Yucom/jxrlib/image/encode/strFwdTransform.c
2020-09-29 14:29:06 -05:00

1111 lines
36 KiB
C

//*@@@+++@@@@******************************************************************
//
// Copyright © Microsoft Corp.
// All rights reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are met:
//
// • Redistributions of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
// • Redistributions in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
// AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
// ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
// LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
// CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
// SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
// CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
// ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
// POSSIBILITY OF SUCH DAMAGE.
//
//*@@@---@@@@******************************************************************
#include "strTransform.h"
#include "encode.h"
/** rotation by pi/8 **/
#define ROTATE1(a, b) (b) -= (((a) + 1) >> 1), (a) += (((b) + 1) >> 1) // this works well too
#define ROTATE2(a, b) (b) -= (((a)*3 + 4) >> 3), (a) += (((b)*3 + 4) >> 3) // this works well too
/** local functions **/
static Void fwdOddOdd(PixelI *, PixelI *, PixelI *, PixelI *);
static Void fwdOddOddPre(PixelI *, PixelI *, PixelI *, PixelI *);
static Void fwdOdd(PixelI *, PixelI *, PixelI *, PixelI *);
static Void strDCT2x2alt(PixelI * a, PixelI * b, PixelI * c, PixelI * d);
static Void strHSTenc1(PixelI *, PixelI *);
static Void strHSTenc(PixelI *, PixelI *, PixelI *, PixelI *);
static Void strHSTenc1_edge (PixelI *pa, PixelI *pd);
//static Void scaleDownUp0(PixelI *, PixelI *);
//static Void scaleDownUp1(PixelI *, PixelI *);
//static Void scaleDownUp2(PixelI *, PixelI *);
//#define FOURBUTTERFLY_ENC_ALT(p, i00, i01, i02, i03, i10, i11, i12, i13, \
// i20, i21, i22, i23, i30, i31, i32, i33) \
// strHSTenc(&p[i00], &p[i01], &p[i02], &p[i03]); \
// strHSTenc(&p[i10], &p[i11], &p[i12], &p[i13]); \
// strHSTenc(&p[i20], &p[i21], &p[i22], &p[i23]); \
// strHSTenc(&p[i30], &p[i31], &p[i32], &p[i33]); \
// strHSTenc1(&p[i00], &p[i03]); \
// strHSTenc1(&p[i10], &p[i13]); \
// strHSTenc1(&p[i20], &p[i23]); \
// strHSTenc1(&p[i30], &p[i33])
/** DCT stuff **/
/** data order before DCT **/
/** 0 1 2 3 **/
/** 4 5 6 7 **/
/** 8 9 10 11 **/
/** 12 13 14 15 **/
/** data order after DCT **/
/** 0 8 4 6 **/
/** 2 10 14 12 **/
/** 1 11 15 13 **/
/** 9 3 7 5 **/
/** reordering should be combined with zigzag scan **/
Void strDCT4x4Stage1(PixelI * p)
{
/** butterfly **/
//FOURBUTTERFLY(p, 0, 4, 8, 12, 1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15);
FOURBUTTERFLY_HARDCODED1(p);
/** top left corner, butterfly => butterfly **/
strDCT2x2up(&p[0], &p[1], &p[2], &p[3]);
/** bottom right corner, pi/8 rotation => pi/8 rotation **/
fwdOddOdd(&p[15], &p[14], &p[13], &p[12]);
/** top right corner, butterfly => pi/8 rotation **/
fwdOdd(&p[5], &p[4], &p[7], &p[6]);
/** bottom left corner, pi/8 rotation => butterfly **/
fwdOdd(&p[10], &p[8], &p[11], &p[9]);
}
Void strDCT4x4SecondStage(PixelI * p)
{
/** butterfly **/
FOURBUTTERFLY(p, 0, 192, 48, 240, 64, 128, 112, 176,16, 208, 32, 224, 80, 144, 96, 160);
/** top left corner, butterfly => butterfly **/
strDCT2x2up(&p[0], &p[64], &p[16], &p[80]);
/** bottom right corner, pi/8 rotation => pi/8 rotation **/
