h264idct_template.c
Go to the documentation of this file.
1 /*
2  * H.264 IDCT
3  * Copyright (c) 2004-2011 Michael Niedermayer <michaelni@gmx.at>
4  *
5  * This file is part of Libav.
6  *
7  * Libav is free software; you can redistribute it and/or
8  * modify it under the terms of the GNU Lesser General Public
9  * License as published by the Free Software Foundation; either
10  * version 2.1 of the License, or (at your option) any later version.
11  *
12  * Libav is distributed in the hope that it will be useful,
13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15  * Lesser General Public License for more details.
16  *
17  * You should have received a copy of the GNU Lesser General Public
18  * License along with Libav; if not, write to the Free Software
19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20  */
21 
28 #include "bit_depth_template.c"
29 
30 #ifndef AVCODEC_H264IDCT_INTERNAL_H
31 #define AVCODEC_H264IDCT_INTERNAL_H
32 //FIXME this table is a duplicate from h264data.h, and will be removed once the tables from, h264 have been split
33 static const uint8_t scan8[16*3]={
34  4+ 1*8, 5+ 1*8, 4+ 2*8, 5+ 2*8,
35  6+ 1*8, 7+ 1*8, 6+ 2*8, 7+ 2*8,
36  4+ 3*8, 5+ 3*8, 4+ 4*8, 5+ 4*8,
37  6+ 3*8, 7+ 3*8, 6+ 4*8, 7+ 4*8,
38  4+ 6*8, 5+ 6*8, 4+ 7*8, 5+ 7*8,
39  6+ 6*8, 7+ 6*8, 6+ 7*8, 7+ 7*8,
40  4+ 8*8, 5+ 8*8, 4+ 9*8, 5+ 9*8,
41  6+ 8*8, 7+ 8*8, 6+ 9*8, 7+ 9*8,
42  4+11*8, 5+11*8, 4+12*8, 5+12*8,
43  6+11*8, 7+11*8, 6+12*8, 7+12*8,
44  4+13*8, 5+13*8, 4+14*8, 5+14*8,
45  6+13*8, 7+13*8, 6+14*8, 7+14*8
46 };
47 #endif
48 
49 void FUNCC(ff_h264_idct_add)(uint8_t *_dst, DCTELEM *_block, int stride)
50 {
51  int i;
52  pixel *dst = (pixel*)_dst;
53  dctcoef *block = (dctcoef*)_block;
54  stride /= sizeof(pixel);
55 
56  block[0] += 1 << 5;
57 
58  for(i=0; i<4; i++){
59  const int z0= block[i + 4*0] + block[i + 4*2];
60  const int z1= block[i + 4*0] - block[i + 4*2];
61  const int z2= (block[i + 4*1]>>1) - block[i + 4*3];
62  const int z3= block[i + 4*1] + (block[i + 4*3]>>1);
63 
64  block[i + 4*0]= z0 + z3;
65  block[i + 4*1]= z1 + z2;
66  block[i + 4*2]= z1 - z2;
67  block[i + 4*3]= z0 - z3;
68  }
69 
70  for(i=0; i<4; i++){
71  const int z0= block[0 + 4*i] + block[2 + 4*i];
72  const int z1= block[0 + 4*i] - block[2 + 4*i];
73  const int z2= (block[1 + 4*i]>>1) - block[3 + 4*i];
74  const int z3= block[1 + 4*i] + (block[3 + 4*i]>>1);
75 
76  dst[i + 0*stride]= av_clip_pixel(dst[i + 0*stride] + ((z0 + z3) >> 6));
77  dst[i + 1*stride]= av_clip_pixel(dst[i + 1*stride] + ((z1 + z2) >> 6));
78  dst[i + 2*stride]= av_clip_pixel(dst[i + 2*stride] + ((z1 - z2) >> 6));
79  dst[i + 3*stride]= av_clip_pixel(dst[i + 3*stride] + ((z0 - z3) >> 6));
80  }
81 }
82 
83 void FUNCC(ff_h264_idct8_add)(uint8_t *_dst, DCTELEM *_block, int stride){
84  int i;
