| 15fb393b |
/*
* H.264 IDCT
* Copyright (c) 2004-2011 Michael Niedermayer <michaelni@gmx.at>
* |
| d18e243f |
* This file is part of FFmpeg. |
| 15fb393b |
* |
| d18e243f |
* FFmpeg is free software; you can redistribute it and/or |
| 15fb393b |
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
* |
| d18e243f |
* FFmpeg is distributed in the hope that it will be useful, |
| 15fb393b |
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public |
| d18e243f |
* License along with FFmpeg; if not, write to the Free Software |
| 15fb393b |
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* H.264 IDCT.
* @author Michael Niedermayer <michaelni@gmx.at>
*/
|
| c45f6295 |
#include "bit_depth_template.c" |
| 1d9c2dc8 |
#include "libavutil/common.h" |
| 49fe2807 |
#include "h264.h" |
| aba70bb5 |
#include "h264idct.h" |
| 15fb393b |
|
| 88bd7fdc |
void FUNCC(ff_h264_idct_add)(uint8_t *_dst, int16_t *_block, int stride) |
| 7b4ee3a2 |
{ |
| 15fb393b |
int i; |
| 5d4bd9cc |
pixel *dst = (pixel*)_dst;
dctcoef *block = (dctcoef*)_block; |
| d18e243f |
stride >>= sizeof(pixel)-1; |
| 15fb393b |
|
| 7b4ee3a2 |
block[0] += 1 << 5; |
| 15fb393b |
for(i=0; i<4; i++){ |
| 7b4ee3a2 |
const int z0= block[i + 4*0] + block[i + 4*2];
const int z1= block[i + 4*0] - block[i + 4*2];
const int z2= (block[i + 4*1]>>1) - block[i + 4*3];
const int z3= block[i + 4*1] + (block[i + 4*3]>>1);
block[i + 4*0]= z0 + z3;
block[i + 4*1]= z1 + z2;
block[i + 4*2]= z1 - z2;
block[i + 4*3]= z0 - z3; |
| 15fb393b |
}
for(i=0; i<4; i++){ |
| 7b4ee3a2 |
const int z0= block[0 + 4*i] + block[2 + 4*i];
const int z1= block[0 + 4*i] - block[2 + 4*i];
const int z2= (block[1 + 4*i]>>1) - block[3 + 4*i];
const int z3= block[1 + 4*i] + (block[3 + 4*i]>>1);
|
| c23acbae |
dst[i + 0*stride]= av_clip_pixel(dst[i + 0*stride] + ((z0 + z3) >> 6));
dst[i + 1*stride]= av_clip_pixel(dst[i + 1*stride] + ((z1 + z2) >> 6));
dst[i + 2*stride]= av_clip_pixel(dst[i + 2*stride] + ((z1 - z2) >> 6));
dst[i + 3*stride]= av_clip_pixel(dst[i + 3*stride] + ((z0 - z3) >> 6)); |
| 15fb393b |
} |
| 1acd7d59 |
memset(block, 0, 16 * sizeof(dctcoef)); |
| 15fb393b |
}
|
| 88bd7fdc |
void FUNCC(ff_h264_idct8_add)(uint8_t *_dst, int16_t *_block, int stride){ |
| 15fb393b |
int i; |
| 5d4bd9cc |
pixel *dst = (pixel*)_dst;
dctcoef *block = (dctcoef*)_block; |
| d18e243f |
stride >>= sizeof(pixel)-1; |
| 15fb393b |
block[0] += 32;
for( i = 0; i < 8; i++ )
{
const int a0 = block[i+0*8] + block[i+4*8];
const int a2 = block[i+0*8] - block[i+4*8];
const int a4 = (block[i+2*8]>>1) - block[i+6*8];
const int a6 = (block[i+6*8]>>1) + block[i+2*8];
const int b0 = a0 + a6;
const int b2 = a2 + a4;
const int b4 = a2 - a4;
const int b6 = a0 - a6;
