x264中psnr的计算代码如下：
static double x264_psnr( double sqe, double size )
{
double mse = sqe / (PIXEL_MAX*PIXEL_MAX * size);
if( mse <= 0.0000000001 ) /* Max 100dB */
return 100;
return -10.0 * log10( mse );
}
××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××
而输入参数sqe即ssd的计算代码为：
uint64_t x264_pixel_ssd_wxh( x264_pixel_function_t *pf, pixel *pix1, intptr_t i_pix1,
pixel *pix2, intptr_t i_pix2, int i_width, int i_height )
{
uint64_t i_ssd = 0;
int y;
int align = !(((intptr_t)pix1 | (intptr_t)pix2 | i_pix1 | i_pix2) & 15);
#define SSD(size) i_ssd += pf->ssd[size]( pix1 + y*i_pix1 + x, i_pix1, \
pix2 + y*i_pix2 + x, i_pix2 );
for( y = 0; y < i_height-15; y += 16 )
{
int x = 0;
if( align )
for( ; x < i_width-15; x += 16 )
SSD(PIXEL_16x16);
for( ; x < i_width-7; x += 8 )
SSD(PIXEL_8x16);
}
if( y < i_height-7 )
for( int x = 0; x < i_width-7; x += 8 )
SSD(PIXEL_8x8);
#undef SSD
#define SSD1 { int d = pix1[y*i_pix1+x] - pix2[y*i_pix2+x]; i_ssd += d*d; }
if( i_width & 7 )
{
for( y = 0; y < (i_height & ~7); y++ )
for( int x = i_width & ~7; x < i_width; x++ )
SSD1;
}
if( i_height & 7 )
{
for( y = i_height & ~7; y < i_height; y++ )
for( int x = 0; x < i_width; x++ )
SSD1;
}
#undef SSD1
return i_ssd;
}
static void pixel_ssd_nv12_core( pixel *pixuv1, intptr_t stride1, pixel *pixuv2, intptr_t stride2,
int width, int height, uint64_t *ssd_u, uint64_t *ssd_v )
{
*ssd_u = 0, *ssd_v = 0;
for( int y = 0; y < height; y++, pixuv1+=stride1, pixuv2+=stride2 )
for( int x = 0; x < width; x++ )
{
int du = pixuv1[2*x] - pixuv2[2*x];
int dv = pixuv1[2*x+1] - pixuv2[2*x+1];
*ssd_u += du*du;
*ssd_v += dv*dv;
}
}
void x264_pixel_ssd_nv12( x264_pixel_function_t *pf, pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2,
int i_width, int i_height, uint64_t *ssd_u, uint64_t *ssd_v )
{
pf->ssd_nv12_core( pix1, i_pix1, pix2, i_pix2, i_width&~7, i_height, ssd_u, ssd_v );
if( i_width&7 )
{
uint64_t tmp[2];
pixel_ssd_nv12_core( pix1+(i_width&~7), i_pix1, pix2+(i_width&~7), i_pix2, i_width&7, i_height, &tmp[0], &tmp[1] );
*ssd_u += tmp[0];
*ssd_v += tmp[1];
}
}
×××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××××8
****************************************************************************
* pixel_satd_WxH: sum of 4x4 Hadamard transformed differences
****************************************************************************/
static NOINLINE int x264_pixel_satd_4x4( pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2 )
{
