Added filter.c/.h and deblocking functions

build/VS2010/HEVC_encoder.vcxproj

@@ -90,6 +90,7 @@
     <ClCompile Include="..\..\src\context.c" />
     <ClCompile Include="..\..\src\encmain.c" />
     <ClCompile Include="..\..\src\encoder.c" />
+    <ClCompile Include="..\..\src\filter.c" />
     <ClCompile Include="..\..\src\intra.c" />
     <ClCompile Include="..\..\src\nal.c" />
     <ClCompile Include="..\..\src\picture.c" />
@@ -101,6 +102,7 @@
     <ClInclude Include="..\..\src\config.h" />
     <ClInclude Include="..\..\src\context.h" />
     <ClInclude Include="..\..\src\encoder.h" />
+    <ClInclude Include="..\..\src\filter.h" />
     <ClInclude Include="..\..\src\global.h" />
     <ClInclude Include="..\..\src\intra.h" />
     <ClInclude Include="..\..\src\nal.h" />

build/VS2010/HEVC_encoder.vcxproj.filters

@@ -45,6 +45,9 @@
     <ClCompile Include="..\..\src\context.c">
       <Filter>Source Files</Filter>
     </ClCompile>
+    <ClCompile Include="..\..\src\filter.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
   </ItemGroup>
   <ItemGroup>
     <ClInclude Include="..\..\src\global.h">
@@ -77,5 +80,8 @@
     <ClInclude Include="..\..\src\context.h">
       <Filter>Header Files</Filter>
     </ClInclude>
+    <ClInclude Include="..\..\src\filter.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
   </ItemGroup>
 </Project>

src/encoder.h

@@ -53,6 +53,10 @@ typedef struct
   picture_list *ref;
   uint8_t QP;
   uint8_t bitdepth;
+
+  /* Filtering */
+  int8_t betaOffset;
+  int8_t tcOffset;
 } encoder_control;
 
 typedef struct
@@ -102,7 +106,7 @@ void init_tables(void);
 
 static uint32_t* g_auiSigLastScan[3][7];
 int8_t  g_aucConvertToBit[LCU_WIDTH+1];
-static int8_t g_uiBitDepth     = 8;
+static int8_t g_bitDepth     = 8;
 static int8_t g_uiBitIncrement = 0;
 
 static const uint8_t g_uiGroupIdx[ 32 ]    = {0,1,2,3,4,4,5,5,6,6,6,6,7,7,7,7,8,8,8,8,8,8,8,8,9,9,9,9,9,9,9,9};

