/* The copyright in this software is being made available under the BSD
* License, included below. This software may be subject to other third party
* and contributor rights, including patent rights, and no such rights are
* granted under this license.
*
* Copyright (c) 2010-2018, ITU/ISO/IEC
* 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.
* * Neither the name of the ITU/ISO/IEC nor the names of its contributors may
* be used to endorse or promote products derived from this software without
* specific prior written permission.
*
* 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.
*/
/** \file EncCu.cpp
\brief Coding Unit (CU) encoder class
*/
#include "EncCu.h"
#include "EncLib.h"
#include "Analyze.h"
#include "AQp.h"
#include "CommonLib/dtrace_codingstruct.h"
#include "CommonLib/Picture.h"
#include "CommonLib/UnitTools.h"
#include "CommonLib/dtrace_buffer.h"
#include <stdio.h>
#include <cmath>
#include <algorithm>
#if ENABLE_WPP_PARALLELISM
#include <mutex>
extern std::recursive_mutex g_cache_mutex;
#endif
//! \ingroup EncoderLib
//! \{
// ====================================================================================================================
// Constructor / destructor / create / destroy
// ====================================================================================================================
void EncCu::create( EncCfg* encCfg )
{
unsigned uiMaxWidth = encCfg->getMaxCUWidth();
unsigned uiMaxHeight = encCfg->getMaxCUHeight();
ChromaFormat chromaFormat = encCfg->getChromaFormatIdc();
bool BTnoRQT = encCfg->getQTBT();
unsigned numWidths = gp_sizeIdxInfo->numWidths();
unsigned numHeights = gp_sizeIdxInfo->numHeights();
unsigned maxMEPart = BTnoRQT ? 1 : NUMBER_OF_PART_SIZES;
m_pTempCS = new CodingStructure** [numWidths];
m_pBestCS = new CodingStructure** [numWidths];
#if JVET_L0266_HMVP
m_pTempMotLUTs = new LutMotionCand**[numWidths];
m_pBestMotLUTs = new LutMotionCand**[numWidths];
m_pSplitTempMotLUTs = new LutMotionCand**[numWidths];
#endif
for( unsigned w = 0; w < numWidths; w++ )
{
m_pTempCS[w] = new CodingStructure* [numHeights];
m_pBestCS[w] = new CodingStructure* [numHeights];
#if JVET_L0266_HMVP
m_pTempMotLUTs[w] = new LutMotionCand*[numHeights];
m_pBestMotLUTs[w] = new LutMotionCand*[numHeights];
m_pSplitTempMotLUTs[w] = new LutMotionCand*[numHeights];
#endif
for( unsigned h = 0; h < numHeights; h++ )
{
unsigned width = gp_sizeIdxInfo->sizeFrom( w );
unsigned height = gp_sizeIdxInfo->sizeFrom( h );
if( ( BTnoRQT || w == h ) && gp_sizeIdxInfo->isCuSize( width ) && gp_sizeIdxInfo->isCuSize( height ) )
{
m_pTempCS[w][h] = new CodingStructure( m_unitCache.cuCache, m_unitCache.puCache, m_unitCache.tuCache );
m_pBestCS[w][h] = new CodingStructure( m_unitCache.cuCache, m_unitCache.puCache, m_unitCache.tuCache );
m_pTempCS[w][h]->create( chromaFormat, Area( 0, 0, width, height ), false );
m_pBestCS[w][h]->create( chromaFormat, Area( 0, 0, width, height ), false );
#if JVET_L0266_HMVP
m_pTempMotLUTs[w][h] = new LutMotionCand ;
m_pBestMotLUTs[w][h] = new LutMotionCand ;
m_pSplitTempMotLUTs[w][h] = new LutMotionCand;
m_pSplitTempMotLUTs[w][h]->currCnt = 0;
m_pSplitTempMotLUTs[w][h]->motionCand = nullptr;
m_pSplitTempMotLUTs[w][h]->motionCand = new MotionInfo[MAX_NUM_HMVP_CANDS];
m_pTempMotLUTs[w][h]->currCnt = 0;
m_pTempMotLUTs[w][h]->motionCand = nullptr;
m_pTempMotLUTs[w][h]->motionCand = new MotionInfo[MAX_NUM_HMVP_CANDS];
m_pBestMotLUTs[w][h]->currCnt = 0;
m_pBestMotLUTs[w][h]->motionCand = nullptr;
m_pBestMotLUTs[w][h]->motionCand = new MotionInfo[MAX_NUM_HMVP_CANDS];
#endif
}
else
{
m_pTempCS[w][h] = nullptr;
m_pBestCS[w][h] = nullptr;
#if JVET_L0266_HMVP
m_pTempMotLUTs[w][h] = nullptr;
m_pBestMotLUTs[w][h] = nullptr;
m_pSplitTempMotLUTs[w][h] = nullptr;
#endif
}
}
}
// WIA: only the weight==height case is relevant without QTBT
m_pImvTempCS = nullptr;
if( IMV_OFF != encCfg->getIMV() && !BTnoRQT )
{
m_pImvTempCS = new CodingStructure**[numWidths];
for( unsigned w = 0; w < numWidths; w++ )
{
unsigned width = gp_sizeIdxInfo->sizeFrom( w );
unsigned height = gp_sizeIdxInfo->sizeFrom( w );
m_pImvTempCS[w] = new CodingStructure*[maxMEPart];
for( unsigned p = 0; p < maxMEPart; p++ )
{
if( gp_sizeIdxInfo->isCuSize( width ) )
{
m_pImvTempCS[w][p] = new CodingStructure( m_unitCache.cuCache, m_unitCache.puCache, m_unitCache.tuCache );
m_pImvTempCS[w][p]->create( chromaFormat, Area( 0, 0, width, height ), false );
}
else
{
m_pImvTempCS[w][p] = nullptr;
}
}
}
}
m_cuChromaQpOffsetIdxPlus1 = 0;
unsigned maxDepth = numWidths + numHeights;
if( encCfg->getQTBT() )
{
m_modeCtrl = new EncModeCtrlMTnoRQT();
}
else
{
THROW( "Unknown partitioner!" );
}
#if REUSE_CU_RESULTS
m_modeCtrl->create( *encCfg );
#endif
#if JVET_L0054_MMVD
for (unsigned ui = 0; ui < MMVD_MRG_MAX_RD_BUF_NUM; ui++)
#else
for( unsigned ui = 0; ui < MRG_MAX_NUM_CANDS; ui++ )
#endif
{
m_acMergeBuffer[ui].create( chromaFormat, Area( 0, 0, uiMaxWidth, uiMaxHeight ) );
}
m_CtxBuffer.resize( maxDepth );
m_CurrCtx = 0;
}
void EncCu::destroy()
{
bool BTnoRQT = m_pcEncCfg->getQTBT();
unsigned maxMEPart = BTnoRQT ? 1 : NUMBER_OF_PART_SIZES;
unsigned numWidths = gp_sizeIdxInfo->numWidths();
unsigned numHeights = gp_sizeIdxInfo->numHeights();
for( unsigned w = 0; w < numWidths; w++ )
{
for( unsigned h = 0; h < numHeights; h++ )
{
if( BTnoRQT || w == h )
{
if( m_pBestCS[w][h] ) m_pBestCS[w][h]->destroy();
if( m_pTempCS[w][h] ) m_pTempCS[w][h]->destroy();
delete m_pBestCS[w][h];
delete m_pTempCS[w][h];
#if JVET_L0266_HMVP
if (m_pTempMotLUTs[w][h])
{
delete[] m_pTempMotLUTs[w][h]->motionCand;
m_pTempMotLUTs[w][h]->motionCand = nullptr;
delete[] m_pTempMotLUTs[w][h];
}
if (m_pBestMotLUTs[w][h])
{
delete[] m_pBestMotLUTs[w][h]->motionCand;
m_pBestMotLUTs[w][h]->motionCand = nullptr;
delete[] m_pBestMotLUTs[w][h];
}
if (m_pSplitTempMotLUTs[w][h])
{
delete[] m_pSplitTempMotLUTs[w][h]->motionCand;
m_pSplitTempMotLUTs[w][h]->motionCand = nullptr;
delete[] m_pSplitTempMotLUTs[w][h];
}
#endif
}
}
delete[] m_pTempCS[w];
delete[] m_pBestCS[w];
#if JVET_L0266_HMVP
delete[] m_pBestMotLUTs[w];
delete[] m_pTempMotLUTs[w];
delete[] m_pSplitTempMotLUTs[w];
#endif
}
delete[] m_pBestCS; m_pBestCS = nullptr;
delete[] m_pTempCS; m_pTempCS = nullptr;
#if JVET_L0266_HMVP
delete[] m_pSplitTempMotLUTs; m_pSplitTempMotLUTs = nullptr;
delete[] m_pBestMotLUTs; m_pBestMotLUTs = nullptr;
delete[] m_pTempMotLUTs; m_pTempMotLUTs = nullptr;
#endif
#if REUSE_CU_RESULTS
m_modeCtrl->destroy();
#endif
delete m_modeCtrl;
m_modeCtrl = nullptr;
// WIA: only the weight==height case is relevant without QTBT
if( m_pImvTempCS )
{
for( unsigned w = 0; w < numWidths; w++ )
{
for( unsigned p = 0; p < maxMEPart; p++ )
{
if( m_pImvTempCS[w][p] ) m_pImvTempCS[w][p]->destroy();
delete m_pImvTempCS[w][p];
}
delete[] m_pImvTempCS[w];
}
delete[] m_pImvTempCS;
m_pImvTempCS = nullptr;
}
#if JVET_L0054_MMVD
for (unsigned ui = 0; ui < MMVD_MRG_MAX_RD_BUF_NUM; ui++)
#else
for( unsigned ui = 0; ui < MRG_MAX_NUM_CANDS; ui++ )
#endif
{
m_acMergeBuffer[ui].destroy();
}
}
EncCu::~EncCu()
{
}
/** \param pcEncLib pointer of encoder class
*/
void EncCu::init( EncLib* pcEncLib, const SPS& sps PARL_PARAM( const int tId ) )
{
m_pcEncCfg = pcEncLib;
m_pcIntraSearch = pcEncLib->getIntraSearch( PARL_PARAM0( tId ) );
m_pcInterSearch = pcEncLib->getInterSearch( PARL_PARAM0( tId ) );
m_pcTrQuant = pcEncLib->getTrQuant( PARL_PARAM0( tId ) );
m_pcRdCost = pcEncLib->getRdCost ( PARL_PARAM0( tId ) );
m_CABACEstimator = pcEncLib->getCABACEncoder( PARL_PARAM0( tId ) )->getCABACEstimator( &sps );
m_CABACEstimator->setEncCu(this);
m_CtxCache = pcEncLib->getCtxCache( PARL_PARAM0( tId ) );
m_pcRateCtrl = pcEncLib->getRateCtrl();
m_pcSliceEncoder = pcEncLib->getSliceEncoder();
#if ENABLE_SPLIT_PARALLELISM || ENABLE_WPP_PARALLELISM
m_pcEncLib = pcEncLib;
m_dataId = tId;
#endif
#if REUSE_CU_RESULTS
DecCu::init( m_pcTrQuant, m_pcIntraSearch, m_pcInterSearch );
#endif
m_modeCtrl->init( m_pcEncCfg, m_pcRateCtrl, m_pcRdCost );
m_pcInterSearch->setModeCtrl( m_modeCtrl );
::memset(m_subMergeBlkSize, 0, sizeof(m_subMergeBlkSize));
::memset(m_subMergeBlkNum, 0, sizeof(m_subMergeBlkNum));
m_prevPOC = MAX_UINT;
m_clearSubMergeStatic = false;
}
// ====================================================================================================================
// Public member functions
// ====================================================================================================================
void EncCu::compressCtu( CodingStructure& cs, const UnitArea& area, const unsigned ctuRsAddr, const int prevQP[], const int currQP[] )
{
m_modeCtrl->initCTUEncoding( *cs.slice );
#if ENABLE_SPLIT_PARALLELISM
if( m_pcEncCfg->getNumSplitThreads() > 1 )
{
for( int jId = 1; jId < NUM_RESERVERD_SPLIT_JOBS; jId++ )
{
EncCu* jobEncCu = m_pcEncLib->getCuEncoder( cs.picture->scheduler.getSplitDataId( jId ) );
CacheBlkInfoCtrl* cacheCtrl = dynamic_cast< CacheBlkInfoCtrl* >( jobEncCu->m_modeCtrl );
if( cacheCtrl )
{
cacheCtrl->init( *cs.slice );
}
}
}
if( auto* cacheCtrl = dynamic_cast<CacheBlkInfoCtrl*>( m_modeCtrl ) ) { cacheCtrl->tick(); }
#endif
// init the partitioning manager
Partitioner *partitioner = PartitionerFactory::get( *cs.slice );
partitioner->initCtu( area, CH_L, *cs.slice );
// init current context pointer
m_CurrCtx = m_CtxBuffer.data();
CodingStructure *tempCS = m_pTempCS[gp_sizeIdxInfo->idxFrom( area.lumaSize().width )][gp_sizeIdxInfo->idxFrom( area.lumaSize().height )];
CodingStructure *bestCS = m_pBestCS[gp_sizeIdxInfo->idxFrom( area.lumaSize().width )][gp_sizeIdxInfo->idxFrom( area.lumaSize().height )];
#if JVET_L0266_HMVP
LutMotionCand *tempMotCandLUTs = m_pTempMotLUTs[gp_sizeIdxInfo->idxFrom(area.lumaSize().width)][gp_sizeIdxInfo->idxFrom(area.lumaSize().height)];
LutMotionCand *bestMotCandLUTs = m_pBestMotLUTs[gp_sizeIdxInfo->idxFrom(area.lumaSize().width)][gp_sizeIdxInfo->idxFrom(area.lumaSize().height)];
cs.slice->copyMotionLUTs(cs.slice->getMotionLUTs(), tempMotCandLUTs);
cs.slice->copyMotionLUTs(cs.slice->getMotionLUTs(), bestMotCandLUTs);
#endif
cs.initSubStructure( *tempCS, partitioner->chType, partitioner->currArea(), false );
cs.initSubStructure( *bestCS, partitioner->chType, partitioner->currArea(), false );
tempCS->currQP[CH_L] = bestCS->currQP[CH_L] =
tempCS->baseQP = bestCS->baseQP = currQP[CH_L];
tempCS->prevQP[CH_L] = bestCS->prevQP[CH_L] = prevQP[CH_L];
xCompressCU( tempCS, bestCS, *partitioner
#if JVET_L0266_HMVP
, tempMotCandLUTs
, bestMotCandLUTs
#endif
);
// all signals were already copied during compression if the CTU was split - at this point only the structures are copied to the top level CS
const bool copyUnsplitCTUSignals = bestCS->cus.size() == 1 && KEEP_PRED_AND_RESI_SIGNALS;
cs.useSubStructure( *bestCS, partitioner->chType, CS::getArea( *bestCS, area, partitioner->chType ), copyUnsplitCTUSignals, false, false, copyUnsplitCTUSignals );
#if JVET_L0266_HMVP
cs.slice->copyMotionLUTs(bestMotCandLUTs, cs.slice->getMotionLUTs());
#endif
if( !cs.pcv->ISingleTree && cs.slice->isIRAP() && cs.pcv->chrFormat != CHROMA_400 )
{
m_CABACEstimator->getCtx() = m_CurrCtx->start;
partitioner->initCtu( area, CH_C, *cs.slice );
cs.initSubStructure( *tempCS, partitioner->chType, partitioner->currArea(), false );
cs.initSubStructure( *bestCS, partitioner->chType, partitioner->currArea(), false );
tempCS->currQP[CH_C] = bestCS->currQP[CH_C] =
tempCS->baseQP = bestCS->baseQP = currQP[CH_C];
tempCS->prevQP[CH_C] = bestCS->prevQP[CH_C] = prevQP[CH_C];
xCompressCU( tempCS, bestCS, *partitioner
#if JVET_L0266_HMVP
, tempMotCandLUTs
, bestMotCandLUTs
#endif
);
const bool copyUnsplitCTUSignals = bestCS->cus.size() == 1 && KEEP_PRED_AND_RESI_SIGNALS;
cs.useSubStructure( *bestCS, partitioner->chType, CS::getArea( *bestCS, area, partitioner->chType ), copyUnsplitCTUSignals, false, false, copyUnsplitCTUSignals );
}
if (m_pcEncCfg->getUseRateCtrl())
{
(m_pcRateCtrl->getRCPic()->getLCU(ctuRsAddr)).m_actualMSE = (double)bestCS->dist / (double)m_pcRateCtrl->getRCPic()->getLCU(ctuRsAddr).m_numberOfPixel;
}
// reset context states and uninit context pointer
m_CABACEstimator->getCtx() = m_CurrCtx->start;
m_CurrCtx = 0;
delete partitioner;
#if ENABLE_SPLIT_PARALLELISM && ENABLE_WPP_PARALLELISM
if( m_pcEncCfg->getNumSplitThreads() > 1 && m_pcEncCfg->getNumWppThreads() > 1 )
{
cs.picture->finishCtuPart( area );
}
#endif
// Ensure that a coding was found
// Selected mode's RD-cost must be not MAX_DOUBLE.