const int a1 = -block[i+3*8] + block[i+5*8] - block[i+7*8] - (block[i+7*8]>>1);
const int a3 = block[i+1*8] + block[i+7*8] - block[i+3*8] - (block[i+3*8]>>1);
const int a5 = -block[i+1*8] + block[i+7*8] + block[i+5*8] + (block[i+5*8]>>1);
const int a7 = block[i+3*8] + block[i+5*8] + block[i+1*8] + (block[i+1*8]>>1);
const int b1 = (a7>>2) + a1;
const int b3 = a3 + (a5>>2);
const int b5 = (a3>>2) - a5;
const int b7 = a7 - (a1>>2);
block[i+0*8] = b0 + b7;
block[i+7*8] = b0 - b7;
block[i+1*8] = b2 + b5;
block[i+6*8] = b2 - b5;
block[i+2*8] = b4 + b3;
block[i+5*8] = b4 - b3;
block[i+3*8] = b6 + b1;
block[i+4*8] = b6 - b1;
}
for( i = 0; i < 8; i++ )
{
const int a0 = block[0+i*8] + block[4+i*8];
const int a2 = block[0+i*8] - block[4+i*8];
const int a4 = (block[2+i*8]>>1) - block[6+i*8];
const int a6 = (block[6+i*8]>>1) + block[2+i*8];
const int b0 = a0 + a6;
const int b2 = a2 + a4;
const int b4 = a2 - a4;
const int b6 = a0 - a6;
const int a1 = -block[3+i*8] + block[5+i*8] - block[7+i*8] - (block[7+i*8]>>1);
const int a3 = block[1+i*8] + block[7+i*8] - block[3+i*8] - (block[3+i*8]>>1);
const int a5 = -block[1+i*8] + block[7+i*8] + block[5+i*8] + (block[5+i*8]>>1);
const int a7 = block[3+i*8] + block[5+i*8] + block[1+i*8] + (block[1+i*8]>>1);
const int b1 = (a7>>2) + a1;
const int b3 = a3 + (a5>>2);
const int b5 = (a3>>2) - a5;
const int b7 = a7 - (a1>>2);
|
| c23acbae |
dst[i + 0*stride] = av_clip_pixel( dst[i + 0*stride] + ((b0 + b7) >> 6) );
dst[i + 1*stride] = av_clip_pixel( dst[i + 1*stride] + ((b2 + b5) >> 6) );
dst[i + 2*stride] = av_clip_pixel( dst[i + 2*stride] + ((b4 + b3) >> 6) );
dst[i + 3*stride] = av_clip_pixel( dst[i + 3*stride] + ((b6 + b1) >> 6) );
dst[i + 4*stride] = av_clip_pixel( dst[i + 4*stride] + ((b6 - b1) >> 6) );
dst[i + 5*stride] = av_clip_pixel( dst[i + 5*stride] + ((b4 - b3) >> 6) );
dst[i + 6*stride] = av_clip_pixel( dst[i + 6*stride] + ((b2 - b5) >> 6) );
dst[i + 7*stride] = av_clip_pixel( dst[i + 7*stride] + ((b0 - b7) >> 6) ); |
| 15fb393b |
} |
| 1acd7d59 |
memset(block, 0, 64 * sizeof(dctcoef)); |
| 15fb393b |
}
// assumes all AC coefs are 0 |
| 1acd7d59 |
void FUNCC(ff_h264_idct_dc_add)(uint8_t *_dst, int16_t *_block, int stride){ |
| 15fb393b |
int i, j; |
| 5d4bd9cc |
pixel *dst = (pixel*)_dst; |
| 1acd7d59 |
dctcoef *block = (dctcoef*)_block;
int dc = (block[0] + 32) >> 6; |
| 5d4bd9cc |
stride /= sizeof(pixel); |
| 1acd7d59 |
block[0] = 0; |
| 15fb393b |
for( j = 0; j < 4; j++ )
{
for( i = 0; i < 4; i++ ) |
| c23acbae |
dst[i] = av_clip_pixel( dst[i] + dc ); |
| 15fb393b |
dst += stride;
}
}
|
| 1acd7d59 |
void FUNCC(ff_h264_idct8_dc_add)(uint8_t *_dst, int16_t *_block, int stride){ |
| 15fb393b |
int i, j; |
| 5d4bd9cc |
pixel *dst = (pixel*)_dst; |
| 1acd7d59 |
dctcoef *block = (dctcoef*)_block;
