Newer
Older

Karsten Suehring
committed
, m_commonCtx ( commonCtx )
{
}
template<uint8_t numIPos>
inline void State::updateState(const ScanInfo &scanInfo, const State *prevStates, const Decision &decision)
{
m_rdCost = decision.rdCost;
if( decision.prevId > -2 )
{
if( decision.prevId >= 0 )
{
const State* prvState = prevStates + decision.prevId;
m_numSigSbb = prvState->m_numSigSbb + !!decision.absLevel;
m_refSbbCtxId = prvState->m_refSbbCtxId;
m_sbbFracBits = prvState->m_sbbFracBits;
m_remRegBins = prvState->m_remRegBins - 1;
m_goRicePar = prvState->m_goRicePar;
if( m_remRegBins >= 3 )
{
TCoeff rem = (decision.absLevel - 4) >> 1;
if( m_goRicePar < 3 && rem > (3<<m_goRicePar)-1 )
{
m_goRicePar++;
}
m_remRegBins -= std::min<TCoeff>( decision.absLevel, 2 );
}

Karsten Suehring
committed
::memcpy( m_absLevelsAndCtxInit, prvState->m_absLevelsAndCtxInit, 48*sizeof(uint8_t) );
}
else
{
m_numSigSbb = 1;
m_refSbbCtxId = -1;
if ( scanInfo.sbbSize == 4 )
{
m_remRegBins = MAX_NUM_REG_BINS_2x2SUBBLOCK - MAX_NUM_GT2_BINS_2x2SUBBLOCK - std::min<TCoeff>( decision.absLevel, 2 );
}
else
{
m_remRegBins = MAX_NUM_REG_BINS_4x4SUBBLOCK - MAX_NUM_GT2_BINS_4x4SUBBLOCK - std::min<TCoeff>( decision.absLevel, 2 );
}
m_goRicePar = ( ((decision.absLevel - 4) >> 1) > (3<<0)-1 ? 1 : 0 );

Karsten Suehring
committed
::memset( m_absLevelsAndCtxInit, 0, 48*sizeof(uint8_t) );
}
uint8_t* levels = reinterpret_cast<uint8_t*>(m_absLevelsAndCtxInit);
levels[ scanInfo.insidePos ] = (uint8_t)std::min<TCoeff>( 255, decision.absLevel );
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
if (m_remRegBins >= 3)
{
TCoeff tinit = m_absLevelsAndCtxInit[8 + scanInfo.nextInsidePos];
TCoeff sumAbs1 = (tinit >> 3) & 31;
TCoeff sumNum = tinit & 7;
#define UPDATE(k) {TCoeff t=levels[scanInfo.nextNbInfoSbb.inPos[k]]; sumAbs1+=std::min<TCoeff>(2+(t&1),t); sumNum+=!!t; }
if (numIPos == 1)
{
UPDATE(0);
}
else if (numIPos == 2)
{
UPDATE(0);
UPDATE(1);
}
else if (numIPos == 3)
{
UPDATE(0);
UPDATE(1);
UPDATE(2);
}
else if (numIPos == 4)
{
UPDATE(0);
UPDATE(1);
UPDATE(2);
UPDATE(3);
}
else if (numIPos == 5)
{
UPDATE(0);
UPDATE(1);
UPDATE(2);
UPDATE(3);
UPDATE(4);
}
#undef UPDATE
TCoeff sumGt1 = sumAbs1 - sumNum;
m_sigFracBits = m_sigFracBitsArray[scanInfo.sigCtxOffsetNext + (sumAbs1 < 5 ? sumAbs1 : 5)];
m_coeffFracBits = m_gtxFracBitsArray[scanInfo.gtxCtxOffsetNext + (sumGt1 < 4 ? sumGt1 : 4)];
}
else
{
TCoeff sumAbs = m_absLevelsAndCtxInit[8 + scanInfo.nextInsidePos] >> 8;
#define UPDATE(k) {TCoeff t=levels[scanInfo.nextNbInfoSbb.inPos[k]]; sumAbs+=t; }
if (numIPos == 1)
{
UPDATE(0);
}
else if (numIPos == 2)
{
UPDATE(0);
UPDATE(1);
}
else if (numIPos == 3)
{
UPDATE(0);
UPDATE(1);
UPDATE(2);
}
else if (numIPos == 4)
{
UPDATE(0);
UPDATE(1);
UPDATE(2);
UPDATE(3);
}
else if (numIPos == 5)
{
UPDATE(0);
UPDATE(1);
UPDATE(2);
UPDATE(3);
UPDATE(4);
}
#undef UPDATE
sumAbs = std::min(31, sumAbs);
m_goRicePar = g_auiGoRiceParsCoeff[sumAbs];
m_goRiceZero = m_goRiceZeroArray[sumAbs];
}