fwdOddOdd(&p[160], &p[224], &p[176], &p[240]);
/** top right corner, butterfly => pi/8 rotation **/
fwdOdd(&p[128], &p[192], &p[144], &p[208]);
/** bottom left corner, pi/8 rotation => butterfly **/
fwdOdd(&p[32], &p[48], &p[96], &p[112]);
}
Void strNormalizeEnc(PixelI* p, Bool bChroma)
{
int i;
if (!bChroma) {
//for (i = 0; i < 256; i += 16) {
// p[i] = (p[i] + 1) >> 2;
//}
}
else {
for (i = 0; i < 256; i += 16) {
p[i] >>= 1;
}
}
}
/** 2x2 DCT with pre-scaling - for use on encoder side **/
Void strDCT2x2dnEnc(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
PixelI a, b, c, d, C, t;
a = (*pa + 0) >> 1;
b = (*pb + 0) >> 1;
C = (*pc + 0) >> 1;
d = (*pd + 0) >> 1;
//PixelI t1, t2;
a += d;
b -= C;
t = ((a - b) >> 1);
c = t - d;
d = t - C;
a -= d;
b += c;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
/** pre filter stuff **/
/** 2-point pre for boundaries **/
Void strPre2(PixelI * pa, PixelI * pb)
{
PixelI a, b;
a = *pa;
b = *pb;
/** rotate **/
b -= ((a + 2) >> 2);
a -= ((b + 1) >> 1);
a -= (b >> 5);
a -= (b >> 9);
a -= (b >> 13);
b -= ((a + 2) >> 2);
*pa = a;
*pb = b;
}
Void strPre2x2(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
PixelI a, b, c, d;
a = *pa;
b = *pb;
c = *pc;
d = *pd;
/** butterflies **/
a += d;
b += c;
d -= (a + 1) >> 1;
c -= (b + 1) >> 1;
/** rotate **/
b -= ((a + 2) >> 2);
a -= ((b + 1) >> 1);
a -= (b >> 5);
a -= (b >> 9);
a -= (b >> 13);
b -= ((a + 2) >> 2);
/** butterflies **/
d += (a + 1) >> 1;
c += (b + 1) >> 1;
a -= d;
b -= c;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
/** 4-point pre for boundaries **/
Void strPre4(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
PixelI a, b, c, d;
a = *pa;
b = *pb;
c = *pc;
d = *pd;
a += d, b += c;
d -= ((a + 1) >> 1), c -= ((b + 1) >> 1);
ROTATE1(c, d);
strHSTenc1_edge(&a, &d); strHSTenc1_edge(&b, &c);
d += ((a + 1) >> 1), c += ((b + 1) >> 1);
a -= d, b -= c;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
/*****************************************************************************************
Input data offsets:
(15)(14)|(10+64)(11+64) p0 (15)(14)|(74)(75)
(13)(12)|( 8+64)( 9+64) (13)(12)|(72)(73)
--------+-------------- --------+--------
( 5)( 4)|( 0+64) (1+64) p1 ( 5)( 4)|(64)(65)
( 7)( 6)|( 2+64) (3+64) ( 7)( 6)|(66)(67)
*****************************************************************************************/
Void strPre4x4Stage1Split(PixelI *p0, PixelI *p1, Int iOffset)
{
PixelI *p2 = p0 + 72 - iOffset;
PixelI *p3 = p1 + 64 - iOffset;
p0 += 12;
p1 += 4;
/** butterfly & scaling **/
strHSTenc(p0 + 0, p2 + 0, p1 + 0, p3 + 0);
strHSTenc(p0 + 1, p2 + 1, p1 + 1, p3 + 1);
strHSTenc(p0 + 2, p2 + 2, p1 + 2, p3 + 2);
strHSTenc(p0 + 3, p2 + 3, p1 + 3, p3 + 3);
strHSTenc1(p0 + 0, p3 + 0);
strHSTenc1(p0 + 1, p3 + 1);
strHSTenc1(p0 + 2, p3 + 2);
strHSTenc1(p0 + 3, p3 + 3);
/** anti diagonal corners: rotation by pi/8 **/
ROTATE1(p1[2], p1[3]);
ROTATE1(p1[0], p1[1]);
ROTATE1(p2[1], p2[3]);
ROTATE1(p2[0], p2[2]);
/** bottom right corner: pi/8 rotation => pi/8 rotation **/
fwdOddOddPre(p3 + 0, p3 + 1, p3 + 2, p3 + 3);
/** butterfly **/
strDCT2x2dn(p0 + 0, p2 + 0, p1 + 0, p3 + 0);
strDCT2x2dn(p0 + 1, p2 + 1, p1 + 1, p3 + 1);
strDCT2x2dn(p0 + 2, p2 + 2, p1 + 2, p3 + 2);
strDCT2x2dn(p0 + 3, p2 + 3, p1 + 3, p3 + 3);
}
Void strPre4x4Stage1(PixelI* p, Int iOffset)
{
strPre4x4Stage1Split(p, p + 16, iOffset);
}
