_width *16 - 16;
538  c->ymax = - y + s->mb_height*16 - 16;
539  }
540  if(range){
541  c->xmin = FFMAX(c->xmin,-range);
542  c->xmax = FFMIN(c->xmax, range);
543  c->ymin = FFMAX(c->ymin,-range);
544  c->ymax = FFMIN(c->ymax, range);
545  }
546 }
547 
548 static inline void init_mv4_ref(MotionEstContext *c){
549  const int stride= c->stride;
550 
551  c->ref[1][0] = c->ref[0][0] + 8;
552  c->ref[2][0] = c->ref[0][0] + 8*stride;
553  c->ref[3][0] = c->ref[2][0] + 8;
554  c->src[1][0] = c->src[0][0] + 8;
555  c->src[2][0] = c->src[0][0] + 8*stride;
556  c->src[3][0] = c->src[2][0] + 8;
557 }
558 
559 static inline int h263_mv4_search(MpegEncContext *s, int mx, int my, int shift)
560 {
561  MotionEstContext * const c= &s->me;
562  const int size= 1;
563  const int h=8;
564  int block;
565  int P[10][2];
566  int dmin_sum=0, mx4_sum=0, my4_sum=0;
567  int same=1;
568  const int stride= c->stride;
570 
571  init_mv4_ref(c);
572 
573  for(block=0; block<4; block++){
574  int mx4, my4;
575  int pred_x4, pred_y4;
576  int dmin4;
577  static const int off[4]= {2, 1, 1, -1};
578  const int mot_stride = s->b8_stride;
579  const int mot_xy = s->block_index[block];
580 
581  P_LEFT[0] = s->current_picture.f.motion_val[0][mot_xy - 1][0];
582  P_LEFT[1] = s->current_picture.f.motion_val[0][mot_xy - 1][1];
583 
584  if(P_LEFT[0] > (c->xmax<<shift)) P_LEFT[0] = (c->xmax<<shift);
585 
586  /* special case for first line */
587  if (s->first_slice_line && block<2) {
588  c->pred_x= pred_x4= P_LEFT[0];
589  c->pred_y= pred_y4= P_LEFT[1];
590  } else {
591  P_TOP[0] = s->current_picture.f.motion_val[0][mot_xy - mot_stride ][0];
592  P_TOP[1] = s->current_picture.f.motion_val[0][mot_xy - mot_stride ][1];
593  P_TOPRIGHT[0] = s->current_picture.f.motion_val[0][mot_xy - mot_stride + off[block]][0];
594  P_TOPRIGHT[1] = s->current_picture.f.motion_val[0][mot_xy - mot_stride + off[block]][1];
595  if(P_TOP[1] > (c->ymax<<shift)) P_TOP[1] = (c->ymax<<shift);
596  if(P_TOPRIGHT[0] < (c->xmin<<shift)) P_TOPRIGHT[0]= (c->xmin<<shift);
597  if(P_TOPRIGHT[0] > (c->xmax<<shift)) P_TOPRIGHT[0]= (c->xmax<<shift);
598  if(P_TOPRIGHT[1] > (c->ymax<<shift)) P_TOPRIGHT[1]= (c->ymax<<shift);
599 
600  P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
601  P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
602 
603  c->pred_x= pred_x4 = P_MEDIAN[0];
604  c->pred_y= pred_y4 = P_MEDIAN[1];
605  }
606  P_MV1[0]= mx;
607  P_MV1[1]= my;
608 
609  dmin4 = epzs_motion_search4(s, &mx4, &my4, P, block, block, s->p_mv_table, (1<<16)>>shift);
610 
611  dmin4= c->sub_motion_search(s, &mx4, &my4, dmin4, block, block, size, h);
612 
613  if(s->dsp.me_sub_cmp[0] != s->dsp.mb_cmp[0]){
614  int dxy;
615  const int offset= ((block&1) + (block>>1)*stride)*8;
616  uint8_t *dest_y = c->scratchpad + offset;
617  if(s->quarter_sample){
618  uint8_t *ref= c->ref[block][0] + (mx4>>2) + (my4>>2)*stride;
619  dxy = ((my4 & 3) << 2) | (mx4 & 3);
620 
621  if(s->no_rounding)
622  s->dsp.put_no_rnd_qpel_pixels_tab[1][dxy](dest_y , ref , stride);
623  else
624  s->dsp.put_qpel_pixels_tab [1][dxy](dest_y , ref , stride);
625  }else{
626  uint8_t *ref= c->ref[block][0] + (mx4>>1) + (my4>>1)*stride;
627  dxy = ((my4 & 1) << 1) | (mx4 & 1);
628 
629  if(s->no_rounding)
630  s->dsp.put_no_rnd_pixels_tab[1][dxy](dest_y , ref , stride, h);
631  else
632  s->dsp.put_pixels_tab [1][dxy](dest_y , ref , stride, h);
633  }
634  dmin_sum+= (mv_penalty[mx4-pred_x4] + mv_penalty[my4-pred_y4])*c->mb_penalty_factor;
635  }else
636  dmin_sum+= dmin4;
637 
638  if(s->quarter_sample){
639  mx4_sum+= mx4/2;
640  my4_sum+= my4/2;
641  }else{
642  mx4_sum+= mx4;
643  my4_sum+= my4;
644  }
645 
646  s->current_picture.f.motion_val[0][s->block_index[block]][0] = mx4;
647  s->current_picture.f.motion_val[0][s->block_index[block]][1] = my4;
648 
649  if(mx4 != mx || my4 != my) same=0;
650  }
651 
652  if(same)
653  return INT_MAX;
654 
655  if(s->dsp.me_sub_cmp[0] != s->dsp.mb_cmp[0]){
656  dmin_sum += s->dsp.mb_cmp[0](s, s->new_picture.f.data[0] + s->mb_x*16 + s->mb_y*16*stride, c->scratchpad, stride, 16);
657  }
658 
659  if(c->avctx->mb_cmp&FF_CMP_CHROMA){
660  int dxy;
661  int mx, my;
662  int offset;
663 
664  mx= ff_h263_round_chroma(mx4_sum);
665  my= ff_h263_round_chroma(my4_sum);
666  dxy = ((my & 1) << 1) | (mx & 1);
667 
668  offset= (s->mb_x*8 + (mx>>1)) + (s->mb_y*8 + (my>>1))*s->uvlinesize;
669 
670  if(s->no_rounding){
671  s->dsp.put_no_rnd_pixels_tab[1][dxy](c->scratchpad , s->last_picture.f.data[1] + offset, s->uvlinesize, 8);
672  s->dsp.put_no_rnd_pixels_tab[1][dxy](c->scratchpad + 8, s->last_picture.f.data[2] + offset, s->uvlinesize, 8);
673  }else{
674  s->dsp.put_pixels_tab [1][dxy](c->scratchpad , s->last_picture.f.data[1] + offset, s->uvlinesize, 8);
675  s->dsp.put_pixels_tab [1][dxy](c->scratchpad + 8, s->last_picture.f.data[2] + offset, s->uvlinesize, 8);
676  }
677 
678  dmin_sum += s->dsp.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);
679  dmin_sum += s->dsp.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);
680  }
681 
682  c->pred_x= mx;
683  c->pred_y= my;
684 
685  switch(c->avctx->mb_cmp&0xFF){
686  /*case FF_CMP_SSE:
687  return dmin_sum+ 32*s->qscale*s->qscale;*/
688  case FF_CMP_RD:
689  return dmin_sum;
690  default:
691  return dmin_sum+ 11*c->mb_penalty_factor;
692  }
693 }
694 
695 static inline void init_interlaced_ref(MpegEncContext *s, int ref_index){
696  MotionEstContext * const c= &s->me;
697 
698  c->ref[1+ref_index][0] = c->ref[0+ref_index][0] + s->linesize;
699  c->src[1][0] = c->src[0][0] + s->linesize;
700  if(c->flags & FLAG_CHROMA){
701  c->ref[1+ref_index][1] = c->ref[0+ref_index][1] + s->uvlinesize;
702  c->ref[1+ref_index][2] = c->ref[0+ref_index][2] + s->uvlinesize;
703  c->src[1][1] = c->src[0][1] + s->uvlinesize;
704  c->src[1][2] = c->src[0][2] + s->uvlinesize;
705  }
706 }
707 
708 static int interlaced_search(MpegEncContext *s, int ref_index,
709  int16_t (*mv_tables[2][2])[2], uint8_t *field_select_tables[2], int mx, int my, int user_field_select)
710 {
711  MotionEstContext * const c= &s->me;
712  const int size=0;
713  const int h=8;
714  int block;
715  int P[10][2];
717  int same=1;
718  const int stride= 2*s->linesize;
719  int dmin_sum= 0;
720  const int mot_stride= s->mb_stride;
721  const int xy= s->mb_x + s->mb_y*mot_stride;
722 
723  c->ymin>>=1;
724  c->ymax>>=1;
725  c->stride<<=1;
726  c->uvstride<<=1;
727  init_interlaced_ref(s, ref_index);
728 
729  for(block=0; block<2; block++){
730  int field_select;
731  int best_dmin= INT_MAX;
732  int best_field= -1;
733 
734  for(field_select=0; field_select<2; field_select++){
735  int dmin, mx_i, my_i;
736  int16_t (*mv_table)[2]= mv_tables[block][field_select];
737 
738  if(user_field_select){
739  assert(field_select==0 || field_select==1);
740  assert(field_select_tables[block][xy]==0 || field_select_tables[block][xy]==1);
741  if(field_select_tables[block][xy] != field_select)
742  continue;
743  }
744 
745  P_LEFT[0] = mv_table[xy - 1][0];
746  P_LEFT[1] = mv_table[xy - 1][1];
747  if(P_LEFT[0] > (c->xmax<<1)) P_LEFT[0] = (c->xmax<<1);
748 
749  c->pred_x= P_LEFT[0];
750  c->pred_y= P_LEFT[1];
751 
752  if(!s->first_slice_line){
753  P_TOP[0] = mv_table[xy - mot_stride][0];