Karsten Suehring
committed
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
}
}
inline void State::updateStateEOS(const ScanInfo &scanInfo, const State *prevStates, const State *skipStates,
const Decision &decision)
{
m_rdCost = decision.rdCost;
if( decision.prevId > -2 )
{
const State* prvState = 0;
if( decision.prevId >= 0 )
{
prvState = ( decision.prevId < 4 ? prevStates : skipStates - 4 ) + decision.prevId;
m_numSigSbb = prvState->m_numSigSbb + !!decision.absLevel;
::memcpy( m_absLevelsAndCtxInit, prvState->m_absLevelsAndCtxInit, 16*sizeof(uint8_t) );
}
else
{
m_numSigSbb = 1;
::memset( m_absLevelsAndCtxInit, 0, 16*sizeof(uint8_t) );
}
reinterpret_cast<uint8_t*>(m_absLevelsAndCtxInit)[ scanInfo.insidePos ] = (uint8_t)std::min<TCoeff>( 255, decision.absLevel );
m_commonCtx.update( scanInfo, prvState, *this );
TCoeff tinit = m_absLevelsAndCtxInit[ 8 + scanInfo.nextInsidePos ];
TCoeff sumNum = tinit & 7;
TCoeff sumAbs1 = ( tinit >> 3 ) & 31;
TCoeff sumGt1 = sumAbs1 - sumNum;
m_sigFracBits = m_sigFracBitsArray[ scanInfo.sigCtxOffsetNext + ( sumAbs1 < 5 ? sumAbs1 : 5 ) ];
m_coeffFracBits = m_gtxFracBitsArray[ scanInfo.gtxCtxOffsetNext + ( sumGt1 < 4 ? sumGt1 : 4 ) ];
}
}
inline void CommonCtx::update(const ScanInfo &scanInfo, const State *prevState, State &currState)
{
uint8_t* sbbFlags = m_currSbbCtx[ currState.m_stateId ].sbbFlags;
uint8_t* levels = m_currSbbCtx[ currState.m_stateId ].levels;
std::size_t setCpSize = m_nbInfo[ scanInfo.scanIdx - 1 ].maxDist * sizeof(uint8_t);
if( prevState && prevState->m_refSbbCtxId >= 0 )
{
::memcpy( sbbFlags, m_prevSbbCtx[prevState->m_refSbbCtxId].sbbFlags, scanInfo.numSbb*sizeof(uint8_t) );
::memcpy( levels + scanInfo.scanIdx, m_prevSbbCtx[prevState->m_refSbbCtxId].levels + scanInfo.scanIdx, setCpSize );
}
else
{
::memset( sbbFlags, 0, scanInfo.numSbb*sizeof(uint8_t) );
::memset( levels + scanInfo.scanIdx, 0, setCpSize );
}
sbbFlags[ scanInfo.sbbPos ] = !!currState.m_numSigSbb;
::memcpy( levels + scanInfo.scanIdx, currState.m_absLevelsAndCtxInit, scanInfo.sbbSize*sizeof(uint8_t) );
const int sigNSbb = ( ( scanInfo.nextSbbRight ? sbbFlags[ scanInfo.nextSbbRight ] : false ) || ( scanInfo.nextSbbBelow ? sbbFlags[ scanInfo.nextSbbBelow ] : false ) ? 1 : 0 );
currState.m_numSigSbb = 0;
if (scanInfo.sbbSize == 4)
{
currState.m_remRegBins = MAX_NUM_REG_BINS_2x2SUBBLOCK - MAX_NUM_GT2_BINS_2x2SUBBLOCK;
}
else
{
currState.m_remRegBins = MAX_NUM_REG_BINS_4x4SUBBLOCK - MAX_NUM_GT2_BINS_4x4SUBBLOCK;
}
currState.m_goRicePar = 0;

