EncGOP.cpp

          , (m_pcCfg->getUsePerceptQPA() && !m_pcCfg->getUseRateCtrl() && pcSlice->getPPS()->getUseDQP() ? m_pcEncLib->getRdCost(PARL_PARAM0(0))->getChromaWeight() : 0.0)
#endif
          , pcPic, uiNumSliceSegments
        );

        //assign ALF slice header
        for (int s = 0; s < uiNumSliceSegments; s++)
        {
           //For the first slice, even if it is lossless, slice level ALF is not disabled and ALF-APS is signaled so that the later lossy slices can use APS of the first slice. 
           //However, if the first slice is lossless, the ALF process is disabled for all of the CTUs ( m_ctuEnableFlag == 0) of that slice which is implemented in the function void EncAdaptiveLoopFilter::ALFProcess.         
      
          if (pcPic->slices[s]->isLossless() && s && m_pcCfg->getCostMode() == COST_LOSSLESS_CODING)
          {
            pcPic->slices[s]->setTileGroupAlfEnabledFlag(COMPONENT_Y, false);
            pcPic->slices[s]->setTileGroupAlfEnabledFlag(COMPONENT_Cb, false);
            pcPic->slices[s]->setTileGroupAlfEnabledFlag(COMPONENT_Cr, false);
          }
          else
          {
            pcPic->slices[s]->setTileGroupAlfEnabledFlag(COMPONENT_Y, cs.slice->getTileGroupAlfEnabledFlag(COMPONENT_Y));
            pcPic->slices[s]->setTileGroupAlfEnabledFlag(COMPONENT_Cb, cs.slice->getTileGroupAlfEnabledFlag(COMPONENT_Cb));
            pcPic->slices[s]->setTileGroupAlfEnabledFlag(COMPONENT_Cr, cs.slice->getTileGroupAlfEnabledFlag(COMPONENT_Cr));

          }
          if (pcPic->slices[s]->getTileGroupAlfEnabledFlag(COMPONENT_Y))
          {
            pcPic->slices[s]->setTileGroupNumAps(cs.slice->getTileGroupNumAps());
            pcPic->slices[s]->setAlfAPSs(cs.slice->getTileGroupApsIdLuma());
          }
          else
          {
            pcPic->slices[s]->setTileGroupNumAps(0);
          }
          pcPic->slices[s]->setAlfAPSs(cs.slice->getAlfAPSs());
          pcPic->slices[s]->setTileGroupApsIdChroma(cs.slice->getTileGroupApsIdChroma());
          pcPic->slices[s]->setTileGroupCcAlfCbApsId(cs.slice->getTileGroupCcAlfCbApsId());
          pcPic->slices[s]->setTileGroupCcAlfCrApsId(cs.slice->getTileGroupCcAlfCrApsId());
          pcPic->slices[s]->m_ccAlfFilterParam      = m_pcALF->getCcAlfFilterParam();
          pcPic->slices[s]->m_ccAlfFilterControl[0] = m_pcALF->getCcAlfControlIdc(COMPONENT_Cb);
          pcPic->slices[s]->m_ccAlfFilterControl[1] = m_pcALF->getCcAlfControlIdc(COMPONENT_Cr);
        }
      }
      DTRACE_UPDATE( g_trace_ctx, ( std::make_pair( "final", 1 ) ) );
      if (m_pcCfg->getUseCompositeRef() && getPrepareLTRef())
      {
        updateCompositeReference(pcSlice, rcListPic, pocCurr);
      }
    }
    else // skip enc picture
    {
      pcSlice->setSliceQpBase( pcSlice->getSliceQp() );

#if ENABLE_QPA
      if (m_pcCfg->getUsePerceptQPA() && !m_pcCfg->getUseRateCtrl() && pcSlice->getPPS()->getUseDQP())
      {
        const double picLambda = pcSlice->getLambdas()[0];

        for (uint32_t ctuRsAddr = 0; ctuRsAddr < numberOfCtusInFrame; ctuRsAddr++)
        {
          pcPic->m_uEnerHpCtu[ctuRsAddr] = picLambda;  // initialize to slice lambda (just for safety)
        }
      }
#endif
      if( pcSlice->getSPS()->getSAOEnabledFlag() )
      {
        m_pcSAO->disabledRate( *pcPic->cs, pcPic->getSAO(1), m_pcCfg->getSaoEncodingRate(), m_pcCfg->getSaoEncodingRateChroma());
      }
      if (pcSlice->getSPS()->getALFEnabledFlag() && (pcSlice->getTileGroupAlfEnabledFlag(COMPONENT_Y) || pcSlice->getTileGroupCcAlfCbEnabledFlag() || pcSlice->getTileGroupCcAlfCrEnabledFlag()))
      {
        // IRAP AU: reset APS map
        {
          int layerIdx = pcSlice->getVPS() == nullptr ? 0 : pcSlice->getVPS()->getGeneralLayerIdx( pcSlice->getPic()->layerId );
          if( !layerIdx && ( pcSlice->getPendingRasInit() || pcSlice->isIDRorBLA() ) )
          {
            // We have to reset all APS on IRAP, but in not encoding case we have to keep the parsed APS of current slice
            // Get active ALF APSs from picture/slice header
            const std::vector<int> sliceApsIdsLuma = pcSlice->getTileGroupApsIdLuma();

            m_pcALF->setApsIdStart( ALF_CTB_MAX_NUM_APS );

            ParameterSetMap<APS>* apsMap = m_pcEncLib->getApsMap();
            apsMap->clear();

            for( int apsId = 0; apsId < ALF_CTB_MAX_NUM_APS; apsId++ )
            {
              int psId = ( apsId << NUM_APS_TYPE_LEN ) + ALF_APS;
              APS* aps = apsMap->getPS( psId );
              if( aps )
              {
                // Check if this APS is currently the active one (used in current slice)
                bool activeAps = false;
                bool activeApsCcAlf = false;
                // Luma
                for( int i = 0; i < sliceApsIdsLuma.size(); i++ )
                {
                  if( aps->getAPSId() == sliceApsIdsLuma[i] )
                  {
                    activeAps = true;
                    break;
                  }
                }
                // Chroma
                activeAps |= aps->getAPSId() == pcSlice->getTileGroupApsIdChroma();
                // CC-ALF
                activeApsCcAlf |= pcSlice->getTileGroupCcAlfCbEnabledFlag() && aps->getAPSId() == pcSlice->getTileGroupCcAlfCbApsId();
                activeApsCcAlf |= pcSlice->getTileGroupCcAlfCrEnabledFlag() && aps->getAPSId() == pcSlice->getTileGroupCcAlfCrApsId();
                if( !activeAps && !activeApsCcAlf )
                {
                  apsMap->clearChangedFlag( psId );
                }
                if( !activeAps  )
                {
                  aps->getAlfAPSParam().reset();
                }
                if( !activeApsCcAlf )
                {
                  aps->getCcAlfAPSParam().reset();
                }
              }
            }
          }
        }