src/filter.c

@@ -0,0 +1,179 @@
+/**
+ *  HEVC Encoder
+ *  - Marko Viitanen ( fador at iki.fi ), Tampere University of Technology, Department of Computer Systems.
+ */
+
+/*! \file filter.c
+    \brief filtering
+    \author Marko Viitanen
+    \date 2013-04
+    
+    Filtering functions
+*/
+
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include "global.h"
+#include "config.h"
+#include "bitstream.h"
+#include "picture.h"
+#include "cabac.h"
+#include "encoder.h"
+#include "filter.h"
+
+const uint8_t tctable_8x8[54] =
+{
+  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,5,5,6,6,7,8,9,10,11,13,14,16,18,20,22,24
+};
+
+const uint8_t betatable_8x8[52] =
+{
+  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,6,7,8,9,10,11,12,13,14,15,16,17,18,20,22,24,26,28,30,32,34,36,38,40,42,44,46,48,50,52,54,56,58,60,62,64
+};
+
+
+void filter_luma( uint8_t* piSrc, int32_t iOffset, int32_t tc , int8_t sw, int8_t bPartPNoFilter, int8_t bPartQNoFilter, int32_t iThrCut, int8_t bFilterSecondP, int8_t bFilterSecondQ)
+{
+  int32_t delta;
+  
+  uint8_t m4  = piSrc[0];
+  uint8_t m3  = piSrc[-iOffset];
+  uint8_t m5  = piSrc[ iOffset];
+  uint8_t m2  = piSrc[-iOffset*2];
+  uint8_t m6  = piSrc[ iOffset*2];
+  uint8_t m1  = piSrc[-iOffset*3];
+  uint8_t m7  = piSrc[ iOffset*3];
+  uint8_t m0  = piSrc[-iOffset*4];
+
+  if (sw)
+  {
+    piSrc[-iOffset]   = CLIP(m3-2*tc, m3+2*tc, ((m1 + 2*m2 + 2*m3 + 2*m4 + m5 + 4) >> 3));
+    piSrc[0]          = CLIP(m4-2*tc, m4+2*tc, ((m2 + 2*m3 + 2*m4 + 2*m5 + m6 + 4) >> 3));
+    piSrc[-iOffset*2] = CLIP(m2-2*tc, m2+2*tc, ((m1 + m2 + m3 + m4 + 2)>>2));
+    piSrc[ iOffset]   = CLIP(m5-2*tc, m5+2*tc, ((m3 + m4 + m5 + m6 + 2)>>2));
+    piSrc[-iOffset*3] = CLIP(m1-2*tc, m1+2*tc, ((2*m0 + 3*m1 + m2 + m3 + m4 + 4 )>>3));
+    piSrc[ iOffset*2] = CLIP(m6-2*tc, m6+2*tc, ((m3 + m4 + m5 + 3*m6 + 2*m7 +4 )>>3));
+  }
+  else
+  {
+    // Weak filter
+    delta = (9*(m4-m3) -3*(m5-m2) + 8)>>4 ;
+
+    if ( abs(delta) < iThrCut )
+    {
+      int32_t tc2 = tc>>1;
+      delta = CLIP(-tc, tc, delta);        
+      piSrc[-iOffset] = CLIP(0,(1 << g_bitDepth)-1,(m3+delta));
+      piSrc[0] = CLIP(0,(1 << g_bitDepth)-1,(m4-delta));
+
+      
+      if(bFilterSecondP)
+      {
+        int32_t delta1 = CLIP(-tc2, tc2, (( ((m1+m3+1)>>1)- m2+delta)>>1));
+        piSrc[-iOffset*2] = CLIP(0,(1 << g_bitDepth)-1,(m2+delta1));
+      }
+      if(bFilterSecondQ)
+      {
+        int32_t delta2 = CLIP(-tc2, tc2, (( ((m6+m4+1)>>1)- m5-delta)>>1));
+        piSrc[ iOffset] = CLIP(0,(1 << g_bitDepth)-1,(m5+delta2));
+      }
+    }
+  }
+
+  if(bPartPNoFilter)
+  {
+    piSrc[-iOffset] = m3;
+    piSrc[-iOffset*2] = m2;
+    piSrc[-iOffset*3] = m1;
+  }
+  if(bPartQNoFilter)
+  {
+    piSrc[0] = m4;
+    piSrc[ iOffset] = m5;
+    piSrc[ iOffset*2] = m6;
+  }
+}
+
+void filter_deblock_edge_luma(encoder_control* encoder, int32_t xpos, int32_t ypos, int8_t depth, int32_t edge, int8_t chroma, int8_t dir)
+{
+  int i,iIdx;
+  int32_t iStride = encoder->in.cur_pic.width;
+  int32_t iOffset = 0;
+  int32_t betaOffsetDiv2 = encoder->betaOffset>>1;