Karsten Suehring
committed
currState.m_refSbbCtxId = currState.m_stateId;
currState.m_sbbFracBits = m_sbbFlagBits[ sigNSbb ];
uint16_t templateCtxInit[16];
const int scanBeg = scanInfo.scanIdx - scanInfo.sbbSize;
const NbInfoOut* nbOut = m_nbInfo + scanBeg;
const uint8_t* absLevels = levels + scanBeg;
for( int id = 0; id < scanInfo.sbbSize; id++, nbOut++ )
{
if( nbOut->num )
{
TCoeff sumAbs = 0, sumAbs1 = 0, sumNum = 0;
#define UPDATE(k) {TCoeff t=absLevels[nbOut->outPos[k]]; sumAbs+=t; sumAbs1+=std::min<TCoeff>(2+(t&1),t); sumNum+=!!t; }

Karsten Suehring
committed
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
UPDATE(0);
if( nbOut->num > 1 )
{
UPDATE(1);
if( nbOut->num > 2 )
{
UPDATE(2);
if( nbOut->num > 3 )
{
UPDATE(3);
if( nbOut->num > 4 )
{
UPDATE(4);
}
}
}
}
#undef UPDATE
templateCtxInit[id] = uint16_t(sumNum) + ( uint16_t(sumAbs1) << 3 ) + ( (uint16_t)std::min<TCoeff>( 127, sumAbs ) << 8 );
}
else
{
templateCtxInit[id] = 0;
}
}
::memset( currState.m_absLevelsAndCtxInit, 0, 16*sizeof(uint8_t) );
::memcpy( currState.m_absLevelsAndCtxInit + 8, templateCtxInit, 16*sizeof(uint16_t) );
}
/*================================================================================*/
/*===== =====*/
/*===== T C Q =====*/
/*===== =====*/
/*================================================================================*/
class DepQuant : private RateEstimator
{
public:
DepQuant();
void quant ( TransformUnit& tu, const CCoeffBuf& srcCoeff, const ComponentID compID, const QpParam& cQP, const double lambda, const Ctx& ctx, TCoeff& absSum );
void dequant ( const TransformUnit& tu, CoeffBuf& recCoeff, const ComponentID compID, const QpParam& cQP ) const;
private:
void xDecideAndUpdate ( const TCoeff absCoeff, const ScanInfo& scanInfo );
void xDecide ( const ScanPosType spt, const TCoeff absCoeff, const int lastOffset, Decision* decisions );

Karsten Suehring
committed
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
private:
CommonCtx m_commonCtx;
State m_allStates[ 12 ];
State* m_currStates;
State* m_prevStates;
State* m_skipStates;
State m_startState;
Quantizer m_quant;
Decision m_trellis[ MAX_TU_SIZE * MAX_TU_SIZE ][ 8 ];
};
#define TINIT(x) {*this,m_commonCtx,x}
DepQuant::DepQuant()
: RateEstimator ()
, m_commonCtx ()
, m_allStates {TINIT(0),TINIT(1),TINIT(2),TINIT(3),TINIT(0),TINIT(1),TINIT(2),TINIT(3),TINIT(0),TINIT(1),TINIT(2),TINIT(3)}
, m_currStates ( m_allStates )
, m_prevStates ( m_currStates + 4 )
, m_skipStates ( m_prevStates + 4 )
, m_startState TINIT(0)
{}
#undef TINIT
void DepQuant::dequant( const TransformUnit& tu, CoeffBuf& recCoeff, const ComponentID compID, const QpParam& cQP ) const
{
m_quant.dequantBlock( tu, compID, cQP, recCoeff );
}
#define DINIT(l,p) {std::numeric_limits<int64_t>::max()>>2,l,p}
static const Decision startDec[8] = {DINIT(-1,-2),DINIT(-1,-2),DINIT(-1,-2),DINIT(-1,-2),DINIT(0,4),DINIT(0,5),DINIT(0,6),DINIT(0,7)};
#undef DINIT
void DepQuant::xDecide( const ScanPosType spt, const TCoeff absCoeff, const int lastOffset, Decision* decisions)

Karsten Suehring
committed
{
::memcpy( decisions, startDec, 8*sizeof(Decision) );
PQData pqData[4];
m_quant.preQuantCoeff( absCoeff, pqData );
m_prevStates[0].checkRdCosts( spt, pqData[0], pqData[2], decisions[0], decisions[2]);
m_prevStates[1].checkRdCosts( spt, pqData[0], pqData[2], decisions[2], decisions[0]);
m_prevStates[2].checkRdCosts( spt, pqData[3], pqData[1], decisions[1], decisions[3]);
m_prevStates[3].checkRdCosts( spt, pqData[3], pqData[1], decisions[3], decisions[1]);