        // Assign tne correct APS to slice and emulate the setting of ALF start APS ID
        int changedApsId = -1;
        for( int apsId = ALF_CTB_MAX_NUM_APS - 1; apsId >= 0; apsId-- )
        {
          ParameterSetMap<APS>* apsMap = m_pcEncLib->getApsMap();
          int psId = ( apsId << NUM_APS_TYPE_LEN ) + ALF_APS;
          APS* aps = apsMap->getPS( psId );
          if( aps )
          {
            // In slice, replace the old APS (from decoder map) with the APS from encoder map due to later checks while bitstream writing
            if( pcSlice->getAlfAPSs() && pcSlice->getAlfAPSs()[apsId] )
            {
              pcSlice->getAlfAPSs()[apsId] = aps;
            }
            if( apsMap->getChangedFlag( psId ) )
              changedApsId = apsId;
          }
        }
        if( changedApsId >= 0 )
          m_pcALF->setApsIdStart( changedApsId );
      }
    }

    pcSlice->freeScaledRefPicList( scaledRefPic );

    if( m_pcCfg->getUseAMaxBT() )
    {
      for( const CodingUnit *cu : pcPic->cs->cus )
      {
        if( !pcSlice->isIntra() )
        {
          m_uiBlkSize[pcSlice->getDepth()] += cu->Y().area();
          m_uiNumBlk [pcSlice->getDepth()]++;
        }
      }
    }

    if( encPic || decPic )
    {
      pcSlice = pcPic->slices[0];

      /////////////////////////////////////////////////////////////////////////////////////////////////// File writing

      // write various parameter sets
      bool writePS = m_bSeqFirst || (m_pcCfg->getReWriteParamSets() && (pcSlice->isIRAP()));
      if (writePS)
      {
        m_pcEncLib->setParamSetChanged(pcSlice->getSPS()->getSPSId(), pcSlice->getPPS()->getPPSId());
      }

      int layerIdx = m_pcEncLib->getVPS() == nullptr ? 0 : m_pcEncLib->getVPS()->getGeneralLayerIdx( m_pcEncLib->getLayerId() );

      // it is assumed that layerIdx equal to 0 is always present
      m_audIrapOrGdrAuFlag = pcSlice->getPicHeader()->getGdrPicFlag() || (pcSlice->isIRAP() && !pcSlice->getPPS()->getMixedNaluTypesInPicFlag());
      if ((( m_pcEncLib->getVPS()->getMaxLayers() > 1 && m_audIrapOrGdrAuFlag) || m_pcCfg->getAccessUnitDelimiter()) && !layerIdx )
      {
        xWriteAccessUnitDelimiter(accessUnit, pcSlice);
      }

      // it is assumed that layerIdx equal to 0 is always present
      actualTotalBits += xWriteParameterSets(accessUnit, pcSlice, writePS, layerIdx);
      if (writePS)
      {
        // create prefix SEI messages at the beginning of the sequence
        CHECK(!(leadingSeiMessages.empty()), "Unspecified error");
        xCreateIRAPLeadingSEIMessages(leadingSeiMessages, pcSlice->getSPS(), pcSlice->getPPS());

        m_bSeqFirst = false;
      }


      //send LMCS APS when LMCSModel is updated. It can be updated even current slice does not enable reshaper.
      //For example, in RA, update is on intra slice, but intra slice may not use reshaper
      if (pcSlice->getSPS()->getUseLmcs())
      {
        //only 1 LMCS data for 1 picture
        int apsId = picHeader->getLmcsAPSId();
        ParameterSetMap<APS> *apsMap = m_pcEncLib->getApsMap();
        APS* aps = apsMap->getPS((apsId << NUM_APS_TYPE_LEN) + LMCS_APS);
        bool writeAPS = aps && apsMap->getChangedFlag((apsId << NUM_APS_TYPE_LEN) + LMCS_APS);
        if (writeAPS)
        {
#if JVET_R0433
          aps->chromaPresentFlag = pcSlice->getSPS()->getChromaFormatIdc() != CHROMA_400;
#endif
          actualTotalBits += xWriteAPS( accessUnit, aps, m_pcEncLib->getLayerId(), true );
          apsMap->clearChangedFlag((apsId << NUM_APS_TYPE_LEN) + LMCS_APS);
          CHECK(aps != picHeader->getLmcsAPS(), "Wrong LMCS APS pointer in compressGOP");
        }
      }

      // only 1 SCALING LIST data for 1 picture
      if( pcSlice->getSPS()->getScalingListFlag() && ( m_pcCfg->getUseScalingListId() == SCALING_LIST_FILE_READ ) )
      {
        int apsId = picHeader->getScalingListAPSId();
        ParameterSetMap<APS> *apsMap = m_pcEncLib->getApsMap();
        APS* aps = apsMap->getPS( ( apsId << NUM_APS_TYPE_LEN ) + SCALING_LIST_APS );
        bool writeAPS = aps && apsMap->getChangedFlag( ( apsId << NUM_APS_TYPE_LEN ) + SCALING_LIST_APS );
        if( writeAPS )
        {
#if JVET_R0433
          aps->chromaPresentFlag = pcSlice->getSPS()->getChromaFormatIdc() != CHROMA_400;
#endif
          actualTotalBits += xWriteAPS( accessUnit, aps, m_pcEncLib->getLayerId(), true );
          apsMap->clearChangedFlag( ( apsId << NUM_APS_TYPE_LEN ) + SCALING_LIST_APS );
          CHECK( aps != picHeader->getScalingListAPS(), "Wrong SCALING LIST APS pointer in compressGOP" );
        }
      }

      if (pcSlice->getSPS()->getALFEnabledFlag() && (pcSlice->getTileGroupAlfEnabledFlag(COMPONENT_Y) || pcSlice->getTileGroupCcAlfCbEnabledFlag() || pcSlice->getTileGroupCcAlfCrEnabledFlag()))
      {
        for (int apsId = 0; apsId < ALF_CTB_MAX_NUM_APS; apsId++)
        {
          ParameterSetMap<APS> *apsMap = m_pcEncLib->getApsMap();