int dc = (block[0] + 32) >> 6;
block[0] = 0; |
| 5d4bd9cc |
stride /= sizeof(pixel); |
| 15fb393b |
for( j = 0; j < 8; j++ )
{
for( i = 0; i < 8; i++ ) |
| c23acbae |
dst[i] = av_clip_pixel( dst[i] + dc ); |
| 15fb393b |
dst += stride;
}
}
|
| 88bd7fdc |
void FUNCC(ff_h264_idct_add16)(uint8_t *dst, const int *block_offset, int16_t *block, int stride, const uint8_t nnzc[15*8]){ |
| 15fb393b |
int i;
for(i=0; i<16; i++){
int nnz = nnzc[ scan8[i] ];
if(nnz){ |
| 5d4bd9cc |
if(nnz==1 && ((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride); |
| 7b4ee3a2 |
else FUNCC(ff_h264_idct_add )(dst + block_offset[i], block + i*16*sizeof(pixel), stride); |
| 15fb393b |
}
}
}
|
| 88bd7fdc |
void FUNCC(ff_h264_idct_add16intra)(uint8_t *dst, const int *block_offset, int16_t *block, int stride, const uint8_t nnzc[15*8]){ |
| 15fb393b |
int i;
for(i=0; i<16; i++){ |
| 7b4ee3a2 |
if(nnzc[ scan8[i] ]) FUNCC(ff_h264_idct_add )(dst + block_offset[i], block + i*16*sizeof(pixel), stride); |
| 5d4bd9cc |
else if(((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride); |
| 15fb393b |
}
}
|
| 88bd7fdc |
void FUNCC(ff_h264_idct8_add4)(uint8_t *dst, const int *block_offset, int16_t *block, int stride, const uint8_t nnzc[15*8]){ |
| 15fb393b |
int i;
for(i=0; i<16; i+=4){
int nnz = nnzc[ scan8[i] ];
if(nnz){ |
| 5d4bd9cc |
if(nnz==1 && ((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct8_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
else FUNCC(ff_h264_idct8_add )(dst + block_offset[i], block + i*16*sizeof(pixel), stride); |
| 15fb393b |
}
}
}
|
| 88bd7fdc |
void FUNCC(ff_h264_idct_add8)(uint8_t **dest, const int *block_offset, int16_t *block, int stride, const uint8_t nnzc[15*8]){ |
| c90b9442 |
int i, j;
for(j=1; j<3; j++){
for(i=j*16; i<j*16+4; i++){
if(nnzc[ scan8[i] ])
FUNCC(ff_h264_idct_add )(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
else if(((dctcoef*)block)[i*16])
FUNCC(ff_h264_idct_dc_add)(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
} |
| 15fb393b |
}
} |
| 231a6df9 |
|
| 88bd7fdc |
void FUNCC(ff_h264_idct_add8_422)(uint8_t **dest, const int *block_offset, int16_t *block, int stride, const uint8_t nnzc[15*8]){ |
| 231a6df9 |
int i, j;
for(j=1; j<3; j++){
for(i=j*16; i<j*16+4; i++){
if(nnzc[ scan8[i] ])
FUNCC(ff_h264_idct_add )(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
else if(((dctcoef*)block)[i*16])
FUNCC(ff_h264_idct_dc_add)(dest[j-1] + block_offset[i], block + i*16*sizeof(pixel), stride);
}
}
for(j=1; j<3; j++){
for(i=j*16+4; i<j*16+8; i++){
if(nnzc[ scan8[i+4] ])
FUNCC(ff_h264_idct_add )(dest[j-1] + block_offset[i+4], block + i*16*sizeof(pixel), stride);
else if(((dctcoef*)block)[i*16])