754  P_TOP[1] = mv_table[xy - mot_stride][1];
755  P_TOPRIGHT[0] = mv_table[xy - mot_stride + 1][0];
756  P_TOPRIGHT[1] = mv_table[xy - mot_stride + 1][1];
757  if(P_TOP[1] > (c->ymax<<1)) P_TOP[1] = (c->ymax<<1);
758  if(P_TOPRIGHT[0] < (c->xmin<<1)) P_TOPRIGHT[0]= (c->xmin<<1);
759  if(P_TOPRIGHT[0] > (c->xmax<<1)) P_TOPRIGHT[0]= (c->xmax<<1);
760  if(P_TOPRIGHT[1] > (c->ymax<<1)) P_TOPRIGHT[1]= (c->ymax<<1);
761 
762  P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
763  P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
764  }
765  P_MV1[0]= mx; //FIXME not correct if block != field_select
766  P_MV1[1]= my / 2;
767 
768  dmin = epzs_motion_search2(s, &mx_i, &my_i, P, block, field_select+ref_index, mv_table, (1<<16)>>1);
769 
770  dmin= c->sub_motion_search(s, &mx_i, &my_i, dmin, block, field_select+ref_index, size, h);
771 
772  mv_table[xy][0]= mx_i;
773  mv_table[xy][1]= my_i;
774 
775  if(s->dsp.me_sub_cmp[0] != s->dsp.mb_cmp[0]){
776  int dxy;
777 
778  //FIXME chroma ME
779  uint8_t *ref= c->ref[field_select+ref_index][0] + (mx_i>>1) + (my_i>>1)*stride;
780  dxy = ((my_i & 1) << 1) | (mx_i & 1);
781 
782  if(s->no_rounding){
783  s->dsp.put_no_rnd_pixels_tab[size][dxy](c->scratchpad, ref , stride, h);
784  }else{
785  s->dsp.put_pixels_tab [size][dxy](c->scratchpad, ref , stride, h);
786  }
787  dmin= s->dsp.mb_cmp[size](s, c->src[block][0], c->scratchpad, stride, h);
788  dmin+= (mv_penalty[mx_i-c->pred_x] + mv_penalty[my_i-c->pred_y] + 1)*c->mb_penalty_factor;
789  }else
790  dmin+= c->mb_penalty_factor; //field_select bits
791 
792  dmin += field_select != block; //slightly prefer same field
793 
794  if(dmin < best_dmin){
795  best_dmin= dmin;
796  best_field= field_select;
797  }
798  }
799  {
800  int16_t (*mv_table)[2]= mv_tables[block][best_field];
801 
802  if(mv_table[xy][0] != mx) same=0; //FIXME check if these checks work and are any good at all
803  if(mv_table[xy][1]&1) same=0;
804  if(mv_table[xy][1]*2 != my) same=0;
805  if(best_field != block) same=0;
806  }
807 
808  field_select_tables[block][xy]= best_field;
809  dmin_sum += best_dmin;
810  }
811 
812  c->ymin<<=1;
813  c->ymax<<=1;
814  c->stride>>=1;
815  c->uvstride>>=1;
816 
817  if(same)
818  return INT_MAX;
819 
820  switch(c->avctx->mb_cmp&0xFF){
821  /*case FF_CMP_SSE:
822  return dmin_sum+ 32*s->qscale*s->qscale;*/
823  case FF_CMP_RD:
824  return dmin_sum;
825  default:
826  return dmin_sum+ 11*c->mb_penalty_factor;
827  }
828 }
829 
830 static void clip_input_mv(MpegEncContext * s, int16_t *mv, int interlaced){
831  int ymax= s->me.ymax>>interlaced;
832  int ymin= s->me.ymin>>interlaced;
833 
834  if(mv[0] < s->me.xmin) mv[0] = s->me.xmin;
835  if(mv[0] > s->me.xmax) mv[0] = s->me.xmax;
836  if(mv[1] < ymin) mv[1] = ymin;
837  if(mv[1] > ymax) mv[1] = ymax;
838 }
839 
840 static inline int check_input_motion(MpegEncContext * s, int mb_x, int mb_y, int p_type){
841  MotionEstContext * const c= &s->me;
843  int mb_xy= mb_x + mb_y*s->mb_stride;
844  int xy= 2*mb_x + 2*mb_y*s->b8_stride;
845  int mb_type= s->current_picture.f.mb_type[mb_xy];
846  int flags= c->flags;
847  int shift= (flags&FLAG_QPEL) + 1;
848  int mask= (1<<shift)-1;
849  int x, y, i;
850  int d=0;
851  me_cmp_func cmpf= s->dsp.sse[0];
852  me_cmp_func chroma_cmpf= s->dsp.sse[1];
853 
854  if(p_type && USES_LIST(mb_type, 1)){
855  av_log(c->avctx, AV_LOG_ERROR, "backward motion vector in P frame\n");
856  return INT_MAX/2;
857  }
858  assert(IS_INTRA(mb_type) || USES_LIST(mb_type,0) || USES_LIST(mb_type,1));
859 
860  for(i=0; i<4; i++){
861  int xy= s->block_index[i];
862  clip_input_mv(s, p->f.motion_val[0][xy], !!IS_INTERLACED(mb_type));
863  clip_input_mv(s, p->f.motion_val[1][xy], !!IS_INTERLACED(mb_type));
864  }
865 
866  if(IS_INTERLACED(mb_type)){
867  int xy2= xy + s->b8_stride;
869  c->stride<<=1;
870  c->uvstride<<=1;
871 
872  if(!(s->flags & CODEC_FLAG_INTERLACED_ME)){
873  av_log(c->avctx, AV_LOG_ERROR, "Interlaced macroblock selected but interlaced motion estimation disabled\n");
874  return INT_MAX/2;
875  }
876 
877  if(USES_LIST(mb_type, 0)){
878  int field_select0= p->f.ref_index[0][4*mb_xy ];
879  int field_select1= p->f.ref_index[0][4*mb_xy+2];
880  assert(field_select0==0 ||field_select0==1);
881  assert(field_select1==0 ||field_select1==1);
882  init_interlaced_ref(s, 0);
883 
884  if(p_type){
885  s->p_field_select_table[0][mb_xy]= field_select0;
886  s->p_field_select_table[1][mb_xy]= field_select1;
887  *(uint32_t*)s->p_field_mv_table[0][field_select0][mb_xy] = *(uint32_t*)p->f.motion_val[0][xy ];
888  *(uint32_t*)s->p_field_mv_table[1][field_select1][mb_xy] = *(uint32_t*)p->f.motion_val[0][xy2];
890  }else{
891  s->b_field_select_table[0][0][mb_xy]= field_select0;
892  s->b_field_select_table[0][1][mb_xy]= field_select1;
893  *(uint32_t*)s->b_field_mv_table[0][0][field_select0][mb_xy] = *(uint32_t*)p->f.motion_val[0][xy ];
894  *(uint32_t*)s->b_field_mv_table[0][1][field_select1][mb_xy] = *(uint32_t*)p->f.motion_val[0][xy2];
896  }
897 
898  x = p->f.motion_val[0][xy ][0];
899  y = p->f.motion_val[0][xy ][1];
900  d = cmp(s, x>>shift, y>>shift, x&mask, y&mask, 0, 8, field_select0, 0, cmpf, chroma_cmpf, flags);
901  x = p->f.motion_val[0][xy2][0];
902  y = p->f.motion_val[0][xy2][1];
903  d+= cmp(s, x>>shift, y>>shift, x&mask, y&mask, 0, 8, field_select1, 1, cmpf, chroma_cmpf, flags);
904  }
905  if(USES_LIST(mb_type, 1)){
906  int field_select0 = p->f.ref_index[1][4 * mb_xy ];
907  int field_select1 = p->f.ref_index[1][4 * mb_xy + 2];
908  assert(field_select0==0 ||field_select0==1);
909  assert(field_select1==0 ||field_select1==1);
910  init_interlaced_ref(s, 2);
911 
912  s->b_field_select_table[1][0][mb_xy]= field_select0;
913  s->b_field_select_table[1][1][mb_xy]= field_select1;
914  *(uint32_t*)s->b_field_mv_table[1][0][field_select0][mb_xy] = *(uint32_t*)p->f.motion_val[1][xy ];
915  *(uint32_t*)s->b_field_mv_table[1][1][field_select1][mb_xy] = *(uint32_t*)p->f.motion_val[1][xy2];
916  if(USES_LIST(mb_type, 0)){
918  }else{
920  }
921 
922  x = p->f.motion_val[1][xy ][0];
923  y = p->f.motion_val[1][xy ][1];
924  d = cmp(s, x>>shift, y>>shift, x&mask, y&mask, 0, 8, field_select0+2, 0, cmpf, chroma_cmpf, flags);
925  x = p->f.motion_val[1][xy2][0];
926  y = p->f.motion_val[1][xy2][1];
927  d+= cmp(s, x>>shift, y>>shift, x&mask, y&mask, 0, 8, field_select1+2, 1, cmpf, chroma_cmpf, flags);
928  //FIXME bidir scores
929  }
930  c->stride>>=1;
931  c->uvstride>>=1;
932  }else if(IS_8X8(mb_type)){
933  if(!(s->flags & CODEC_FLAG_4MV)){
934  av_log(c->avctx, AV_LOG_ERROR, "4MV macroblock selected but 4MV encoding disabled\n");
935  return INT_MAX/2;
936  }
937  cmpf= s->dsp.sse[1];
938  chroma_cmpf= s->dsp.sse[1];
939  init_mv4_ref(c);
940  for(i=0; i<4; i++){
941  xy= s->block_index[i];
942  x= p->f.motion_val[0][xy][0];
943  y= p->f.motion_val[0][xy][1];
944  d+= cmp(s, x>>shift, y>>shift, x&mask, y&mask, 1, 8, i, i, cmpf, chroma_cmpf, flags);
945  }
947  }else{
948  if(USES_LIST(mb_type, 0)){
949  if(p_type){
950  *(uint32_t*)s->p_mv_table[mb_xy] = *(uint32_t*)p->f.motion_val[0][xy];
952  }else if(USES_LIST(mb_type, 1)){
953  *(uint32_t*)s->b_bidir_forw_mv_table[mb_xy] = *(uint32_t*)p->f.motion_val[0][xy];
954  *(uint32_t*)s->b_bidir_back_mv_table[mb_xy] = *(uint32_t*)p->f.motion_val[1][xy];