+  int32_t tcOffsetDiv2   = encoder->tcOffset>>1;
+  const int8_t scu_width       = (LCU_WIDTH>>MAX_DEPTH);  
+  const int8_t scu_width_log2  = TOBITS(scu_width);
+  int8_t uiNumParts = 1;
+  int8_t uiBs       = 1; /* Filter strength */
+  /* ToDo: support 10+bits */
+  uint8_t* src      = &encoder->in.cur_pic.yRecData[xpos+ypos*iStride];
+  uint8_t* piTmpSrc = src;
+  int32_t iSrcStep;
+  CU_info* cu = &encoder->in.cur_pic.CU[0][(xpos>>scu_width_log2) + (ypos>>scu_width_log2)*(encoder->in.width>>scu_width_log2)];
+  
+  if(dir == EDGE_VER)
+  {
+    iOffset = 1;
+    iSrcStep = iStride;
+    piTmpSrc += edge*scu_width;
+  }
+  else
+  {
+    iOffset = iStride;
+    iSrcStep = 1;
+    piTmpSrc += edge*scu_width*iStride;
+  }
+  
+  /* For each subpart */
+  for(iIdx = 0; iIdx < uiNumParts; iIdx++)
+  {
+    int32_t iQP            = encoder->QP;
+    int32_t iBitdepthScale = 1 << (g_bitDepth-8);
+    int32_t iIndexTC       = CLIP(0, 51+2, (int32_t)(iQP + 2*(uiBs-1) + (tcOffsetDiv2 << 1)));
+    int32_t iIndexB        = CLIP(0, 51, iQP + (betaOffsetDiv2 << 1));
+    int32_t iTc            = tctable_8x8[iIndexTC]*iBitdepthScale;
+    int32_t iBeta          = betatable_8x8[iIndexB]*iBitdepthScale;
+    int32_t iSideThreshold = (iBeta+(iBeta>>1))>>3;
+    int32_t iThrCut        = iTc*10;
+    uint32_t uiBlocksInPart= scu_width / 4 ? scu_width / 4 : 1;
+    uint32_t iBlkIdx;
+
+    for (iBlkIdx = 0; iBlkIdx < uiBlocksInPart; iBlkIdx++)
+    {
+      uint8_t* piTmpSrcShift;
+      int32_t dp0,dq0,dp3,dq3,d0,d3,dp,dq,d;
+
+      /* Check conditions for filtering */
+      piTmpSrcShift = piTmpSrc+iSrcStep*(iIdx*scu_width+iBlkIdx*4+0);
+      dp0 = abs( piTmpSrcShift[-iOffset*3] - 2*piTmpSrcShift[-iOffset*2] + piTmpSrcShift[-iOffset] );
+      piTmpSrcShift = piTmpSrc+iSrcStep*(iIdx*scu_width+iBlkIdx*4+0);
+      dq0 = abs( piTmpSrcShift[0] - 2*piTmpSrcShift[iOffset] + piTmpSrcShift[iOffset*2] );
+      piTmpSrcShift = piTmpSrc+iSrcStep*(iIdx*scu_width+iBlkIdx*4+3);
+      dp3 = abs( piTmpSrcShift[-iOffset*3] - 2*piTmpSrcShift[-iOffset*2] + piTmpSrcShift[-iOffset] );
+      piTmpSrcShift = piTmpSrc+iSrcStep*(iIdx*scu_width+iBlkIdx*4+3);
+      dq3 = abs( piTmpSrcShift[0] - 2*piTmpSrcShift[iOffset] + piTmpSrcShift[iOffset*2] );
+      d0 = dp0 + dq0;
+      d3 = dp3 + dq3;        
+      dp = dp0 + dp3;
+      dq = dq0 + dq3;
+      d  =  d0 + d3;
+
+      #if ENABLE_PCM == 1
+      //ToDo: add PCM deblocking
+      #endif
+      if (d < iBeta)
+      { 
+        int8_t bFilterP = (dp < iSideThreshold)?1:0;
+        int8_t bFilterQ = (dq < iSideThreshold)?1:0;          
+        int8_t sw = 0;// xUseStrongFiltering( iOffset, 2*d0, iBeta, iTc, piTmpSrc+iSrcStep*(iIdx*uiPelsInPart+iBlkIdx*4+0))
+                            //&& xUseStrongFiltering( iOffset, 2*d3, iBeta, iTc, piTmpSrc+iSrcStep*(iIdx*uiPelsInPart+iBlkIdx*4+3));          
+        for (i = 0; i < 8/2; i++)
+        {
+          filter_luma( piTmpSrc+iSrcStep*(iIdx*scu_width+iBlkIdx*4+i), iOffset, iTc, sw, 0, 0, iThrCut, bFilterP, bFilterQ);
+        }
+      }
+    }
+  }
+}