FUNCC(ff_h264_idct_dc_add)(dest[j-1] + block_offset[i+4], block + i*16*sizeof(pixel), stride);
}
}
}
|
| 15fb393b |
/**
* IDCT transforms the 16 dc values and dequantizes them. |
| 66e2dd11 |
* @param qmul quantization parameter |
| 15fb393b |
*/ |
| 88bd7fdc |
void FUNCC(ff_h264_luma_dc_dequant_idct)(int16_t *_output, int16_t *_input, int qmul){ |
| 15fb393b |
#define stride 16
int i;
int temp[16];
static const uint8_t x_offset[4]={0, 2*stride, 8*stride, 10*stride}; |
| 5d4bd9cc |
dctcoef *input = (dctcoef*)_input;
dctcoef *output = (dctcoef*)_output; |
| 15fb393b |
for(i=0; i<4; i++){
const int z0= input[4*i+0] + input[4*i+1];
const int z1= input[4*i+0] - input[4*i+1];
const int z2= input[4*i+2] - input[4*i+3];
const int z3= input[4*i+2] + input[4*i+3];
temp[4*i+0]= z0+z3;
temp[4*i+1]= z0-z3;
temp[4*i+2]= z1-z2;
temp[4*i+3]= z1+z2;
}
for(i=0; i<4; i++){
const int offset= x_offset[i];
const int z0= temp[4*0+i] + temp[4*2+i];
const int z1= temp[4*0+i] - temp[4*2+i];
const int z2= temp[4*1+i] - temp[4*3+i];
const int z3= temp[4*1+i] + temp[4*3+i];
output[stride* 0+offset]= ((((z0 + z3)*qmul + 128 ) >> 8));
output[stride* 1+offset]= ((((z1 + z2)*qmul + 128 ) >> 8));
output[stride* 4+offset]= ((((z1 - z2)*qmul + 128 ) >> 8));
output[stride* 5+offset]= ((((z0 - z3)*qmul + 128 ) >> 8));
}
#undef stride
}
|
| 88bd7fdc |
void FUNCC(ff_h264_chroma422_dc_dequant_idct)(int16_t *_block, int qmul){ |
| 231a6df9 |
const int stride= 16*2;
const int xStride= 16;
int i;
int temp[8];
static const uint8_t x_offset[2]={0, 16}; |
| 76741b0e |
dctcoef *block = (dctcoef*)_block; |
| 231a6df9 |
for(i=0; i<4; i++){
temp[2*i+0] = block[stride*i + xStride*0] + block[stride*i + xStride*1];
temp[2*i+1] = block[stride*i + xStride*0] - block[stride*i + xStride*1];
}
for(i=0; i<2; i++){
const int offset= x_offset[i];
const int z0= temp[2*0+i] + temp[2*2+i];
const int z1= temp[2*0+i] - temp[2*2+i];
const int z2= temp[2*1+i] - temp[2*3+i];
const int z3= temp[2*1+i] + temp[2*3+i];
block[stride*0+offset]= ((z0 + z3)*qmul + 128) >> 8;
block[stride*1+offset]= ((z1 + z2)*qmul + 128) >> 8;
block[stride*2+offset]= ((z1 - z2)*qmul + 128) >> 8;
block[stride*3+offset]= ((z0 - z3)*qmul + 128) >> 8;
}
}
|
| 88bd7fdc |
void FUNCC(ff_h264_chroma_dc_dequant_idct)(int16_t *_block, int qmul){ |
| 15fb393b |
const int stride= 16*2;
const int xStride= 16;
int a,b,c,d,e; |
| 5d4bd9cc |
dctcoef *block = (dctcoef*)_block; |
| 15fb393b |
a= block[stride*0 + xStride*0];
b= block[stride*0 + xStride*1];
c= block[stride*1 + xStride*0];
d= block[stride*1 + xStride*1];
e= a-b;
a= a+b;
b= c-d;
c= c+d;
block[stride*0 + xStride*0]= ((a+c)*qmul) >> 7;
block[stride*0 + xStride*1]= ((e+b)*qmul) >> 7;
block[stride*1 + xStride*0]= ((a-c)*qmul) >> 7;
block[stride*1 + xStride*1]= ((e-b)*qmul) >> 7;
} |