Karsten Suehring
committed
if( spt==SCAN_EOCSBB )
{
m_skipStates[0].checkRdCostSkipSbb( decisions[0] );
m_skipStates[1].checkRdCostSkipSbb( decisions[1] );
m_skipStates[2].checkRdCostSkipSbb( decisions[2] );
m_skipStates[3].checkRdCostSkipSbb( decisions[3] );
}
m_startState.checkRdCostStart( lastOffset, pqData[0], decisions[0] );
m_startState.checkRdCostStart( lastOffset, pqData[2], decisions[2] );
}
void DepQuant::xDecideAndUpdate( const TCoeff absCoeff, const ScanInfo& scanInfo )
{
Decision* decisions = m_trellis[ scanInfo.scanIdx ];
std::swap( m_prevStates, m_currStates );
xDecide( scanInfo.spt, absCoeff, lastOffset(scanInfo.scanIdx), decisions);

Karsten Suehring
committed
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
if( scanInfo.scanIdx )
{
if( scanInfo.eosbb )
{
m_commonCtx.swap();
m_currStates[0].updateStateEOS( scanInfo, m_prevStates, m_skipStates, decisions[0] );
m_currStates[1].updateStateEOS( scanInfo, m_prevStates, m_skipStates, decisions[1] );
m_currStates[2].updateStateEOS( scanInfo, m_prevStates, m_skipStates, decisions[2] );
m_currStates[3].updateStateEOS( scanInfo, m_prevStates, m_skipStates, decisions[3] );
::memcpy( decisions+4, decisions, 4*sizeof(Decision) );
}
else
{
switch( scanInfo.nextNbInfoSbb.num )
{
case 0:
m_currStates[0].updateState<0>( scanInfo, m_prevStates, decisions[0] );
m_currStates[1].updateState<0>( scanInfo, m_prevStates, decisions[1] );
m_currStates[2].updateState<0>( scanInfo, m_prevStates, decisions[2] );
m_currStates[3].updateState<0>( scanInfo, m_prevStates, decisions[3] );
break;
case 1:
m_currStates[0].updateState<1>( scanInfo, m_prevStates, decisions[0] );
m_currStates[1].updateState<1>( scanInfo, m_prevStates, decisions[1] );
m_currStates[2].updateState<1>( scanInfo, m_prevStates, decisions[2] );
m_currStates[3].updateState<1>( scanInfo, m_prevStates, decisions[3] );
break;
case 2:
m_currStates[0].updateState<2>( scanInfo, m_prevStates, decisions[0] );
m_currStates[1].updateState<2>( scanInfo, m_prevStates, decisions[1] );
m_currStates[2].updateState<2>( scanInfo, m_prevStates, decisions[2] );
m_currStates[3].updateState<2>( scanInfo, m_prevStates, decisions[3] );
break;
case 3:
m_currStates[0].updateState<3>( scanInfo, m_prevStates, decisions[0] );
m_currStates[1].updateState<3>( scanInfo, m_prevStates, decisions[1] );
m_currStates[2].updateState<3>( scanInfo, m_prevStates, decisions[2] );
m_currStates[3].updateState<3>( scanInfo, m_prevStates, decisions[3] );
break;
case 4:
m_currStates[0].updateState<4>( scanInfo, m_prevStates, decisions[0] );
m_currStates[1].updateState<4>( scanInfo, m_prevStates, decisions[1] );
m_currStates[2].updateState<4>( scanInfo, m_prevStates, decisions[2] );
m_currStates[3].updateState<4>( scanInfo, m_prevStates, decisions[3] );
break;
default:
m_currStates[0].updateState<5>( scanInfo, m_prevStates, decisions[0] );
m_currStates[1].updateState<5>( scanInfo, m_prevStates, decisions[1] );
m_currStates[2].updateState<5>( scanInfo, m_prevStates, decisions[2] );
m_currStates[3].updateState<5>( scanInfo, m_prevStates, decisions[3] );
}
}
if( scanInfo.spt == SCAN_SOCSBB )