CHECK( bestCS->cus.empty() , "No possible encoding found" );
CHECK( bestCS->cus[0]->partSize == NUMBER_OF_PART_SIZES , "No possible encoding found" );
CHECK( bestCS->cus[0]->predMode == NUMBER_OF_PREDICTION_MODES, "No possible encoding found" );
CHECK( bestCS->cost == MAX_DOUBLE , "No possible encoding found" );
}
// ====================================================================================================================
// Protected member functions
// ====================================================================================================================
static int xCalcHADs8x8_ISlice(const Pel *piOrg, const int iStrideOrg)
{
int k, i, j, jj;
int diff[64], m1[8][8], m2[8][8], m3[8][8], iSumHad = 0;
for (k = 0; k < 64; k += 8)
{
diff[k + 0] = piOrg[0];
diff[k + 1] = piOrg[1];
diff[k + 2] = piOrg[2];
diff[k + 3] = piOrg[3];
diff[k + 4] = piOrg[4];
diff[k + 5] = piOrg[5];
diff[k + 6] = piOrg[6];
diff[k + 7] = piOrg[7];
piOrg += iStrideOrg;
}
//horizontal
for (j = 0; j < 8; j++)
{
jj = j << 3;
m2[j][0] = diff[jj ] + diff[jj + 4];
m2[j][1] = diff[jj + 1] + diff[jj + 5];
m2[j][2] = diff[jj + 2] + diff[jj + 6];
m2[j][3] = diff[jj + 3] + diff[jj + 7];
m2[j][4] = diff[jj ] - diff[jj + 4];
m2[j][5] = diff[jj + 1] - diff[jj + 5];
m2[j][6] = diff[jj + 2] - diff[jj + 6];
m2[j][7] = diff[jj + 3] - diff[jj + 7];
m1[j][0] = m2[j][0] + m2[j][2];
m1[j][1] = m2[j][1] + m2[j][3];
m1[j][2] = m2[j][0] - m2[j][2];
m1[j][3] = m2[j][1] - m2[j][3];
m1[j][4] = m2[j][4] + m2[j][6];
m1[j][5] = m2[j][5] + m2[j][7];
m1[j][6] = m2[j][4] - m2[j][6];
m1[j][7] = m2[j][5] - m2[j][7];
m2[j][0] = m1[j][0] + m1[j][1];
m2[j][1] = m1[j][0] - m1[j][1];
m2[j][2] = m1[j][2] + m1[j][3];
m2[j][3] = m1[j][2] - m1[j][3];
m2[j][4] = m1[j][4] + m1[j][5];
m2[j][5] = m1[j][4] - m1[j][5];
m2[j][6] = m1[j][6] + m1[j][7];
m2[j][7] = m1[j][6] - m1[j][7];
}
//vertical
for (i = 0; i < 8; i++)
{
m3[0][i] = m2[0][i] + m2[4][i];
m3[1][i] = m2[1][i] + m2[5][i];
m3[2][i] = m2[2][i] + m2[6][i];
m3[3][i] = m2[3][i] + m2[7][i];
m3[4][i] = m2[0][i] - m2[4][i];
m3[5][i] = m2[1][i] - m2[5][i];
m3[6][i] = m2[2][i] - m2[6][i];
m3[7][i] = m2[3][i] - m2[7][i];
m1[0][i] = m3[0][i] + m3[2][i];
m1[1][i] = m3[1][i] + m3[3][i];
m1[2][i] = m3[0][i] - m3[2][i];
m1[3][i] = m3[1][i] - m3[3][i];
m1[4][i] = m3[4][i] + m3[6][i];
m1[5][i] = m3[5][i] + m3[7][i];
m1[6][i] = m3[4][i] - m3[6][i];
m1[7][i] = m3[5][i] - m3[7][i];
m2[0][i] = m1[0][i] + m1[1][i];
m2[1][i] = m1[0][i] - m1[1][i];
m2[2][i] = m1[2][i] + m1[3][i];
m2[3][i] = m1[2][i] - m1[3][i];
m2[4][i] = m1[4][i] + m1[5][i];
m2[5][i] = m1[4][i] - m1[5][i];
m2[6][i] = m1[6][i] + m1[7][i];
m2[7][i] = m1[6][i] - m1[7][i];
}
for (i = 0; i < 8; i++)
{
for (j = 0; j < 8; j++)
{
iSumHad += abs(m2[i][j]);
}
}
iSumHad -= abs(m2[0][0]);
iSumHad = (iSumHad + 2) >> 2;
return(iSumHad);
}
int EncCu::updateCtuDataISlice(const CPelBuf buf)
{
int xBl, yBl;
const int iBlkSize = 8;
const Pel* pOrgInit = buf.buf;
int iStrideOrig = buf.stride;
int iSumHad = 0;
for( yBl = 0; ( yBl + iBlkSize ) <= buf.height; yBl += iBlkSize )
{
for( xBl = 0; ( xBl + iBlkSize ) <= buf.width; xBl += iBlkSize )
{
const Pel* pOrg = pOrgInit + iStrideOrig*yBl + xBl;
iSumHad += xCalcHADs8x8_ISlice( pOrg, iStrideOrig );
}
}
return( iSumHad );
}
#if JVET_L0266_HMVP
bool EncCu::xCheckBestMode( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
#else
void EncCu::xCheckBestMode( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
#endif
{
#if JVET_L0266_HMVP
bool bestCSUpdated = false;
#endif
if( !tempCS->cus.empty() )
{
if( tempCS->cus.size() == 1 )
{
const CodingUnit& cu = *tempCS->cus.front();
CHECK( cu.skip && !cu.firstPU->mergeFlag, "Skip flag without a merge flag is not allowed!" );
}
#if WCG_EXT
DTRACE_BEST_MODE( tempCS, bestCS, m_pcRdCost->getLambda( true ) );
#else
DTRACE_BEST_MODE( tempCS, bestCS, m_pcRdCost->getLambda() );
#endif
if( m_modeCtrl->useModeResult( encTestMode, tempCS, partitioner ) )
{
if( tempCS->cus.size() == 1 )
{
// if tempCS is not a split-mode
CodingUnit &cu = *tempCS->cus.front();
if( CU::isLosslessCoded( cu ) && !cu.ipcm )
{
xFillPCMBuffer( cu );
}
}
std::swap( tempCS, bestCS );
// store temp best CI for next CU coding
m_CurrCtx->best = m_CABACEstimator->getCtx();
#if JVET_L0266_HMVP
bestCSUpdated = true;
#endif
}
}
// reset context states
m_CABACEstimator->getCtx() = m_CurrCtx->start;
#if JVET_L0266_HMVP
return bestCSUpdated;
#endif
}
void EncCu::xCompressCU( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner
#if JVET_L0266_HMVP
, LutMotionCand *&tempMotCandLUTs
, LutMotionCand *&bestMotCandLUTs
#endif
)
{
#if ENABLE_SPLIT_PARALLELISM
CHECK( m_dataId != tempCS->picture->scheduler.getDataId(), "Working in the wrong dataId!" );
if( m_pcEncCfg->getNumSplitThreads() != 1 && tempCS->picture->scheduler.getSplitJobId() == 0 )
{
if( m_modeCtrl->isParallelSplit( *tempCS, partitioner ) )
{
m_modeCtrl->setParallelSplit( true );
xCompressCUParallel( tempCS, bestCS, partitioner );
return;
}
}
#endif
Slice& slice = *tempCS->slice;
const PPS &pps = *tempCS->pps;
const SPS &sps = *tempCS->sps;
const uint32_t uiLPelX = tempCS->area.Y().lumaPos().x;
const uint32_t uiTPelY = tempCS->area.Y().lumaPos().y;
const unsigned wIdx = gp_sizeIdxInfo->idxFrom( partitioner.currArea().lwidth() );
const UnitArea currCsArea = clipArea( CS::getArea( *bestCS, bestCS->area, partitioner.chType ), *tempCS->picture );
if( m_pImvTempCS && !slice.isIntra() )
{
const unsigned maxMEPart = tempCS->pcv->only2Nx2N ? 1 : NUMBER_OF_PART_SIZES;
for( unsigned p = 0; p < maxMEPart; p++ )
{
tempCS->initSubStructure( *m_pImvTempCS[wIdx][p], partitioner.chType, partitioner.currArea(), false );
}
}
m_modeCtrl->initCULevel( partitioner, *tempCS );
m_CurrCtx->start = m_CABACEstimator->getCtx();
m_cuChromaQpOffsetIdxPlus1 = 0;
if( slice.getUseChromaQpAdj() )
{
int lgMinCuSize = sps.getLog2MinCodingBlockSize() +
std::max<int>( 0, sps.getLog2DiffMaxMinCodingBlockSize() - int( pps.getPpsRangeExtension().getDiffCuChromaQpOffsetDepth() ) );
m_cuChromaQpOffsetIdxPlus1 = ( ( uiLPelX >> lgMinCuSize ) + ( uiTPelY >> lgMinCuSize ) ) % ( pps.getPpsRangeExtension().getChromaQpOffsetListLen() + 1 );
}
if( !m_modeCtrl->anyMode() )
{
m_modeCtrl->finishCULevel( partitioner );
return;
}
#if JVET_L0266_HMVP
if (!slice.isIntra())
{
tempCS->slice->copyMotionLUTs(tempMotCandLUTs, tempCS->slice->getMotionLUTs());
}
#endif
DTRACE_UPDATE( g_trace_ctx, std::make_pair( "cux", uiLPelX ) );
DTRACE_UPDATE( g_trace_ctx, std::make_pair( "cuy", uiTPelY ) );
DTRACE_UPDATE( g_trace_ctx, std::make_pair( "cuw", tempCS->area.lwidth() ) );
DTRACE_UPDATE( g_trace_ctx, std::make_pair( "cuh", tempCS->area.lheight() ) );
DTRACE( g_trace_ctx, D_COMMON, "@(%4d,%4d) [%2dx%2d]\n", tempCS->area.lx(), tempCS->area.ly(), tempCS->area.lwidth(), tempCS->area.lheight() );
do
{
const EncTestMode currTestMode = m_modeCtrl->currTestMode();
#if SHARP_LUMA_DELTA_QP
if( m_pcEncCfg->getLumaLevelToDeltaQPMapping().isEnabled() && partitioner.currDepth <= pps.getMaxCuDQPDepth() )
{
#if ENABLE_SPLIT_PARALLELISM
CHECK( tempCS->picture->scheduler.getSplitJobId() > 0, "Changing lambda is only allowed in the master thread!" );
#endif
if (currTestMode.qp >= 0)
{
updateLambda(&slice, currTestMode.qp);
}
}
#endif
if( currTestMode.type == ETM_INTER_ME )
{
if( ( currTestMode.opts & ETO_IMV ) != 0 )
{
xCheckRDCostInterIMV(tempCS, bestCS, partitioner, currTestMode);
}
else
{
xCheckRDCostInter( tempCS, bestCS, partitioner, currTestMode );
}
}
else if( currTestMode.type == ETM_AFFINE )
{
xCheckRDCostAffineMerge2Nx2N( tempCS, bestCS, partitioner, currTestMode );
}
#if REUSE_CU_RESULTS
else if( currTestMode.type == ETM_RECO_CACHED )
{
xReuseCachedResult( tempCS, bestCS, partitioner );
}
#endif
else if( currTestMode.type == ETM_MERGE_SKIP )
{
xCheckRDCostMerge2Nx2N( tempCS, bestCS, partitioner, currTestMode );
#if JVET_L0054_MMVD
CodingUnit* cu = bestCS->getCU(partitioner.chType);
cu->mmvdSkip = cu->skip == false ? false : cu->mmvdSkip;
#endif
}
else if( currTestMode.type == ETM_INTRA )
{
xCheckRDCostIntra( tempCS, bestCS, partitioner, currTestMode );
}
else if( currTestMode.type == ETM_IPCM )
{
xCheckIntraPCM( tempCS, bestCS, partitioner, currTestMode );
}
else if( isModeSplit( currTestMode ) )
{
xCheckModeSplit( tempCS, bestCS, partitioner, currTestMode
#if JVET_L0266_HMVP
, tempMotCandLUTs
, bestMotCandLUTs
, partitioner.currArea()
#endif
);
}
else
{
THROW( "Don't know how to handle mode: type = " << currTestMode.type << ", size = " << currTestMode.partSize << ", options = " << currTestMode.opts );
}
} while( m_modeCtrl->nextMode( *tempCS, partitioner ) );
//////////////////////////////////////////////////////////////////////////
// Finishing CU
#if ENABLE_SPLIT_PARALLELISM
if( bestCS->cus.empty() )
{
CHECK( bestCS->cost != MAX_DOUBLE, "Cost should be maximal if no encoding found" );
CHECK( bestCS->picture->scheduler.getSplitJobId() == 0, "Should always get a result in serial case" );
m_modeCtrl->finishCULevel( partitioner );
return;
}
#endif
// set context states
m_CABACEstimator->getCtx() = m_CurrCtx->best;
// QP from last processed CU for further processing
bestCS->prevQP[partitioner.chType] = bestCS->cus.back()->qp;
#if JVET_L0266_HMVP
if (!slice.isIntra() && bestCS->cus.size() == 1 && bestCS->cus.back()->predMode == MODE_INTER && bestCS->area == *bestCS->cus.back())
{
bestCS->slice->updateMotionLUTs(bestMotCandLUTs, (*bestCS->cus.back()));
}
#endif
bestCS->picture->getRecoBuf( currCsArea ).copyFrom( bestCS->getRecoBuf( currCsArea ) );
m_modeCtrl->finishCULevel( partitioner );
#if ENABLE_SPLIT_PARALLELISM
if( tempCS->picture->scheduler.getSplitJobId() == 0 && m_pcEncCfg->getNumSplitThreads() != 1 )
{
tempCS->picture->finishParallelPart( currCsArea );
}
#endif
// Assert if Best prediction mode is NONE
// Selected mode's RD-cost must be not MAX_DOUBLE.