85  pixel *dst = (pixel*)_dst;
86  dctcoef *block = (dctcoef*)_block;
87  stride /= sizeof(pixel);
88 
89  block[0] += 32;
90 
91  for( i = 0; i < 8; i++ )
92  {
93  const int a0 = block[i+0*8] + block[i+4*8];
94  const int a2 = block[i+0*8] - block[i+4*8];
95  const int a4 = (block[i+2*8]>>1) - block[i+6*8];
96  const int a6 = (block[i+6*8]>>1) + block[i+2*8];
97 
98  const int b0 = a0 + a6;
99  const int b2 = a2 + a4;
100  const int b4 = a2 - a4;
101  const int b6 = a0 - a6;
102 
103  const int a1 = -block[i+3*8] + block[i+5*8] - block[i+7*8] - (block[i+7*8]>>1);
104  const int a3 = block[i+1*8] + block[i+7*8] - block[i+3*8] - (block[i+3*8]>>1);
105  const int a5 = -block[i+1*8] + block[i+7*8] + block[i+5*8] + (block[i+5*8]>>1);
106  const int a7 = block[i+3*8] + block[i+5*8] + block[i+1*8] + (block[i+1*8]>>1);
107 
108  const int b1 = (a7>>2) + a1;
109  const int b3 = a3 + (a5>>2);
110  const int b5 = (a3>>2) - a5;
111  const int b7 = a7 - (a1>>2);
112 
113  block[i+0*8] = b0 + b7;
114  block[i+7*8] = b0 - b7;
115  block[i+1*8] = b2 + b5;
116  block[i+6*8] = b2 - b5;
117  block[i+2*8] = b4 + b3;
118  block[i+5*8] = b4 - b3;
119  block[i+3*8] = b6 + b1;
120  block[i+4*8] = b6 - b1;
121  }
122  for( i = 0; i < 8; i++ )
123  {
124  const int a0 = block[0+i*8] + block[4+i*8];
125  const int a2 = block[0+i*8] - block[4+i*8];
126  const int a4 = (block[2+i*8]>>1) - block[6+i*8];
127  const int a6 = (block[6+i*8]>>1) + block[2+i*8];
128 
129  const int b0 = a0 + a6;
130  const int b2 = a2 + a4;
131  const int b4 = a2 - a4;
132  const int b6 = a0 - a6;
133 
134  const int a1 = -block[3+i*8] + block[5+i*8] - block[7+i*8] - (block[7+i*8]>>1);
135  const int a3 = block[1+i*8] + block[7+i*8] - block[3+i*8] - (block[3+i*8]>>1);
136  const int a5 = -block[1+i*8] + block[7+i*8] + block[5+i*8] + (block[5+i*8]>>1);
137  const int a7 = block[3+i*8] + block[5+i*8] + block[1+i*8] + (block[1+i*8]>>1);
138 
139  const int b1 = (a7>>2) + a1;
140  const int b3 = a3 + (a5>>2);
141  const int b5 = (a3>>2) - a5;
142  const int b7 = a7 - (a1>>2);
143 
144  dst[i + 0*stride] = av_clip_pixel( dst[i + 0*stride] + ((b0 + b7) >> 6) );
145  dst[i + 1*stride] = av_clip_pixel( dst[i + 1*stride] + ((b2 + b5) >> 6) );
146  dst[i + 2*stride] = av_clip_pixel( dst[i + 2*stride] + ((b4 + b3) >> 6) );
147  dst[i + 3*stride] = av_clip_pixel( dst[i + 3*stride] + ((b6 + b1) >> 6) );
148  dst[i + 4*stride] = av_clip_pixel( dst[i + 4*stride] + ((b6 - b1) >> 6) );
149  dst[i + 5*stride] = av_clip_pixel( dst[i + 5*stride] + ((b4 - b3) >> 6) );
150  dst[i + 6*stride] = av_clip_pixel( dst[i + 6*stride] + ((b2 - b5) >> 6) );
151  dst[i + 7*stride] = av_clip_pixel( dst[i + 7*stride] + ((b0 - b7) >> 6) );
152  }
153 }
154 
155 // assumes all AC coefs are 0
156 void FUNCC(ff_h264_idct_dc_add)(uint8_t *_dst, DCTELEM *block, int stride){