/*****************************************************************************************
Input data offsets:
(15)(14)|(10+32)(11+32) p0 (15)(14)|(42)(43)
(13)(12)|( 8+32)( 9+32) (13)(12)|(40)(41)
--------+-------------- --------+--------
( 5)( 4)|( 0+32)( 1+32) p1 ( 5)( 4)|(32)(33)
( 7)( 6)|( 2+32)( 3+32) ( 7)( 6)|(34)(35)
*****************************************************************************************/
Void strPre4x4Stage2Split(PixelI* p0, PixelI* p1)
{
/** butterfly **/
strHSTenc(p0 - 96, p0 + 96, p1 - 112, p1 + 80);
strHSTenc(p0 - 32, p0 + 32, p1 - 48, p1 + 16);
strHSTenc(p0 - 80, p0 + 112, p1 - 128, p1 + 64);
strHSTenc(p0 - 16, p0 + 48, p1 - 64, p1 + 0);
strHSTenc1(p0 - 96, p1 + 80);
strHSTenc1(p0 - 32, p1 + 16);
strHSTenc1(p0 - 80, p1 + 64);
strHSTenc1(p0 - 16, p1 + 0);
/** anti diagonal corners: rotation **/
ROTATE1(p1[-48], p1[-112]);
ROTATE1(p1[-64], p1[-128]);
ROTATE1(p0[112], p0[ 96]);
ROTATE1(p0[ 48], p0[ 32]);
/** bottom right corner: pi/8 rotation => pi/8 rotation **/
fwdOddOddPre(p1 + 0, p1 + 64, p1 + 16, p1 + 80);
/** butterfly **/
strDCT2x2dn(p0 - 96, p1 - 112, p0 + 96, p1 + 80);
strDCT2x2dn(p0 - 32, p1 - 48, p0 + 32, p1 + 16);
strDCT2x2dn(p0 - 80, p1 - 128, p0 + 112, p1 + 64);
strDCT2x2dn(p0 - 16, p1 - 64, p0 + 48, p1 + 0);
}
/**
Hadamard+Scale transform
for some strange reason, breaking up the function into two blocks, strHSTenc1 and strHSTenc
seems to work faster
**/
static Void strHSTenc(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
/** different realization : does rescaling as well! **/
PixelI a, b, c, d;
a = *pa;
b = *pb;
d = *pc;
c = *pd;
a += c;
b -= d;
c = ((a - b) >> 1) - c;
d += (b >> 1);
b += c;
a -= (d * 3 + 4) >> 3;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
static Void strHSTenc1(PixelI *pa, PixelI *pd)
{
/** different realization : does rescaling as well! **/
PixelI a, d;
a = *pa;
d = *pd;
d -= (a >> 7);
d += (a >> 10);
//a -= (d * 3 + 4) >> 3;
d -= (a * 3 + 0) >> 4;
a -= (d * 3 + 0) >> 3;
d = (a >> 1) - d;
a -= d;
*pa = a;
*pd = d;
}
static Void strHSTenc1_edge (PixelI *pa, PixelI *pd)
{
/** different realizion as compared to scaling operator for 2D case **/
PixelI a, d;
a = *pa;
d = -(*pd); // Negative sign needed here for 1D scaling case to ensure correct scaling.
a -= d;
d += (a >> 1);
a -= (d * 3 + 4) >> 3;
// End new operations
//Scaling modification of adding 7/1024 in two steps (without multiplication by 7).
d -= (a >> 7);
d += (a >> 10);
d -= (a * 3 + 0) >> 4;
a -= (d * 3 + 0) >> 3;
d = (a >> 1) - d;
a -= d;
*pa = a;
*pd = d;
}
/** Kron(Rotate(pi/8), Rotate(pi/8)) **/\
static Void fwdOddOdd(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
PixelI a, b, c, d, t1, t2;
a = *pa;
b = -*pb;
c = -*pc;
d = *pd;
/** butterflies **/
d += a;
c -= b;
a -= (t1 = d >> 1);
b += (t2 = c >> 1);
/** rotate pi/4 **/
a += (b * 3 + 4) >> 3;
b -= (a * 3 + 3) >> 2;
a += (b * 3 + 3) >> 3;
/** butterflies **/
b -= t2;
a += t1;
c += b;
d -= a;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
/** Kron(Rotate(pi/8), Rotate(pi/8)) **/
static Void fwdOddOddPre(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
PixelI a, b, c, d, t1, t2;
a = *pa;
b = *pb;
c = *pc;
d = *pd;
/** butterflies **/
d += a;
c -= b;
a -= (t1 = d >> 1);
b += (t2 = c >> 1);
/** rotate pi/4 **/
a += (b * 3 + 4) >> 3;
b -= (a * 3 + 2) >> 2;
a += (b * 3 + 6) >> 3;
/** butterflies **/
b -= t2;
a += t1;
c += b;