956  }else{
957  *(uint32_t*)s->b_forw_mv_table[mb_xy] = *(uint32_t*)p->f.motion_val[0][xy];
959  }
960  x = p->f.motion_val[0][xy][0];
961  y = p->f.motion_val[0][xy][1];
962  d = cmp(s, x>>shift, y>>shift, x&mask, y&mask, 0, 16, 0, 0, cmpf, chroma_cmpf, flags);
963  }else if(USES_LIST(mb_type, 1)){
964  *(uint32_t*)s->b_back_mv_table[mb_xy] = *(uint32_t*)p->f.motion_val[1][xy];
966 
967  x = p->f.motion_val[1][xy][0];
968  y = p->f.motion_val[1][xy][1];
969  d = cmp(s, x>>shift, y>>shift, x&mask, y&mask, 0, 16, 2, 0, cmpf, chroma_cmpf, flags);
970  }else
972  }
973  return d;
974 }
975 
977  int mb_x, int mb_y)
978 {
979  MotionEstContext * const c= &s->me;
980  uint8_t *pix, *ppix;
981  int sum, mx, my, dmin;
982  int varc;
983  int vard;
984  int P[10][2];
985  const int shift= 1+s->quarter_sample;
986  int mb_type=0;
987  Picture * const pic= &s->current_picture;
988 
989  init_ref(c, s->new_picture.f.data, s->last_picture.f.data, NULL, 16*mb_x, 16*mb_y, 0);
990 
991  assert(s->quarter_sample==0 || s->quarter_sample==1);
992  assert(s->linesize == c->stride);
993  assert(s->uvlinesize == c->uvstride);
994 
999 
1000  get_limits(s, 16*mb_x, 16*mb_y);
1001  c->skip=0;
1002 
1003  /* intra / predictive decision */
1004  pix = c->src[0][0];
1005  sum = s->dsp.pix_sum(pix, s->linesize);
1006  varc = s->dsp.pix_norm1(pix, s->linesize) - (((unsigned)sum*sum)>>8) + 500;
1007 
1008  pic->mb_mean[s->mb_stride * mb_y + mb_x] = (sum+128)>>8;
1009  pic->mb_var [s->mb_stride * mb_y + mb_x] = (varc+128)>>8;
1010  c->mb_var_sum_temp += (varc+128)>>8;
1011 
1012  if(c->avctx->me_threshold){
1013  vard= check_input_motion(s, mb_x, mb_y, 1);
1014 
1015  if((vard+128)>>8 < c->avctx->me_threshold){
1016  int p_score= FFMIN(vard, varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*100);
1017  int i_score= varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*20;
1018  pic->mc_mb_var[s->mb_stride * mb_y + mb_x] = (vard+128)>>8;
1019  c->mc_mb_var_sum_temp += (vard+128)>>8;
1020  c->scene_change_score+= ff_sqrt(p_score) - ff_sqrt(i_score);
1021  return;
1022  }
1023  if((vard+128)>>8 < c->avctx->mb_threshold)
1024  mb_type= s->mb_type[mb_x + mb_y*s->mb_stride];
1025  }
1026 
1027  switch(s->me_method) {
1028  case ME_ZERO:
1029  default:
1030  mx = 0;
1031  my = 0;
1032  dmin = 0;
1033  break;
1034  case ME_X1:
1035  case ME_EPZS:
1036  {
1037  const int mot_stride = s->b8_stride;
1038  const int mot_xy = s->block_index[0];
1039 
1040  P_LEFT[0] = s->current_picture.f.motion_val[0][mot_xy - 1][0];
1041  P_LEFT[1] = s->current_picture.f.motion_val[0][mot_xy - 1][1];
1042 
1043  if(P_LEFT[0] > (c->xmax<<shift)) P_LEFT[0] = (c->xmax<<shift);
1044 
1045  if(!s->first_slice_line) {
1046  P_TOP[0] = s->current_picture.f.motion_val[0][mot_xy - mot_stride ][0];
1047  P_TOP[1] = s->current_picture.f.motion_val[0][mot_xy - mot_stride ][1];
1048  P_TOPRIGHT[0] = s->current_picture.f.motion_val[0][mot_xy - mot_stride + 2][0];
1049  P_TOPRIGHT[1] = s->current_picture.f.motion_val[0][mot_xy - mot_stride + 2][1];
1050  if(P_TOP[1] > (c->ymax<<shift)) P_TOP[1] = (c->ymax<<shift);
1051  if(P_TOPRIGHT[0] < (c->xmin<<shift)) P_TOPRIGHT[0]= (c->xmin<<shift);
1052  if(P_TOPRIGHT[1] > (c->ymax<<shift)) P_TOPRIGHT[1]= (c->ymax<<shift);
1053 
1054  P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1055  P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1056 
1057  if(s->out_format == FMT_H263){
1058  c->pred_x = P_MEDIAN[0];
1059  c->pred_y = P_MEDIAN[1];
1060  }else { /* mpeg1 at least */
1061  c->pred_x= P_LEFT[0];
1062  c->pred_y= P_LEFT[1];
1063  }
1064  }else{
1065  c->pred_x= P_LEFT[0];
1066  c->pred_y= P_LEFT[1];
1067  }
1068 
1069  }
1070  dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, 0, s->p_mv_table, (1<<16)>>shift, 0, 16);
1071 
1072  break;
1073  }
1074 
1075  /* At this point (mx,my) are full-pell and the relative displacement */
1076  ppix = c->ref[0][0] + (my * s->linesize) + mx;
1077 
1078  vard = s->dsp.sse[0](NULL, pix, ppix, s->linesize, 16);
1079 
1080  pic->mc_mb_var[s->mb_stride * mb_y + mb_x] = (vard+128)>>8;
1081 // pic->mb_cmp_score[s->mb_stride * mb_y + mb_x] = dmin;
1082  c->mc_mb_var_sum_temp += (vard+128)>>8;
1083 
1084  if(mb_type){
1085  int p_score= FFMIN(vard, varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*100);
1086  int i_score= varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*20;
1087  c->scene_change_score+= ff_sqrt(p_score) - ff_sqrt(i_score);
1088 
1089  if(mb_type == CANDIDATE_MB_TYPE_INTER){
1090  c->sub_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1091  set_p_mv_tables(s, mx, my, 1);
1092  }else{
1093  mx <<=shift;
1094  my <<=shift;
1095  }
1096  if(mb_type == CANDIDATE_MB_TYPE_INTER4V){
1097  h263_mv4_search(s, mx, my, shift);
1098 
1099  set_p_mv_tables(s, mx, my, 0);
1100  }
1101  if(mb_type == CANDIDATE_MB_TYPE_INTER_I){
1102  interlaced_search(s, 0, s->p_field_mv_table, s->p_field_select_table, mx, my, 1);
1103  }
1104  }else if(c->avctx->mb_decision > FF_MB_DECISION_SIMPLE){
1105  int p_score= FFMIN(vard, varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*100);
1106  int i_score= varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*20;
1107  c->scene_change_score+= ff_sqrt(p_score) - ff_sqrt(i_score);
1108 
1109  if (vard*2 + 200*256 > varc)
1110  mb_type|= CANDIDATE_MB_TYPE_INTRA;
1111  if (varc*2 + 200*256 > vard || s->qscale > 24){
1112 // if (varc*2 + 200*256 + 50*(s->lambda2>>FF_LAMBDA_SHIFT) > vard){
1113  mb_type|= CANDIDATE_MB_TYPE_INTER;
1114  c->sub_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1115  if(s->flags&CODEC_FLAG_MV0)
1116  if(mx || my)
1117  mb_type |= CANDIDATE_MB_TYPE_SKIPPED; //FIXME check difference
1118  }else{
1119  mx <<=shift;
1120  my <<=shift;
1121  }
1122  if((s->flags&CODEC_FLAG_4MV)
1123  && !c->skip && varc>50<<8 && vard>10<<8){
1124  if(h263_mv4_search(s, mx, my, shift) < INT_MAX)
1125  mb_type|=CANDIDATE_MB_TYPE_INTER4V;
1126 
1127  set_p_mv_tables(s, mx, my, 0);
1128  }else
1129  set_p_mv_tables(s, mx, my, 1);
1131  && !c->skip){ //FIXME varc/d checks
1132  if(interlaced_search(s, 0, s->p_field_mv_table, s->p_field_select_table, mx, my, 0) < INT_MAX)
1133  mb_type |= CANDIDATE_MB_TYPE_INTER_I;
1134  }
1135  }else{
1136  int intra_score, i;
1137  mb_type= CANDIDATE_MB_TYPE_INTER;
1138 
1139  dmin= c->sub_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1140  if(c->avctx->me_sub_cmp != c->avctx->mb_cmp && !c->skip)
1141  dmin= get_mb_score(s, mx, my, 0, 0, 0, 16, 1);
1142 
1143  if((s->flags&CODEC_FLAG_4MV)
1144  && !c->skip && varc>50<<8 && vard>10<<8){
1145  int dmin4= h263_mv4_search(s, mx, my, shift);
1146  if(dmin4 < dmin){
1147  mb_type= CANDIDATE_MB_TYPE_INTER4V;
1148  dmin=dmin4;
1149  }
1150  }
1152  && !c->skip){ //FIXME varc/d checks
1153  int dmin_i= interlaced_search(s, 0, s->p_field_mv_table, s->p_field_select_table, mx, my, 0);
1154  if(dmin_i < dmin){
1155  mb_type = CANDIDATE_MB_TYPE_INTER_I;
1156  dmin= dmin_i;
1157  }
1158  }
1159 
1160 // pic->mb_cmp_score[s->mb_stride * mb_y + mb_x] = dmin;
1161  set_p_mv_tables(s, mx, my, mb_type!=CANDIDATE_MB_TYPE_INTER4V);
1162 
1163  /* get intra luma score */
1164  if((c->avctx->mb_cmp&0xFF)==FF_CMP_SSE){
1165  intra_score= varc - 500;