CHECK( bestCS->cus.empty() , "No possible encoding found" );
CHECK( bestCS->cus[0]->partSize == NUMBER_OF_PART_SIZES , "No possible encoding found" );
CHECK( bestCS->cus[0]->predMode == NUMBER_OF_PREDICTION_MODES, "No possible encoding found" );
CHECK( bestCS->cost == MAX_DOUBLE , "No possible encoding found" );
}
#if SHARP_LUMA_DELTA_QP
void EncCu::updateLambda( Slice* slice, double dQP )
{
#if WCG_EXT
int NumberBFrames = ( m_pcEncCfg->getGOPSize() - 1 );
int SHIFT_QP = 12;
double dLambda_scale = 1.0 - Clip3( 0.0, 0.5, 0.05*(double)(slice->getPic()->fieldPic ? NumberBFrames/2 : NumberBFrames) );
int bitdepth_luma_qp_scale = 6
* (slice->getSPS()->getBitDepth(CHANNEL_TYPE_LUMA) - 8
- DISTORTION_PRECISION_ADJUSTMENT(slice->getSPS()->getBitDepth(CHANNEL_TYPE_LUMA)));
double qp_temp = (double) dQP + bitdepth_luma_qp_scale - SHIFT_QP;
double dQPFactor = m_pcEncCfg->getGOPEntry( m_pcSliceEncoder->getGopId() ).m_QPFactor;
if( slice->getSliceType() == I_SLICE )
{
if( m_pcEncCfg->getIntraQpFactor() >= 0.0 /*&& m_pcEncCfg->getGOPEntry( m_pcSliceEncoder->getGopId() ).m_sliceType != I_SLICE*/ )
{
dQPFactor = m_pcEncCfg->getIntraQpFactor();
}
else
{
if( m_pcEncCfg->getLambdaFromQPEnable() )
{
dQPFactor = 0.57;
}
else
{
dQPFactor = 0.57*dLambda_scale;
}
}
}
else if( m_pcEncCfg->getLambdaFromQPEnable() )
{
dQPFactor = 0.57*dQPFactor;
}
double dLambda = dQPFactor*pow( 2.0, qp_temp/3.0 );
int depth = slice->getDepth();
if( !m_pcEncCfg->getLambdaFromQPEnable() && depth>0 )
{
int qp_temp_slice = slice->getSliceQp() + bitdepth_luma_qp_scale - SHIFT_QP; // avoid lambda over adjustment, use slice_qp here
dLambda *= Clip3( 2.00, 4.00, (qp_temp_slice / 6.0) ); // (j == B_SLICE && p_cur_frm->layer != 0 )
}
if( !m_pcEncCfg->getUseHADME() && slice->getSliceType( ) != I_SLICE )
{
dLambda *= 0.95;
}
const int temporalId = m_pcEncCfg->getGOPEntry( m_pcSliceEncoder->getGopId() ).m_temporalId;
const std::vector<double> &intraLambdaModifiers = m_pcEncCfg->getIntraLambdaModifier();
double lambdaModifier;
if( slice->getSliceType( ) != I_SLICE || intraLambdaModifiers.empty())
{
lambdaModifier = m_pcEncCfg->getLambdaModifier(temporalId);
}
else
{
lambdaModifier = intraLambdaModifiers[(temporalId < intraLambdaModifiers.size()) ? temporalId : (intraLambdaModifiers.size() - 1)];
}
dLambda *= lambdaModifier;
int qpBDoffset = slice->getSPS()->getQpBDOffset(CHANNEL_TYPE_LUMA);
int iQP = Clip3(-qpBDoffset, MAX_QP, (int)floor(dQP + 0.5));
m_pcSliceEncoder->setUpLambda(slice, dLambda, iQP);
#else
int iQP = (int)dQP;
const double oldQP = (double)slice->getSliceQpBase();
const double oldLambda = m_pcSliceEncoder->calculateLambda (slice, m_pcSliceEncoder->getGopId(), slice->getDepth(), oldQP, oldQP, iQP);
const double newLambda = oldLambda * pow (2.0, (dQP - oldQP) / 3.0);
#if RDOQ_CHROMA_LAMBDA
const double chromaLambda = newLambda / m_pcRdCost->getChromaWeight();
const double lambdaArray[MAX_NUM_COMPONENT] = {newLambda, chromaLambda, chromaLambda};
m_pcTrQuant->setLambdas (lambdaArray);
#else
m_pcTrQuant->setLambda (newLambda);
#endif
m_pcRdCost->setLambda( newLambda, slice->getSPS()->getBitDepths() );
#endif
}
#endif
#if ENABLE_SPLIT_PARALLELISM
//#undef DEBUG_PARALLEL_TIMINGS
//#define DEBUG_PARALLEL_TIMINGS 1
void EncCu::xCompressCUParallel( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner )
{
const unsigned wIdx = gp_sizeIdxInfo->idxFrom( partitioner.currArea().lwidth() );
const unsigned hIdx = gp_sizeIdxInfo->idxFrom( partitioner.currArea().lheight() );
Picture* picture = tempCS->picture;
int numJobs = m_modeCtrl->getNumParallelJobs( *bestCS, partitioner );
bool jobUsed [NUM_RESERVERD_SPLIT_JOBS];
std::fill( jobUsed, jobUsed + NUM_RESERVERD_SPLIT_JOBS, false );
const UnitArea currArea = CS::getArea( *tempCS, partitioner.currArea(), partitioner.chType );
#if ENABLE_WPP_PARALLELISM
const int wppTId = picture->scheduler.getWppThreadId();
#endif
const bool doParallel = !m_pcEncCfg->getForceSingleSplitThread();
#if _MSC_VER && ENABLE_WPP_PARALLELISM
#pragma omp parallel for schedule(dynamic,1) num_threads(NUM_SPLIT_THREADS_IF_MSVC) if(doParallel)
#else
omp_set_num_threads( m_pcEncCfg->getNumSplitThreads() );
#pragma omp parallel for schedule(dynamic,1) if(doParallel)
#endif
for( int jId = 1; jId <= numJobs; jId++ )
{
// thread start
#if ENABLE_WPP_PARALLELISM
picture->scheduler.setWppThreadId( wppTId );
#endif
picture->scheduler.setSplitThreadId();
picture->scheduler.setSplitJobId( jId );
Partitioner* jobPartitioner = PartitionerFactory::get( *tempCS->slice );
EncCu* jobCuEnc = m_pcEncLib->getCuEncoder( picture->scheduler.getSplitDataId( jId ) );
auto* jobBlkCache = dynamic_cast<CacheBlkInfoCtrl*>( jobCuEnc->m_modeCtrl );
jobPartitioner->copyState( partitioner );
jobCuEnc ->copyState( this, *jobPartitioner, currArea, true );
if( jobBlkCache )
{
jobBlkCache->tick();
}
CodingStructure *&jobBest = jobCuEnc->m_pBestCS[wIdx][hIdx];
CodingStructure *&jobTemp = jobCuEnc->m_pTempCS[wIdx][hIdx];
jobUsed[jId] = true;
jobCuEnc->xCompressCU( jobTemp, jobBest, *jobPartitioner );
delete jobPartitioner;
picture->scheduler.setSplitJobId( 0 );
// thread stop
}
picture->scheduler.setSplitThreadId( 0 );
int bestJId = 0;
double bestCost = bestCS->cost;
for( int jId = 1; jId <= numJobs; jId++ )
{
EncCu* jobCuEnc = m_pcEncLib->getCuEncoder( picture->scheduler.getSplitDataId( jId ) );
if( jobUsed[jId] && jobCuEnc->m_pBestCS[wIdx][hIdx]->cost < bestCost )
{
bestCost = jobCuEnc->m_pBestCS[wIdx][hIdx]->cost;
bestJId = jId;
}
}
if( bestJId > 0 )
{
copyState( m_pcEncLib->getCuEncoder( picture->scheduler.getSplitDataId( bestJId ) ), partitioner, currArea, false );
m_CurrCtx->best = m_CABACEstimator->getCtx();
tempCS = m_pTempCS[wIdx][hIdx];
bestCS = m_pBestCS[wIdx][hIdx];
}
const int bitDepthY = tempCS->sps->getBitDepth( CH_L );
const UnitArea clipdArea = clipArea( currArea, *picture );
CHECK( calcCheckSum( picture->getRecoBuf( clipdArea.Y() ), bitDepthY ) != calcCheckSum( bestCS->getRecoBuf( clipdArea.Y() ), bitDepthY ), "Data copied incorrectly!" );
picture->finishParallelPart( currArea );
if( auto *blkCache = dynamic_cast<CacheBlkInfoCtrl*>( m_modeCtrl ) )
{
for( int jId = 1; jId <= numJobs; jId++ )
{
if( !jobUsed[jId] || jId == bestJId ) continue;
auto *jobBlkCache = dynamic_cast<CacheBlkInfoCtrl*>( m_pcEncLib->getCuEncoder( picture->scheduler.getSplitDataId( jId ) )->m_modeCtrl );
CHECK( !jobBlkCache, "If own mode controller has blk info cache capability so should all other mode controllers!" );
blkCache->CacheBlkInfoCtrl::copyState( *jobBlkCache, partitioner.currArea() );
}
blkCache->tick();
}
}
void EncCu::copyState( EncCu* other, Partitioner& partitioner, const UnitArea& currArea, const bool isDist )
{
const unsigned wIdx = gp_sizeIdxInfo->idxFrom( partitioner.currArea().lwidth () );
const unsigned hIdx = gp_sizeIdxInfo->idxFrom( partitioner.currArea().lheight() );
if( isDist )
{
other->m_pBestCS[wIdx][hIdx]->initSubStructure( *m_pBestCS[wIdx][hIdx], partitioner.chType, partitioner.currArea(), false );
other->m_pTempCS[wIdx][hIdx]->initSubStructure( *m_pTempCS[wIdx][hIdx], partitioner.chType, partitioner.currArea(), false );
}
else
{
CodingStructure* dst = m_pBestCS[wIdx][hIdx];
const CodingStructure *src = other->m_pBestCS[wIdx][hIdx];
bool keepResi = KEEP_PRED_AND_RESI_SIGNALS;
dst->useSubStructure( *src, partitioner.chType, currArea, KEEP_PRED_AND_RESI_SIGNALS, true, keepResi, keepResi );
dst->cost = src->cost;
dst->dist = src->dist;
dst->fracBits = src->fracBits;
dst->features = src->features;
}
if( isDist )
{
m_CurrCtx = m_CtxBuffer.data();
}
m_pcInterSearch->copyState( *other->m_pcInterSearch );
m_modeCtrl ->copyState( *other->m_modeCtrl, partitioner.currArea() );
m_pcRdCost ->copyState( *other->m_pcRdCost );
m_pcTrQuant ->copyState( *other->m_pcTrQuant );
m_CABACEstimator->getCtx() = other->m_CABACEstimator->getCtx();
}
#endif
void EncCu::xCheckModeSplit(CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode
#if JVET_L0266_HMVP
, LutMotionCand* &tempMotCandLUTs
, LutMotionCand* &bestMotCandLUTs
, UnitArea parArea
#endif
)
{
const int qp = encTestMode.qp;
const PPS &pps = *tempCS->pps;
const Slice &slice = *tempCS->slice;
const bool bIsLosslessMode = false; // False at this level. Next level down may set it to true.