d -= a;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
/** Kron(Rotate(pi/8), [1 1; 1 -1]/sqrt(2)) **/
/** [a b c d] => [D C A B] **/
Void fwdOdd(PixelI *pa, PixelI *pb, PixelI *pc, PixelI *pd)
{
PixelI a, b, c, d;
a = *pa;
b = *pb;
c = *pc;
d = *pd;
/** butterflies **/
b -= c;
a += d;
c += (b + 1) >> 1;
d = ((a + 1) >> 1) - d;
/** rotate pi/8 **/
ROTATE2(a, b);
ROTATE2(c, d);
/** butterflies **/
d += (b) >> 1;
c -= (a + 1) >> 1;
b -= d;
a += c;
*pa = a;
*pb = b;
*pc = c;
*pd = d;
}
/*************************************************************************
Top-level function to tranform possible part of a macroblock
*************************************************************************/
Void transformMacroblock(CWMImageStrCodec * pSC)
{
OVERLAP olOverlap = pSC->WMISCP.olOverlap;
COLORFORMAT cfColorFormat = pSC->m_param.cfColorFormat;
Bool left = (pSC->cColumn == 0), right = (pSC->cColumn == pSC->cmbWidth);
Bool top = (pSC->cRow == 0), bottom = (pSC->cRow == pSC->cmbHeight);
Bool leftORright = (left || right), topORbottom = (top || bottom);
Bool topORleft = (left || top);// rightORbottom = (right || bottom);
Bool leftAdjacentColumn = (pSC->cColumn == 1), rightAdjacentColumn = (pSC->cColumn == pSC->cmbWidth - 1);
// Bool topAdjacentRow = (pSC->cRow == 1), bottomAdjacentRow = (pSC->cRow == pSC->cmbHeight - 1);
PixelI * p = NULL;// * pt = NULL;
Int i, j;
Int iNumChromaFullPlanes = (Int)((YUV_420 == cfColorFormat || YUV_422 == cfColorFormat) ?
1 : pSC->m_param.cNumChannels);
#define mbX pSC->mbX
#define mbY pSC->mbY
#define tileX pSC->tileX
#define tileY pSC->tileY
#define bVertTileBoundary pSC->bVertTileBoundary
#define bHoriTileBoundary pSC->bHoriTileBoundary
#define bOneMBLeftVertTB pSC->bOneMBLeftVertTB
#define bOneMBRightVertTB pSC->bOneMBRightVertTB
#define iPredBefore pSC->iPredBefore
#define iPredAfter pSC->iPredAfter
if (pSC->WMISCP.bUseHardTileBoundaries) {
//Add tile location information
if (pSC->cColumn == 0) {
bVertTileBoundary = FALSE;
tileY = 0;
}
bOneMBLeftVertTB = bOneMBRightVertTB = FALSE;
if(tileY > 0 && tileY <= pSC->WMISCP.cNumOfSliceMinus1H && (pSC->cColumn - 1) == pSC->WMISCP.uiTileY[tileY])
bOneMBRightVertTB = TRUE;
if(tileY < pSC->WMISCP.cNumOfSliceMinus1H && pSC->cColumn == pSC->WMISCP.uiTileY[tileY + 1]) {
bVertTileBoundary = TRUE;
tileY++;
}
else
bVertTileBoundary = FALSE;
if(tileY < pSC->WMISCP.cNumOfSliceMinus1H && (pSC->cColumn + 1) == pSC->WMISCP.uiTileY[tileY + 1])
bOneMBLeftVertTB = TRUE;
if (pSC->cRow == 0) {
bHoriTileBoundary = FALSE;
tileX = 0;
}
else if(mbY != pSC->cRow && tileX < pSC->WMISCP.cNumOfSliceMinus1V && pSC->cRow == pSC->WMISCP.uiTileX[tileX + 1]) {
bHoriTileBoundary = TRUE;
tileX++;
}
else if(mbY != pSC->cRow)
bHoriTileBoundary = FALSE;
}
else {
bVertTileBoundary = FALSE;
bHoriTileBoundary = FALSE;
bOneMBLeftVertTB = FALSE;
bOneMBRightVertTB = FALSE;
}
mbX = pSC->cColumn, mbY = pSC->cRow;
//================================================================
// 400_Y, 444_YUV
for(i = 0; i < iNumChromaFullPlanes; ++i)
{
PixelI* const p0 = pSC->p0MBbuffer[i];//(0 == i ? pSC->pY0 : (1 == i ? pSC->pU0 : pSC->pV0));
PixelI* const p1 = pSC->p1MBbuffer[i];//(0 == i ? pSC->pY1 : (1 == i ? pSC->pU1 : pSC->pV1));
//================================
// first level overlap
if(OL_NONE != olOverlap)
{
/* Corner operations */
if ((top || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p1 + 0, p1 + 1, p1 + 2, p1 + 3);