src/filter.h

@@ -0,0 +1,22 @@
+/**
+ *  HEVC Encoder
+ *  - Marko Viitanen ( fador at iki.fi ), Tampere University of Technology, Department of Computer Systems 2012.
+ */
+
+/*! \file filter.h
+    \brief filter
+    \author Marko Viitanen
+    \date 2013-04
+    
+    Filtering function headers
+*/
+#ifndef __FILTER_H
+#define __FILTER_H
+
+#define EDGE_VER 0
+#define EDGE_HOR 1
+
+
+void filter_deblock_edge_luma(encoder_control* encoder, int32_t xpos, int32_t ypos, int8_t depth, int32_t edge, int8_t chroma, int8_t dir);
+
+#endif

src/intra.c

@@ -83,7 +83,7 @@ int8_t intra_getBlockMode(picture* pic,uint32_t xCtb, uint32_t yCtb, uint8_t dep
 int16_t intra_getDCPred(int16_t* pic, uint16_t picwidth,uint32_t xpos, uint32_t ypos, uint8_t width)
 {
   int32_t i, iSum = 0;
-  int16_t pDcVal = 1<<(g_uiBitDepth-1);  
+  int16_t pDcVal = 1<<(g_bitDepth-1);  
   
   /* Average of pixels on top and left */
   for (i = -picwidth; i < width-picwidth ; i++)
@@ -367,7 +367,7 @@ void intra_buildReferenceBorder(picture* pic, int32_t xCtb, int32_t yCtb,int16_t
   int32_t leftColumn;  /*!< left column iterator */
   int16_t val;         /*!< variable to store extrapolated value */
   int32_t i;           /*!< index iterator */
-  int16_t dcVal        = 1<<(g_uiBitDepth-1); /*!< default predictor value */
+  int16_t dcVal        = 1<<(g_bitDepth-1); /*!< default predictor value */
   int32_t topRow;      /*!< top row iterator */
   int32_t srcWidth     = (pic->width>>(chroma?1:0)); /*!< source picture width */
   int32_t srcHeight    = (pic->height>>(chroma?1:0));/*!< source picture height */
@@ -546,7 +546,7 @@ void intra_getAngularPred(int16_t* pSrc, int32_t srcStride, int16_t* rpDst, int3
     {
       for (k=0;k<blkSize;k++)
       {
-        pDst[k*dstStride] = CLIP(0, (1<<g_uiBitDepth)-1, pDst[k*dstStride] + (( refSide[k+1] - refSide[0] ) >> 1) );
+        pDst[k*dstStride] = CLIP(0, (1<<g_bitDepth)-1, pDst[k*dstStride] + (( refSide[k+1] - refSide[0] ) >> 1) );
       }
     }
   }
@@ -636,7 +636,7 @@ void intra_DCPredFiltering(uint8_t* pSrc, int32_t iSrcStride, uint8_t* rpDst, in
 */
 void intra_getPlanarPred(int16_t* src,int32_t srcstride, uint32_t xpos, uint32_t ypos,uint32_t width, int16_t* dst,int32_t dststride)
 {
-  int16_t pDcVal = 1<<(g_uiBitDepth-1);
+  int16_t pDcVal = 1<<(g_bitDepth-1);
   int32_t k, l, bottomLeft, topRight;
   int32_t horPred;
   int32_t leftColumn[LCU_WIDTH+1], topRow[LCU_WIDTH+1], bottomRow[LCU_WIDTH+1], rightColumn[LCU_WIDTH+1];

src/picture.c

@@ -131,173 +131,173 @@
 //Calculates image PSNR value
 double imagePSNR(uint8_t *frame1, uint8_t *frame2, uint32_t x, uint32_t y)
 {   
-    double MSE=0.0;
+  uint64_t MSE=0;
-    double MSEtemp=0.0;
+  uint64_t MSEtemp=0;
-    double psnr=0.0;
+  double psnr=0.0;
-    int32_t index;
+  int32_t index;
 