const int oldPrevQp = tempCS->prevQP[partitioner.chType];
const uint32_t currDepth = partitioner.currDepth;
#if JVET_L0266_HMVP
const unsigned wParIdx = gp_sizeIdxInfo->idxFrom(parArea.lwidth());
const unsigned hParIdx = gp_sizeIdxInfo->idxFrom(parArea.lheight());
tempCS->slice->copyMotionLUTs(tempMotCandLUTs, m_pSplitTempMotLUTs[wParIdx][hParIdx]);
#endif
const PartSplit split = getPartSplit( encTestMode );
CHECK( split == CU_DONT_SPLIT, "No proper split provided!" );
tempCS->initStructData( qp, bIsLosslessMode );
m_CABACEstimator->getCtx() = m_CurrCtx->start;
if( tempCS->sps->getSpsNext().getUseQTBT() )
{
const TempCtx ctxStartSP( m_CtxCache, SubCtx( Ctx::SplitFlag, m_CABACEstimator->getCtx() ) );
const TempCtx ctxStartBT( m_CtxCache, SubCtx( Ctx::BTSplitFlag, m_CABACEstimator->getCtx() ) );
m_CABACEstimator->resetBits();
if( partitioner.getImplicitSplit( *tempCS ) != CU_QUAD_SPLIT )
{
if( partitioner.canSplit( CU_QUAD_SPLIT, *tempCS ) )
{
m_CABACEstimator->split_cu_flag( split == CU_QUAD_SPLIT, *tempCS, partitioner );
}
if( split != CU_QUAD_SPLIT )
{
m_CABACEstimator->split_cu_mode_mt( split, *tempCS, partitioner );
}
}
const double factor = ( tempCS->currQP[partitioner.chType] > 30 ? 1.1 : 1.075 );
const double cost = m_pcRdCost->calcRdCost( uint64_t( m_CABACEstimator->getEstFracBits() + ( ( bestCS->fracBits ) / factor ) ), Distortion( bestCS->dist / factor ) );
m_CABACEstimator->getCtx() = SubCtx( Ctx::SplitFlag, ctxStartSP );
m_CABACEstimator->getCtx() = SubCtx( Ctx::BTSplitFlag, ctxStartBT );
if( cost > bestCS->cost )
{
xCheckBestMode( tempCS, bestCS, partitioner, encTestMode );
return;
}
}
partitioner.splitCurrArea( split, *tempCS );
m_CurrCtx++;
tempCS->getRecoBuf().fill( 0 );
do
{
const auto &subCUArea = partitioner.currArea();
if( tempCS->picture->Y().contains( subCUArea.lumaPos() ) )
{
const unsigned wIdx = gp_sizeIdxInfo->idxFrom( subCUArea.lwidth () );
const unsigned hIdx = gp_sizeIdxInfo->idxFrom( subCUArea.lheight() );
CodingStructure *tempSubCS = m_pTempCS[wIdx][hIdx];
CodingStructure *bestSubCS = m_pBestCS[wIdx][hIdx];
tempCS->initSubStructure( *tempSubCS, partitioner.chType, subCUArea, false );
tempCS->initSubStructure( *bestSubCS, partitioner.chType, subCUArea, false );
#if JVET_L0266_HMVP
LutMotionCand *tempSubMotCandLUTs = m_pTempMotLUTs[wIdx][hIdx];
LutMotionCand *bestSubMotCandLUTs = m_pBestMotLUTs[wIdx][hIdx];
tempCS->slice->copyMotionLUTs(tempMotCandLUTs, tempSubMotCandLUTs);
tempCS->slice->copyMotionLUTs(tempMotCandLUTs, bestSubMotCandLUTs);
#endif
xCompressCU( tempSubCS, bestSubCS, partitioner
#if JVET_L0266_HMVP
, tempSubMotCandLUTs
, bestSubMotCandLUTs
#endif
);
if( bestSubCS->cost == MAX_DOUBLE )
{
CHECK( split == CU_QUAD_SPLIT, "Split decision reusing cannot skip quad split" );
tempCS->cost = MAX_DOUBLE;
m_CurrCtx--;
partitioner.exitCurrSplit();
#if JVET_L0266_HMVP
bool bestCSUpdated =
#endif
xCheckBestMode( tempCS, bestCS, partitioner, encTestMode );
#if JVET_L0266_HMVP
if (bestCSUpdated)
{
std::swap(tempMotCandLUTs, bestMotCandLUTs);
}
#endif
return;
}
bool keepResi = KEEP_PRED_AND_RESI_SIGNALS;
tempCS->useSubStructure( *bestSubCS, partitioner.chType, CS::getArea( *tempCS, subCUArea, partitioner.chType ), KEEP_PRED_AND_RESI_SIGNALS, true, keepResi, keepResi );
#if JVET_L0266_HMVP
tempCS->slice->copyMotionLUTs(bestSubMotCandLUTs, tempMotCandLUTs);
#endif
if(currDepth < pps.getMaxCuDQPDepth())
{
tempCS->prevQP[partitioner.chType] = bestSubCS->prevQP[partitioner.chType];
}
tempSubCS->releaseIntermediateData();
bestSubCS->releaseIntermediateData();
}
} while( partitioner.nextPart( *tempCS ) );
partitioner.exitCurrSplit();
m_CurrCtx--;
// Finally, generate split-signaling bits for RD-cost check
const PartSplit implicitSplit = partitioner.getImplicitSplit( *tempCS );
{
bool enforceQT = implicitSplit == CU_QUAD_SPLIT;
#if HM_QTBT_REPRODUCE_FAST_LCTU_BUG
// LARGE CTU bug
if( m_pcEncCfg->getQTBT() && m_pcEncCfg->getUseFastLCTU() )
{
unsigned minDepth = 0;
unsigned maxDepth = g_aucLog2[tempCS->sps->getSpsNext().getCTUSize()] - g_aucLog2[tempCS->sps->getSpsNext().getMinQTSize( slice.getSliceType(), partitioner.chType )];
if( auto ad = dynamic_cast<AdaptiveDepthPartitioner*>( &partitioner ) )
{
ad->setMaxMinDepth( minDepth, maxDepth, *tempCS );
}
if( minDepth > partitioner.currQtDepth )
{
// enforce QT
enforceQT = true;
}
}
#endif
if( !enforceQT )
{
m_CABACEstimator->resetBits();
if( partitioner.canSplit( CU_QUAD_SPLIT, *tempCS ) )
{
m_CABACEstimator->split_cu_flag( split == CU_QUAD_SPLIT, *tempCS, partitioner );
}
if( split != CU_QUAD_SPLIT )
{
m_CABACEstimator->split_cu_mode_mt( split, *tempCS, partitioner );
}
tempCS->fracBits += m_CABACEstimator->getEstFracBits(); // split bits
}
}
tempCS->cost = m_pcRdCost->calcRdCost( tempCS->fracBits, tempCS->dist );
// Check Delta QP bits for splitted structure
xCheckDQP( *tempCS, partitioner, true );
// If the configuration being tested exceeds the maximum number of bytes for a slice / slice-segment, then
// a proper RD evaluation cannot be performed. Therefore, termination of the
// slice/slice-segment must be made prior to this CTU.
// This can be achieved by forcing the decision to be that of the rpcTempCU.
// The exception is each slice / slice-segment must have at least one CTU.
if (bestCS->cost != MAX_DOUBLE)
{
#if HEVC_TILES_WPP
const TileMap& tileMap = *tempCS->picture->tileMap;
#endif
#if HEVC_TILES_WPP || HEVC_DEPENDENT_SLICES
const uint32_t CtuAddr = CU::getCtuAddr( *bestCS->getCU( partitioner.chType ) );
#endif
const bool isEndOfSlice = slice.getSliceMode() == FIXED_NUMBER_OF_BYTES
&& ((slice.getSliceBits() + CS::getEstBits(*bestCS)) > slice.getSliceArgument() << 3)
#if HEVC_TILES_WPP
&& CtuAddr != tileMap.getCtuTsToRsAddrMap(slice.getSliceCurStartCtuTsAddr())
#endif
#if HEVC_DEPENDENT_SLICES
&& CtuAddr != tileMap.getCtuTsToRsAddrMap(slice.getSliceSegmentCurStartCtuTsAddr());
#else
;
#endif
#if HEVC_DEPENDENT_SLICES
const bool isEndOfSliceSegment = slice.getSliceSegmentMode() == FIXED_NUMBER_OF_BYTES
&& ((slice.getSliceSegmentBits() + CS::getEstBits(*bestCS)) > slice.getSliceSegmentArgument() << 3)
&& CtuAddr != tileMap.getCtuTsToRsAddrMap(slice.getSliceSegmentCurStartCtuTsAddr());
// Do not need to check slice condition for slice-segment since a slice-segment is a subset of a slice.
if (isEndOfSlice || isEndOfSliceSegment)
#else
if(isEndOfSlice)
#endif
{
bestCS->cost = MAX_DOUBLE;
}
}
// RD check for sub partitioned coding structure.
#if JVET_L0266_HMVP
bool bestCSUpdated =
#endif
xCheckBestMode( tempCS, bestCS, partitioner, encTestMode );
#if JVET_L0266_HMVP
if (!slice.isIntra())
{
if (bestCSUpdated)
{
std::swap(tempMotCandLUTs, bestMotCandLUTs);
}
tempCS->slice->copyMotionLUTs(m_pSplitTempMotLUTs[wParIdx][hParIdx], tempMotCandLUTs);
}
#endif
tempCS->releaseIntermediateData();
tempCS->prevQP[partitioner.chType] = oldPrevQp;
}
void EncCu::xCheckRDCostIntra( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
{
double bestInterCost = m_modeCtrl->getBestInterCost();
double costSize2Nx2NemtFirstPass = m_modeCtrl->getEmtSize2Nx2NFirstPassCost();
double costSizeNxNemtFirstPass = MAX_DOUBLE;
bool skipSecondEmtPass = m_modeCtrl->getSkipSecondEMTPass();
const SPS &sps = *tempCS->sps;
const PPS &pps = *tempCS->pps;
const CodingUnit *bestCU = bestCS->getCU( partitioner.chType );
const int maxSizeEMT = pps.pcv->noRQT ? EMT_INTRA_MAX_CU_WITH_QTBT : EMT_INTRA_MAX_CU;
#if HM_EMT_NSST_AS_IN_JEM
uint8_t considerEmtSecondPass = ( sps.getSpsNext().getUseIntraEMT() && isLuma( partitioner.chType ) && partitioner.currArea().lwidth() <= maxSizeEMT && partitioner.currArea().lheight() <= maxSizeEMT ) ? 1 : 0;
#else
uint8_t considerEmtSecondPass = ( sps.getSpsNext().getUseIntraEMT() && isLuma( partitioner.chType ) && partitioner.currArea().lwidth() <= maxSizeEMT && partitioner.currArea().lheight() <= maxSizeEMT && nsstIdx == 0 ) ? 1 : 0;
#endif
Distortion interHad = m_modeCtrl->getInterHad();
for( uint8_t emtCuFlag = 0; emtCuFlag <= considerEmtSecondPass; emtCuFlag++ )
{
//Possible early EMT tests interruptions
//2) Second EMT pass. This "if clause" is necessary because of the NSST and PDPC "for loops".