          APS* aps = apsMap->getPS((apsId << NUM_APS_TYPE_LEN) + ALF_APS);
          bool writeAPS = aps && apsMap->getChangedFlag((apsId << NUM_APS_TYPE_LEN) + ALF_APS);
          if (!aps && pcSlice->getAlfAPSs() && pcSlice->getAlfAPSs()[apsId])
          {
            writeAPS = true;
            aps = pcSlice->getAlfAPSs()[apsId]; // use asp from slice header
            *apsMap->allocatePS((apsId << NUM_APS_TYPE_LEN) + ALF_APS) = *aps; //allocate and cpy
            m_pcALF->setApsIdStart( apsId );
          }
          else if (pcSlice->getTileGroupCcAlfCbEnabledFlag() && !aps && apsId == pcSlice->getTileGroupCcAlfCbApsId())
          {
            writeAPS = true;
            aps = apsMap->getPS((pcSlice->getTileGroupCcAlfCbApsId() << NUM_APS_TYPE_LEN) + ALF_APS);
          }
          else if (pcSlice->getTileGroupCcAlfCrEnabledFlag() && !aps && apsId == pcSlice->getTileGroupCcAlfCrApsId())
          {
            writeAPS = true;
            aps = apsMap->getPS((pcSlice->getTileGroupCcAlfCrApsId() << NUM_APS_TYPE_LEN) + ALF_APS);
          }

          if (writeAPS )
          {
#if JVET_R0433
            aps->chromaPresentFlag = pcSlice->getSPS()->getChromaFormatIdc() != CHROMA_400;
#endif
            actualTotalBits += xWriteAPS( accessUnit, aps, m_pcEncLib->getLayerId(), true );
            apsMap->clearChangedFlag((apsId << NUM_APS_TYPE_LEN) + ALF_APS);
            CHECK(aps != pcSlice->getAlfAPSs()[apsId] && apsId != pcSlice->getTileGroupCcAlfCbApsId() && apsId != pcSlice->getTileGroupCcAlfCrApsId(), "Wrong APS pointer in compressGOP");
          }
        }
      }

      // reset presence of BP SEI indication
      m_bufferingPeriodSEIPresentInAU = false;
      // create prefix SEI associated with a picture
      xCreatePerPictureSEIMessages(iGOPid, leadingSeiMessages, nestedSeiMessages, pcSlice);

      // pcSlice is currently slice 0.
      std::size_t binCountsInNalUnits   = 0; // For implementation of cabac_zero_word stuffing (section 7.4.3.10)
      std::size_t numBytesInVclNalUnits = 0; // For implementation of cabac_zero_word stuffing (section 7.4.3.10)
#if JVET_Q0406_CABAC_ZERO
      std::size_t sumZeroWords          = 0; // sum of cabac_zero_word inserted per sub-picture
      std::vector<EncBitstreamParams> subPicStats (pcPic->cs->pps->getNumSubPics());
#endif

      for(uint32_t sliceSegmentIdxCount = 0; sliceSegmentIdxCount < pcPic->cs->pps->getNumSlicesInPic(); sliceSegmentIdxCount++ )
      {
        pcSlice = pcPic->slices[sliceSegmentIdxCount];
        if(sliceSegmentIdxCount > 0 && pcSlice->getSliceType()!= I_SLICE)
        {
          pcSlice->checkColRefIdx(sliceSegmentIdxCount, pcPic);
        }
        m_pcSliceEncoder->setSliceSegmentIdx(sliceSegmentIdxCount);

        pcSlice->setRPL0(pcPic->slices[0]->getRPL0());
        pcSlice->setRPL1(pcPic->slices[0]->getRPL1());
        pcSlice->setRPL0idx(pcPic->slices[0]->getRPL0idx());
        pcSlice->setRPL1idx(pcPic->slices[0]->getRPL1idx());

        picHeader->setNoOutputBeforeRecoveryFlag( false );
        if (pcSlice->isIRAP())
        {
          if (pcSlice->getNalUnitType() >= NAL_UNIT_CODED_SLICE_IDR_W_RADL && pcSlice->getNalUnitType() <= NAL_UNIT_CODED_SLICE_IDR_N_LP)
          {
            picHeader->setNoOutputBeforeRecoveryFlag( true );
          }
          //the inference for NoOutputPriorPicsFlag
          // KJS: This cannot happen at the encoder
          if( !m_bFirst && ( pcSlice->isIRAP() || pcSlice->getNalUnitType() >= NAL_UNIT_CODED_SLICE_GDR ) && picHeader->getNoOutputBeforeRecoveryFlag() )
          {
            if (pcSlice->getNalUnitType() == NAL_UNIT_CODED_SLICE_CRA || pcSlice->getNalUnitType() >= NAL_UNIT_CODED_SLICE_GDR)
            {
#if JVET_S0193_NO_OUTPUT_PRIOR_PIC
              pcSlice->setNoOutputOfPriorPicsFlag(true);
#else
              picHeader->setNoOutputOfPriorPicsFlag(true);
#endif
            }
          }
        }

        // code picture header before first slice
        if(sliceSegmentIdxCount == 0)
        {
          // code RPL in picture header or slice headers
          if( !m_pcCfg->getSliceLevelRpl() && (!pcSlice->getIdrPicFlag() || pcSlice->getSPS()->getIDRRefParamListPresent()) )
          {
            picHeader->setRPL0idx(pcSlice->getRPL0idx());
            picHeader->setRPL1idx(pcSlice->getRPL1idx());
            picHeader->setRPL0(pcSlice->getRPL0());
            picHeader->setRPL1(pcSlice->getRPL1());
            *picHeader->getLocalRPL0() = *pcSlice->getLocalRPL0();
            *picHeader->getLocalRPL1() = *pcSlice->getLocalRPL1();
          }