if ((top || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p1 - 59, p1 - 60, p1 - 57, p1 - 58);
if ((bottom || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p0 + 48 + 10, p0 + 48 + 11, p0 + 48 + 8, p0 + 48 + 9);
if ((bottom || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p0 - 1, p0 - 2, p0 - 3, p0 - 4);
if(!right && !bottom)
{
if (top || bHoriTileBoundary)
{
for (j = ((left || bVertTileBoundary) ? 0 : -64); j < 192; j += 64)
{
p = p1 + j;
strPre4(p + 5, p + 4, p + 64, p + 65);
strPre4(p + 7, p + 6, p + 66, p + 67);
p = NULL;
}
}
else
{
for (j = ((left || bVertTileBoundary) ? 0 : -64); j < 192; j += 64)
{
strPre4x4Stage1Split(p0 + 48 + j, p1 + j, 0);
}
}
if (left || bVertTileBoundary)
{
if (!top && !bHoriTileBoundary)
{
strPre4(p0 + 58, p0 + 56, p1 + 0, p1 + 2);
strPre4(p0 + 59, p0 + 57, p1 + 1, p1 + 3);
}
for (j = -64; j < -16; j += 16)
{
p = p1 + j;
strPre4(p + 74, p + 72, p + 80, p + 82);
strPre4(p + 75, p + 73, p + 81, p + 83);
p = NULL;
}
}
else
{
for (j = -64; j < -16; j += 16)
{
strPre4x4Stage1(p1 + j, 0);
}
}
strPre4x4Stage1(p1 + 0, 0);
strPre4x4Stage1(p1 + 16, 0);
strPre4x4Stage1(p1 + 32, 0);
strPre4x4Stage1(p1 + 64, 0);
strPre4x4Stage1(p1 + 80, 0);
strPre4x4Stage1(p1 + 96, 0);
strPre4x4Stage1(p1 + 128, 0);
strPre4x4Stage1(p1 + 144, 0);
strPre4x4Stage1(p1 + 160, 0);
}
if (bottom || bHoriTileBoundary)
{
for (j = ((left || bVertTileBoundary) ? 48 : -16); j < (right ? -16 : 240); j += 64)
{
p = p0 + j;
strPre4(p + 15, p + 14, p + 74, p + 75);
strPre4(p + 13, p + 12, p + 72, p + 73);
p = NULL;
}
}
if ((right || bVertTileBoundary) && !bottom)
{
if (!top && !bHoriTileBoundary)
{
strPre4(p0 - 1, p0 - 3, p1 - 59, p1 - 57);
strPre4(p0 - 2, p0 - 4, p1 - 60, p1 - 58);
}
for (j = -64; j < -16; j += 16)
{
p = p1 + j;
strPre4(p + 15, p + 13, p + 21, p + 23);
strPre4(p + 14, p + 12, p + 20, p + 22);
p = NULL;
}
}
}
//================================
// first level transform
if (!top)
{
for (j = (left ? 48 : -16); j < (right ? 48 : 240); j += 64)
{
strDCT4x4Stage1(p0 + j);
}
}
if (!bottom)
{
for (j = (left ? 0 : -64); j < (right ? 0 : 192); j += 64)
{
strDCT4x4Stage1(p1 + j + 0);
strDCT4x4Stage1(p1 + j + 16);
strDCT4x4Stage1(p1 + j + 32);
}
}
//================================
// second level overlap
if (OL_TWO == olOverlap)
{
/* Corner operations */
if ((top || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p1 + 0, p1 + 64, p1 + 0 + 16, p1 + 64 + 16);
if ((top || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p1 - 128, p1 - 64, p1 - 128 + 16, p1 - 64 + 16);
if ((bottom || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p0 + 32, p0 + 96, p0 + 32 + 16, p0 + 96 + 16);
if ((bottom || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p0 - 96, p0 - 32, p0 - 96 + 16, p0 - 32 + 16);
if ((leftORright || bVertTileBoundary) && (!topORbottom && !bHoriTileBoundary))
{
if (left || bVertTileBoundary) {
j = 0;
strPre4(p0 + j + 32, p0 + j + 48, p1 + j + 0, p1 + j + 16);
strPre4(p0 + j + 96, p0 + j + 112, p1 + j + 64, p1 + j + 80);
}
if (right || bVertTileBoundary) {
j = -128;
strPre4(p0 + j + 32, p0 + j + 48, p1 + j + 0, p1 + j + 16);
strPre4(p0 + j + 96, p0 + j + 112, p1 + j + 64, p1 + j + 80);
}
}
if (!leftORright && !bVertTileBoundary)
{
if (topORbottom || bHoriTileBoundary)
{
if (top || bHoriTileBoundary) {
p = p1;
strPre4(p - 128, p - 64, p + 0, p + 64);
strPre4(p - 112, p - 48, p + 16, p + 80);
p = NULL;