if( emtCuFlag && skipSecondEmtPass )
{
continue;
}
//3) if interHad is 0, only try further modes if some intra mode was already better than inter
if( m_pcEncCfg->getUsePbIntraFast() && !tempCS->slice->isIntra() && bestCU && CU::isInter( *bestCS->getCU( partitioner.chType ) ) && interHad == 0 )
{
continue;
}
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
CodingUnit &cu = tempCS->addCU( CS::getArea( *tempCS, tempCS->area, partitioner.chType ), partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = encTestMode.partSize;
cu.predMode = MODE_INTRA;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
//cu.ipcm = false;
cu.emtFlag = emtCuFlag;
CU::addPUs( cu );
tempCS->interHad = interHad;
if( isLuma( partitioner.chType ) )
{
m_pcIntraSearch->estIntraPredLumaQT( cu, partitioner );
if (m_pcEncCfg->getUsePbIntraFast() && tempCS->dist == std::numeric_limits<Distortion>::max()
&& tempCS->interHad == 0)
{
interHad = 0;
// JEM assumes only perfect reconstructions can from now on beat the inter mode
m_modeCtrl->enforceInterHad( 0 );
continue;
}
if( !CS::isDualITree( *tempCS ) )
{
cu.cs->picture->getRecoBuf( cu.Y() ).copyFrom( cu.cs->getRecoBuf( COMPONENT_Y ) );
}
}
if( tempCS->area.chromaFormat != CHROMA_400 && ( partitioner.chType == CHANNEL_TYPE_CHROMA || !CS::isDualITree( *tempCS ) ) )
{
m_pcIntraSearch->estIntraPredChromaQT( cu, partitioner );
}
cu.rootCbf = false;
for( uint32_t t = 0; t < getNumberValidTBlocks( *cu.cs->pcv ); t++ )
{
cu.rootCbf |= cu.firstTU->cbf[t] != 0;
}
// Get total bits for current mode: encode CU
m_CABACEstimator->resetBits();
if( pps.getTransquantBypassEnabledFlag() )
{
m_CABACEstimator->cu_transquant_bypass_flag( cu );
}
if( !cu.cs->slice->isIntra()
)
{
m_CABACEstimator->cu_skip_flag ( cu );
}
m_CABACEstimator->pred_mode ( cu );
m_CABACEstimator->cu_pred_data ( cu );
m_CABACEstimator->pcm_data ( cu );
// Encode Coefficients
CUCtx cuCtx;
cuCtx.isDQPCoded = true;
cuCtx.isChromaQpAdjCoded = true;
m_CABACEstimator->cu_residual( cu, partitioner, cuCtx );
tempCS->fracBits = m_CABACEstimator->getEstFracBits();
tempCS->cost = m_pcRdCost->calcRdCost(tempCS->fracBits, tempCS->dist);
xEncodeDontSplit( *tempCS, partitioner );
xCheckDQP( *tempCS, partitioner );
// we save the cost of the modes for the first EMT pass
if( !emtCuFlag ) static_cast< double& >( cu.partSize == SIZE_2Nx2N ? costSize2Nx2NemtFirstPass : costSizeNxNemtFirstPass ) = tempCS->cost;
#if WCG_EXT
DTRACE_MODE_COST( *tempCS, m_pcRdCost->getLambda( true ) );
#else
DTRACE_MODE_COST( *tempCS, m_pcRdCost->getLambda() );
#endif
xCheckBestMode( tempCS, bestCS, partitioner, encTestMode );
//now we check whether the second pass of SIZE_2Nx2N and the whole Intra SIZE_NxN should be skipped or not
if( !emtCuFlag && !tempCS->slice->isIntra() && bestCU && bestCU->predMode != MODE_INTRA && cu.partSize == SIZE_2Nx2N && m_pcEncCfg->getFastInterEMT() )
{
const double thEmtInterFastSkipIntra = 1.4; // Skip checking Intra if "2Nx2N using DCT2" is worse than best Inter mode
if( costSize2Nx2NemtFirstPass > thEmtInterFastSkipIntra * bestInterCost )
{
skipSecondEmtPass = true;
m_modeCtrl->setSkipSecondEMTPass( true );
break;
}
}
} //for emtCuFlag
}
void EncCu::xCheckIntraPCM(CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
{
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = SIZE_2Nx2N;
cu.predMode = MODE_INTRA;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
cu.ipcm = true;
tempCS->addPU(tempCS->area, partitioner.chType);
tempCS->addTU( tempCS->area, partitioner.chType );
m_pcIntraSearch->IPCMSearch(*tempCS, partitioner);
m_CABACEstimator->getCtx() = m_CurrCtx->start;
m_CABACEstimator->resetBits();
if( tempCS->pps->getTransquantBypassEnabledFlag() )
{
m_CABACEstimator->cu_transquant_bypass_flag( cu );
}
if( !cu.cs->slice->isIntra()
)
{
m_CABACEstimator->cu_skip_flag ( cu );
}
m_CABACEstimator->pred_mode ( cu );
m_CABACEstimator->pcm_data ( cu );
tempCS->fracBits = m_CABACEstimator->getEstFracBits();
tempCS->cost = m_pcRdCost->calcRdCost(tempCS->fracBits, tempCS->dist);
xEncodeDontSplit( *tempCS, partitioner );
xCheckDQP( *tempCS, partitioner );
#if WCG_EXT
DTRACE_MODE_COST( *tempCS, m_pcRdCost->getLambda( true ) );
#else
DTRACE_MODE_COST( *tempCS, m_pcRdCost->getLambda() );
#endif
xCheckBestMode( tempCS, bestCS, partitioner, encTestMode );
}
void EncCu::xCheckDQP( CodingStructure& cs, Partitioner& partitioner, bool bKeepCtx )
{
CHECK( bKeepCtx && cs.cus.size() <= 1 && partitioner.getImplicitSplit( cs ) == CU_DONT_SPLIT, "bKeepCtx should only be set in split case" );
CHECK( !bKeepCtx && cs.cus.size() > 1, "bKeepCtx should never be set for non-split case" );
if( !cs.pps->getUseDQP() )
{
return;
}
if( bKeepCtx && partitioner.currDepth != cs.pps->getMaxCuDQPDepth() )
{
return;
}
if( !bKeepCtx && partitioner.currDepth > cs.pps->getMaxCuDQPDepth() )
{
return;
}
CodingUnit* cuFirst = cs.getCU( partitioner.chType );
CHECK( !cuFirst, "No CU available" );
bool hasResidual = false;
for( const auto &cu : cs.cus )
{
if( cu->rootCbf )
{
hasResidual = true;
break;
}
}
int predQP = CU::predictQP( *cuFirst, cs.prevQP[partitioner.chType] );
if( hasResidual )
{
TempCtx ctxTemp( m_CtxCache );
if( !bKeepCtx ) ctxTemp = SubCtx( Ctx::DeltaQP, m_CABACEstimator->getCtx() );
m_CABACEstimator->resetBits();
m_CABACEstimator->cu_qp_delta( *cuFirst, predQP, cuFirst->qp );
cs.fracBits += m_CABACEstimator->getEstFracBits(); // dQP bits
cs.cost = m_pcRdCost->calcRdCost(cs.fracBits, cs.dist);
if( !bKeepCtx ) m_CABACEstimator->getCtx() = SubCtx( Ctx::DeltaQP, ctxTemp );
// NOTE: reset QPs for CUs without residuals up to first coded CU
for( const auto &cu : cs.cus )
{
if( cu->rootCbf )
{
break;
}
cu->qp = predQP;
}
}
else
{
// No residuals: reset CU QP to predicted value
for( const auto &cu : cs.cus )
{
cu->qp = predQP;
}
}
}
void EncCu::xFillPCMBuffer( CodingUnit &cu )
{
const ChromaFormat format = cu.chromaFormat;
const uint32_t numberValidComponents = getNumberValidComponents(format);
for( auto &tu : CU::traverseTUs( cu ) )
{
for( uint32_t ch = 0; ch < numberValidComponents; ch++ )
{
const ComponentID compID = ComponentID( ch );
const CompArea &compArea = tu.blocks[ compID ];
const CPelBuf source = tu.cs->getOrgBuf( compArea );
PelBuf destination = tu.getPcmbuf( compID );
destination.copyFrom( source );
}
}
}
void EncCu::xCheckRDCostMerge2Nx2N( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
{
const Slice &slice = *tempCS->slice;
CHECK( slice.getSliceType() == I_SLICE, "Merge modes not available for I-slices" );
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
MergeCtx mergeCtx;
const SPS &sps = *tempCS->sps;
if( sps.getSpsNext().getUseSubPuMvp() )
{
Size bufSize = g_miScaling.scale( tempCS->area.lumaSize() );
mergeCtx.subPuMvpMiBuf = MotionBuf( m_SubPuMiBuf, bufSize );
}
{
// first get merge candidates
CodingUnit cu( tempCS->area );
cu.cs = tempCS;
cu.partSize = SIZE_2Nx2N;
cu.predMode = MODE_INTER;
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap(tempCS->area.lumaPos());
#endif
PredictionUnit pu( tempCS->area );
pu.cu = &cu;
pu.cs = tempCS;
PU::getInterMergeCandidates(pu, mergeCtx
#if JVET_L0054_MMVD
, 0
#endif
);
PU::restrictBiPredMergeCands(pu, mergeCtx);
#if JVET_L0054_MMVD
PU::getInterMMVDMergeCandidates(pu, mergeCtx);
#endif
}
#if JVET_L0054_MMVD
bool candHasNoResidual[MRG_MAX_NUM_CANDS + MMVD_ADD_NUM];
for (uint32_t ui = 0; ui < MRG_MAX_NUM_CANDS + MMVD_ADD_NUM; ui++)
{
candHasNoResidual[ui] = false;
}
bool bestIsSkip = false;
bool bestIsMMVDSkip = true;
PelUnitBuf acMergeRealBuffer[MMVD_MRG_MAX_RD_BUF_NUM];
PelUnitBuf * acMergeTempBuffer[MMVD_MRG_MAX_RD_NUM];
PelUnitBuf * singleMergeTempBuffer;
int insertPos;
unsigned uiNumMrgSATDCand = mergeCtx.numValidMergeCand + MMVD_ADD_NUM;
static_vector<unsigned, MRG_MAX_NUM_CANDS + MMVD_ADD_NUM> RdModeList;
bool mrgTempBufSet = false;
for (unsigned i = 0; i < MRG_MAX_NUM_CANDS + MMVD_ADD_NUM; i++)
{
RdModeList.push_back(i);
}
const UnitArea localUnitArea(tempCS->area.chromaFormat, Area(0, 0, tempCS->area.Y().width, tempCS->area.Y().height));
for (unsigned i = 0; i < MMVD_MRG_MAX_RD_BUF_NUM; i++)
{
acMergeRealBuffer[i] = m_acMergeBuffer[i].getBuf(localUnitArea);
if (i < MMVD_MRG_MAX_RD_NUM)
{
acMergeTempBuffer[i] = acMergeRealBuffer + i;
}
else
{
singleMergeTempBuffer = acMergeRealBuffer + i;
}
}
#else
bool candHasNoResidual[MRG_MAX_NUM_CANDS];
for (uint32_t ui = 0; ui < mergeCtx.numValidMergeCand; ui++)
{
candHasNoResidual[ui] = false;
}
bool bestIsSkip = false;
unsigned uiNumMrgSATDCand = mergeCtx.numValidMergeCand;
PelUnitBuf acMergeBuffer [ MRG_MAX_NUM_CANDS ];
static_vector<unsigned, MRG_MAX_NUM_CANDS> RdModeList;
bool mrgTempBufSet = false;
#if DMVR_JVET_LOW_LATENCY_K0217
Mv refinedMvdL0[MRG_MAX_NUM_CANDS];
#endif
for( unsigned i = 0; i < MRG_MAX_NUM_CANDS; i++ )
{
RdModeList.push_back( i );
}
#endif
if( m_pcEncCfg->getUseFastMerge() )
{
uiNumMrgSATDCand = NUM_MRG_SATD_CAND;
bestIsSkip = false;
if( auto blkCache = dynamic_cast< CacheBlkInfoCtrl* >( m_modeCtrl ) )
{
bestIsSkip = blkCache->isSkip( tempCS->area );
#if JVET_L0054_MMVD
bestIsMMVDSkip = blkCache->isMMVDSkip(tempCS->area);
#endif
}
#if JVET_L0054_MMVD
static_vector<double, MRG_MAX_NUM_CANDS + MMVD_ADD_NUM> candCostList;
#else
static_vector<double, MRG_MAX_NUM_CANDS> candCostList;
#endif
// 1. Pass: get SATD-cost for selected candidates and reduce their count
if( !bestIsSkip )
{
RdModeList.clear();
mrgTempBufSet = true;
const double sqrtLambdaForFirstPass = m_pcRdCost->getMotionLambda( encTestMode.lossless );
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = SIZE_2Nx2N;
//cu.affine
cu.predMode = MODE_INTER;
//cu.LICFlag
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
//cu.emtFlag is set below
PredictionUnit &pu = tempCS->addPU( cu, partitioner.chType );
DistParam distParam;
const bool bUseHadamard= !encTestMode.lossless;
m_pcRdCost->setDistParam (distParam, tempCS->getOrgBuf().Y(), m_acMergeBuffer[0].Y(), sps.getBitDepth (CHANNEL_TYPE_LUMA), COMPONENT_Y, bUseHadamard);