          // code DBLK in picture header or slice headers
          if( !m_pcCfg->getSliceLevelDblk() )
          {
            picHeader->setDeblockingFilterOverrideFlag   ( pcSlice->getDeblockingFilterOverrideFlag()   );
            picHeader->setDeblockingFilterDisable        ( pcSlice->getDeblockingFilterDisable()        );
            picHeader->setDeblockingFilterBetaOffsetDiv2 ( pcSlice->getDeblockingFilterBetaOffsetDiv2() );
            picHeader->setDeblockingFilterTcOffsetDiv2   ( pcSlice->getDeblockingFilterTcOffsetDiv2()   );
            picHeader->setDeblockingFilterCbBetaOffsetDiv2( pcSlice->getDeblockingFilterCbBetaOffsetDiv2() );
            picHeader->setDeblockingFilterCbTcOffsetDiv2  ( pcSlice->getDeblockingFilterCbTcOffsetDiv2() );
            picHeader->setDeblockingFilterCrBetaOffsetDiv2( pcSlice->getDeblockingFilterCrBetaOffsetDiv2() );
            picHeader->setDeblockingFilterCrTcOffsetDiv2  ( pcSlice->getDeblockingFilterCrTcOffsetDiv2() );
          }

          if (!m_pcCfg->getSliceLevelDeltaQp())
          {
            picHeader->setQpDelta(pcSlice->getSliceQp() - (pcSlice->getPPS()->getPicInitQPMinus26() + 26));
          }

          // code SAO parameters in picture header or slice headers
          if( !m_pcCfg->getSliceLevelSao() )
          {
            picHeader->setSaoEnabledFlag(CHANNEL_TYPE_LUMA,   pcSlice->getSaoEnabledFlag(CHANNEL_TYPE_LUMA  ));
            picHeader->setSaoEnabledFlag(CHANNEL_TYPE_CHROMA, pcSlice->getSaoEnabledFlag(CHANNEL_TYPE_CHROMA));
          }

          // code ALF parameters in picture header or slice headers
          if( !m_pcCfg->getSliceLevelAlf() )
          {
            picHeader->setAlfEnabledFlag(COMPONENT_Y,  pcSlice->getTileGroupAlfEnabledFlag(COMPONENT_Y ) );
            picHeader->setAlfEnabledFlag(COMPONENT_Cb, pcSlice->getTileGroupAlfEnabledFlag(COMPONENT_Cb) );
            picHeader->setAlfEnabledFlag(COMPONENT_Cr, pcSlice->getTileGroupAlfEnabledFlag(COMPONENT_Cr) );
            picHeader->setNumAlfAps(pcSlice->getTileGroupNumAps());
            picHeader->setAlfAPSs(pcSlice->getTileGroupApsIdLuma());
            picHeader->setAlfApsIdChroma(pcSlice->getTileGroupApsIdChroma());
            picHeader->setCcAlfEnabledFlag(COMPONENT_Cb, pcSlice->getTileGroupCcAlfCbEnabledFlag());
            picHeader->setCcAlfEnabledFlag(COMPONENT_Cr, pcSlice->getTileGroupCcAlfCrEnabledFlag());
            picHeader->setCcAlfCbApsId(pcSlice->getTileGroupCcAlfCbApsId());
            picHeader->setCcAlfCrApsId(pcSlice->getTileGroupCcAlfCrApsId());
          }

          // code WP parameters in picture header or slice headers
          if (!m_pcCfg->getSliceLevelWp())
          {
            picHeader->setWpScaling(pcSlice->getWpScalingAll());
            picHeader->setNumL0Weights(pcSlice->getNumRefIdx(REF_PIC_LIST_0));
            picHeader->setNumL0Weights(pcSlice->getNumRefIdx(REF_PIC_LIST_1));
          }

          pcPic->cs->picHeader->setPic(pcPic);
          pcPic->cs->picHeader->setValid();
          if (pcPic->cs->pps->getNumSlicesInPic() > 1 || !m_pcCfg->getEnablePictureHeaderInSliceHeader())
          {
            pcSlice->setPictureHeaderInSliceHeader(false);
            actualTotalBits += xWritePicHeader(accessUnit, pcPic->cs->picHeader);
          }
          else
          {
            pcSlice->setPictureHeaderInSliceHeader(true);
          }
          if (pcSlice->getSPS()->getProfileTierLevel()->getConstraintInfo()->getPicHeaderInSliceHeaderConstraintFlag())
          {
            CHECK(pcSlice->getPictureHeaderInSliceHeader() == false, "PH shall be present in SH, when pic_header_in_slice_header_constraint_flag is equal to 1");
          }
        }
        pcSlice->setPicHeader( pcPic->cs->picHeader );
        pcSlice->setNalUnitLayerId( m_pcEncLib->getLayerId() );

        for ( uint32_t ui = 0 ; ui < numSubstreams; ui++ )
        {
          substreamsOut[ui].clear();
        }

        /* start slice NALunit */
        OutputNALUnit nalu( pcSlice->getNalUnitType(), m_pcEncLib->getLayerId(), pcSlice->getTLayer() );
        m_HLSWriter->setBitstream( &nalu.m_Bitstream );


        tmpBitsBeforeWriting = m_HLSWriter->getNumberOfWrittenBits();
        m_HLSWriter->codeSliceHeader( pcSlice );
        actualHeadBits += ( m_HLSWriter->getNumberOfWrittenBits() - tmpBitsBeforeWriting );

        pcSlice->setFinalized(true);

        pcSlice->resetNumberOfSubstream( );
        pcSlice->setNumSubstream( pcSlice->getSPS(), pcSlice->getPPS() );
        pcSlice->clearSubstreamSizes(  );
#if JVET_Q0406_CABAC_ZERO
        const int subpicIdx = pcPic->cs->pps->getSubPicIdxFromSubPicId(pcSlice->getSliceSubPicId());
#endif
        {
          uint32_t numBinsCoded = 0;
          m_pcSliceEncoder->encodeSlice(pcPic, &(substreamsOut[0]), numBinsCoded);
          binCountsInNalUnits+=numBinsCoded;
#if JVET_Q0406_CABAC_ZERO
          subPicStats[subpicIdx].numBinsWritten += numBinsCoded;
#endif
        }
        {
          // Construct the final bitstream by concatenating substreams.
          // The final bitstream is either nalu.m_Bitstream or pcBitstreamRedirect;
          // Complete the slice header info.
          m_HLSWriter->setBitstream( &nalu.m_Bitstream );
          m_HLSWriter->codeTilesWPPEntryPoint( pcSlice );

          // Append substreams...
          OutputBitstream *pcOut = pcBitstreamRedirect;
          const int numSubstreamsToCode = pcSlice->getNumberOfSubstream() + 1;

          for ( uint32_t ui = 0 ; ui < numSubstreamsToCode; ui++ )
          {
            pcOut->addSubstream(&(substreamsOut[ui]));
          }
        }