157  int i, j;
158  int dc = (((dctcoef*)block)[0] + 32) >> 6;
159  pixel *dst = (pixel*)_dst;
160  stride /= sizeof(pixel);
161  for( j = 0; j < 4; j++ )
162  {
163  for( i = 0; i < 4; i++ )
164  dst[i] = av_clip_pixel( dst[i] + dc );
165  dst += stride;
166  }
167 }
168 
169 void FUNCC(ff_h264_idct8_dc_add)(uint8_t *_dst, DCTELEM *block, int stride){
170  int i, j;
171  int dc = (((dctcoef*)block)[0] + 32) >> 6;
172  pixel *dst = (pixel*)_dst;
173  stride /= sizeof(pixel);
174  for( j = 0; j < 8; j++ )
175  {
176  for( i = 0; i < 8; i++ )
177  dst[i] = av_clip_pixel( dst[i] + dc );
178  dst += stride;
179  }
180 }
181 
182 void FUNCC(ff_h264_idct_add16)(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[15*8]){
183  int i;
184  for(i=0; i<16; i++){
185  int nnz = nnzc[ scan8[i] ];
186  if(nnz){
187  if(nnz==1 && ((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
188  else FUNCC(ff_h264_idct_add )(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
189  }
190  }
191 }
192 
193 void FUNCC(ff_h264_idct_add16intra)(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[15*8]){
194  int i;
195  for(i=0; i<16; i++){
196  if(nnzc[ scan8[i] ]) FUNCC(ff_h264_idct_add )(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
197  else if(((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
198  }
199 }
200 
201 void FUNCC(ff_h264_idct8_add4)(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[15*8]){
202  int i;
203  for(i=0; i<16; i+=4){
204  int nnz = nnzc[ scan8[i] ];
205  if(nnz){
206  if(nnz==1 && ((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct8_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
207  else FUNCC(ff_h264_idct8_add )(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
208  }
209  }
210 }
211 
212 void FUNCC(ff_h264_idct_add8)(uint8_t **dest, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[15*8]){
213  int i, j;
214  for(j=1; j<3; j++){
215  for(i=j*16; i<j*16+4; i++){
216  if(nnzc[ scan8[i] ])
217  FUNCC(ff_h264_idct_add )(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
218  else if(((dctcoef*)block)[i*16])
219  FUNCC(ff_h264_idct_dc_add)(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
220  }
221  }
222 }
223 
224 void FUNCC(ff_h264_idct_add8_422)(uint8_t **dest, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[15*8]){
225  int i, j;
226 
227  for(j=1; j<3; j++){
228  for(i=j*16; i<j*16+4; i++){
229  if(nnzc[ scan8[i] ])
230  FUNCC(ff_h264_idct_add )(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
231  else if(((dctcoef*)block)[i*16])
232  FUNCC(ff_h264_idct_dc_add)(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
233  }
234  }
235 
236  for(j=1; j<3; j++){
237  for(i=j*16+4; i<j*16+8; i++){
238  if(nnzc[ scan8[i+4] ])
239  FUNCC(ff_h264_idct_add )(dest[j-1] + block_offset[i+4], block + i*16*sizeof(pixel), stride);