const UnitArea localUnitArea( tempCS->area.chromaFormat, Area( 0, 0, tempCS->area.Y().width, tempCS->area.Y().height) );
for( uint32_t uiMergeCand = 0; uiMergeCand < mergeCtx.numValidMergeCand; uiMergeCand++ )
{
#if !JVET_L0054_MMVD
acMergeBuffer[uiMergeCand] = m_acMergeBuffer[uiMergeCand].getBuf( localUnitArea );
#endif
mergeCtx.setMergeInfo( pu, uiMergeCand );
PU::spanMotionInfo( pu, mergeCtx );
#if JVET_L0054_MMVD
distParam.cur = singleMergeTempBuffer->Y();
m_pcInterSearch->motionCompensation(pu, *singleMergeTempBuffer);
#else
distParam.cur = acMergeBuffer[uiMergeCand].Y();
m_pcInterSearch->motionCompensation( pu, acMergeBuffer[uiMergeCand] );
#endif
if( mergeCtx.interDirNeighbours[uiMergeCand] == 3 && mergeCtx.mrgTypeNeighbours[uiMergeCand] == MRG_TYPE_DEFAULT_N )
{
mergeCtx.mvFieldNeighbours[2*uiMergeCand].mv = pu.mv[0];
mergeCtx.mvFieldNeighbours[2*uiMergeCand+1].mv = pu.mv[1];
#if DMVR_JVET_LOW_LATENCY_K0217
refinedMvdL0[uiMergeCand] = pu.mvd[0];
#endif
}
Distortion uiSad = distParam.distFunc(distParam);
uint32_t uiBitsCand = uiMergeCand + 1;
if( uiMergeCand == tempCS->slice->getMaxNumMergeCand() - 1 )
{
uiBitsCand--;
}
#if JVET_L0054_MMVD
uiBitsCand++; // for mmvd_flag
#endif
double cost = (double)uiSad + (double)uiBitsCand * sqrtLambdaForFirstPass;
#if JVET_L0054_MMVD
insertPos = -1;
updateCandList(uiMergeCand, cost, RdModeList, candCostList, uiNumMrgSATDCand, &insertPos);
if (insertPos != -1)
{
if (insertPos == RdModeList.size() - 1)
{
swap(singleMergeTempBuffer, acMergeTempBuffer[insertPos]);
}
else
{
for (uint32_t i = uint32_t(RdModeList.size()) - 1; i > insertPos; i--)
{
swap(acMergeTempBuffer[i - 1], acMergeTempBuffer[i]);
}
swap(singleMergeTempBuffer, acMergeTempBuffer[insertPos]);
}
}
#else
updateCandList( uiMergeCand, cost, RdModeList, candCostList, uiNumMrgSATDCand );
#endif
CHECK( std::min( uiMergeCand + 1, uiNumMrgSATDCand ) != RdModeList.size(), "" );
}
#if JVET_L0054_MMVD
cu.mmvdSkip = true;
int tempNum = 0;
tempNum = MMVD_ADD_NUM;
bool allowDirection[4] = { true, true, true, true };
for (uint32_t mergeCand = mergeCtx.numValidMergeCand; mergeCand < mergeCtx.numValidMergeCand + tempNum; mergeCand++)
{
const int mmvdMergeCand = mergeCand - mergeCtx.numValidMergeCand;
int bitsBaseIdx = 0;
int bitsRefineStep = 0;
int bitsDirection = 2;
int bitsCand = 0;
int baseIdx;
int refineStep;
int direction;
baseIdx = mmvdMergeCand / MMVD_MAX_REFINE_NUM;
refineStep = (mmvdMergeCand - (baseIdx * MMVD_MAX_REFINE_NUM)) / 4;
direction = (mmvdMergeCand - baseIdx * MMVD_MAX_REFINE_NUM - refineStep * 4) % 4;
if (refineStep == 0)
{
allowDirection[direction] = true;
}
if (allowDirection[direction] == false)
{
continue;
}
bitsBaseIdx = baseIdx + 1;
if (baseIdx == MMVD_BASE_MV_NUM - 1)
{
bitsBaseIdx--;
}
bitsRefineStep = refineStep + 1;
if (refineStep == MMVD_REFINE_STEP - 1)
{
bitsRefineStep--;
}
bitsCand = bitsBaseIdx + bitsRefineStep + bitsDirection;
bitsCand++; // for mmvd_flag
#if !JVET_L0054_MMVD
acMergeBuffer[mergeCand] = m_acMergeBuffer[mergeCand].getBuf(localUnitArea);
#endif
mergeCtx.setMmvdMergeCandiInfo(pu, mmvdMergeCand);
PU::spanMotionInfo(pu, mergeCtx);
#if JVET_L0054_MMVD
distParam.cur = singleMergeTempBuffer->Y();
m_pcInterSearch->motionCompensation(pu, *singleMergeTempBuffer);
#else
distParam.cur = acMergeBuffer[mergeCand].Y();
m_pcInterSearch->motionCompensation(pu, acMergeBuffer[mergeCand]);
#endif
Distortion uiSad = distParam.distFunc(distParam);
#if !JVET_L0054_MMVD
uint32_t bitsCand = mergeCand + 1;
if (mergeCand == tempCS->slice->getMaxNumMergeCand() - 1)
{
bitsCand--;
}
#endif
double cost = (double)uiSad + (double)bitsCand * sqrtLambdaForFirstPass;
#if JVET_L0054_MMVD
allowDirection[direction] = cost > 1.3 * candCostList[0] ? 0 : 1;
#endif
#if JVET_L0054_MMVD
insertPos = -1;
updateCandList(mergeCand, cost, RdModeList, candCostList, uiNumMrgSATDCand, &insertPos);
if (insertPos != -1)
{
for (int i = int(RdModeList.size()) - 1; i > insertPos; i--)
{
swap(acMergeTempBuffer[i - 1], acMergeTempBuffer[i]);
}
swap(singleMergeTempBuffer, acMergeTempBuffer[insertPos]);
}
#else
updateCandList(mergeCand, cost, RdModeList, candCostList, uiNumMrgSATDCand);
#endif
#if !JVET_L0054_MMVD
CHECK(std::min(mergeCand + 1, uiNumMrgSATDCand) != RdModeList.size(), "");
#endif
}
#endif
// Try to limit number of candidates using SATD-costs
for( uint32_t i = 1; i < uiNumMrgSATDCand; i++ )
{
if( candCostList[i] > MRG_FAST_RATIO * candCostList[0] )
{
uiNumMrgSATDCand = i;
break;
}
}
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
}
else
{
#if JVET_L0054_MMVD
if (bestIsMMVDSkip)
{
uiNumMrgSATDCand = mergeCtx.numValidMergeCand + MMVD_ADD_NUM;
}
else
{
uiNumMrgSATDCand = mergeCtx.numValidMergeCand;
}
#else
uiNumMrgSATDCand = mergeCtx.numValidMergeCand;
#endif
}
}
const uint32_t iteration = encTestMode.lossless ? 1 : 2;
// 2. Pass: check candidates using full RD test
for( uint32_t uiNoResidualPass = 0; uiNoResidualPass < iteration; uiNoResidualPass++ )
{
for( uint32_t uiMrgHADIdx = 0; uiMrgHADIdx < uiNumMrgSATDCand; uiMrgHADIdx++ )
{
uint32_t uiMergeCand = RdModeList[uiMrgHADIdx];
#if JVET_L0054_MMVD
if (((uiNoResidualPass != 0) && candHasNoResidual[uiMrgHADIdx])
#else
if( ( (uiNoResidualPass != 0) && candHasNoResidual[uiMergeCand] )
#endif
|| ( (uiNoResidualPass == 0) && bestIsSkip ) )
{
continue;
}
// first get merge candidates
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = SIZE_2Nx2N;
//cu.affine
cu.predMode = MODE_INTER;
//cu.LICFlag
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
PredictionUnit &pu = tempCS->addPU( cu, partitioner.chType );
#if JVET_L0054_MMVD
if (uiMergeCand >= mergeCtx.numValidMergeCand)
{
cu.mmvdSkip = true;
mergeCtx.setMmvdMergeCandiInfo(pu, uiMergeCand - mergeCtx.numValidMergeCand);
}
else
{
cu.mmvdSkip = false;
mergeCtx.setMergeInfo(pu, uiMergeCand);
}
#else
mergeCtx.setMergeInfo( pu, uiMergeCand );
#endif
PU::spanMotionInfo( pu, mergeCtx );
if( mrgTempBufSet )
{
#if DMVR_JVET_LOW_LATENCY_K0217
pu.mvd[0] = refinedMvdL0[uiMergeCand];
#endif
#if JVET_L0054_MMVD
tempCS->getPredBuf().copyFrom(*acMergeTempBuffer[uiMrgHADIdx]);
#else
tempCS->getPredBuf().copyFrom( acMergeBuffer[ uiMergeCand ]);
#endif
}
else
{
m_pcInterSearch->motionCompensation( pu );
}
#if JVET_L0054_MMVD
xEncodeInterResidual(tempCS, bestCS, partitioner, encTestMode, uiNoResidualPass
, NULL
, 1
, uiNoResidualPass == 0 ? &candHasNoResidual[uiMrgHADIdx] : NULL);
#else
xEncodeInterResidual( tempCS, bestCS, partitioner, encTestMode, uiNoResidualPass
, NULL
, 1
, uiNoResidualPass == 0 ? &candHasNoResidual[uiMergeCand] : NULL );
#endif
if( m_pcEncCfg->getUseFastDecisionForMerge() && !bestIsSkip )
{
bestIsSkip = bestCS->getCU( partitioner.chType )->rootCbf == 0;
}
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
}// end loop uiMrgHADIdx
if( uiNoResidualPass == 0 && m_pcEncCfg->getUseEarlySkipDetection() )
{
const CodingUnit &bestCU = *bestCS->getCU( partitioner.chType );
const PredictionUnit &bestPU = *bestCS->getPU( partitioner.chType );
if( bestCU.rootCbf == 0 )
{
if( bestPU.mergeFlag )
{
m_modeCtrl->setEarlySkipDetected();
}
else if( m_pcEncCfg->getMotionEstimationSearchMethod() != MESEARCH_SELECTIVE )
{
int absolute_MV = 0;
for( uint32_t uiRefListIdx = 0; uiRefListIdx < 2; uiRefListIdx++ )
{
if( slice.getNumRefIdx( RefPicList( uiRefListIdx ) ) > 0 )
{
absolute_MV += bestPU.mvd[uiRefListIdx].getAbsHor() + bestPU.mvd[uiRefListIdx].getAbsVer();
}
}
if( absolute_MV == 0 )
{
m_modeCtrl->setEarlySkipDetected();
}
}
}
}
}
}
void EncCu::xCheckRDCostAffineMerge2Nx2N( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
{
if( m_modeCtrl->getFastDeltaQp() )
{
return;
}
if ( bestCS->area.lumaSize().width < 8 || bestCS->area.lumaSize().height < 8 )
{
return;
}
#if JVET_L0632_AFFINE_MERGE
const Slice &slice = *tempCS->slice;
CHECK( slice.getSliceType() == I_SLICE, "Affine Merge modes not available for I-slices" );
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
AffineMergeCtx affineMergeCtx;
const SPS &sps = *tempCS->sps;
#if JVET_L0369_SUBBLOCK_MERGE
MergeCtx mrgCtx;
if ( sps.getSpsNext().getUseSubPuMvp() )
{
Size bufSize = g_miScaling.scale( tempCS->area.lumaSize() );
mrgCtx.subPuMvpMiBuf = MotionBuf( m_SubPuMiBuf, bufSize );
affineMergeCtx.mrgCtx = &mrgCtx;
}
#endif
{
// first get merge candidates
CodingUnit cu( tempCS->area );
cu.cs = tempCS;
cu.partSize = SIZE_2Nx2N;
cu.predMode = MODE_INTER;
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
PredictionUnit pu( tempCS->area );
pu.cu = &cu;
pu.cs = tempCS;
PU::getAffineMergeCand( pu, affineMergeCtx );
if ( affineMergeCtx.numValidMergeCand <= 0 )
{
return;
}
}
bool candHasNoResidual[AFFINE_MRG_MAX_NUM_CANDS];
for ( uint32_t ui = 0; ui < affineMergeCtx.numValidMergeCand; ui++ )
{
candHasNoResidual[ui] = false;
}
bool bestIsSkip = false;
uint32_t uiNumMrgSATDCand = affineMergeCtx.numValidMergeCand;
PelUnitBuf acMergeBuffer[AFFINE_MRG_MAX_NUM_CANDS];
static_vector<uint32_t, AFFINE_MRG_MAX_NUM_CANDS> RdModeList;
bool mrgTempBufSet = false;
for ( uint32_t i = 0; i < AFFINE_MRG_MAX_NUM_CANDS; i++ )
{
RdModeList.push_back( i );
}
if ( m_pcEncCfg->getUseFastMerge() )
{
uiNumMrgSATDCand = std::min( NUM_AFF_MRG_SATD_CAND, affineMergeCtx.numValidMergeCand );
bestIsSkip = false;
if ( auto blkCache = dynamic_cast<CacheBlkInfoCtrl*>(m_modeCtrl) )
{
bestIsSkip = blkCache->isSkip( tempCS->area );
}
static_vector<double, AFFINE_MRG_MAX_NUM_CANDS> candCostList;
// 1. Pass: get SATD-cost for selected candidates and reduce their count
if ( !bestIsSkip )
{
RdModeList.clear();
mrgTempBufSet = true;
const double sqrtLambdaForFirstPass = m_pcRdCost->getMotionLambda( encTestMode.lossless );
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
cu.partSize = SIZE_2Nx2N;
cu.affine = true;
cu.predMode = MODE_INTER;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
PredictionUnit &pu = tempCS->addPU( cu, partitioner.chType );
DistParam distParam;
const bool bUseHadamard = !encTestMode.lossless;
m_pcRdCost->setDistParam( distParam, tempCS->getOrgBuf().Y(), m_acMergeBuffer[0].Y(), sps.getBitDepth( CHANNEL_TYPE_LUMA ), COMPONENT_Y, bUseHadamard );