}
if (bottom || bHoriTileBoundary) {
p = p0 + 32;
strPre4(p - 128, p - 64, p + 0, p + 64);
strPre4(p - 112, p - 48, p + 16, p + 80);
p = NULL;
}
}
else
{
strPre4x4Stage2Split(p0, p1);
}
}
}
//================================
// second level transform
if (!topORleft){
if (pSC->m_param.bScaledArith) {
strNormalizeEnc(p0 - 256, (i != 0));
}
strDCT4x4SecondStage(p0 - 256);
}
}
//================================================================
// 420_UV
for(i = 0; i < (YUV_420 == cfColorFormat? 2 : 0); ++i)
{
PixelI* const p0 = pSC->p0MBbuffer[1 + i];//(0 == i ? pSC->pU0 : pSC->pV0);
PixelI* const p1 = pSC->p1MBbuffer[1 + i];//(0 == i ? pSC->pU1 : pSC->pV1);
//================================
// first level overlap (420_UV)
if (OL_NONE != olOverlap)
{
/* Corner operations */
if ((top || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p1 + 0, p1 + 1, p1 + 2, p1 + 3);
if ((top || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p1 - 27, p1 - 28, p1 - 25, p1 - 26);
if ((bottom || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p0 + 16 + 10, p0 + 16 + 11, p0 + 16 + 8, p0 + 16 + 9);
if ((bottom || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p0 - 1, p0 - 2, p0 - 3, p0 - 4);
if(!right && !bottom)
{
if (top || bHoriTileBoundary)
{
for (j = ((left || bVertTileBoundary) ? 0 : -32); j < 32; j += 32)
{
p = p1 + j;
strPre4(p + 5, p + 4, p + 32, p + 33);
strPre4(p + 7, p + 6, p + 34, p + 35);
p = NULL;
}
}
else
{
for (j = ((left || bVertTileBoundary) ? 0: -32); j < 32; j += 32)
{
strPre4x4Stage1Split(p0 + 16 + j, p1 + j, 32);
}
}
if (left || bVertTileBoundary)
{
if (!top && !bHoriTileBoundary)
{
strPre4(p0 + 26, p0 + 24, p1 + 0, p1 + 2);
strPre4(p0 + 27, p0 + 25, p1 + 1, p1 + 3);
}
strPre4(p1 + 10, p1 + 8, p1 + 16, p1 + 18);
strPre4(p1 + 11, p1 + 9, p1 + 17, p1 + 19);
}
else if (!bVertTileBoundary)
{
strPre4x4Stage1(p1 - 32, 32);
}
strPre4x4Stage1(p1, 32);
}
if (bottom || bHoriTileBoundary)
{
for (j = ((left || bVertTileBoundary) ? 16: -16); j < (right ? -16: 32); j += 32)
{
p = p0 + j;
strPre4(p + 15, p + 14, p + 42, p + 43);
strPre4(p + 13, p + 12, p + 40, p + 41);
p = NULL;
}
}
if ((right || bVertTileBoundary) && !bottom)
{
if (!top && !bHoriTileBoundary)
{
strPre4(p0 - 1, p0 - 3, p1 - 27, p1 - 25);
strPre4(p0 - 2, p0 - 4, p1 - 28, p1 - 26);
}
strPre4(p1 - 17, p1 - 19, p1 - 11, p1 - 9);
strPre4(p1 - 18, p1 - 20, p1 - 12, p1 - 10);
}
}
//================================
// first level transform (420_UV)
if (!top)
{
for (j = (left ? 16 : -16); j < (right ? 16 : 48); j += 32)
{
strDCT4x4Stage1(p0 + j);
}
}
if (!bottom)
{
for (j = (left ? 0 : -32); j < (right ? 0 : 32); j += 32)
{
strDCT4x4Stage1(p1 + j);
}
}
//================================
// second level overlap (420_UV)
if (OL_TWO == olOverlap)
{
if ((leftAdjacentColumn || bOneMBRightVertTB) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p1 - 64 + 0, *(p1 - 64 + 32));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (top || bHoriTileBoundary))
iPredBefore[i][0] = *(p1 + 0);
if ((right || bVertTileBoundary) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p1 - 64 + 32, iPredBefore[i][0]);
if ((leftAdjacentColumn || bOneMBRightVertTB) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p0 - 64 + 16, *(p0 - 64 + 48));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (bottom || bHoriTileBoundary))
iPredBefore[i][1] = *(p0 + 16);