        // If current NALU is the first NALU of slice (containing slice header) and more NALUs exist (due to multiple dependent slices) then buffer it.
        // If current NALU is the last NALU of slice and a NALU was buffered, then (a) Write current NALU (b) Update an write buffered NALU at approproate location in NALU list.
        bool bNALUAlignedWrittenToList    = false; // used to ensure current NALU is not written more than once to the NALU list.
        xAttachSliceDataToNalUnit(nalu, pcBitstreamRedirect);
        accessUnit.push_back(new NALUnitEBSP(nalu));
        actualTotalBits += uint32_t(accessUnit.back()->m_nalUnitData.str().size()) * 8;
        numBytesInVclNalUnits += (std::size_t)(accessUnit.back()->m_nalUnitData.str().size());
#if JVET_Q0406_CABAC_ZERO
        subPicStats[subpicIdx].numBytesInVclNalUnits += (std::size_t)(accessUnit.back()->m_nalUnitData.str().size());
#endif
        bNALUAlignedWrittenToList = true;

        if (!bNALUAlignedWrittenToList)
        {
          nalu.m_Bitstream.writeAlignZero();
          accessUnit.push_back(new NALUnitEBSP(nalu));
        }

        if( ( m_pcCfg->getPictureTimingSEIEnabled() || m_pcCfg->getDecodingUnitInfoSEIEnabled() ) &&
        ((pcSlice->getSPS()->getGeneralHrdParameters()->getGeneralNalHrdParametersPresentFlag())
          || (pcSlice->getSPS()->getGeneralHrdParameters()->getGeneralVclHrdParametersPresentFlag())) &&
          (pcSlice->getSPS()->getGeneralHrdParameters()->getGeneralDecodingUnitHrdParamsPresentFlag()))
        {
          uint32_t numNalus = 0;
          uint32_t numRBSPBytes = 0;
          for (AccessUnit::const_iterator it = accessUnit.begin(); it != accessUnit.end(); it++)
          {
            numRBSPBytes += uint32_t((*it)->m_nalUnitData.str().size());
            numNalus ++;
          }
          duData.push_back(DUData());
          duData.back().accumBitsDU = ( numRBSPBytes << 3 );
          duData.back().accumNalsDU = numNalus;
        }
#if JVET_Q0406_CABAC_ZERO
        if (pcSlice->isLastSliceInSubpic())
        {
          // Check picture level encoding constraints/requirements
          ProfileLevelTierFeatures profileLevelTierFeatures;
          profileLevelTierFeatures.extractPTLInformation(*(pcSlice->getSPS()));
          sumZeroWords += cabac_zero_word_padding(pcSlice, pcPic, subPicStats[subpicIdx].numBinsWritten, subPicStats[subpicIdx].numBytesInVclNalUnits, 0,
                                                  accessUnit.back()->m_nalUnitData, m_pcCfg->getCabacZeroWordPaddingEnabled(), profileLevelTierFeatures);
        }
#endif
      } // end iteration over slices

      {
        // Check picture level encoding constraints/requirements
        ProfileLevelTierFeatures profileLevelTierFeatures;
        profileLevelTierFeatures.extractPTLInformation(*(pcSlice->getSPS()));
        validateMinCrRequirements(profileLevelTierFeatures, numBytesInVclNalUnits, pcPic, m_pcCfg);
        // cabac_zero_words processing
#if JVET_Q0406_CABAC_ZERO
        cabac_zero_word_padding(pcSlice, pcPic, binCountsInNalUnits, numBytesInVclNalUnits, sumZeroWords, accessUnit.back()->m_nalUnitData, m_pcCfg->getCabacZeroWordPaddingEnabled(), profileLevelTierFeatures);
#else
        cabac_zero_word_padding(pcSlice, pcPic, binCountsInNalUnits, numBytesInVclNalUnits, accessUnit.back()->m_nalUnitData, m_pcCfg->getCabacZeroWordPaddingEnabled(), profileLevelTierFeatures);
#endif
      }

      //-- For time output for each slice
      auto elapsed = std::chrono::steady_clock::now() - beforeTime;
      auto encTime = std::chrono::duration_cast<std::chrono::seconds>( elapsed ).count();

      std::string digestStr;
      if (m_pcCfg->getDecodedPictureHashSEIType()!=HASHTYPE_NONE)
      {
        SEIDecodedPictureHash *decodedPictureHashSei = new SEIDecodedPictureHash();
        PelUnitBuf recoBuf = pcPic->cs->getRecoBuf();
        m_seiEncoder.initDecodedPictureHashSEI(decodedPictureHashSei, recoBuf, digestStr, pcSlice->getSPS()->getBitDepths());
        trailingSeiMessages.push_back(decodedPictureHashSei);
      }

      m_pcCfg->setEncodedFlag(iGOPid, true);

      double PSNR_Y;
      xCalculateAddPSNRs(isField, isTff, iGOPid, pcPic, accessUnit, rcListPic, encTime, snr_conversion, printFrameMSE, &PSNR_Y, isEncodeLtRef );


      xWriteTrailingSEIMessages(trailingSeiMessages, accessUnit, pcSlice->getTLayer());

      printHash(m_pcCfg->getDecodedPictureHashSEIType(), digestStr);

      if ( m_pcCfg->getUseRateCtrl() )
      {
        double avgQP     = m_pcRateCtrl->getRCPic()->calAverageQP();
        double avgLambda = m_pcRateCtrl->getRCPic()->calAverageLambda();
        if ( avgLambda < 0.0 )
        {
          avgLambda = lambda;
        }

        m_pcRateCtrl->getRCPic()->updateAfterPicture( actualHeadBits, actualTotalBits, avgQP, avgLambda, pcSlice->isIRAP());
        m_pcRateCtrl->getRCPic()->addToPictureLsit( m_pcRateCtrl->getPicList() );