const UnitArea localUnitArea( tempCS->area.chromaFormat, Area( 0, 0, tempCS->area.Y().width, tempCS->area.Y().height ) );
for ( uint32_t uiMergeCand = 0; uiMergeCand < affineMergeCtx.numValidMergeCand; uiMergeCand++ )
{
acMergeBuffer[uiMergeCand] = m_acMergeBuffer[uiMergeCand].getBuf( localUnitArea );
// set merge information
pu.interDir = affineMergeCtx.interDirNeighbours[uiMergeCand];
pu.mergeFlag = true;
pu.mergeIdx = uiMergeCand;
cu.affineType = affineMergeCtx.affineType[uiMergeCand];
#if JVET_L0646_GBI
cu.GBiIdx = affineMergeCtx.GBiIdx[uiMergeCand];
#endif
#if JVET_L0369_SUBBLOCK_MERGE
pu.mergeType = affineMergeCtx.mergeType[uiMergeCand];
if ( pu.mergeType == MRG_TYPE_SUBPU_ATMVP )
{
pu.refIdx[0] = affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 0][0].refIdx;
pu.refIdx[1] = affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 1][0].refIdx;
PU::spanMotionInfo( pu, mrgCtx );
}
else
{
#endif
PU::setAllAffineMvField( pu, affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 0], REF_PIC_LIST_0 );
PU::setAllAffineMvField( pu, affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 1], REF_PIC_LIST_1 );
PU::spanMotionInfo( pu );
#if JVET_L0369_SUBBLOCK_MERGE
}
#endif
distParam.cur = acMergeBuffer[uiMergeCand].Y();
m_pcInterSearch->motionCompensation( pu, acMergeBuffer[uiMergeCand] );
Distortion uiSad = distParam.distFunc( distParam );
uint32_t uiBitsCand = uiMergeCand + 1;
if ( uiMergeCand == tempCS->slice->getMaxNumAffineMergeCand() - 1 )
{
uiBitsCand--;
}
double cost = (double)uiSad + (double)uiBitsCand * sqrtLambdaForFirstPass;
updateCandList( uiMergeCand, cost, RdModeList, candCostList, uiNumMrgSATDCand );
CHECK( std::min( uiMergeCand + 1, uiNumMrgSATDCand ) != RdModeList.size(), "" );
}
// Try to limit number of candidates using SATD-costs
for ( uint32_t i = 1; i < uiNumMrgSATDCand; i++ )
{
if ( candCostList[i] > MRG_FAST_RATIO * candCostList[0] )
{
uiNumMrgSATDCand = i;
break;
}
}
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
}
else
{
uiNumMrgSATDCand = affineMergeCtx.numValidMergeCand;
}
}
const uint32_t iteration = encTestMode.lossless ? 1 : 2;
// 2. Pass: check candidates using full RD test
for ( uint32_t uiNoResidualPass = 0; uiNoResidualPass < iteration; uiNoResidualPass++ )
{
for ( uint32_t uiMrgHADIdx = 0; uiMrgHADIdx < uiNumMrgSATDCand; uiMrgHADIdx++ )
{
uint32_t uiMergeCand = RdModeList[uiMrgHADIdx];
if ( ((uiNoResidualPass != 0) && candHasNoResidual[uiMergeCand])
|| ((uiNoResidualPass == 0) && bestIsSkip) )
{
continue;
}
// first get merge candidates
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
cu.partSize = SIZE_2Nx2N;
cu.affine = true;
cu.predMode = MODE_INTER;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
PredictionUnit &pu = tempCS->addPU( cu, partitioner.chType );
// set merge information
pu.mergeFlag = true;
pu.mergeIdx = uiMergeCand;
pu.interDir = affineMergeCtx.interDirNeighbours[uiMergeCand];
cu.affineType = affineMergeCtx.affineType[uiMergeCand];
#if JVET_L0646_GBI
cu.GBiIdx = affineMergeCtx.GBiIdx[uiMergeCand];
#endif
#if JVET_L0369_SUBBLOCK_MERGE
pu.mergeType = affineMergeCtx.mergeType[uiMergeCand];
if ( pu.mergeType == MRG_TYPE_SUBPU_ATMVP )
{
pu.refIdx[0] = affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 0][0].refIdx;
pu.refIdx[1] = affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 1][0].refIdx;
PU::spanMotionInfo( pu, mrgCtx );
}
else
{
#endif
PU::setAllAffineMvField( pu, affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 0], REF_PIC_LIST_0 );
PU::setAllAffineMvField( pu, affineMergeCtx.mvFieldNeighbours[(uiMergeCand << 1) + 1], REF_PIC_LIST_1 );
PU::spanMotionInfo( pu );
#if JVET_L0369_SUBBLOCK_MERGE
}
#endif
if ( mrgTempBufSet )
{
tempCS->getPredBuf().copyFrom( acMergeBuffer[uiMergeCand] );
}
else
{
m_pcInterSearch->motionCompensation( pu );
}
xEncodeInterResidual( tempCS, bestCS, partitioner, encTestMode, uiNoResidualPass, NULL, true, ((uiNoResidualPass == 0) ? &candHasNoResidual[uiMergeCand] : NULL) );
if ( m_pcEncCfg->getUseFastDecisionForMerge() && !bestIsSkip )
{
bestIsSkip = bestCS->getCU( partitioner.chType )->rootCbf == 0;
}
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
}// end loop uiMrgHADIdx
if ( uiNoResidualPass == 0 && m_pcEncCfg->getUseEarlySkipDetection() )
{
const CodingUnit &bestCU = *bestCS->getCU( partitioner.chType );
const PredictionUnit &bestPU = *bestCS->getPU( partitioner.chType );
if ( bestCU.rootCbf == 0 )
{
if ( bestPU.mergeFlag )
{
m_modeCtrl->setEarlySkipDetected();
}
else if ( m_pcEncCfg->getMotionEstimationSearchMethod() != MESEARCH_SELECTIVE )
{
int absolute_MV = 0;
for ( uint32_t uiRefListIdx = 0; uiRefListIdx < 2; uiRefListIdx++ )
{
if ( slice.getNumRefIdx( RefPicList( uiRefListIdx ) ) > 0 )
{
absolute_MV += bestPU.mvd[uiRefListIdx].getAbsHor() + bestPU.mvd[uiRefListIdx].getAbsVer();
}
}
if ( absolute_MV == 0 )
{
m_modeCtrl->setEarlySkipDetected();
}
}
}
}
}
#else
MvField affineMvField[2][3];
unsigned char interDirNeighbours;
int numValidMergeCand;
bool hasNoResidual = false;
#if JVET_L0646_GBI
uint8_t gbiIdx = GBI_DEFAULT;
#endif
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = encTestMode.partSize;
cu.affine = true;
cu.predMode = MODE_INTER;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
CU::addPUs( cu );
cu.firstPU->mergeFlag = true;
cu.firstPU->mergeIdx = 0;
#if JVET_L0646_GBI
PU::getAffineMergeCand( *cu.firstPU, affineMvField, interDirNeighbours, gbiIdx, numValidMergeCand );
#else
PU::getAffineMergeCand( *cu.firstPU, affineMvField, interDirNeighbours, numValidMergeCand );
#endif
if( numValidMergeCand == -1 )
{
return;
}
cu.firstPU->interDir = interDirNeighbours;
PU::setAllAffineMvField( *cu.firstPU, affineMvField[REF_PIC_LIST_0], REF_PIC_LIST_0 );
PU::setAllAffineMvField( *cu.firstPU, affineMvField[REF_PIC_LIST_1], REF_PIC_LIST_1 );
#if JVET_L0646_GBI
cu.GBiIdx = gbiIdx;
#endif
PU::spanMotionInfo( *cu.firstPU );
m_pcInterSearch->motionCompensation( cu );
xEncodeInterResidual(tempCS, bestCS, partitioner, encTestMode, 0
, NULL
, 1
, &hasNoResidual);
if( ! (encTestMode.lossless || hasNoResidual) )
{
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
tempCS->copyStructure( *bestCS, partitioner.chType );
tempCS->getPredBuf().copyFrom( bestCS->getPredBuf() );
xEncodeInterResidual(tempCS, bestCS, partitioner, encTestMode, 1
, NULL
, 1
, &hasNoResidual);
}
#endif
}
void EncCu::xCheckRDCostInter( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
{
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
#if JVET_L0646_GBI
m_pcInterSearch->setAffineModeSelected(false);
if( tempCS->slice->getCheckLDC() )
{
m_bestGbiCost[0] = m_bestGbiCost[1] = std::numeric_limits<double>::max();
m_bestGbiIdx[0] = m_bestGbiIdx[1] = -1;
}
m_pcInterSearch->resetBufferedUniMotions();
int gbiLoopNum = (tempCS->slice->isInterB() ? GBI_NUM : 1);
gbiLoopNum = (tempCS->sps->getSpsNext().getUseGBi() ? gbiLoopNum : 1);
if( tempCS->area.lwidth() * tempCS->area.lheight() < GBI_SIZE_CONSTRAINT )
{
gbiLoopNum = 1;
}
double curBestCost = bestCS->cost;
double equGBiCost = MAX_DOUBLE;
for( int gbiLoopIdx = 0; gbiLoopIdx < gbiLoopNum; gbiLoopIdx++ )
{
if( m_pcEncCfg->getUseGBiFast() )
{
auto blkCache = dynamic_cast< CacheBlkInfoCtrl* >(m_modeCtrl);
if( blkCache )
{
bool isBestInter = blkCache->getInter(bestCS->area);
uint8_t bestGBiIdx = blkCache->getGbiIdx(bestCS->area);
if( isBestInter && g_GbiSearchOrder[gbiLoopIdx] != GBI_DEFAULT && g_GbiSearchOrder[gbiLoopIdx] != bestGBiIdx )
{
continue;
}
}
}
if( !tempCS->slice->getCheckLDC() )
{
if( gbiLoopIdx != 0 && gbiLoopIdx != 3 && gbiLoopIdx != 4 )
{
continue;
}
}
#endif
CodingUnit &cu = tempCS->addCU( tempCS->area, partitioner.chType );
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = encTestMode.partSize;
//cu.affine
cu.predMode = MODE_INTER;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
CU::addPUs( cu );
#if JVET_L0646_GBI
cu.GBiIdx = g_GbiSearchOrder[gbiLoopIdx];
uint8_t gbiIdx = cu.GBiIdx;
bool testGbi = (gbiIdx != GBI_DEFAULT);
#endif
m_pcInterSearch->predInterSearch( cu, partitioner );
const unsigned wIdx = gp_sizeIdxInfo->idxFrom( tempCS->area.lwidth () );
#if JVET_L0646_GBI
gbiIdx = CU::getValidGbiIdx(cu);
if( testGbi && gbiIdx == GBI_DEFAULT ) // Enabled GBi but the search results is uni.
{
tempCS->initStructData(encTestMode.qp, encTestMode.lossless);
continue;
}
CHECK(!(testGbi || (!testGbi && gbiIdx == GBI_DEFAULT)), " !( bTestGbi || (!bTestGbi && gbiIdx == GBI_DEFAULT ) )");
bool isEqualUni = false;
if( m_pcEncCfg->getUseGBiFast() )
{
if( cu.firstPU->interDir != 3 && testGbi == 0 )
{
isEqualUni = true;
}
}
#endif
xEncodeInterResidual( tempCS, bestCS, partitioner, encTestMode, 0
, m_pImvTempCS ? m_pImvTempCS[wIdx][encTestMode.partSize] : NULL
, 1
, 0
#if JVET_L0646_GBI
, &equGBiCost
#endif
);
#if JVET_L0646_GBI
if( g_GbiSearchOrder[gbiLoopIdx] == GBI_DEFAULT )
m_pcInterSearch->setAffineModeSelected((bestCS->cus.front()->affine && !(bestCS->cus.front()->firstPU->mergeFlag)));
tempCS->initStructData(encTestMode.qp, encTestMode.lossless);
double skipTH = MAX_DOUBLE;
skipTH = (m_pcEncCfg->getUseGBiFast() ? 1.05 : MAX_DOUBLE);
if( equGBiCost > curBestCost * skipTH )
{
break;
}
if( m_pcEncCfg->getUseGBiFast() )
{
if( isEqualUni == true && m_pcEncCfg->getIntraPeriod() == -1 )
{
break;
}
}
if( g_GbiSearchOrder[gbiLoopIdx] == GBI_DEFAULT && xIsGBiSkip(cu) && m_pcEncCfg->getUseGBiFast() )
{
break;
}
} // for( UChar gbiLoopIdx = 0; gbiLoopIdx < gbiLoopNum; gbiLoopIdx++ )
#endif
}
bool EncCu::xCheckRDCostInterIMV( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode )
{
int iIMV = int( ( encTestMode.opts & ETO_IMV ) >> ETO_IMV_SHIFT );
#if JVET_L0646_GBI
m_pcInterSearch->setAffineModeSelected(false);
#endif
// Only int-Pel, 4-Pel and fast 4-Pel allowed
CHECK( iIMV != 1 && iIMV != 2 && iIMV != 3, "Unsupported IMV Mode" );
// Fast 4-Pel Mode
EncTestMode encTestModeBase = encTestMode; // copy for clearing non-IMV options
encTestModeBase.opts = EncTestModeOpts( encTestModeBase.opts & ETO_IMV ); // clear non-IMV options (is that intended?)