-    //Calculate MSE
+  //Calculate MSE
-    for(index=x*y-1;index>=0;index--)
+  for(index=x*y-1;index>=0;index--)
-    {
+  {
-        MSEtemp=abs(frame1[index]-frame2[index]);
+    MSEtemp=frame1[index]-frame2[index];
-        MSE+=MSEtemp*MSEtemp;
+    MSE+=MSEtemp*MSEtemp;
-    }
+  }
-    MSE/=x*y;
+  MSE/=x*y;
 
-    //Avoid division by zero
+  //Avoid division by zero
-    if(MSE==0) return 99.0;
+  if(MSE==0) return 99.0;
 
-    //The PSNR
+  //The PSNR
-    psnr=10*log10(PSNRMAX/MSE);
+  psnr=10*log10(PSNRMAX/MSE);
 
-    //Thats it.
+  //Thats it.
-    return psnr;
+  return psnr;
 }
 
 //Sum of Absolute Difference for block
 uint32_t SAD(uint8_t *block,uint8_t* block2, uint32_t x, uint32_t y)
 {
-    uint32_t i;
+  uint32_t i;
-    uint32_t sum=0;
+  uint32_t sum=0;
-    for(i=0;i<x*y;i+=4)
+  for(i=0;i<x*y;i+=4)
-    {
+  {
-        sum+=abs((int32_t)block[i]-(int32_t)block2[i]);
+    sum+=abs((int32_t)block[i]-(int32_t)block2[i]);
-        sum+=abs((int32_t)block[i+1]-(int32_t)block2[i+1]);
+    sum+=abs((int32_t)block[i+1]-(int32_t)block2[i+1]);
-        sum+=abs((int32_t)block[i+2]-(int32_t)block2[i+2]);
+    sum+=abs((int32_t)block[i+2]-(int32_t)block2[i+2]);
-        sum+=abs((int32_t)block[i+3]-(int32_t)block2[i+3]);
+    sum+=abs((int32_t)block[i+3]-(int32_t)block2[i+3]);
-    }
+  }
 
-    return sum;    
+  return sum;    
 }
 
 uint32_t SAD64x64(int16_t *block,uint32_t stride1,int16_t* block2, uint32_t stride2)
 {
-    int32_t i,ii,y,x;
+  int32_t i,ii,y,x;
-    uint32_t sum=0;
+  uint32_t sum=0;
-    for(y=0;y<64;y++)
+  for(y=0;y<64;y++)
+  {
+    i = y*stride1; 
+    ii = y*stride2;
+    for(x = 0; x < 64;x++)
     {
-      i = y*stride1; 
+      sum+=abs((int32_t)block[i+x]-(int32_t)block2[ii+x]);
-      ii = y*stride2;
-      for(x = 0; x < 64;x++)
-      {
-        sum+=abs((int32_t)block[i+x]-(int32_t)block2[ii+x]);
-      }
-
     }
 