sum2_t tmp[4][2];
sum2_t a0, a1, a2, a3, b0, b1;
sum2_t sum = 0;
for( int i = 0; i < 4; i++, pix1 += i_pix1, pix2 += i_pix2 )
{
a0 = pix1[0] - pix2[0];
a1 = pix1[1] - pix2[1];
b0 = (a0+a1) + ((a0-a1)<<BITS_PER_SUM);
a2 = pix1[2] - pix2[2];
a3 = pix1[3] - pix2[3];
b1 = (a2+a3) + ((a2-a3)<<BITS_PER_SUM);
tmp[i][0] = b0 + b1;
tmp[i][1] = b0 - b1;
}
for( int i = 0; i < 2; i++ )
{
HADAMARD4( a0, a1, a2, a3, tmp[0][i], tmp[1][i], tmp[2][i], tmp[3][i] );
a0 = abs2(a0) + abs2(a1) + abs2(a2) + abs2(a3);
sum += ((sum_t)a0) + (a0>>BITS_PER_SUM);
}
return sum >> 1;
}
static NOINLINE int x264_pixel_satd_8x4( pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2 )
{
sum2_t tmp[4][4];
sum2_t a0, a1, a2, a3;
sum2_t sum = 0;
for( int i = 0; i < 4; i++, pix1 += i_pix1, pix2 += i_pix2 )
{
a0 = (pix1[0] - pix2[0]) + ((sum2_t)(pix1[4] - pix2[4]) << BITS_PER_SUM);
a1 = (pix1[1] - pix2[1]) + ((sum2_t)(pix1[5] - pix2[5]) << BITS_PER_SUM);
a2 = (pix1[2] - pix2[2]) + ((sum2_t)(pix1[6] - pix2[6]) << BITS_PER_SUM);
a3 = (pix1[3] - pix2[3]) + ((sum2_t)(pix1[7] - pix2[7]) << BITS_PER_SUM);
HADAMARD4( tmp[i][0], tmp[i][1], tmp[i][2], tmp[i][3], a0,a1,a2,a3 );
}
for( int i = 0; i < 4; i++ )
{
HADAMARD4( a0, a1, a2, a3, tmp[0][i], tmp[1][i], tmp[2][i], tmp[3][i] );
sum += abs2(a0) + abs2(a1) + abs2(a2) + abs2(a3);
}
return (((sum_t)sum) + (sum>>BITS_PER_SUM)) >> 1;
}
#define PIXEL_SATD_C( w, h, sub )\
static int x264_pixel_satd_##w##x##h( pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2 )\
{\
int sum = sub( pix1, i_pix1, pix2, i_pix2 )\
+ sub( pix1+4*i_pix1, i_pix1, pix2+4*i_pix2, i_pix2 );\
if( w==16 )\
sum+= sub( pix1+8, i_pix1, pix2+8, i_pix2 )\
+ sub( pix1+8+4*i_pix1, i_pix1, pix2+8+4*i_pix2, i_pix2 );\
if( h==16 )\
sum+= sub( pix1+8*i_pix1, i_pix1, pix2+8*i_pix2, i_pix2 )\
+ sub( pix1+12*i_pix1, i_pix1, pix2+12*i_pix2, i_pix2 );\
if( w==16 && h==16 )\
sum+= sub( pix1+8+8*i_pix1, i_pix1, pix2+8+8*i_pix2, i_pix2 )\
+ sub( pix1+8+12*i_pix1, i_pix1, pix2+8+12*i_pix2, i_pix2 );\
return sum;\
}
PIXEL_SATD_C( 16, 16, x264_pixel_satd_8x4 )
PIXEL_SATD_C( 16, 8, x264_pixel_satd_8x4 )
PIXEL_SATD_C( 8, 16, x264_pixel_satd_8x4 )
PIXEL_SATD_C( 8, 8, x264_pixel_satd_8x4 )
PIXEL_SATD_C( 4, 16, x264_pixel_satd_4x4 )
PIXEL_SATD_C( 4, 8, x264_pixel_satd_4x4 )
static NOINLINE int sa8d_8x8( pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2 )
{
sum2_t tmp[8][4];
sum2_t a0, a1, a2, a3, a4, a5, a6, a7, b0, b1, b2, b3;
sum2_t sum = 0;