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
CodingStructure* pcCUInfo2Reuse = nullptr;
if( m_pImvTempCS && encTestMode.partSize != SIZE_2Nx2N && ( ( encTestMode.opts & ETO_FORCE_MERGE ) == 0 ) )
{
pcCUInfo2Reuse = m_pImvTempCS[gp_sizeIdxInfo->idxFrom( tempCS->area.lwidth() )][encTestMode.partSize];
if( pcCUInfo2Reuse != nullptr )
{
CHECK( tempCS->area != pcCUInfo2Reuse->area, " mismatch" );
tempCS->copyStructure( *pcCUInfo2Reuse, partitioner.chType );
tempCS->fracBits = 0;
tempCS->dist = 0;
tempCS->cost = MAX_DOUBLE;
}
}
#if JVET_L0646_GBI
m_pcInterSearch->resetBufferedUniMotions();
int gbiLoopNum = (tempCS->slice->isInterB() ? GBI_NUM : 1);
gbiLoopNum = (pcCUInfo2Reuse != NULL ? 1 : gbiLoopNum);
gbiLoopNum = (tempCS->slice->getSPS()->getSpsNext().getUseGBi() ? gbiLoopNum : 1);
if( tempCS->area.lwidth() * tempCS->area.lheight() < GBI_SIZE_CONSTRAINT )
{
gbiLoopNum = 1;
}
double curBestCost = bestCS->cost;
double equGBiCost = MAX_DOUBLE;
for( int gbiLoopIdx = 0; gbiLoopIdx < gbiLoopNum; gbiLoopIdx++ )
{
if( m_pcEncCfg->getUseGBiFast() )
{
auto blkCache = dynamic_cast< CacheBlkInfoCtrl* >(m_modeCtrl);
if( blkCache )
{
bool isBestInter = blkCache->getInter(bestCS->area);
uint8_t bestGBiIdx = blkCache->getGbiIdx(bestCS->area);
if( isBestInter && g_GbiSearchOrder[gbiLoopIdx] != GBI_DEFAULT && g_GbiSearchOrder[gbiLoopIdx] != bestGBiIdx )
{
continue;
}
}
}
if( !tempCS->slice->getCheckLDC() )
{
if( gbiLoopIdx != 0 && gbiLoopIdx != 3 && gbiLoopIdx != 4 )
{
continue;
}
}
if( m_pcEncCfg->getUseGBiFast() && tempCS->slice->getCheckLDC() && g_GbiSearchOrder[gbiLoopIdx] != GBI_DEFAULT
&& (m_bestGbiIdx[0] >= 0 && g_GbiSearchOrder[gbiLoopIdx] != m_bestGbiIdx[0])
&& (m_bestGbiIdx[1] >= 0 && g_GbiSearchOrder[gbiLoopIdx] != m_bestGbiIdx[1]))
{
continue;
}
#endif
CodingUnit &cu = ( pcCUInfo2Reuse != nullptr ) ? *tempCS->getCU( partitioner.chType ) : tempCS->addCU( tempCS->area, partitioner.chType );
if( pcCUInfo2Reuse == nullptr )
{
partitioner.setCUData( cu );
cu.slice = tempCS->slice;
#if HEVC_TILES_WPP
cu.tileIdx = tempCS->picture->tileMap->getTileIdxMap( tempCS->area.lumaPos() );
#endif
cu.skip = false;
#if JVET_L0054_MMVD
cu.mmvdSkip = false;
#endif
cu.partSize = encTestMode.partSize;
//cu.affine
cu.predMode = MODE_INTER;
cu.transQuantBypass = encTestMode.lossless;
cu.chromaQpAdj = cu.transQuantBypass ? 0 : m_cuChromaQpOffsetIdxPlus1;
cu.qp = encTestMode.qp;
CU::addPUs( cu );
}
else
{
CHECK( cu.partSize != encTestMode.partSize, "Mismatch" );
CHECK( cu.skip, "Mismatch" );
CHECK( cu.qtDepth != partitioner.currQtDepth, "Mismatch" );
CHECK( cu.btDepth != partitioner.currBtDepth, "Mismatch" );
CHECK( cu.mtDepth != partitioner.currMtDepth, "Mismatch" );
CHECK( cu.depth != partitioner.currDepth, "Mismatch" );
}
cu.imv = iIMV > 1 ? 2 : 1;
cu.emtFlag = false;
#if JVET_L0646_GBI
bool testGbi;
uint8_t gbiIdx;
#endif
if( pcCUInfo2Reuse != nullptr )
{
// reuse the motion info from pcCUInfo2Reuse
CU::resetMVDandMV2Int( cu, m_pcInterSearch );
#if JVET_L0646_GBI
CHECK(cu.GBiIdx < 0 || cu.GBiIdx >= GBI_NUM, "cu.GBiIdx < 0 || cu.GBiIdx >= GBI_NUM");
gbiIdx = CU::getValidGbiIdx(cu);
testGbi = (gbiIdx != GBI_DEFAULT);
#endif
if( !CU::hasSubCUNonZeroMVd( cu ) )
{
m_modeCtrl->useModeResult( encTestModeBase, tempCS, partitioner );
return false;
}
else
{
m_pcInterSearch->motionCompensation( cu );
}
}
else
{
#if JVET_L0646_GBI
cu.GBiIdx = g_GbiSearchOrder[gbiLoopIdx];
gbiIdx = cu.GBiIdx;
testGbi = (gbiIdx != GBI_DEFAULT);
#endif
m_pcInterSearch->predInterSearch( cu, partitioner );
#if JVET_L0646_GBI
gbiIdx = CU::getValidGbiIdx(cu);
#endif
}
#if JVET_L0646_GBI
if( testGbi && gbiIdx == GBI_DEFAULT ) // Enabled GBi but the search results is uni.
{
tempCS->initStructData(encTestMode.qp, encTestMode.lossless);
continue;
}
CHECK(!(testGbi || (!testGbi && gbiIdx == GBI_DEFAULT)), " !( bTestGbi || (!bTestGbi && gbiIdx == GBI_DEFAULT ) )");
bool isEqualUni = false;
if( m_pcEncCfg->getUseGBiFast() )
{
if( cu.firstPU->interDir != 3 && testGbi == 0 )
{
isEqualUni = true;
}
}
#endif
if( !CU::hasSubCUNonZeroMVd( cu ) )
{
m_modeCtrl->useModeResult( encTestModeBase, tempCS, partitioner );
return false;
}
xEncodeInterResidual( tempCS, bestCS, partitioner, encTestModeBase, 0
, NULL
, true
, 0
#if JVET_L0646_GBI
, &equGBiCost
#endif
);
#if JVET_L0646_GBI
tempCS->initStructData(encTestMode.qp, encTestMode.lossless);
double skipTH = MAX_DOUBLE;
skipTH = (m_pcEncCfg->getUseGBiFast() ? 1.05 : MAX_DOUBLE);
if( equGBiCost > curBestCost * skipTH )
{
break;
}
if( m_pcEncCfg->getUseGBiFast() )
{
if( isEqualUni == true && m_pcEncCfg->getIntraPeriod() == -1 )
{
break;
}
}
if( g_GbiSearchOrder[gbiLoopIdx] == GBI_DEFAULT && xIsGBiSkip(cu) && m_pcEncCfg->getUseGBiFast() )
{
break;
}
} // for( UChar gbiLoopIdx = 0; gbiLoopIdx < gbiLoopNum; gbiLoopIdx++ )
#endif
return true;
}
void EncCu::xEncodeInterResidual( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner, const EncTestMode& encTestMode, int residualPass
, CodingStructure* imvCS
, int emtMode
, bool* bestHasNonResi
#if JVET_L0646_GBI
, double* equGBiCost
#endif
)
{
if( residualPass == 1 && encTestMode.lossless )
{
return;
}
CodingUnit* cu = tempCS->getCU( partitioner.chType );
double bestCostInternal = MAX_DOUBLE;
double bestCost = bestCS->cost;
const SPS& sps = *tempCS->sps;
const int maxSizeEMT = tempCS->pcv->noRQT ? EMT_INTER_MAX_CU_WITH_QTBT : EMT_INTER_MAX_CU;
bool swapped = false; // avoid unwanted data copy
bool reloadCU = false;
const bool considerEmtSecondPass = emtMode && sps.getSpsNext().getUseInterEMT() && partitioner.currArea().lwidth() <= maxSizeEMT && partitioner.currArea().lheight() <= maxSizeEMT;
int minEMTMode = 0;
int maxEMTMode = (considerEmtSecondPass?1:0);
if( emtMode == 2 )
{
minEMTMode = maxEMTMode = (cu->emtFlag?1:0);
}
for( int curEmtMode = minEMTMode; curEmtMode <= maxEMTMode; curEmtMode++ )
{
if( reloadCU )
{
if( bestCost == bestCS->cost ) //The first EMT pass didn't become the bestCS, so we clear the TUs generated
{
tempCS->clearTUs();
}
else if( false == swapped )
{
tempCS->initStructData( encTestMode.qp, encTestMode.lossless );
tempCS->copyStructure( *bestCS, partitioner.chType );
tempCS->getPredBuf().copyFrom( bestCS->getPredBuf() );
bestCost = bestCS->cost;
cu = tempCS->getCU( partitioner.chType );
swapped = true;
}
else
{
tempCS->clearTUs();
bestCost = bestCS->cost;
cu = tempCS->getCU( partitioner.chType );
}
//we need to restart the distortion for the new tempCS, the bit count and the cost
tempCS->dist = 0;
tempCS->fracBits = 0;
tempCS->cost = MAX_DOUBLE;
}
reloadCU = true; // enable cu reloading
cu->skip = false;
cu->emtFlag = curEmtMode;
const bool skipResidual = residualPass == 1;
m_pcInterSearch->encodeResAndCalcRdInterCU( *tempCS, partitioner, skipResidual );
xEncodeDontSplit( *tempCS, partitioner );
xCheckDQP( *tempCS, partitioner );
#if JVET_L0646_GBI
if( ETM_INTER_ME == encTestMode.type )
{
if( equGBiCost != NULL )
{
if( tempCS->cost < (*equGBiCost) && cu->GBiIdx == GBI_DEFAULT )
{
(*equGBiCost) = tempCS->cost;
}
}
else
{
CHECK(equGBiCost == NULL, "equGBiCost == NULL");
}
if( tempCS->slice->getCheckLDC() && !cu->imv && cu->GBiIdx != GBI_DEFAULT && tempCS->cost < m_bestGbiCost[1] )
{
if( tempCS->cost < m_bestGbiCost[0] )
{
m_bestGbiCost[1] = m_bestGbiCost[0];
m_bestGbiCost[0] = tempCS->cost;
m_bestGbiIdx[1] = m_bestGbiIdx[0];
m_bestGbiIdx[0] = cu->GBiIdx;
}
else
{
m_bestGbiCost[1] = tempCS->cost;
m_bestGbiIdx[1] = cu->GBiIdx;
}
}
}
#endif
double emtFirstPassCost = tempCS->cost;
if( imvCS && (tempCS->cost < imvCS->cost) )
{
if( imvCS->cost != MAX_DOUBLE )
{
imvCS->initStructData( encTestMode.qp, encTestMode.lossless );
}
imvCS->copyStructure( *tempCS, partitioner.chType );
}
if( NULL != bestHasNonResi && (bestCostInternal > tempCS->cost) )
{
bestCostInternal = tempCS->cost;
*bestHasNonResi = !cu->rootCbf;
}
#if WCG_EXT
DTRACE_MODE_COST( *tempCS, m_pcRdCost->getLambda( true ) );
#else
DTRACE_MODE_COST( *tempCS, m_pcRdCost->getLambda() );
#endif
xCheckBestMode( tempCS, bestCS, partitioner, encTestMode );
//now we check whether the second pass should be skipped or not
if( !curEmtMode && maxEMTMode )
{
const double thresholdToSkipEmtSecondPass = 1.1; // Skip checking EMT transforms
const bool bCond1 = !cu->firstTU->cbf[COMPONENT_Y];
const bool bCond3 = emtFirstPassCost > ( bestCost * thresholdToSkipEmtSecondPass );
if( m_pcEncCfg->getFastInterEMT() && (bCond1 || bCond3 ) )
{
maxEMTMode = 0; // do not test EMT
}
}
} //end emt loop
}
void EncCu::xEncodeDontSplit( CodingStructure &cs, Partitioner &partitioner )
{
m_CABACEstimator->resetBits();
{
if( partitioner.canSplit( CU_QUAD_SPLIT, cs ) )
{
m_CABACEstimator->split_cu_flag( false, cs, partitioner );
}
if( partitioner.canSplit( CU_MT_SPLIT, cs ) )
{
m_CABACEstimator->split_cu_mode_mt( CU_DONT_SPLIT, cs, partitioner );
}
}
cs.fracBits += m_CABACEstimator->getEstFracBits(); // split bits
cs.cost = m_pcRdCost->calcRdCost( cs.fracBits, cs.dist );
}
#if REUSE_CU_RESULTS
void EncCu::xReuseCachedResult( CodingStructure *&tempCS, CodingStructure *&bestCS, Partitioner &partitioner )
{
const SPS &sps = *tempCS->sps;
BestEncInfoCache* bestEncCache = dynamic_cast<BestEncInfoCache*>( m_modeCtrl );
CHECK( !bestEncCache, "If this mode is chosen, mode controller has to implement the mode caching capabilities" );
EncTestMode cachedMode;
if( bestEncCache->setCsFrom( *tempCS, cachedMode, partitioner ) )
{
CodingUnit& cu = *tempCS->cus.front();
partitioner.setCUData( cu );
if( CU::isIntra( cu ) )
{
xReconIntraQT( cu );
}
else
{
xDeriveCUMV( cu );
xReconInter( cu );
}
Distortion finalDistortion = 0;
const int numValidComponents = getNumberValidComponents( tempCS->area.chromaFormat );
for( int comp = 0; comp < numValidComponents; comp++ )
{
const ComponentID compID = ComponentID( comp );
if( CS::isDualITree( *tempCS ) && toChannelType( compID ) != partitioner.chType )
{
continue;
}
CPelBuf reco = tempCS->getRecoBuf( compID );
CPelBuf org = tempCS->getOrgBuf ( compID );
#if WCG_EXT
if( m_pcEncCfg->getLumaLevelToDeltaQPMapping().isEnabled() )
{
const CPelBuf orgLuma = tempCS->getOrgBuf(tempCS->area.blocks[COMPONENT_Y]);
finalDistortion += m_pcRdCost->getDistPart( org, reco, sps.getBitDepth( toChannelType( compID ) ), compID, DF_SSE_WTD, &orgLuma );
}
else
#endif
finalDistortion += m_pcRdCost->getDistPart( org, reco, sps.getBitDepth( toChannelType( compID ) ), compID, DF_SSE );
}
m_CABACEstimator->getCtx() = m_CurrCtx->start;
m_CABACEstimator->resetBits();
CUCtx cuCtx;
cuCtx.isDQPCoded = true;
cuCtx.isChromaQpAdjCoded = true;
m_CABACEstimator->coding_unit( cu, partitioner, cuCtx );
tempCS->dist = finalDistortion;
tempCS->fracBits = m_CABACEstimator->getEstFracBits();
tempCS->cost = m_pcRdCost->calcRdCost( tempCS->fracBits, tempCS->dist );
xEncodeDontSplit( *tempCS, partitioner );
xCheckDQP ( *tempCS, partitioner );
xCheckBestMode ( tempCS, bestCS, partitioner, cachedMode );
}
else
{
THROW( "Should never happen!" );
}
}
#endif
//! \}