        m_pcRateCtrl->getRCSeq()->updateAfterPic( actualTotalBits );
        if ( !pcSlice->isIRAP() )
        {
          m_pcRateCtrl->getRCGOP()->updateAfterPicture( actualTotalBits );
        }
        else    // for intra picture, the estimated bits are used to update the current status in the GOP
        {
          m_pcRateCtrl->getRCGOP()->updateAfterPicture( estimatedBits );
        }
  #if U0132_TARGET_BITS_SATURATION
        if (m_pcRateCtrl->getCpbSaturationEnabled())
        {
          m_pcRateCtrl->updateCpbState(actualTotalBits);
          msg( NOTICE, " [CPB %6d bits]", m_pcRateCtrl->getCpbState() );
        }
  #endif
      }
      xCreateFrameFieldInfoSEI( leadingSeiMessages, pcSlice, isField );
      xCreatePictureTimingSEI( m_pcCfg->getEfficientFieldIRAPEnabled() ? effFieldIRAPMap.GetIRAPGOPid() : 0, leadingSeiMessages, nestedSeiMessages, duInfoSeiMessages, pcSlice, isField, duData );

      if (m_pcCfg->getScalableNestingSEIEnabled())
      {
        const SPS* sps = pcSlice->getSPS();
        const PPS* pps = pcSlice->getPPS();

        std::vector<uint16_t> subpicIDs;
        if (sps->getSubPicInfoPresentFlag())
        {
          if(sps->getSubPicIdMappingExplicitlySignalledFlag())
          {
            if(sps->getSubPicIdMappingInSpsFlag())
            {
              subpicIDs = sps->getSubPicIds();
            }
            else
            {
              subpicIDs = pps->getSubPicIds();
            }
          }
          else
          {
            const int numSubPics = sps->getNumSubPics();
            subpicIDs.resize(numSubPics);
            for (int i = 0 ; i < numSubPics; i++)
            {
              subpicIDs[i] = (uint16_t) i;
            }
          }
        }
        xCreateScalableNestingSEI(leadingSeiMessages, nestedSeiMessages, subpicIDs);
      }

      xWriteLeadingSEIMessages( leadingSeiMessages, duInfoSeiMessages, accessUnit, pcSlice->getTLayer(), pcSlice->getSPS(), duData );
      xWriteDuSEIMessages( duInfoSeiMessages, accessUnit, pcSlice->getTLayer(), duData );

      m_AUWriterIf->outputAU( accessUnit );

      msg( NOTICE, "\n" );
      fflush( stdout );
    }


    DTRACE_UPDATE( g_trace_ctx, ( std::make_pair( "final", 0 ) ) );

    pcPic->reconstructed = true;
    m_bFirst = false;
    m_iNumPicCoded++;
    if (!(m_pcCfg->getUseCompositeRef() && isEncodeLtRef))
    {
      for( int i = pcSlice->getTLayer() ; i < pcSlice->getSPS()->getMaxTLayers() ; i ++ )
      {
        m_totalCoded[i]++;
      }
    }
    /* logging: insert a newline at end of picture period */

    if (m_pcCfg->getEfficientFieldIRAPEnabled())
    {
      iGOPid=effFieldIRAPMap.restoreGOPid(iGOPid);
    }

    pcPic->destroyTempBuffers();
    pcPic->cs->destroyCoeffs();
    pcPic->cs->releaseIntermediateData();
  } // iGOPid-loop

  delete pcBitstreamRedirect;

  CHECK( m_iNumPicCoded > 1, "Unspecified error" );
}

void EncGOP::printOutSummary( uint32_t uiNumAllPicCoded, bool isField, const bool printMSEBasedSNR, const bool printSequenceMSE, const bool printHexPsnr, const bool printRprPSNR, const BitDepths &bitDepths )
{
#if ENABLE_QPA
  const bool    useWPSNR = m_pcEncLib->getUseWPSNR();
#endif
#if WCG_WPSNR
  const bool    useLumaWPSNR = m_pcEncLib->getLumaLevelToDeltaQPMapping().isEnabled() || (m_pcCfg->getLmcs() && m_pcCfg->getReshapeSignalType() == RESHAPE_SIGNAL_PQ);
#endif

  if( m_pcCfg->getDecodeBitstream(0).empty() && m_pcCfg->getDecodeBitstream(1).empty() && !m_pcCfg->useFastForwardToPOC() )
  {
    CHECK( !( uiNumAllPicCoded == m_gcAnalyzeAll.getNumPic() ), "Unspecified error" );
  }

  //--CFG_KDY
  const int rateMultiplier=(isField?2:1);
  m_gcAnalyzeAll.setFrmRate( m_pcCfg->getFrameRate()*rateMultiplier / (double)m_pcCfg->getTemporalSubsampleRatio());
  m_gcAnalyzeI.setFrmRate( m_pcCfg->getFrameRate()*rateMultiplier / (double)m_pcCfg->getTemporalSubsampleRatio());
  m_gcAnalyzeP.setFrmRate( m_pcCfg->getFrameRate()*rateMultiplier / (double)m_pcCfg->getTemporalSubsampleRatio());
  m_gcAnalyzeB.setFrmRate( m_pcCfg->getFrameRate()*rateMultiplier / (double)m_pcCfg->getTemporalSubsampleRatio());
#if WCG_WPSNR
  if (useLumaWPSNR)
  {
    m_gcAnalyzeWPSNR.setFrmRate(m_pcCfg->getFrameRate()*rateMultiplier / (double)m_pcCfg->getTemporalSubsampleRatio());
  }
#endif

  const ChromaFormat chFmt = m_pcCfg->getChromaFormatIdc();

  //-- all
  msg( INFO, "\n" );
  msg( DETAILS,"\nSUMMARY --------------------------------------------------------\n" );
#if JVET_O0756_CALCULATE_HDRMETRICS
  const bool calculateHdrMetrics = m_pcEncLib->getCalcluateHdrMetrics();
#endif
#if ENABLE_QPA
  m_gcAnalyzeAll.printOut( 'a', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, printRprPSNR, bitDepths, useWPSNR
#if JVET_O0756_CALCULATE_HDRMETRICS
                          , calculateHdrMetrics
#endif
                          );
#else
  m_gcAnalyzeAll.printOut('a', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, bitDepths
#if JVET_O0756_CALCULATE_HDRMETRICS
                          , calculateHdrMetrics
#endif
                          );
#endif
  msg( DETAILS, "\n\nI Slices--------------------------------------------------------\n" );
  m_gcAnalyzeI.printOut( 'i', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, printRprPSNR, bitDepths );

  msg( DETAILS, "\n\nP Slices--------------------------------------------------------\n" );
  m_gcAnalyzeP.printOut( 'p', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, printRprPSNR, bitDepths );

  msg( DETAILS, "\n\nB Slices--------------------------------------------------------\n" );
  m_gcAnalyzeB.printOut( 'b', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, printRprPSNR, bitDepths );