-    return sum;    
+  }
+
+  return sum;    
 }
 
 uint32_t SAD32x32(int16_t *block,uint32_t stride1,int16_t* block2, uint32_t stride2)
 {
-    int32_t i,ii,y;
+  int32_t i,ii,y;
-    uint32_t sum=0;
+  uint32_t sum=0;
-    for(y=0;y<32;y++)
+  for(y=0;y<32;y++)
-    {
+  {
-      i = y*stride1; 
+    i = y*stride1; 
-      ii = y*stride2;
+    ii = y*stride2;
-      sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
+    sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
-      sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
+    sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
-      sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
+    sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
-      sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
+    sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
-      sum+=abs((int32_t)block[i+4]-(int32_t)block2[ii+4]);
+    sum+=abs((int32_t)block[i+4]-(int32_t)block2[ii+4]);
-      sum+=abs((int32_t)block[i+5]-(int32_t)block2[ii+5]);
+    sum+=abs((int32_t)block[i+5]-(int32_t)block2[ii+5]);
-      sum+=abs((int32_t)block[i+6]-(int32_t)block2[ii+6]);
+    sum+=abs((int32_t)block[i+6]-(int32_t)block2[ii+6]);
-      sum+=abs((int32_t)block[i+7]-(int32_t)block2[ii+7]);
+    sum+=abs((int32_t)block[i+7]-(int32_t)block2[ii+7]);
-      sum+=abs((int32_t)block[i+8]-(int32_t)block2[ii+8]);
+    sum+=abs((int32_t)block[i+8]-(int32_t)block2[ii+8]);
-      sum+=abs((int32_t)block[i+9]-(int32_t)block2[ii+9]);
+    sum+=abs((int32_t)block[i+9]-(int32_t)block2[ii+9]);
-      sum+=abs((int32_t)block[i+10]-(int32_t)block2[ii+10]);
+    sum+=abs((int32_t)block[i+10]-(int32_t)block2[ii+10]);
-      sum+=abs((int32_t)block[i+11]-(int32_t)block2[ii+11]);
+    sum+=abs((int32_t)block[i+11]-(int32_t)block2[ii+11]);
-      sum+=abs((int32_t)block[i+12]-(int32_t)block2[ii+12]);
+    sum+=abs((int32_t)block[i+12]-(int32_t)block2[ii+12]);
-      sum+=abs((int32_t)block[i+13]-(int32_t)block2[ii+13]);
+    sum+=abs((int32_t)block[i+13]-(int32_t)block2[ii+13]);
-      sum+=abs((int32_t)block[i+14]-(int32_t)block2[ii+14]);
+    sum+=abs((int32_t)block[i+14]-(int32_t)block2[ii+14]);
-      sum+=abs((int32_t)block[i+15]-(int32_t)block2[ii+15]);
+    sum+=abs((int32_t)block[i+15]-(int32_t)block2[ii+15]);
-      sum+=abs((int32_t)block[i+16]-(int32_t)block2[ii+16]);
+    sum+=abs((int32_t)block[i+16]-(int32_t)block2[ii+16]);
-      sum+=abs((int32_t)block[i+17]-(int32_t)block2[ii+17]);
+    sum+=abs((int32_t)block[i+17]-(int32_t)block2[ii+17]);
-      sum+=abs((int32_t)block[i+18]-(int32_t)block2[ii+18]);
+    sum+=abs((int32_t)block[i+18]-(int32_t)block2[ii+18]);
-      sum+=abs((int32_t)block[i+19]-(int32_t)block2[ii+19]);
+    sum+=abs((int32_t)block[i+19]-(int32_t)block2[ii+19]);
-      sum+=abs((int32_t)block[i+20]-(int32_t)block2[ii+20]);
+    sum+=abs((int32_t)block[i+20]-(int32_t)block2[ii+20]);
-      sum+=abs((int32_t)block[i+21]-(int32_t)block2[ii+21]);
+    sum+=abs((int32_t)block[i+21]-(int32_t)block2[ii+21]);
-      sum+=abs((int32_t)block[i+22]-(int32_t)block2[ii+22]);
+    sum+=abs((int32_t)block[i+22]-(int32_t)block2[ii+22]);
-      sum+=abs((int32_t)block[i+23]-(int32_t)block2[ii+23]);
+    sum+=abs((int32_t)block[i+23]-(int32_t)block2[ii+23]);
-      sum+=abs((int32_t)block[i+24]-(int32_t)block2[ii+24]);
+    sum+=abs((int32_t)block[i+24]-(int32_t)block2[ii+24]);
-      sum+=abs((int32_t)block[i+25]-(int32_t)block2[ii+25]);
+    sum+=abs((int32_t)block[i+25]-(int32_t)block2[ii+25]);
-      sum+=abs((int32_t)block[i+26]-(int32_t)block2[ii+26]);
+    sum+=abs((int32_t)block[i+26]-(int32_t)block2[ii+26]);
-      sum+=abs((int32_t)block[i+27]-(int32_t)block2[ii+27]);
+    sum+=abs((int32_t)block[i+27]-(int32_t)block2[ii+27]);
-      sum+=abs((int32_t)block[i+28]-(int32_t)block2[ii+28]);
+    sum+=abs((int32_t)block[i+28]-(int32_t)block2[ii+28]);
-      sum+=abs((int32_t)block[i+29]-(int32_t)block2[ii+29]);
+    sum+=abs((int32_t)block[i+29]-(int32_t)block2[ii+29]);
-      sum+=abs((int32_t)block[i+30]-(int32_t)block2[ii+30]);
+    sum+=abs((int32_t)block[i+30]-(int32_t)block2[ii+30]);
-      sum+=abs((int32_t)block[i+31]-(int32_t)block2[ii+31]);
+    sum+=abs((int32_t)block[i+31]-(int32_t)block2[ii+31]);
-    }
+  }
 