if ((right || bVertTileBoundary) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p0 - 64 + 48, iPredBefore[i][1]);
if ((leftORright || bVertTileBoundary) && !topORbottom && !bHoriTileBoundary)
{
if (left || bVertTileBoundary)
strPre2(p0 + 0 + 16, p1 + 0);
if (right || bVertTileBoundary)
strPre2(p0 + -32 + 16, p1 + -32);
}
if (!leftORright)
{
if ((topORbottom || bHoriTileBoundary) && !bVertTileBoundary)
{
if (top || bHoriTileBoundary)
strPre2(p1 - 32, p1);
if (bottom || bHoriTileBoundary)
strPre2(p0 + 16 - 32, p0 + 16);
}
else if (!topORbottom && !bHoriTileBoundary && !bVertTileBoundary)
strPre2x2(p0 - 16, p0 + 16, p1 - 32, p1);
}
if ((leftAdjacentColumn || bOneMBRightVertTB) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p1 - 64 + 0, *(p1 - 64 + 32));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (top || bHoriTileBoundary))
iPredAfter[i][0] = *(p1 + 0);
if ((right || bVertTileBoundary) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p1 - 64 + 32, iPredAfter[i][0]);
if ((leftAdjacentColumn || bOneMBRightVertTB) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p0 - 64 + 16, *(p0 - 64 + 48));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (bottom || bHoriTileBoundary))
iPredAfter[i][1] = *(p0 + 16);
if ((right || bVertTileBoundary) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p0 - 64 + 48, iPredAfter[i][1]);
}
//================================
// second level transform (420_UV)
if (!topORleft)
{
if (!pSC->m_param.bScaledArith) {
strDCT2x2dn(p0 - 64, p0 - 32, p0 - 48, p0 - 16);
}
else {
strDCT2x2dnEnc(p0 - 64, p0 - 32, p0 - 48, p0 - 16);
}
}
}
//================================================================
// 422_UV
for(i = 0; i < (YUV_422 == cfColorFormat? 2 : 0); ++i)
{
PixelI* const p0 = pSC->p0MBbuffer[1 + i];//(0 == i ? pSC->pU0 : pSC->pV0);
PixelI* const p1 = pSC->p1MBbuffer[1 + i];//(0 == i ? pSC->pU1 : pSC->pV1);
//================================
// first level overlap (422_UV)
if (OL_NONE != olOverlap)
{
/* Corner operations */
if ((top || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p1 + 0, p1 + 1, p1 + 2, p1 + 3);
if ((top || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p1 - 59, p1 - 60, p1 - 57, p1 - 58);
if ((bottom || bHoriTileBoundary) && (left || bVertTileBoundary))
strPre4(p0 + 48 + 10, p0 + 48 + 11, p0 + 48 + 8, p0 + 48 + 9);
if ((bottom || bHoriTileBoundary) && (right || bVertTileBoundary))
strPre4(p0 - 1, p0 - 2, p0 - 3, p0 - 4);
if(!right && !bottom)
{
if (top || bHoriTileBoundary)
{
for (j = ((left || bVertTileBoundary) ? 0 : -64); j < 64; j += 64)
{
p = p1 + j;
strPre4(p + 5, p + 4, p + 64, p + 65);
strPre4(p + 7, p + 6, p + 66, p + 67);
p = NULL;
}
}
else
{
for (j = ((left || bVertTileBoundary) ? 0: -64); j < 64; j += 64)
{
strPre4x4Stage1Split(p0 + 48 + j, p1 + j, 0);
}
}
if (left || bVertTileBoundary)
{
if (!top && !bHoriTileBoundary)
{
strPre4(p0 + 58, p0 + 56, p1 + 0, p1 + 2);
strPre4(p0 + 59, p0 + 57, p1 + 1, p1 + 3);
}
for (j = 0; j < 48; j += 16)
{
p = p1 + j;
strPre4(p + 10, p + 8, p + 16, p + 18);
strPre4(p + 11, p + 9, p + 17, p + 19);
p = NULL;
}
}
else if (!bVertTileBoundary)
{
for (j = -64; j < -16; j += 16)
{
strPre4x4Stage1(p1 + j, 0);
}
}
strPre4x4Stage1(p1 + 0, 0);
strPre4x4Stage1(p1 + 16, 0);
strPre4x4Stage1(p1 + 32, 0);
}
if (bottom || bHoriTileBoundary)
{
for (j = ((left || bVertTileBoundary) ? 48: -16); j < (right ? -16: 112); j += 64)
{
p = p0 + j;