1166  }else{
1167  unsigned mean = (sum+128)>>8;
1168  mean*= 0x01010101;
1169 
1170  for(i=0; i<16; i++){
1171  *(uint32_t*)(&c->scratchpad[i*s->linesize+ 0]) = mean;
1172  *(uint32_t*)(&c->scratchpad[i*s->linesize+ 4]) = mean;
1173  *(uint32_t*)(&c->scratchpad[i*s->linesize+ 8]) = mean;
1174  *(uint32_t*)(&c->scratchpad[i*s->linesize+12]) = mean;
1175  }
1176 
1177  intra_score= s->dsp.mb_cmp[0](s, c->scratchpad, pix, s->linesize, 16);
1178  }
1179  intra_score += c->mb_penalty_factor*16;
1180 
1181  if(intra_score < dmin){
1182  mb_type= CANDIDATE_MB_TYPE_INTRA;
1183  s->current_picture.f.mb_type[mb_y*s->mb_stride + mb_x] = CANDIDATE_MB_TYPE_INTRA; //FIXME cleanup
1184  }else
1185  s->current_picture.f.mb_type[mb_y*s->mb_stride + mb_x] = 0;
1186 
1187  {
1188  int p_score= FFMIN(vard, varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*100);
1189  int i_score= varc-500+(s->lambda2>>FF_LAMBDA_SHIFT)*20;
1190  c->scene_change_score+= ff_sqrt(p_score) - ff_sqrt(i_score);
1191  }
1192  }
1193 
1194  s->mb_type[mb_y*s->mb_stride + mb_x]= mb_type;
1195 }
1196 
1198  int mb_x, int mb_y)
1199 {
1200  MotionEstContext * const c= &s->me;
1201  int mx, my, dmin;
1202  int P[10][2];
1203  const int shift= 1+s->quarter_sample;
1204  const int xy= mb_x + mb_y*s->mb_stride;
1205  init_ref(c, s->new_picture.f.data, s->last_picture.f.data, NULL, 16*mb_x, 16*mb_y, 0);
1206 
1207  assert(s->quarter_sample==0 || s->quarter_sample==1);
1208 
1211 
1212  get_limits(s, 16*mb_x, 16*mb_y);
1213  c->skip=0;
1214 
1215  P_LEFT[0] = s->p_mv_table[xy + 1][0];
1216  P_LEFT[1] = s->p_mv_table[xy + 1][1];
1217 
1218  if(P_LEFT[0] < (c->xmin<<shift)) P_LEFT[0] = (c->xmin<<shift);
1219 
1220  /* special case for first line */
1221  if (s->first_slice_line) {
1222  c->pred_x= P_LEFT[0];
1223  c->pred_y= P_LEFT[1];
1224  P_TOP[0]= P_TOPRIGHT[0]= P_MEDIAN[0]=
1225  P_TOP[1]= P_TOPRIGHT[1]= P_MEDIAN[1]= 0; //FIXME
1226  } else {
1227  P_TOP[0] = s->p_mv_table[xy + s->mb_stride ][0];
1228  P_TOP[1] = s->p_mv_table[xy + s->mb_stride ][1];
1229  P_TOPRIGHT[0] = s->p_mv_table[xy + s->mb_stride - 1][0];
1230  P_TOPRIGHT[1] = s->p_mv_table[xy + s->mb_stride - 1][1];
1231  if(P_TOP[1] < (c->ymin<<shift)) P_TOP[1] = (c->ymin<<shift);
1232  if(P_TOPRIGHT[0] > (c->xmax<<shift)) P_TOPRIGHT[0]= (c->xmax<<shift);
1233  if(P_TOPRIGHT[1] < (c->ymin<<shift)) P_TOPRIGHT[1]= (c->ymin<<shift);
1234 
1235  P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1236  P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1237 
1238  c->pred_x = P_MEDIAN[0];
1239  c->pred_y = P_MEDIAN[1];
1240  }
1241 
1242  dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, 0, s->p_mv_table, (1<<16)>>shift, 0, 16);
1243 
1244  s->p_mv_table[xy][0] = mx<<shift;
1245  s->p_mv_table[xy][1] = my<<shift;
1246 
1247  return dmin;
1248 }
1249 
1251  int mb_x, int mb_y, int16_t (*mv_table)[2], int ref_index, int f_code)
1252 {
1253  MotionEstContext * const c= &s->me;
1254  int mx, my, dmin;
1255  int P[10][2];
1256  const int shift= 1+s->quarter_sample;
1257  const int mot_stride = s->mb_stride;
1258  const int mot_xy = mb_y*mot_stride + mb_x;
1259  uint8_t * const mv_penalty= c->mv_penalty[f_code] + MAX_MV;
1260  int mv_scale;
1261 
1266 
1267  get_limits(s, 16*mb_x, 16*mb_y);
1268 
1269  switch(s->me_method) {
1270  case ME_ZERO:
1271  default:
1272  mx = 0;
1273  my = 0;
1274  dmin = 0;
1275  break;
1276  case ME_X1:
1277  case ME_EPZS:
1278  P_LEFT[0] = mv_table[mot_xy - 1][0];
1279  P_LEFT[1] = mv_table[mot_xy - 1][1];
1280 
1281  if (P_LEFT[0] > (c->xmax << shift)) P_LEFT[0] = (c->xmax << shift);
1282 
1283  /* special case for first line */
1284  if (!s->first_slice_line) {
1285  P_TOP[0] = mv_table[mot_xy - mot_stride ][0];
1286  P_TOP[1] = mv_table[mot_xy - mot_stride ][1];
1287  P_TOPRIGHT[0] = mv_table[mot_xy - mot_stride + 1][0];
1288  P_TOPRIGHT[1] = mv_table[mot_xy - mot_stride + 1][1];
1289  if (P_TOP[1] > (c->ymax << shift)) P_TOP[1] = (c->ymax << shift);
1290  if (P_TOPRIGHT[0] < (c->xmin << shift)) P_TOPRIGHT[0] = (c->xmin << shift);
1291  if (P_TOPRIGHT[1] > (c->ymax << shift)) P_TOPRIGHT[1] = (c->ymax << shift);
1292 
1293  P_MEDIAN[0] = mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1294  P_MEDIAN[1] = mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1295  }
1296  c->pred_x = P_LEFT[0];
1297  c->pred_y = P_LEFT[1];
1298 
1299  if(mv_table == s->b_forw_mv_table){
1300  mv_scale= (s->pb_time<<16) / (s->pp_time<<shift);
1301  }else{
1302  mv_scale= ((s->pb_time - s->pp_time)<<16) / (s->pp_time<<shift);
1303  }
1304 
1305  dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, ref_index, s->p_mv_table, mv_scale, 0, 16);
1306 
1307  break;
1308  }
1309 
1310  dmin= c->sub_motion_search(s, &mx, &my, dmin, 0, ref_index, 0, 16);
1311 
1312  if(c->avctx->me_sub_cmp != c->avctx->mb_cmp && !c->skip)
1313  dmin= get_mb_score(s, mx, my, 0, ref_index, 0, 16, 1);
1314 
1315 // s->mb_type[mb_y*s->mb_width + mb_x]= mb_type;
1316  mv_table[mot_xy][0]= mx;
1317  mv_table[mot_xy][1]= my;
1318 
1319  return dmin;
1320 }
1321 
1322 static inline int check_bidir_mv(MpegEncContext * s,
1323  int motion_fx, int motion_fy,
1324  int motion_bx, int motion_by,
1325  int pred_fx, int pred_fy,
1326  int pred_bx, int pred_by,
1327  int size, int h)
1328 {
1329  //FIXME optimize?
1330  //FIXME better f_code prediction (max mv & distance)
1331  //FIXME pointers
1332  MotionEstContext * const c= &s->me;
1333  uint8_t * const mv_penalty_f= c->mv_penalty[s->f_code] + MAX_MV; // f_code of the prev frame
1334  uint8_t * const mv_penalty_b= c->mv_penalty[s->b_code] + MAX_MV; // f_code of the prev frame
1335  int stride= c->stride;
1336  uint8_t *dest_y = c->scratchpad;
1337  uint8_t *ptr;
1338  int dxy;
1339  int src_x, src_y;
1340  int fbmin;
1341  uint8_t **src_data= c->src[0];
1342  uint8_t **ref_data= c->ref[0];
1343  uint8_t **ref2_data= c->ref[2];
1344 
1345  if(s->quarter_sample){
1346  dxy = ((motion_fy & 3) << 2) | (motion_fx & 3);
1347  src_x = motion_fx >> 2;
1348  src_y = motion_fy >> 2;
1349 
1350  ptr = ref_data[0] + (src_y * stride) + src_x;
1351  s->dsp.put_qpel_pixels_tab[0][dxy](dest_y , ptr , stride);
1352 
1353  dxy = ((motion_by & 3) << 2) | (motion_bx & 3);
1354  src_x = motion_bx >> 2;
1355  src_y = motion_by >> 2;
1356 
1357  ptr = ref2_data[0] + (src_y * stride) + src_x;
1358  s->dsp.avg_qpel_pixels_tab[size][dxy](dest_y , ptr , stride);
1359  }else{
1360  dxy = ((motion_fy & 1) << 1) | (motion_fx & 1);
1361  src_x = motion_fx >> 1;
1362  src_y = motion_fy >> 1;
1363 
1364  ptr = ref_data[0] + (src_y * stride) + src_x;
1365  s->dsp.put_pixels_tab[size][dxy](dest_y , ptr , stride, h);
1366 
1367  dxy = ((motion_by & 1) << 1) | (motion_bx & 1);
1368  src_x = motion_bx >> 1;
1369  src_y = motion_by >> 1;
1370 
1371  ptr = ref2_data[0] + (src_y * stride) + src_x;
1372  s->dsp.avg_pixels_tab[size][dxy](dest_y , ptr , stride, h);
1373  }
1374 
1375  fbmin = (mv_penalty_f[motion_fx-pred_fx] + mv_penalty_f[motion_fy-pred_fy])*c->mb_penalty_factor
1376  +(mv_penalty_b[motion_bx-pred_bx] + mv_penalty_b[motion_by-pred_by])*c->mb_penalty_factor
1377  + s->dsp.mb_cmp[size](s, src_data[0], dest_y, stride, h); //FIXME new_pic
1378 
1379  if(c->avctx->mb_cmp&FF_CMP_CHROMA){
1380  }
1381  //FIXME CHROMA !!!