-    return sum;    
+  return sum;    
 }
 
 
 uint32_t SAD16x16(int16_t *block,uint32_t stride1,int16_t* block2, uint32_t stride2)
 {
-    int32_t i,ii,y;
+  int32_t i,ii,y;
-    uint32_t sum=0;
+  uint32_t sum=0;
-    for(y=0;y<16;y++)
+  for(y=0;y<16;y++)
-    {
+  {
-      i = y*stride1; 
+    i = y*stride1; 
-      ii = y*stride2;
+    ii = y*stride2;
-      sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
+    sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
-      sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
+    sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
-      sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
+    sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
-      sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
+    sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
-      sum+=abs((int32_t)block[i+4]-(int32_t)block2[ii+4]);
+    sum+=abs((int32_t)block[i+4]-(int32_t)block2[ii+4]);
-      sum+=abs((int32_t)block[i+5]-(int32_t)block2[ii+5]);
+    sum+=abs((int32_t)block[i+5]-(int32_t)block2[ii+5]);
-      sum+=abs((int32_t)block[i+6]-(int32_t)block2[ii+6]);
+    sum+=abs((int32_t)block[i+6]-(int32_t)block2[ii+6]);
-      sum+=abs((int32_t)block[i+7]-(int32_t)block2[ii+7]);
+    sum+=abs((int32_t)block[i+7]-(int32_t)block2[ii+7]);
-      sum+=abs((int32_t)block[i+8]-(int32_t)block2[ii+8]);
+    sum+=abs((int32_t)block[i+8]-(int32_t)block2[ii+8]);
-      sum+=abs((int32_t)block[i+9]-(int32_t)block2[ii+9]);
+    sum+=abs((int32_t)block[i+9]-(int32_t)block2[ii+9]);
-      sum+=abs((int32_t)block[i+10]-(int32_t)block2[ii+10]);
+    sum+=abs((int32_t)block[i+10]-(int32_t)block2[ii+10]);
-      sum+=abs((int32_t)block[i+11]-(int32_t)block2[ii+11]);
+    sum+=abs((int32_t)block[i+11]-(int32_t)block2[ii+11]);
-      sum+=abs((int32_t)block[i+12]-(int32_t)block2[ii+12]);
+    sum+=abs((int32_t)block[i+12]-(int32_t)block2[ii+12]);
-      sum+=abs((int32_t)block[i+13]-(int32_t)block2[ii+13]);
+    sum+=abs((int32_t)block[i+13]-(int32_t)block2[ii+13]);
-      sum+=abs((int32_t)block[i+14]-(int32_t)block2[ii+14]);
+    sum+=abs((int32_t)block[i+14]-(int32_t)block2[ii+14]);
-      sum+=abs((int32_t)block[i+15]-(int32_t)block2[ii+15]);
+    sum+=abs((int32_t)block[i+15]-(int32_t)block2[ii+15]);
-    }
+  }
 