for( int i = 0; i < 8; i++, pix1 += i_pix1, pix2 += i_pix2 )
{
a0 = pix1[0] - pix2[0];
a1 = pix1[1] - pix2[1];
b0 = (a0+a1) + ((a0-a1)<<BITS_PER_SUM);
a2 = pix1[2] - pix2[2];
a3 = pix1[3] - pix2[3];
b1 = (a2+a3) + ((a2-a3)<<BITS_PER_SUM);
a4 = pix1[4] - pix2[4];
a5 = pix1[5] - pix2[5];
b2 = (a4+a5) + ((a4-a5)<<BITS_PER_SUM);
a6 = pix1[6] - pix2[6];
a7 = pix1[7] - pix2[7];
b3 = (a6+a7) + ((a6-a7)<<BITS_PER_SUM);
HADAMARD4( tmp[i][0], tmp[i][1], tmp[i][2], tmp[i][3], b0,b1,b2,b3 );
}
for( int i = 0; i < 4; i++ )
{
HADAMARD4( a0, a1, a2, a3, tmp[0][i], tmp[1][i], tmp[2][i], tmp[3][i] );
HADAMARD4( a4, a5, a6, a7, tmp[4][i], tmp[5][i], tmp[6][i], tmp[7][i] );
b0 = abs2(a0+a4) + abs2(a0-a4);
b0 += abs2(a1+a5) + abs2(a1-a5);
b0 += abs2(a2+a6) + abs2(a2-a6);
b0 += abs2(a3+a7) + abs2(a3-a7);
sum += (sum_t)b0 + (b0>>BITS_PER_SUM);
}
return sum;
}
static int x264_pixel_sa8d_8x8( pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2 )
{
int sum = sa8d_8x8( pix1, i_pix1, pix2, i_pix2 );
return (sum+2)>>2;
}
static int x264_pixel_sa8d_16x16( pixel *pix1, intptr_t i_pix1, pixel *pix2, intptr_t i_pix2 )
{
int sum = sa8d_8x8( pix1, i_pix1, pix2, i_pix2 )
+ sa8d_8x8( pix1+8, i_pix1, pix2+8, i_pix2 )
+ sa8d_8x8( pix1+8*i_pix1, i_pix1, pix2+8*i_pix2, i_pix2 )
+ sa8d_8x8( pix1+8+8*i_pix1, i_pix1, pix2+8+8*i_pix2, i_pix2 );
return (sum+2)>>2;
}
static NOINLINE uint64_t pixel_hadamard_ac( pixel *pix, intptr_t stride )
{
sum2_t tmp[32];
sum2_t a0, a1, a2, a3, dc;
sum2_t sum4 = 0, sum8 = 0;
for( int i = 0; i < 8; i++, pix+=stride )
{
sum2_t *t = tmp + (i&3) + (i&4)*4;
a0 = (pix[0]+pix[1]) + ((sum2_t)(pix[0]-pix[1])<<BITS_PER_SUM);
a1 = (pix[2]+pix[3]) + ((sum2_t)(pix[2]-pix[3])<<BITS_PER_SUM);
t[0] = a0 + a1;
t[4] = a0 - a1;
a2 = (pix[4]+pix[5]) + ((sum2_t)(pix[4]-pix[5])<<BITS_PER_SUM);
a3 = (pix[6]+pix[7]) + ((sum2_t)(pix[6]-pix[7])<<BITS_PER_SUM);
t[8] = a2 + a3;
t[12] = a2 - a3;
}
for( int i = 0; i < 8; i++ )
{
HADAMARD4( a0, a1, a2, a3, tmp[i*4+0], tmp[i*4+1], tmp[i*4+2], tmp[i*4+3] );
tmp[i*4+0] = a0;
tmp[i*4+1] = a1;
tmp[i*4+2] = a2;
tmp[i*4+3] = a3;
sum4 += abs2(a0) + abs2(a1) + abs2(a2) + abs2(a3);
}
for( int i = 0; i < 8; i++ )
{
HADAMARD4( a0,a1,a2,a3, tmp[i], tmp[8+i], tmp[16+i], tmp[24+i] );
sum8 += abs2(a0) + abs2(a1) + abs2(a2) + abs2(a3);
}
dc = (sum_t)(tmp[0] + tmp[8] + tmp[16] + tmp[24]);
sum4 = (sum_t)sum4 + (sum4>>BITS_PER_SUM) - dc;
sum8 = (sum_t)sum8 + (sum8>>BITS_PER_SUM) - dc;
return ((uint64_t)sum8<<32) + sum4;
}