1382 
1383  return fbmin;
1384 }
1385 
1386 /* refine the bidir vectors in hq mode and return the score in both lq & hq mode*/
1387 static inline int bidir_refine(MpegEncContext * s, int mb_x, int mb_y)
1388 {
1389  MotionEstContext * const c= &s->me;
1390  const int mot_stride = s->mb_stride;
1391  const int xy = mb_y *mot_stride + mb_x;
1392  int fbmin;
1393  int pred_fx= s->b_bidir_forw_mv_table[xy-1][0];
1394  int pred_fy= s->b_bidir_forw_mv_table[xy-1][1];
1395  int pred_bx= s->b_bidir_back_mv_table[xy-1][0];
1396  int pred_by= s->b_bidir_back_mv_table[xy-1][1];
1397  int motion_fx= s->b_bidir_forw_mv_table[xy][0]= s->b_forw_mv_table[xy][0];
1398  int motion_fy= s->b_bidir_forw_mv_table[xy][1]= s->b_forw_mv_table[xy][1];
1399  int motion_bx= s->b_bidir_back_mv_table[xy][0]= s->b_back_mv_table[xy][0];
1400  int motion_by= s->b_bidir_back_mv_table[xy][1]= s->b_back_mv_table[xy][1];
1401  const int flags= c->sub_flags;
1402  const int qpel= flags&FLAG_QPEL;
1403  const int shift= 1+qpel;
1404  const int xmin= c->xmin<<shift;
1405  const int ymin= c->ymin<<shift;
1406  const int xmax= c->xmax<<shift;
1407  const int ymax= c->ymax<<shift;
1408 #define HASH(fx,fy,bx,by) ((fx)+17*(fy)+63*(bx)+117*(by))
1409 #define HASH8(fx,fy,bx,by) ((uint8_t)HASH(fx,fy,bx,by))
1410  int hashidx= HASH(motion_fx,motion_fy, motion_bx, motion_by);
1411  uint8_t map[256] = { 0 };
1412 
1413  map[hashidx&255] = 1;
1414 
1415  fbmin= check_bidir_mv(s, motion_fx, motion_fy,
1416  motion_bx, motion_by,
1417  pred_fx, pred_fy,
1418  pred_bx, pred_by,
1419  0, 16);
1420 
1421  if(s->avctx->bidir_refine){
1422  int end;
1423  static const uint8_t limittab[5]={0,8,32,64,80};
1424  const int limit= limittab[s->avctx->bidir_refine];
1425  static const int8_t vect[][4]={
1426 { 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},
1427 
1428 { 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},
1429 { 0, 1, 0, 1}, { 0,-1, 0,-1}, { 1, 0, 1, 0}, {-1, 0,-1, 0},
1430 { 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},
1431 { 0,-1, 0, 1}, { 0, 1, 0,-1}, {-1, 0, 1, 0}, { 1, 0,-1, 0},
1432 
1433 { 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},
1434 { 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},
1435 { 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},
1436 { 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},
1437 
1438 { 1, 1, 1, 1}, {-1,-1,-1,-1},
1439 { 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},
1440 { 1, 1,-1,-1}, {-1,-1, 1, 1}, { 1,-1,-1, 1}, {-1, 1, 1,-1}, { 1,-1, 1,-1}, {-1, 1,-1, 1},
1441  };
1442  static const uint8_t hash[]={
1443 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),
1444 
1445 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),
1446 HASH8( 0, 1, 0, 1), HASH8( 0,-1, 0,-1), HASH8( 1, 0, 1, 0), HASH8(-1, 0,-1, 0),
1447 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),
1448 HASH8( 0,-1, 0, 1), HASH8( 0, 1, 0,-1), HASH8(-1, 0, 1, 0), HASH8( 1, 0,-1, 0),
1449 
1450 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),
1451 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),
1452 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),
1453 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),
1454 
1455 HASH8( 1, 1, 1, 1), HASH8(-1,-1,-1,-1),
1456 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),
1457 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),
1458 };
1459 
1460 #define CHECK_BIDIR(fx,fy,bx,by)\
1461  if( !map[(hashidx+HASH(fx,fy,bx,by))&255]\
1462  &&(fx<=0 || motion_fx+fx<=xmax) && (fy<=0 || motion_fy+fy<=ymax) && (bx<=0 || motion_bx+bx<=xmax) && (by<=0 || motion_by+by<=ymax)\
1463  &&(fx>=0 || motion_fx+fx>=xmin) && (fy>=0 || motion_fy+fy>=ymin) && (bx>=0 || motion_bx+bx>=xmin) && (by>=0 || motion_by+by>=ymin)){\
1464  int score;\
1465  map[(hashidx+HASH(fx,fy,bx,by))&255] = 1;\
1466  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);\
1467  if(score < fbmin){\
1468  hashidx += HASH(fx,fy,bx,by);\
1469  fbmin= score;\
1470  motion_fx+=fx;\
1471  motion_fy+=fy;\
1472  motion_bx+=bx;\
1473  motion_by+=by;\
1474  end=0;\
1475  }\
1476  }
1477 #define CHECK_BIDIR2(a,b,c,d)\
1478 CHECK_BIDIR(a,b,c,d)\
1479 CHECK_BIDIR(-(a),-(b),-(c),-(d))
1480 
1481  do{
1482  int i;
1483  int borderdist=0;
1484  end=1;
1485 
1486  CHECK_BIDIR2(0,0,0,1)
1487  CHECK_BIDIR2(0,0,1,0)
1488  CHECK_BIDIR2(0,1,0,0)
1489  CHECK_BIDIR2(1,0,0,0)
1490 
1491  for(i=8; i<limit; i++){
1492  int fx= motion_fx+vect[i][0];
1493  int fy= motion_fy+vect[i][1];
1494  int bx= motion_bx+vect[i][2];
1495  int by= motion_by+vect[i][3];
1496  if(borderdist<=0){
1497  int a= (xmax - FFMAX(fx,bx))|(FFMIN(fx,bx) - xmin);
1498  int b= (ymax - FFMAX(fy,by))|(FFMIN(fy,by) - ymin);
1499  if((a|b) < 0)
1500  map[(hashidx+hash[i])&255] = 1;
1501  }
1502  if(!map[(hashidx+hash[i])&255]){
1503  int score;
1504  map[(hashidx+hash[i])&255] = 1;
1505  score= check_bidir_mv(s, fx, fy, bx, by, pred_fx, pred_fy, pred_bx, pred_by, 0, 16);
1506  if(score < fbmin){
1507  hashidx += hash[i];
1508  fbmin= score;
1509  motion_fx=fx;
1510  motion_fy=fy;
1511  motion_bx=bx;
1512  motion_by=by;
1513  end=0;
1514  borderdist--;
1515  if(borderdist<=0){
1516  int a= FFMIN(xmax - FFMAX(fx,bx), FFMIN(fx,bx) - xmin);
1517  int b= FFMIN(ymax - FFMAX(fy,by), FFMIN(fy,by) - ymin);
1518  borderdist= FFMIN(a,b);
1519  }
1520  }
1521  }
1522  }
1523  }while(!end);
1524  }
1525 
1526  s->b_bidir_forw_mv_table[xy][0]= motion_fx;
1527  s->b_bidir_forw_mv_table[xy][1]= motion_fy;
1528  s->b_bidir_back_mv_table[xy][0]= motion_bx;
1529  s->b_bidir_back_mv_table[xy][1]= motion_by;
1530 
1531  return fbmin;
1532 }
1533 
1534 static inline int direct_search(MpegEncContext * s, int mb_x, int mb_y)
1535 {
1536  MotionEstContext * const c= &s->me;
1537  int P[10][2];
1538  const int mot_stride = s->mb_stride;
1539  const int mot_xy = mb_y*mot_stride + mb_x;
1540  const int shift= 1+s->quarter_sample;
1541  int dmin, i;
1542  const int time_pp= s->pp_time;
1543  const int time_pb= s->pb_time;
1544  int mx, my, xmin, xmax, ymin, ymax;
1545  int16_t (*mv_table)[2]= s->b_direct_mv_table;
1546 
1547  c->current_mv_penalty= c->mv_penalty[1] + MAX_MV;
1548  ymin= xmin=(-32)>>shift;
1549  ymax= xmax= 31>>shift;
1550 
1551  if (IS_8X8(s->next_picture.f.mb_type[mot_xy])) {
1552  s->mv_type= MV_TYPE_8X8;
1553  }else{
1554  s->mv_type= MV_TYPE_16X16;
1555  }
1556 
1557  for(i=0; i<4; i++){
1558  int index= s->block_index[i];
1559  int min, max;
1560 
1561  c->co_located_mv[i][0] = s->next_picture.f.motion_val[0][index][0];
1562  c->co_located_mv[i][1] = s->next_picture.f.motion_val[0][index][1];
1563  c->direct_basis_mv[i][0]= c->co_located_mv[i][0]*time_pb/time_pp + ((i& 1)<<(shift+3));
1564  c->direct_basis_mv[i][1]= c->co_located_mv[i][1]*time_pb/time_pp + ((i>>1)<<(shift+3));
1565 // c->direct_basis_mv[1][i][0]= c->co_located_mv[i][0]*(time_pb - time_pp)/time_pp + ((i &1)<<(shift+3);
1566 // c->direct_basis_mv[1][i][1]= c->co_located_mv[i][1]*(time_pb - time_pp)/time_pp + ((i>>1)<<(shift+3);
1567 
1568  max= FFMAX(c->direct_basis_mv[i][0], c->direct_basis_mv[i][0] - c->co_located_mv[i][0])>>shift;
1569  min= FFMIN(c->direct_basis_mv[i][0], c->direct_basis_mv[i][0] - c->co_located_mv[i][0])>>shift;
1570  max+= 16*mb_x + 1; // +-1 is for the simpler rounding
1571  min+= 16*mb_x - 1;
1572  xmax= FFMIN(xmax, s->width - max);
1573  xmin= FFMAX(xmin, - 16 - min);
1574 
1575  max= FFMAX(c->direct_basis_mv[i][1], c->direct_basis_mv[i][1] - c->co_located_mv[i][1])>>shift;
1576  min= FFMIN(c->direct_basis_mv[i][1], c->direct_basis_mv[i][1] - c->co_located_mv[i][1])>>shift;
1577  max+= 16*mb_y + 1; // +-1 is for the simpler rounding
1578  min+= 16*mb_y - 1;
1579  ymax= FFMIN(ymax, s->height - max);
1580  ymin= FFMAX(ymin, - 16 - min);
1581 
1582  if(s->mv_type == MV_TYPE_16X16) break;
1583  }
1584 
1585  assert(xmax <= 15 && ymax <= 15 && xmin >= -16 && ymin >= -16);
1586 
1587  if(xmax < 0 || xmin >0 || ymax < 0 || ymin > 0){
1588  s->b_direct_mv_table[mot_xy][0]= 0;
1589  s->b_direct_mv_table[mot_xy][1]= 0;
1590 
1591  return 256*256*256*64;
1592  }
1593 
1594  c->xmin= xmin;
1595  c->ymin= ymin;
1596  c->xmax= xmax;
1597  c->ymax= ymax;
1598  c->flags |= FLAG_DIRECT;
1599  c->sub_flags |= FLAG_DIRECT;
1600  c->pred_x=0;
1601  c->pred_y=0;
1602 
1603  P_LEFT[0] = av_clip(mv_table[mot_xy - 1][0], xmin<<shift, xmax<<shift);
1604  P_LEFT[1] = av_clip(mv_table[mot_xy - 1][1], ymin<<shift, ymax<<shift);
1605 
1606  /* special case for first line */
1607  if (!s->first_slice_line) { //FIXME maybe allow this over thread boundary as it is clipped
1608  P_TOP[0] = av_clip(mv_table[mot_xy - mot_stride ][0], xmin<<shift, xmax<<shift);
1609  P_TOP[1] = av_clip(mv_table[mot_xy - mot_stride ][1], ymin<<shift, ymax<<shift);