#if WCG_WPSNR
  if (useLumaWPSNR)
  {
    msg(DETAILS, "\nWPSNR SUMMARY --------------------------------------------------------\n");
    m_gcAnalyzeWPSNR.printOut( 'w', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, printRprPSNR, bitDepths, useLumaWPSNR );
  }
#endif
  if (!m_pcCfg->getSummaryOutFilename().empty())
  {
    m_gcAnalyzeAll.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryOutFilename());
  }

  if (!m_pcCfg->getSummaryPicFilenameBase().empty())
  {
    m_gcAnalyzeI.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryPicFilenameBase()+"I.txt");
    m_gcAnalyzeP.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryPicFilenameBase()+"P.txt");
    m_gcAnalyzeB.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryPicFilenameBase()+"B.txt");
  }

#if WCG_WPSNR
  if (!m_pcCfg->getSummaryOutFilename().empty() && useLumaWPSNR)
  {
    m_gcAnalyzeWPSNR.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryOutFilename());
  }
#endif
  if(isField)
  {
    //-- interlaced summary
    m_gcAnalyzeAll_in.setFrmRate( m_pcCfg->getFrameRate() / (double)m_pcCfg->getTemporalSubsampleRatio());
    m_gcAnalyzeAll_in.setBits(m_gcAnalyzeAll.getBits());
    // prior to the above statement, the interlace analyser does not contain the correct total number of bits.

    msg( INFO,"\n\nSUMMARY INTERLACED ---------------------------------------------\n" );
#if ENABLE_QPA
    m_gcAnalyzeAll_in.printOut( 'a', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, printRprPSNR, bitDepths, useWPSNR );
#else
    m_gcAnalyzeAll_in.printOut('a', chFmt, printMSEBasedSNR, printSequenceMSE, printHexPsnr, bitDepths);
#endif
    if (!m_pcCfg->getSummaryOutFilename().empty())
    {
      m_gcAnalyzeAll_in.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryOutFilename());
#if WCG_WPSNR
      if (useLumaWPSNR)
      {
        m_gcAnalyzeWPSNR.printSummary(chFmt, printSequenceMSE, printHexPsnr, bitDepths, m_pcCfg->getSummaryOutFilename());
      }
#endif
    }
  }

  msg( DETAILS,"\nRVM: %.3lf\n", xCalculateRVM() );
}

#if W0038_DB_OPT
uint64_t EncGOP::preLoopFilterPicAndCalcDist( Picture* pcPic )
{
  CodingStructure& cs = *pcPic->cs;
  m_pcLoopFilter->loopFilterPic( cs );

  const CPelUnitBuf picOrg = pcPic->getRecoBuf();
  const CPelUnitBuf picRec = cs.getRecoBuf();

  uint64_t uiDist = 0;
  for( uint32_t comp = 0; comp < (uint32_t)picRec.bufs.size(); comp++)
  {
    const ComponentID compID = ComponentID(comp);
    const uint32_t rshift = 2 * DISTORTION_PRECISION_ADJUSTMENT(cs.sps->getBitDepth(toChannelType(compID)));
#if ENABLE_QPA
    CHECK( rshift >= 8, "shifts greater than 7 are not supported." );
#endif
    uiDist += xFindDistortionPlane( picOrg.get(compID), picRec.get(compID), rshift );
  }
  return uiDist;
}
#endif

// ====================================================================================================================
// Protected member functions
// ====================================================================================================================
void EncGOP::xInitGOP( int iPOCLast, int iNumPicRcvd, bool isField
  , bool isEncodeLtRef
)
{
  CHECK(!( iNumPicRcvd > 0 ), "Unspecified error");
  //  Exception for the first frames
  if ((isField && (iPOCLast == 0 || iPOCLast == 1)) || (!isField && (iPOCLast == 0)) || isEncodeLtRef)
  {
    m_iGopSize    = 1;
  }
  else
  {
    m_iGopSize    = m_pcCfg->getGOPSize();
  }
  CHECK(!(m_iGopSize > 0), "Unspecified error");

  return;
}


void EncGOP::xGetBuffer( PicList&                  rcListPic,
                         std::list<PelUnitBuf*>&   rcListPicYuvRecOut,
                         int                       iNumPicRcvd,
                         int                       iTimeOffset,
                         Picture*&                 rpcPic,
                         int                       pocCurr,
                         bool                      isField )
{
  int i;
  //  Rec. output
  std::list<PelUnitBuf*>::iterator     iterPicYuvRec = rcListPicYuvRecOut.end();

  if (isField && pocCurr > 1 && m_iGopSize!=1)
  {
    iTimeOffset--;
  }

  int multipleFactor = m_pcCfg->getUseCompositeRef() ? 2 : 1;
  for (i = 0; i < (iNumPicRcvd * multipleFactor - iTimeOffset + 1); i += multipleFactor)
  {
    iterPicYuvRec--;
  }

  //  Current pic.
  PicList::iterator        iterPic       = rcListPic.begin();
  while (iterPic != rcListPic.end())
  {
    rpcPic = *(iterPic);
    if( rpcPic->getPOC() == pocCurr && rpcPic->layerId == m_pcEncLib->getLayerId() )
    {
      break;
    }
    iterPic++;
  }

  CHECK(!(rpcPic != NULL), "Unspecified error");
  CHECK(!(rpcPic->getPOC() == pocCurr), "Unspecified error");

  (**iterPicYuvRec) = rpcPic->getRecoBuf();
  return;
}

#if ENABLE_QPA

#ifndef BETA
  #define BETA 0.5 // value between 0.0 and 1; use 0.0 to obtain traditional PSNR
#endif

static inline double calcWeightedSquaredError(const CPelBuf& org,        const CPelBuf& rec,
                                              double &sumAct,            const uint32_t bitDepth,
                                              const uint32_t imageWidth, const uint32_t imageHeight,
                                              const uint32_t offsetX,    const uint32_t offsetY,
                                              int blockWidth,            int blockHeight)
{
  const int    O = org.stride;
  const int    R = rec.stride;
  const Pel   *o = org.bufAt(offsetX, offsetY);
  const Pel   *r = rec.bufAt(offsetX, offsetY);
  const int yAct = offsetY > 0 ? 0 : 1;
  const int xAct = offsetX > 0 ? 0 : 1;

  if (offsetY + (uint32_t)blockHeight > imageHeight) blockHeight = imageHeight - offsetY;
  if (offsetX + (uint32_t)blockWidth  > imageWidth ) blockWidth  = imageWidth  - offsetX;

  const int hAct = offsetY + (uint32_t)blockHeight < imageHeight ? blockHeight : blockHeight - 1;
  const int wAct = offsetX + (uint32_t)blockWidth  < imageWidth  ? blockWidth  : blockWidth  - 1;
  uint64_t ssErr = 0; // sum of squared diffs
  uint64_t saAct = 0; // sum of abs. activity
  double msAct;
  int x, y;