Karsten Suehring
committed
{
std::swap( m_prevStates, m_skipStates );
}
}
}
void DepQuant::quant( TransformUnit& tu, const CCoeffBuf& srcCoeff, const ComponentID compID, const QpParam& cQP, const double lambda, const Ctx& ctx, TCoeff& absSum )
{
CHECKD( tu.cs->sps->getSpsRangeExtension().getExtendedPrecisionProcessingFlag(), "ext precision is not supported" );

Karsten Suehring
committed
//===== reset / pre-init =====
const TUParameters& tuPars = *g_Rom.getTUPars( tu.blocks[compID], compID );

Karsten Suehring
committed
m_quant.initQuantBlock ( tu, compID, cQP, lambda );
TCoeff* qCoeff = tu.getCoeffs( compID ).buf;
const TCoeff* tCoeff = srcCoeff.buf;
const int numCoeff = tu.blocks[compID].area();
::memset( tu.getCoeffs( compID ).buf, 0x00, numCoeff*sizeof(TCoeff) );
absSum = 0;
//===== find first test position =====
int firstTestPos = numCoeff - 1;
const TCoeff thres = m_quant.getLastThreshold();
for( ; firstTestPos >= 0; firstTestPos-- )
{
if( abs( tCoeff[ tuPars.m_scanId2BlkPos[firstTestPos] ] ) > thres )

Karsten Suehring
committed
{
break;
}
}
if( firstTestPos < 0 )
{
return;
}
//===== real init =====
RateEstimator::initCtx( tuPars, tu, compID, ctx.getFracBitsAcess() );
m_commonCtx.reset( tuPars, *this );

Karsten Suehring
committed
for( int k = 0; k < 12; k++ )
{
m_allStates[k].init();
}
m_startState.init();
//===== populate trellis =====
for( int scanIdx = firstTestPos; scanIdx >= 0; scanIdx-- )
{
const ScanInfo& scanInfo = tuPars.m_scanInfo[ scanIdx ];
xDecideAndUpdate( abs( tCoeff[ scanInfo.rasterPos ] ), scanInfo );
}

Karsten Suehring
committed
//===== find best path =====
Decision decision = { std::numeric_limits<int64_t>::max(), -1, -2 };
int64_t minPathCost = 0;
for( int8_t stateId = 0; stateId < 4; stateId++ )
{
int64_t pathCost = m_trellis[0][stateId].rdCost;
if( pathCost < minPathCost )
{
decision.prevId = stateId;
minPathCost = pathCost;
}
}
//===== backward scanning =====
int scanIdx = 0;
for( ; decision.prevId >= 0; scanIdx++ )
{
decision = m_trellis[ scanIdx ][ decision.prevId ];
int32_t blkpos = tuPars.m_scanId2BlkPos[ scanIdx ];

Karsten Suehring
committed
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
qCoeff[ blkpos ] = ( tCoeff[ blkpos ] < 0 ? -decision.absLevel : decision.absLevel );
absSum += decision.absLevel;
}
}
}; // namespace DQIntern
//===== interface class =====
DepQuant::DepQuant( const Quant* other, bool enc ) : QuantRDOQ( other )
{
const DepQuant* dq = dynamic_cast<const DepQuant*>( other );
CHECK( other && !dq, "The DepQuant cast must be successfull!" );
p = new DQIntern::DepQuant();
if( enc )
{
DQIntern::g_Rom.init();
}
}
DepQuant::~DepQuant()
{
delete static_cast<DQIntern::DepQuant*>(p);
}
void DepQuant::quant( TransformUnit &tu, const ComponentID &compID, const CCoeffBuf &pSrc, TCoeff &uiAbsSum, const QpParam &cQP, const Ctx& ctx )
{
if( tu.cs->slice->getDepQuantEnabledFlag() )
{
static_cast<DQIntern::DepQuant*>(p)->quant( tu, pSrc, compID, cQP, Quant::m_dLambda, ctx, uiAbsSum );
}
else
{
QuantRDOQ::quant( tu, compID, pSrc, uiAbsSum, cQP, ctx );
}
}
void DepQuant::dequant( const TransformUnit &tu, CoeffBuf &dstCoeff, const ComponentID &compID, const QpParam &cQP )
{
if( tu.cs->slice->getDepQuantEnabledFlag() )
{
static_cast<DQIntern::DepQuant*>(p)->dequant( tu, dstCoeff, compID, cQP );
}
else
{
QuantRDOQ::dequant( tu, dstCoeff, compID, cQP );
}
}