strPre4(p + 15, p + 14, p + 74, p + 75);
strPre4(p + 13, p + 12, p + 72, p + 73);
p = NULL;
}
}
if ((right || bVertTileBoundary) && !bottom)
{
if (!top && !bHoriTileBoundary)
{
strPre4(p0 - 1, p0 - 3, p1 - 59, p1 - 57);
strPre4(p0 - 2, p0 - 4, p1 - 60, p1 - 58);
}
for (j = -64; j < -16; j += 16)
{
p = p1 + j;
strPre4(p + 15, p + 13, p + 21, p + 23);
strPre4(p + 14, p + 12, p + 20, p + 22);
p = NULL;
}
}
}
//================================
// first level transform (422_UV)
if (!top)
{
for (j = (left ? 48 : -16); j < (right ? 48 : 112); j += 64)
{
strDCT4x4Stage1(p0 + j);
}
}
if (!bottom)
{
for (j = (left ? 0 : -64); j < (right ? 0 : 64); j += 64)
{
strDCT4x4Stage1(p1 + j + 0);
strDCT4x4Stage1(p1 + j + 16);
strDCT4x4Stage1(p1 + j + 32);
}
}
//================================
// second level overlap (422_UV)
if (OL_TWO == olOverlap)
{
if ((leftAdjacentColumn || bOneMBRightVertTB) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p1 - 128 + 0, *(p1 - 128 + 64));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (top || bHoriTileBoundary))
iPredBefore[i][0] = *(p1 + 0);
if ((right || bVertTileBoundary) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p1 - 128 + 64, iPredBefore[i][0]);
if ((leftAdjacentColumn || bOneMBRightVertTB) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p0 - 128 + 48, *(p0 - 128 + 112));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (bottom || bHoriTileBoundary))
iPredBefore[i][1] = *(p0 + 48);
if ((right || bVertTileBoundary) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_DIFF(p0 - 128 + 112, iPredBefore[i][1]);
if (!bottom)
{
if (leftORright || bVertTileBoundary)
{
if (!top && !bHoriTileBoundary)
{
if (left || bVertTileBoundary)
strPre2(p0 + 48 + 0, p1 + 0);
if (right || bVertTileBoundary)
strPre2(p0 + 48 + -64, p1 + -64);
}
if (left || bVertTileBoundary)
strPre2(p1 + 16, p1 + 16 + 16);
if (right || bVertTileBoundary)
strPre2(p1 + -48, p1 + -48 + 16);
}
if (!leftORright && !bVertTileBoundary)
{
if (top || bHoriTileBoundary)
strPre2(p1 - 64, p1);
else
strPre2x2(p0 - 16, p0 + 48, p1 - 64, p1);
strPre2x2(p1 - 48, p1 + 16, p1 - 32, p1 + 32);
}
}
if ((bottom || bHoriTileBoundary) && (!leftORright && !bVertTileBoundary))
strPre2(p0 - 16, p0 + 48);
if ((leftAdjacentColumn || bOneMBRightVertTB) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p1 - 128 + 0, *(p1 - 128 + 64));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (top || bHoriTileBoundary))
iPredAfter[i][0] = *(p1 + 0);
if ((right || bVertTileBoundary) && (top || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p1 - 128 + 64, iPredAfter[i][0]);
if ((leftAdjacentColumn || bOneMBRightVertTB) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p0 - 128 + 48, *(p0 - 128 + 112));
if ((rightAdjacentColumn || bOneMBLeftVertTB) && (bottom || bHoriTileBoundary))
iPredAfter[i][1] = *(p0 + 48);
if ((right || bVertTileBoundary) && (bottom || bHoriTileBoundary))
COMPUTE_CORNER_PRED_ADD(p0 - 128 + 112, iPredAfter[i][1]);
}
//================================
// second level transform (422_UV)
if (!topORleft)
{
if (!pSC->m_param.bScaledArith) {
strDCT2x2dn(p0 - 128, p0 - 64, p0 - 112, p0 - 48);
strDCT2x2dn(p0 - 96, p0 - 32, p0 - 80, p0 - 16);
}
else {
strDCT2x2dnEnc(p0 - 128, p0 - 64, p0 - 112, p0 - 48);
strDCT2x2dnEnc(p0 - 96, p0 - 32, p0 - 80, p0 - 16);
}
// 1D lossless HT
p0[- 96] -= p0[-128];
p0[-128] += ((p0[-96] + 1) >> 1);
}
}
assert(NULL == p);
}