  // calculate image differences and activity
  for (y = 0; y < blockHeight; y++)  // error
  {
    for (x = 0; x < blockWidth; x++)
    {
      const     int64_t iDiff = (int64_t)o[y*O + x] - (int64_t)r[y*R + x];
      ssErr += uint64_t(iDiff * iDiff);
    }
  }
  if (wAct <= xAct || hAct <= yAct) return (double)ssErr;

  for (y = yAct; y < hAct; y++)   // activity
  {
    for (x = xAct; x < wAct; x++)
    {
      const int f = 12 * (int)o[y*O + x] - 2 * ((int)o[y*O + x-1] + (int)o[y*O + x+1] + (int)o[(y-1)*O + x] + (int)o[(y+1)*O + x])
                       - (int)o[(y-1)*O + x-1] - (int)o[(y-1)*O + x+1] - (int)o[(y+1)*O + x-1] - (int)o[(y+1)*O + x+1];
      saAct += abs(f);
    }
  }

  // calculate weight (mean squared activity)
  msAct = (double)saAct / (double(wAct - xAct) * double(hAct - yAct));

  // lower limit, accounts for high-pass gain
  if (msAct < double(1 << (bitDepth - 4))) msAct = double(1 << (bitDepth - 4));

  msAct *= msAct; // because ssErr is squared

  sumAct += msAct; // includes high-pass gain

  // calculate activity weighted error square
  return (double)ssErr * pow(msAct, -1.0 * BETA);
}
#endif // ENABLE_QPA

uint64_t EncGOP::xFindDistortionPlane(const CPelBuf& pic0, const CPelBuf& pic1, const uint32_t rshift
#if ENABLE_QPA
                                    , const uint32_t chromaShiftHor /*= 0*/, const uint32_t chromaShiftVer /*= 0*/
#endif
                                      )
{
  uint64_t uiTotalDiff;
  const  Pel*  pSrc0 = pic0.bufAt(0, 0);
  const  Pel*  pSrc1 = pic1.bufAt(0, 0);

  CHECK(pic0.width  != pic1.width , "Unspecified error");
  CHECK(pic0.height != pic1.height, "Unspecified error");

  if( rshift > 0 )
  {
#if ENABLE_QPA
    const   uint32_t  BD = rshift;      // image bit-depth
    if (BD >= 8)
    {
      const uint32_t   W = pic0.width;  // image width
      const uint32_t   H = pic0.height; // image height
      const double     R = double(W * H) / (1920.0 * 1080.0);
      const uint32_t   B = Clip3<uint32_t>(0, 128 >> chromaShiftVer, 4 * uint32_t(16.0 * sqrt(R) + 0.5)); // WPSNR block size in integer multiple of 4 (for SIMD, = 64 at full-HD)

      uint32_t x, y;

      if (B < 4) // image is too small to use WPSNR, resort to traditional PSNR
      {
        uiTotalDiff = 0;
        for (y = 0; y < H; y++)
        {
          for (x = 0; x < W; x++)
          {
            const           int64_t iDiff = (int64_t)pSrc0[x] - (int64_t)pSrc1[x];
            uiTotalDiff += uint64_t(iDiff * iDiff);
          }
          pSrc0 += pic0.stride;
          pSrc1 += pic1.stride;
        }
        return uiTotalDiff;
      }

      double wmse = 0.0, sumAct = 0.0; // compute activity normalized SNR value

      for (y = 0; y < H; y += B)
      {
        for (x = 0; x < W; x += B)
        {
          wmse += calcWeightedSquaredError(pic1,   pic0,
                                           sumAct, BD,
                                           W,      H,
                                           x,      y,
                                           B,      B);
        }
      }

      // integer weighted distortion
      sumAct = 16.0 * sqrt ((3840.0 * 2160.0) / double((W << chromaShiftHor) * (H << chromaShiftVer))) * double(1 << BD);

      return (wmse <= 0.0) ? 0 : uint64_t(wmse * pow(sumAct, BETA) + 0.5);
    }
#endif // ENABLE_QPA
    uiTotalDiff = 0;
    for (int y = 0; y < pic0.height; y++)
    {
      for (int x = 0; x < pic0.width; x++)
      {
        Intermediate_Int iTemp = pSrc0[x] - pSrc1[x];
        uiTotalDiff += uint64_t((iTemp * iTemp) >> rshift);
      }
      pSrc0 += pic0.stride;
      pSrc1 += pic1.stride;
    }
  }
  else
  {
    uiTotalDiff = 0;
    for (int y = 0; y < pic0.height; y++)
    {
      for (int x = 0; x < pic0.width; x++)
      {
        Intermediate_Int iTemp = pSrc0[x] - pSrc1[x];
        uiTotalDiff += uint64_t(iTemp * iTemp);
      }
      pSrc0 += pic0.stride;
      pSrc1 += pic1.stride;
    }
  }

  return uiTotalDiff;
}
#if WCG_WPSNR
double EncGOP::xFindDistortionPlaneWPSNR(const CPelBuf& pic0, const CPelBuf& pic1, const uint32_t rshift, const CPelBuf& picLuma0,
  ComponentID compID, const ChromaFormat chfmt    )
{
  const bool    useLumaWPSNR = m_pcEncLib->getLumaLevelToDeltaQPMapping().isEnabled() || (m_pcCfg->getLmcs() && m_pcCfg->getReshapeSignalType() == RESHAPE_SIGNAL_PQ);
  if (!useLumaWPSNR)
  {
    return 0;
  }

  double uiTotalDiffWPSNR;
  const  Pel*  pSrc0 = pic0.bufAt(0, 0);
  const  Pel*  pSrc1 = pic1.bufAt(0, 0);
  const  Pel*  pSrcLuma = picLuma0.bufAt(0, 0);