1610  P_TOPRIGHT[0] = av_clip(mv_table[mot_xy - mot_stride + 1 ][0], xmin<<shift, xmax<<shift);
1611  P_TOPRIGHT[1] = av_clip(mv_table[mot_xy - mot_stride + 1 ][1], ymin<<shift, ymax<<shift);
1612 
1613  P_MEDIAN[0]= mid_pred(P_LEFT[0], P_TOP[0], P_TOPRIGHT[0]);
1614  P_MEDIAN[1]= mid_pred(P_LEFT[1], P_TOP[1], P_TOPRIGHT[1]);
1615  }
1616 
1617  dmin = ff_epzs_motion_search(s, &mx, &my, P, 0, 0, mv_table, 1<<(16-shift), 0, 16);
1618  if(c->sub_flags&FLAG_QPEL)
1619  dmin = qpel_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1620  else
1621  dmin = hpel_motion_search(s, &mx, &my, dmin, 0, 0, 0, 16);
1622 
1623  if(c->avctx->me_sub_cmp != c->avctx->mb_cmp && !c->skip)
1624  dmin= get_mb_score(s, mx, my, 0, 0, 0, 16, 1);
1625 
1626  get_limits(s, 16*mb_x, 16*mb_y); //restore c->?min/max, maybe not needed
1627 
1628  mv_table[mot_xy][0]= mx;
1629  mv_table[mot_xy][1]= my;
1630  c->flags &= ~FLAG_DIRECT;
1631  c->sub_flags &= ~FLAG_DIRECT;
1632 
1633  return dmin;
1634 }
1635 
1637  int mb_x, int mb_y)
1638 {
1639  MotionEstContext * const c= &s->me;
1640  const int penalty_factor= c->mb_penalty_factor;
1641  int fmin, bmin, dmin, fbmin, bimin, fimin;
1642  int type=0;
1643  const int xy = mb_y*s->mb_stride + mb_x;
1645  s->next_picture.f.data, 16 * mb_x, 16 * mb_y, 2);
1646 
1647  get_limits(s, 16*mb_x, 16*mb_y);
1648 
1649  c->skip=0;
1650 
1651  if (s->codec_id == AV_CODEC_ID_MPEG4 && s->next_picture.f.mbskip_table[xy]) {
1652  int score= direct_search(s, mb_x, mb_y); //FIXME just check 0,0
1653 
1654  score= ((unsigned)(score*score + 128*256))>>16;
1655  c->mc_mb_var_sum_temp += score;
1656  s->current_picture.mc_mb_var[mb_y*s->mb_stride + mb_x] = score; //FIXME use SSE
1657  s->mb_type[mb_y*s->mb_stride + mb_x]= CANDIDATE_MB_TYPE_DIRECT0;
1658 
1659  return;
1660  }
1661 
1662  if(c->avctx->me_threshold){
1663  int vard= check_input_motion(s, mb_x, mb_y, 0);
1664 
1665  if((vard+128)>>8 < c->avctx->me_threshold){
1666 // pix = c->src[0][0];
1667 // sum = s->dsp.pix_sum(pix, s->linesize);
1668 // varc = s->dsp.pix_norm1(pix, s->linesize) - (((unsigned)(sum*sum))>>8) + 500;
1669 
1670 // pic->mb_var [s->mb_stride * mb_y + mb_x] = (varc+128)>>8;
1671  s->current_picture.mc_mb_var[s->mb_stride * mb_y + mb_x] = (vard+128)>>8;
1672 /* pic->mb_mean [s->mb_stride * mb_y + mb_x] = (sum+128)>>8;
1673  c->mb_var_sum_temp += (varc+128)>>8;*/
1674  c->mc_mb_var_sum_temp += (vard+128)>>8;
1675 /* if (vard <= 64<<8 || vard < varc) {
1676  c->scene_change_score+= ff_sqrt(vard) - ff_sqrt(varc);
1677  }else{
1678  c->scene_change_score+= s->qscale * s->avctx->scenechange_factor;
1679  }*/
1680  return;
1681  }
1682  if((vard+128)>>8 < c->avctx->mb_threshold){
1683  type= s->mb_type[mb_y*s->mb_stride + mb_x];
1684  if(type == CANDIDATE_MB_TYPE_DIRECT){
1685  direct_search(s, mb_x, mb_y);
1686  }
1687  if(type == CANDIDATE_MB_TYPE_FORWARD || type == CANDIDATE_MB_TYPE_BIDIR){
1688  c->skip=0;
1689  ff_estimate_motion_b(s, mb_x, mb_y, s->b_forw_mv_table, 0, s->f_code);
1690  }
1691  if(type == CANDIDATE_MB_TYPE_BACKWARD || type == CANDIDATE_MB_TYPE_BIDIR){
1692  c->skip=0;
1693  ff_estimate_motion_b(s, mb_x, mb_y, s->b_back_mv_table, 2, s->b_code);
1694  }
1696  c->skip=0;
1698  interlaced_search(s, 0,
1700  s->b_forw_mv_table[xy][0], s->b_forw_mv_table[xy][1], 1);
1701  }
1703  c->skip=0;
1705  interlaced_search(s, 2,
1707  s->b_back_mv_table[xy][0], s->b_back_mv_table[xy][1], 1);
1708  }
1709  return;
1710  }
1711  }
1712 
1713  if (s->codec_id == AV_CODEC_ID_MPEG4)
1714  dmin= direct_search(s, mb_x, mb_y);
1715  else
1716  dmin= INT_MAX;
1717 //FIXME penalty stuff for non mpeg4
1718  c->skip=0;
1719  fmin= ff_estimate_motion_b(s, mb_x, mb_y, s->b_forw_mv_table, 0, s->f_code) + 3*penalty_factor;
1720 
1721  c->skip=0;
1722  bmin= ff_estimate_motion_b(s, mb_x, mb_y, s->b_back_mv_table, 2, s->b_code) + 2*penalty_factor;
1723  av_dlog(s, " %d %d ", s->b_forw_mv_table[xy][0], s->b_forw_mv_table[xy][1]);
1724 
1725  c->skip=0;
1726  fbmin= bidir_refine(s, mb_x, mb_y) + penalty_factor;
1727  av_dlog(s, "%d %d %d %d\n", dmin, fmin, bmin, fbmin);
1728 
1730 //FIXME mb type penalty
1731  c->skip=0;
1733  fimin= interlaced_search(s, 0,
1735  s->b_forw_mv_table[xy][0], s->b_forw_mv_table[xy][1], 0);
1737  bimin= interlaced_search(s, 2,
1739  s->b_back_mv_table[xy][0], s->b_back_mv_table[xy][1], 0);
1740  }else
1741  fimin= bimin= INT_MAX;
1742 
1743  {
1744  int score= fmin;
1746 
1747  if (dmin <= score){
1748  score = dmin;
1749  type = CANDIDATE_MB_TYPE_DIRECT;
1750  }
1751  if(bmin<score){
1752  score=bmin;
1754  }
1755  if(fbmin<score){
1756  score=fbmin;
1758  }
1759  if(fimin<score){
1760  score=fimin;
1762  }
1763  if(bimin<score){
1764  score=bimin;
1766  }
1767 
1768  score= ((unsigned)(score*score + 128*256))>>16;
1769  c->mc_mb_var_sum_temp += score;
1770  s->current_picture.mc_mb_var[mb_y*s->mb_stride + mb_x] = score; //FIXME use SSE
1771  }
1772 
1775  if(fimin < INT_MAX)
1777  if(bimin < INT_MAX)
1779  if(fimin < INT_MAX && bimin < INT_MAX){
1780  type |= CANDIDATE_MB_TYPE_BIDIR_I;
1781  }
1782  //FIXME something smarter
1783  if(dmin>256*256*16) type&= ~CANDIDATE_MB_TYPE_DIRECT; //do not try direct mode if it is invalid for this MB
1784  if(s->codec_id == AV_CODEC_ID_MPEG4 && type&CANDIDATE_MB_TYPE_DIRECT && s->flags&CODEC_FLAG_MV0 && *(uint32_t*)s->b_direct_mv_table[xy])
1785  type |= CANDIDATE_MB_TYPE_DIRECT0;
1786  }
1787 
1788  s->mb_type[mb_y*s->mb_stride + mb_x]= type;
1789 }
1790 
1791 /* find best f_code for ME which do unlimited searches */
1792 int ff_get_best_fcode(MpegEncContext * s, int16_t (*mv_table)[2], int type)
1793 {
1794  if(s->me_method>=ME_EPZS){
1795  int score[8];
1796  int i, y, range= s->avctx->me_range ? s->avctx->me_range : (INT_MAX/2);
1797  uint8_t * fcode_tab= s->fcode_tab;
1798  int best_fcode=-1;
1799  int best_score=-10000000;
1800 
1801  if(s->msmpeg4_version)
1802  range= FFMIN(range, 16);
1804  range= FFMIN(range, 256);
1805 
1806  for(i=0; i<8; i++) score[i]= s->mb_num*(8-i);
1807 
1808  for(y=0; y<s->mb_height; y++){
1809  int x;
1810  int xy= y*s->mb_stride;
1811  for(x=0; x<s->mb_width; x++){
1812  if(s->mb_type[xy] & type){
1813  int mx= mv_table[xy][0];
1814  int my= mv_table[xy][1];
1815  int fcode= FFMAX(fcode_tab[mx + MAX_MV],
1816  fcode_tab[my + MAX_MV]);
1817  int j;
1818 
1819  if(mx >= range || mx < -range ||
1820  my >= range || my < -range)
1821  continue;
1822 
1823  for(j=0; j<fcode && j<8; j++){
1825  score[j]-= 170;
1826  }
1827  }
1828  xy++;
1829  }
1830  }
1831 
1832  for(i=1; i<8; i++){
1833  if(score[i] > best_score){
1834  best_score= score[i];
1835  best_fcode= i;
1836  }
1837  }
1838 
1839  return best_fcode;
1840  }else{
1841  return 1;
1842  }
1843 }
1844 
1846 {
1847  MotionEstContext * const c= &s->me;
1848  const int f_code= s->f_code;
1849  int y, range;
1850  assert(s->pict_type==AV_PICTURE_TYPE_P);
1851 
1852  range = (((s->out_format == FMT_MPEG1 || s->msmpeg4_version) ? 8 : 16) << f_code);
1853 
1854  assert(range <= 16 || !s->msmpeg4_version);
1855  assert(range <=256 || !(s->codec_id == AV_CODEC_ID_MPEG2VIDEO && s->avctx->strict_std_compliance >= FF_COMPLIANCE_NORMAL));
1856 
1857  if(c->avctx->me_range && range > c->avctx->me_range) range= c->avctx->me_range;
1858 
1859  if(s->flags&CODEC_FLAG_4MV){
1860  const int wrap= s->b8_stride;
1861 
1862  /* clip / convert to intra 8x8 type MVs */
1863  for(y=0; y<s->mb_height; y++){
1864  int xy= y*2*wrap;
1865  int i= y*s->mb_stride;
1866  int x;
1867 
1868  for(x=0; x<s->mb_width; x++){
1870  int block;
1871  for(block=0; block<4; block++){
1872  int off= (block& 1) + (block>>1)*wrap;
1873  int mx = s->current_picture.f.motion_val[0][ xy + off ][0];
1874  int my = s->current_picture.f.motion_val[0][ xy + off ][1];
1875 
1876  if( mx >=range || mx <-range
1877  || my >=range || my <-range){
1878  s->mb_type[i] &= ~CANDIDATE_MB_TYPE_INTER4V;
1881  }
1882  }
1883  }
1884  xy+=2;
1885  i++;
1886  }
1887  }
1888  }
1889 }
1890 
1895 void ff_fix_long_mvs(MpegEncContext * s, uint8_t *field_select_table, int field_select,
1896  int16_t (*mv_table)[2], int f_code, int type, int truncate)
1897 {
1898  MotionEstContext * const c= &s->me;
1899  int y, h_range, v_range;
1900 
1901  // RAL: 8 in MPEG-1, 16 in MPEG-4
1902  int range = (((s->out_format == FMT_MPEG1 || s->msmpeg4_version) ? 8 : 16) << f_code);
1903 
1904  if(c->avctx->me_range && range > c->avctx->me_range) range= c->avctx->me_range;
1905 
1906  h_range= range;
1907  v_range= field_select_table ? range>>1 : range;
1908 
1909  /* clip / convert to intra 16x16 type MVs */
1910  for(y=0; y<s->mb_height; y++){
1911  int x;
1912  int xy= y*s->mb_stride;
1913  for(x=0; x<s->mb_width; x++){
1914  if (s->mb_type[xy] & type){ // RAL: "type" test added...