-    return sum;    
+  return sum;    
 }
 
 
 uint32_t SAD8x8(int16_t *block,uint32_t stride1,int16_t* block2, uint32_t stride2)
 {
-    int32_t i,ii,y;
+  int32_t i,ii,y;
-    uint32_t sum=0;
+  uint32_t sum=0;
-    for(y=0;y<8;y++)
+  for(y=0;y<8;y++)
-    {
+  {
-      i = y*stride1; 
+    i = y*stride1; 
-      ii = y*stride2;
+    ii = y*stride2;
-      sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
+    sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
-      sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
+    sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
-      sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
+    sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
-      sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
+    sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
-      sum+=abs((int32_t)block[i+4]-(int32_t)block2[ii+4]);
+    sum+=abs((int32_t)block[i+4]-(int32_t)block2[ii+4]);
-      sum+=abs((int32_t)block[i+5]-(int32_t)block2[ii+5]);
+    sum+=abs((int32_t)block[i+5]-(int32_t)block2[ii+5]);
-      sum+=abs((int32_t)block[i+6]-(int32_t)block2[ii+6]);
+    sum+=abs((int32_t)block[i+6]-(int32_t)block2[ii+6]);
-      sum+=abs((int32_t)block[i+7]-(int32_t)block2[ii+7]);
+    sum+=abs((int32_t)block[i+7]-(int32_t)block2[ii+7]);
-    }
+  }
 
-    return sum;    
+  return sum;    
 }
 
 uint32_t SAD4x4(int16_t *block,uint32_t stride1,int16_t* block2, uint32_t stride2)
 {
-    int32_t i,ii,y;
+  int32_t i,ii,y;
-    uint32_t sum=0;
+  uint32_t sum=0;
-    for(y=0;y<4;y++)
+  for(y=0;y<4;y++)
-    {
+  {
-      i = y*stride1; 
+    i = y*stride1; 
-      ii = y*stride2;
+    ii = y*stride2;
-      sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
+    sum+=abs((int32_t)block[i]-(int32_t)block2[ii]);
-      sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
+    sum+=abs((int32_t)block[i+1]-(int32_t)block2[ii+1]);
-      sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
+    sum+=abs((int32_t)block[i+2]-(int32_t)block2[ii+2]);
-      sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
+    sum+=abs((int32_t)block[i+3]-(int32_t)block2[ii+3]);
-    }
+  }
 
-    return sum;    
+  return sum;
 }

src/transform.c

@@ -714,7 +714,7 @@ void itransform2d(int16_t *block,int16_t *coeff, int8_t blockSize, int8_t uiMode
   //(Int bitDepth, Short *coeff,Short *block, Int iWidth, Int iHeight, UInt uiMode)
 {
   int32_t shift_1st = 7;
-  int32_t shift_2nd = 12 - (g_uiBitDepth-8);
+  int32_t shift_2nd = 12 - (g_bitDepth-8);
   int16_t tmp[LCU_WIDTH*LCU_WIDTH];
 
   if( blockSize == 4)
@@ -797,7 +797,7 @@ void quant(encoder_control* encoder, int16_t* pSrc, int16_t* pDes, int32_t iWidt
   
   int32_t *piQuantCoeff = g_quant_coeff[uiLog2TrSize-2][scalingListType][/*m_cQP.m_iRem*/qpScaled%6];
 
-  uint32_t uiBitDepth = g_uiBitDepth;
+  uint32_t uiBitDepth = g_bitDepth;
 
   int32_t iTransformShift = /*MAX_TR_DYNAMIC_RANGE*/15 - uiBitDepth - uiLog2TrSize; // Represents scaling through forward transform
   int32_t iQBits = QUANT_SHIFT + /*cQpBase.m_iPer +*/qpScaled/6 + iTransformShift;
@@ -831,7 +831,7 @@ void dequant(encoder_control* encoder, int16_t* piQCoef, int16_t* piCoef, int32_
   uint32_t uiLog2TrSize = g_aucConvertToBit[ iWidth ] + 2;
   int16_t clipQCoef;
   int32_t n;
-  int32_t iTransformShift = 15 - g_uiBitDepth - (g_aucConvertToBit[ iWidth ] + 2);
+  int32_t iTransformShift = 15 - g_bitDepth - (g_aucConvertToBit[ iWidth ] + 2);
   int32_t qpScaled;
   int32_t iQpBase = encoder->QP;
   int32_t scalingListType = (/*pcCU->isint32_tra(uiAbsPartIdx)*/0 ? 0 : 3) + (int8_t)("\0\3\1\2"[eTType]);