240  else if(((dctcoef*)block)[i*16])
241  FUNCC(ff_h264_idct_dc_add)(dest[j-1] + block_offset[i+4], block + i*16*sizeof(pixel), stride);
242  }
243  }
244 }
245 
250 void FUNCC(ff_h264_luma_dc_dequant_idct)(DCTELEM *_output, DCTELEM *_input, int qmul){
251 #define stride 16
252  int i;
253  int temp[16];
254  static const uint8_t x_offset[4]={0, 2*stride, 8*stride, 10*stride};
255  dctcoef *input = (dctcoef*)_input;
256  dctcoef *output = (dctcoef*)_output;
257 
258  for(i=0; i<4; i++){
259  const int z0= input[4*i+0] + input[4*i+1];
260  const int z1= input[4*i+0] - input[4*i+1];
261  const int z2= input[4*i+2] - input[4*i+3];
262  const int z3= input[4*i+2] + input[4*i+3];
263 
264  temp[4*i+0]= z0+z3;
265  temp[4*i+1]= z0-z3;
266  temp[4*i+2]= z1-z2;
267  temp[4*i+3]= z1+z2;
268  }
269 
270  for(i=0; i<4; i++){
271  const int offset= x_offset[i];
272  const int z0= temp[4*0+i] + temp[4*2+i];
273  const int z1= temp[4*0+i] - temp[4*2+i];
274  const int z2= temp[4*1+i] - temp[4*3+i];
275  const int z3= temp[4*1+i] + temp[4*3+i];
276 
277  output[stride* 0+offset]= ((((z0 + z3)*qmul + 128 ) >> 8));
278  output[stride* 1+offset]= ((((z1 + z2)*qmul + 128 ) >> 8));
279  output[stride* 4+offset]= ((((z1 - z2)*qmul + 128 ) >> 8));
280  output[stride* 5+offset]= ((((z0 - z3)*qmul + 128 ) >> 8));
281  }
282 #undef stride
283 }
284 
286  const int stride= 16*2;
287  const int xStride= 16;
288  int i;
289  int temp[8];
290  static const uint8_t x_offset[2]={0, 16};
291  dctcoef *block = (dctcoef*)_block;
292 
293  for(i=0; i<4; i++){
294  temp[2*i+0] = block[stride*i + xStride*0] + block[stride*i + xStride*1];
295  temp[2*i+1] = block[stride*i + xStride*0] - block[stride*i + xStride*1];
296  }
297 
298  for(i=0; i<2; i++){
299  const int offset= x_offset[i];
300  const int z0= temp[2*0+i] + temp[2*2+i];
301  const int z1= temp[2*0+i] - temp[2*2+i];
302  const int z2= temp[2*1+i] - temp[2*3+i];
303  const int z3= temp[2*1+i] + temp[2*3+i];
304 
305  block[stride*0+offset]= ((z0 + z3)*qmul + 128) >> 8;
306  block[stride*1+offset]= ((z1 + z2)*qmul + 128) >> 8;
307  block[stride*2+offset]= ((z1 - z2)*qmul + 128) >> 8;
308  block[stride*3+offset]= ((z0 - z3)*qmul + 128) >> 8;
309  }
310 }
311 
313  const int stride= 16*2;
314  const int xStride= 16;
315  int a,b,c,d,e;
316  dctcoef *block = (dctcoef*)_block;
317 
318  a= block[stride*0 + xStride*0];
319  b= block[stride*0 + xStride*1];
320  c= block[stride*1 + xStride*0];
321  d= block[stride*1 + xStride*1];
322 
323  e= a-b;
324  a= a+b;
325  b= c-d;
326  c= c+d;
327 
328  block[stride*0 + xStride*0]= ((a+c)*qmul) >> 7;
329  block[stride*0 + xStride*1]= ((e+b)*qmul) >> 7;
330  block[stride*1 + xStride*0]= ((a-c)*qmul) >> 7;
331  block[stride*1 + xStride*1]= ((e-b)*qmul) >> 7;
332 }