1915  if(field_select_table==NULL || field_select_table[xy] == field_select){
1916  if( mv_table[xy][0] >=h_range || mv_table[xy][0] <-h_range
1917  || mv_table[xy][1] >=v_range || mv_table[xy][1] <-v_range){
1918 
1919  if(truncate){
1920  if (mv_table[xy][0] > h_range-1) mv_table[xy][0]= h_range-1;
1921  else if(mv_table[xy][0] < -h_range ) mv_table[xy][0]= -h_range;
1922  if (mv_table[xy][1] > v_range-1) mv_table[xy][1]= v_range-1;
1923  else if(mv_table[xy][1] < -v_range ) mv_table[xy][1]= -v_range;
1924  }else{
1925  s->mb_type[xy] &= ~type;
1926  s->mb_type[xy] |= CANDIDATE_MB_TYPE_INTRA;
1927  mv_table[xy][0]=
1928  mv_table[xy][1]= 0;
1929  }
1930  }
1931  }
1932  }
1933  xy++;
1934  }
1935  }
1936 }
uint8_t * scratchpad
data area for the ME algo, so that the ME does not need to malloc/free
Definition: mpegvideo.h:160
static int minima_cmp(const void *a, const void *b)
Definition: motion_est.c:70
static int cmp_qpel(MpegEncContext *s, const int x, const int y, const int subx, const int suby, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags)
Definition: motion_est.c:277
static unsigned update_map_generation(MotionEstContext *c)
Definition: motion_est.c:53
op_pixels_func put_pixels_tab[4][4]
Halfpel motion compensation with rounding (a+b+1)>>1.
Definition: dsputil.h:259
void ff_estimate_b_frame_motion(MpegEncContext *s, int mb_x, int mb_y)
Definition: motion_est.c:1636
static int epzs_motion_search2(MpegEncContext *s, int *mx_ptr, int *my_ptr, int P[10][2], int src_index, int ref_index, int16_t(*last_mv)[2], int ref_mv_scale)
int size
#define P_TOPRIGHT
Definition: motion_est.c:44
static int check_bidir_mv(MpegEncContext *s, int motion_fx, int motion_fy, int motion_bx, int motion_by, int pred_fx, int pred_fy, int pred_bx, int pred_by, int size, int h)
Definition: motion_est.c:1322
int skip
set if ME is skipped for the current MB
Definition: mpegvideo.h:157
int16_t(* p_mv_table)[2]
MV table (1MV per MB) p-frame encoding.
Definition: mpegvideo.h:373
#define ME_MAP_SHIFT
Definition: mpegvideo.h:66
static void get_limits(MpegEncContext *s, int x, int y)
get fullpel ME search limits.
Definition: motion_est.c:515
uint8_t * fcode_tab
smallest fcode needed for each MV
Definition: mpegvideo.h:402
void ff_estimate_p_frame_motion(MpegEncContext *s, int mb_x, int mb_y)
Definition: motion_est.c:976
uint8_t * mb_mean
Table for MB luminance.
Definition: mpegvideo.h:145
static void init_mv4_ref(MotionEstContext *c)
Definition: motion_est.c:548
#define CANDIDATE_MB_TYPE_BACKWARD_I
Definition: mpegvideo.h:428
#define P_LEFT
Definition: motion_est.c:42
uint16_t * mb_var
Table for MB variances.
Definition: mpegvideo.h:143
#define CANDIDATE_MB_TYPE_BIDIR
Definition: mpegvideo.h:424
uint8_t * current_mv_penalty
Definition: mpegvideo.h:201
static int hash(int head, const int add)
Hash function adding character.
Definition: lzwenc.c:74
#define CANDIDATE_MB_TYPE_INTRA
Definition: mpegvideo.h:415
int msmpeg4_version
0=not msmpeg4, 1=mp41, 2=mp42, 3=mp43/divx3 4=wmv1/7 5=wmv2/8
Definition: mpegvideo.h:616
int(* me_cmp_func)(void *s, uint8_t *blk1, uint8_t *blk2, int line_size, int h)
Definition: dsputil.h:176
enum AVCodecID codec_id
Definition: mpegvideo.h:227
static int ff_estimate_motion_b(MpegEncContext *s, int mb_x, int mb_y, int16_t(*mv_table)[2], int ref_index, int f_code)
Definition: motion_est.c:1250
enhanced predictive zonal search
Definition: avcodec.h:517
static int cmp_simple(MpegEncContext *s, const int x, const int y, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func)
Definition: motion_est.c:223
int sub_penalty_factor
Definition: mpegvideo.h:174
me_cmp_func sse[6]
Definition: dsputil.h:225
int16_t(*[2][2] p_field_mv_table)[2]
MV table (2MV per MB) interlaced p-frame encoding.
Definition: mpegvideo.h:379
static int bidir_refine(MpegEncContext *s, int mb_x, int mb_y)
Definition: motion_est.c:1387
#define wrap(func)
Definition: w64xmmtest.h:70
mpegvideo header.
int pre_penalty_factor
Definition: mpegvideo.h:168
av_dlog(ac->avr,"%d samples - audio_convert: %s to %s (%s)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt), use_generic?ac->func_descr_generic:ac->func_descr)
int scene_change_score
Definition: mpegvideo.h:194
int mb_num
number of MBs of a picture
Definition: mpegvideo.h:252
int stride
Definition: mace.c:144
int qscale
QP.
Definition: mpegvideo.h:342
int16_t(* b_back_mv_table)[2]
MV table (1MV per MB) backward mode b-frame encoding.
Definition: mpegvideo.h:375
#define P_MV1
Definition: motion_est.c:46
uint8_t * ref[4][4]
Definition: mpegvideo.h:188
int(* pix_sum)(uint8_t *pix, int line_size)
Definition: dsputil.h:220
static int cmp_internal(MpegEncContext *s, const int x, const int y, const int subx, const int suby, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags)
Definition: motion_est.c:239
#define USES_LIST(a, list)
does this mb use listX, note does not work if subMBs
Definition: mpegvideo.h:126
int y
Definition: motion_est.c:66
op_pixels_func(* hpel_put)[4]
Definition: mpegvideo.h:196
uint8_t
enum OutputFormat out_format
output format
Definition: mpegvideo.h:219
int me_range
maximum motion estimation search range in subpel units If 0 then no limit.
Definition: avcodec.h:1833
#define FLAG_DIRECT
Definition: motion_est.c:79
#define b
Definition: input.c:52
int pre_dia_size
ME prepass diamond size & shape.
Definition: avcodec.h:1800
Motion estimation context.
Definition: mpegvideo.h:155
qpel_mc_func(* qpel_put)[16]
Definition: mpegvideo.h:198
me_cmp_func me_pre_cmp[6]
Definition: dsputil.h:239
int no_rounding
apply no rounding to motion compensation (MPEG4, msmpeg4, ...) for b-frames rounding mode is always 0...
Definition: mpegvideo.h:407
int me_cmp
motion estimation comparison function
Definition: avcodec.h:1731
Picture current_picture
copy of the current picture structure.
Definition: mpegvideo.h:313
me_cmp_func me_cmp[6]
Definition: dsputil.h:240
op_pixels_func avg_pixels_tab[4][4]
Halfpel motion compensation with rounding (a+b+1)>>1.
Definition: dsputil.h:271
int16_t(* b_bidir_forw_mv_table)[2]
MV table (1MV per MB) bidir mode b-frame encoding.
Definition: mpegvideo.h:376
uint8_t(* mv_penalty)[MAX_MV *2+1]
amount of bits needed to encode a MV
Definition: mpegvideo.h:200
static int get_flags(MotionEstContext *c, int direct, int chroma)
Definition: motion_est.c:99
void ff_set_cmp(DSPContext *c, me_cmp_func *cmp, int type)
Definition: dsputil.c:1741
static int flags
Definition: log.c:42
int mb_threshold
Macroblock threshold below which the user specified macroblock types will be used.
Definition: avcodec.h:1944
uint16_t pp_time
time distance between the last 2 p,s,i frames
Definition: mpegvideo.h:538
op_pixels_func(* hpel_avg)[4]
Definition: mpegvideo.h:197
int ff_epzs_motion_search(MpegEncContext *s, int *mx_ptr, int *my_ptr, int P[10][2], int src_index, int ref_index, int16_t(*last_mv)[2], int ref_mv_scale, int size, int h)
int mb_height
number of MBs horizontally & vertically
Definition: mpegvideo.h:247
#define IS_INTERLACED(a)
Definition: mpegvideo.h:112
#define CHECK_BIDIR2(a, b, c, d)
static float t
qpel_mc_func put_qpel_pixels_tab[2][16]
Definition: dsputil.h:312
#define CANDIDATE_MB_TYPE_FORWARD
Definition: mpegvideo.h:422
#define r
Definition: input.c:51
no search, that is use 0,0 vector whenever one is needed
Definition: avcodec.h:513
reserved for experiments
Definition: avcodec.h:518
#define CANDIDATE_MB_TYPE_DIRECT
Definition: mpegvideo.h:421
int height
Definition: motion_est.c:65
static int ff_h263_round_chroma(int x)
Definition: mpegvideo.h:846
static const uint16_t mask[17]
Definition: lzw.c:38
static int epzs_motion_search4(MpegEncContext *s, int *mx_ptr, int *my_ptr, int P[10][2], int src_index, int ref_index, int16_t(*last_mv)[2], int ref_mv_scale)
static int no_sub_motion_search(MpegEncContext *s, int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h)
#define LOAD_COMMON
qpel_mc_func avg_qpel_pixels_tab[2][16]
Definition: dsputil.h:313
int me_sub_cmp
subpixel motion estimation comparison function
Definition: avcodec.h:1737
static uint8_t fcode_tab[MAX_MV *2+1]
Minimal fcode that a motion vector component would need.
Definition: ituh263enc.c:54
int unrestricted_mv
mv can point outside of the coded picture
Definition: mpegvideo.h:358
#define P_MEDIAN
Definition: motion_est.c:45
int flags
CODEC_FLAG_*.
Definition: avcodec.h:1434
static void init_interlaced_ref(MpegEncContext *s, int ref_index)
Definition: motion_est.c:695
#define CANDIDATE_MB_TYPE_INTER
Definition: mpegvideo.h:416
void av_log(void *avcl, int level, const char *fmt,...)
Definition: log.c:146
struct Minima Minima
me_cmp_func mb_cmp[6]
Definition: dsputil.h:242
int quarter_sample
1->qpel, 0->half pel ME/MC
Definition: mpegvideo.h:557
uint16_t * mb_type
Table for candidate MB types for encoding.
Definition: mpegvideo.h:414
#define IS_INTRA(a)
Definition: mpegvideo.h:108
op_pixels_func put_no_rnd_pixels_tab[4][4]
Halfpel motion compensation with no rounding (a+b)>>1.
Definition: dsputil.h:283
uint8_t *[2][2] b_field_select_table
Definition: mpegvideo.h:382
int off
Definition: dsputil_bfin.c:28
int checked
Definition: motion_est.c:67
uint8_t * src[4][4]
Definition: mpegvideo.h:187
#define FLAG_CHROMA
Definition: motion_est.c:78
#define IS_8X8(a)
Definition: mpegvideo.h:118
Motion estimation template.
static DCTELEM block[64]
Definition: dct-test.c:169
#define FLAG_QPEL
Definition: motion_est.c:77
int me_method
ME algorithm.
Definition: mpegvideo.h:383
static int cmp_fpel_internal(MpegEncContext *s, const int x, const int y, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags)
Definition: motion_est.c:229
Picture new_picture
copy of the source picture structure for encoding.
Definition: mpegvideo.h:307
static int hpel_motion_search(MpegEncContext *s, int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h)
#define P_TOP
Definition: motion_est.c:43
Picture * current_picture_ptr
pointer to the current picture
Definition: mpegvideo.h:317
Picture.
Definition: mpegvideo.h:94
#define ME_MAP_MV_BITS
Definition: mpegvideo.h:67
#define CANDIDATE_MB_TYPE_INTER4V
Definition: mpegvideo.h:417
unsigned map_generation
Definition: mpegvideo.h:167
static int interlaced_search(MpegEncContext *s, int ref_index, int16_t(*mv_tables[2][2])[2], uint8_t *field_select_tables[2], int mx, int my, int user_field_select)
Definition: motion_est.c:708
static av_always_inline int cmp(MpegEncContext *s, const int x, const int y, const int subx, const int suby, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags)
compares a block (either a full macroblock or a partition thereof) against a proposed motion-compensa...
Definition: motion_est.c:252
MotionEstContext me
Definition: mpegvideo.h:405
int mb_decision
macroblock decision mode
Definition: avcodec.h:1882
#define ME_MAP_SIZE
Definition: mpegvideo.h:65
#define CANDIDATE_MB_TYPE_BACKWARD
Definition: mpegvideo.h:423
#define CANDIDATE_MB_TYPE_SKIPPED
Definition: mpegvideo.h:418
static int get_mb_score(MpegEncContext *s, int mx, int my, int src_index, int ref_index, int size, int h, int add_rate)
preferred ID for MPEG-1/2 video decoding
Definition: avcodec.h:100
int block_index[6]
index to current MB in block based arrays with edges
Definition: mpegvideo.h:433
static int qpel_motion_search(MpegEncContext *s, int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h)
static void clip_input_mv(MpegEncContext *s, int16_t *mv, int interlaced)
Definition: motion_est.c:830
int penalty_factor
an estimate of the bits required to code a given mv value, e.g.
Definition: mpegvideo.h:169
#define HASH8(fx, fy, bx, by)
static const int8_t mv[256][2]
Definition: 4xm.c:73
uint32_t * mb_type
macroblock type table mb_type_base + mb_width + 2
Definition: avcodec.h:1180
#define MV_TYPE_16X16
1 vector for the whole mb
Definition: mpegvideo.h:389
int first_slice_line
used in mpeg4 too to handle resync markers
Definition: mpegvideo.h:614
static int check_input_motion(MpegEncContext *s, int mb_x, int mb_y, int p_type)
Definition: motion_est.c:840
NULL
Definition: eval.c:52
uint16_t * mc_mb_var
Table for motion compensated MB variances.
Definition: mpegvideo.h:144
int bidir_refine
Definition: avcodec.h:2001
unsigned int lambda2
(lambda*lambda) >> FF_LAMBDA_SHIFT
Definition: mpegvideo.h:345
external API header
int me_threshold
Motion estimation threshold below which no motion estimation is performed, but instead the user speci...
Definition: avcodec.h:1937
AVCodecContext * avctx
Definition: mpegvideo.h:156
enum AVCodecID codec_id
Definition: avcodec.h:1350
static av_const unsigned int ff_sqrt(unsigned int a)
Definition: mathops.h:198
#define CHECK_SAD_HALF_MV(suffix, x, y)
Definition: motion_est.c:375
static int get_penalty_factor(int lambda, int lambda2, int type)
Definition: dsputil.h:539
int height
picture size. must be a multiple of 16
Definition: mpegvideo.h:215
int16_t(*[2] motion_val)[2]
motion vector table
Definition: avcodec.h:1172
static int direct_search(MpegEncContext *s, int mb_x, int mb_y)
Definition: motion_est.c:1534
void ff_fix_long_p_mvs(MpegEncContext *s)
Definition: motion_est.c:1845
int16_t(*[2][2][2] b_field_mv_table)[2]
MV table (4MV per MB) interlaced b-frame encoding.
Definition: mpegvideo.h:380
int index
Definition: gxfenc.c:72
static av_always_inline int cmp_direct_inline(MpegEncContext *s, const int x, const int y, const int subx, const int suby, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, int qpel)
Definition: motion_est.c:105
#define mid_pred
Definition: mathops.h:94
int8_t * ref_index[2]
motion reference frame index the order in which these are stored can depend on the codec...
Definition: avcodec.h:1195
DSPContext dsp
pointers for accelerated dsp functions
Definition: mpegvideo.h:361
int(* pix_norm1)(uint8_t *pix, int line_size)
Definition: dsputil.h:221
#define MAX_MV
Definition: mpegvideo.h:59
int f_code
forward MV resolution
Definition: mpegvideo.h:363
int ff_pre_estimate_p_frame_motion(MpegEncContext *s, int mb_x, int mb_y)
Definition: motion_est.c:1197
#define CANDIDATE_MB_TYPE_BIDIR_I
Definition: mpegvideo.h:429
int pict_type
AV_PICTURE_TYPE_I, AV_PICTURE_TYPE_P, AV_PICTURE_TYPE_B, ...
Definition: mpegvideo.h:349
static void set_p_mv_tables(MpegEncContext *s, int mx, int my, int mv4)
Definition: motion_est.c:488
#define CANDIDATE_MB_TYPE_FORWARD_I
Definition: mpegvideo.h:427
static int sad_hpel_motion_search(MpegEncContext *s, int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h)
Definition: motion_est.c:382
int16_t(* b_bidir_back_mv_table)[2]
MV table (1MV per MB) bidir mode b-frame encoding.
Definition: mpegvideo.h:377
int ff_init_me(MpegEncContext *s)
Definition: motion_est.c:296
uint8_t *[2] p_field_select_table
Definition: mpegvideo.h:381
int16_t(* b_direct_mv_table)[2]
MV table (1MV per MB) direct mode b-frame encoding.
Definition: mpegvideo.h:378
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: avcodec.h:997
qpel_mc_func(* qpel_avg)[16]
Definition: mpegvideo.h:199
int mc_mb_var_sum_temp
Definition: mpegvideo.h:192
int16_t(* b_forw_mv_table)[2]
MV table (1MV per MB) forward mode b-frame encoding.
Definition: mpegvideo.h:374
int b8_stride
2*mb_width+1 used for some 8x8 block arrays to allow simple addressing
Definition: mpegvideo.h:249
MpegEncContext.
Definition: mpegvideo.h:211
struct AVCodecContext * avctx
Definition: mpegvideo.h:213
static uint8_t mv_penalty[MAX_FCODE+1][MAX_MV *2+1]
Table of number of bits a motion vector component needs.
Definition: ituh263enc.c:49
int mb_cmp
macroblock comparison function (not supported yet)
Definition: avcodec.h:1743
int direct_basis_mv[4][2]
Definition: mpegvideo.h:159
int mb_stride
mb_width+1 used for some arrays to allow simple addressing of left & top MBs without sig11 ...
Definition: mpegvideo.h:248
Picture last_picture
copy of the previous picture structure.
Definition: mpegvideo.h:295
Bi-dir predicted.
Definition: avutil.h:247
int co_located_mv[4][2]
mv from last P-frame for direct mode ME
Definition: mpegvideo.h:158
uint32_t * map
map to avoid duplicate evaluations
Definition: mpegvideo.h:165
static int h263_mv4_search(MpegEncContext *s, int mx, int my, int shift)
Definition: motion_est.c:559
DSP utils.
#define CANDIDATE_MB_TYPE_DIRECT0
Definition: mpegvideo.h:431
int dia_size
ME diamond size & shape.
Definition: avcodec.h:1772
static void init_ref(MotionEstContext *c, uint8_t *src[3], uint8_t *ref[3], uint8_t *ref2[3], int x, int y, int ref_index)
Definition: motion_est.c:81
static int cmp_hpel(MpegEncContext *s, const int x, const int y, const int subx, const int suby, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, const int flags)
Definition: motion_est.c:267
int ff_get_best_fcode(MpegEncContext *s, int16_t(*mv_table)[2], int type)
Definition: motion_est.c:1792
static int zero_cmp(void *s, uint8_t *a, uint8_t *b, int stride, int h)
Definition: motion_est.c:289
Picture next_picture
copy of the next picture structure.
Definition: mpegvideo.h:301
int linesize
line size, in bytes, may be different from width
Definition: mpegvideo.h:253
uint8_t * mbskip_table
mbskip_table[mb]>=1 if MB didn't change stride= mb_width = (width+15)>>4
Definition: avcodec.h:1158
void ff_fix_long_mvs(MpegEncContext *s, uint8_t *field_select_table, int field_select, int16_t(*mv_table)[2], int f_code, int type, int truncate)
Definition: motion_est.c:1895
struct AVFrame f
Definition: mpegvideo.h:95
int flags
AVCodecContext.flags (HQ, MV4, ...)
Definition: mpegvideo.h:230
qpel_mc_func put_no_rnd_qpel_pixels_tab[2][16]
Definition: dsputil.h:314
uint8_t * temp
Definition: mpegvideo.h:163
static av_always_inline int cmp_inline(MpegEncContext *s, const int x, const int y, const int subx, const int suby, const int size, const int h, int ref_index, int src_index, me_cmp_func cmp_func, me_cmp_func chroma_cmp_func, int qpel, int chroma)
Definition: motion_est.c:177
me_cmp_func me_sub_cmp[6]
Definition: dsputil.h:241
#define MV_TYPE_8X8
4 vectors (h263, mpeg4 4MV)
Definition: mpegvideo.h:390
int b_code
backward MV resolution for B Frames (mpeg4)
Definition: mpegvideo.h:364
int me_pre_cmp
motion estimation prepass comparison function
Definition: avcodec.h:1793
float min
int uvlinesize
line size, for chroma in bytes, may be different from width
Definition: mpegvideo.h:254
int(* sub_motion_search)(struct MpegEncContext *s, int *mx_ptr, int *my_ptr, int dmin, int src_index, int ref_index, int size, int h)
Definition: mpegvideo.h:202
int x
Definition: motion_est.c:66
int strict_std_compliance
strictly follow the standard (MPEG4, ...).
Definition: avcodec.h:2547
static void zero_hpel(uint8_t *a, const uint8_t *b, int stride, int h)
Definition: motion_est.c:293
#define CANDIDATE_MB_TYPE_INTER_I
Definition: mpegvideo.h:426
Predicted.
Definition: avutil.h:246
unsigned int lambda
lagrange multipler used in rate distortion
Definition: mpegvideo.h:344
#define HASH(fx, fy, bx, by)
uint16_t pb_time
time distance between the last b and p,s,i frame
Definition: mpegvideo.h:539