// 30.10.2002 corrected qpel chroma rounding // 04.10.2002 added qpel support to MBMotionCompensation // 01.05.2002 updated MBMotionCompensationBVOP // 14.04.2002 bframe compensation #include #include "../encoder.h" #include "../utils/mbfunctions.h" #include "../image/interpolate8x8.h" #include "../image/reduced.h" #include "../utils/timer.h" #include "motion.h" #ifndef ABS #define ABS(X) (((X)>0)?(X):-(X)) #endif #ifndef SIGN #define SIGN(X) (((X)>0)?1:-1) #endif #ifndef RSHIFT #define RSHIFT(a,b) ((a) > 0 ? ((a) + (1<<((b)-1)))>>(b) : ((a) + (1<<((b)-1))-1)>>(b)) #endif /* assume b>0 */ #ifndef RDIV #define RDIV(a,b) (((a)>0 ? (a) + ((b)>>1) : (a) - ((b)>>1))/(b)) #endif /* This is borrowed from decoder.c */ static __inline int gmc_sanitize(int value, int quarterpel, int fcode) { int length = 1 << (fcode+4); // if (quarterpel) value *= 2; if (value < -length) return -length; else if (value >= length) return length-1; else return value; } /* And this is borrowed from bitstream.c until we find a common solution */ static uint32_t __inline log2bin(uint32_t value) { /* Changed by Chenm001 */ #if !defined(_MSC_VER) int n = 0; while (value) { value >>= 1; n++; } return n; #else __asm { bsr eax, value inc eax } #endif } static __inline void compensate16x16_interpolate(int16_t * const dct_codes, uint8_t * const cur, const uint8_t * const ref, const uint8_t * const refh, const uint8_t * const refv, const uint8_t * const refhv, uint8_t * const tmp, uint32_t x, uint32_t y, const int32_t dx, const int32_t dy, const int32_t stride, const int quarterpel, const int reduced_resolution, const int32_t rounding) { const uint8_t * ptr; if (!reduced_resolution) { if(quarterpel) { if ((dx&3) | (dy&3)) { interpolate16x16_quarterpel(tmp - y * stride - x, (uint8_t *) ref, tmp + 32, tmp + 64, tmp + 96, x, y, dx, dy, stride, rounding); ptr = tmp; } else ptr = ref + (y + dy/4)*stride + x + dx/4; // fullpixel position } else ptr = get_ref(ref, refh, refv, refhv, x, y, 1, dx, dy, stride); transfer_8to16sub(dct_codes, cur + y * stride + x, ptr, stride); transfer_8to16sub(dct_codes+64, cur + y * stride + x + 8, ptr + 8, stride); transfer_8to16sub(dct_codes+128, cur + y * stride + x + 8*stride, ptr + 8*stride, stride); transfer_8to16sub(dct_codes+192, cur + y * stride + x + 8*stride+8, ptr + 8*stride + 8, stride); } else { //reduced_resolution x *= 2; y *= 2; ptr = get_ref(ref, refh, refv, refhv, x, y, 1, dx, dy, stride); filter_18x18_to_8x8(dct_codes, cur+y*stride + x, stride); filter_diff_18x18_to_8x8(dct_codes, ptr, stride); filter_18x18_to_8x8(dct_codes+64, cur+y*stride + x + 16, stride); filter_diff_18x18_to_8x8(dct_codes+64, ptr + 16, stride); filter_18x18_to_8x8(dct_codes+128, cur+(y+16)*stride + x, stride); filter_diff_18x18_to_8x8(dct_codes+128, ptr + 16*stride, stride); filter_18x18_to_8x8(dct_codes+192, cur+(y+16)*stride + x + 16, stride); filter_diff_18x18_to_8x8(dct_codes+192, ptr + 16*stride + 16, stride); transfer32x32_copy(cur + y*stride + x, ptr, stride); } } static __inline void compensate8x8_interpolate( int16_t * const dct_codes, uint8_t * const cur, const uint8_t * const ref, const uint8_t * const refh, const uint8_t * const refv, const uint8_t * const refhv, uint8_t * const tmp, uint32_t x, uint32_t y, const int32_t dx, const int32_t dy, const int32_t stride, const int32_t quarterpel, const int reduced_resolution, const int32_t rounding) { const uint8_t * ptr; if (!reduced_resolution) { if(quarterpel) { if ((dx&3) | (dy&3)) { interpolate8x8_quarterpel(tmp - y*stride - x, (uint8_t *) ref, tmp + 32, tmp + 64, tmp + 96, x, y, dx, dy, stride, rounding); ptr = tmp; } else ptr = ref + (y + dy/4)*stride + x + dx/4; // fullpixel position } else ptr = get_ref(ref, refh, refv, refhv, x, y, 1, dx, dy, stride); transfer_8to16sub(dct_codes, cur + y * stride + x, ptr, stride); } else { //reduced_resolution x *= 2; y *= 2; ptr = get_ref(ref, refh, refv, refhv, x, y, 1, dx, dy, stride); filter_18x18_to_8x8(dct_codes, cur+y*stride + x, stride); filter_diff_18x18_to_8x8(dct_codes, ptr, stride); transfer16x16_copy(cur + y*stride + x, ptr, stride); } } static __inline void compensate16x16_interpolate_ro(int16_t * const dct_codes, const uint8_t * const cur, const uint8_t * const ref, const uint8_t * const refh, const uint8_t * const refv, const uint8_t * const refhv, uint8_t * const tmp, const uint32_t x, const uint32_t y, const int32_t dx, const int32_t dy, const int32_t stride, const int quarterpel) { const uint8_t * ptr; if(quarterpel) { if ((dx&3) | (dy&3)) { interpolate16x16_quarterpel(tmp - y * stride - x, (uint8_t *) ref, tmp + 32, tmp + 64, tmp + 96, x, y, dx, dy, stride, 0); ptr = tmp; } else ptr = ref + (y + dy/4)*stride + x + dx/4; // fullpixel position } else ptr = get_ref(ref, refh, refv, refhv, x, y, 1, dx, dy, stride); transfer_8to16subro(dct_codes, cur + y * stride + x, ptr, stride); transfer_8to16subro(dct_codes+64, cur + y * stride + x + 8, ptr + 8, stride); transfer_8to16subro(dct_codes+128, cur + y * stride + x + 8*stride, ptr + 8*stride, stride); transfer_8to16subro(dct_codes+192, cur + y * stride + x + 8*stride+8, ptr + 8*stride + 8, stride); } /* XXX: slow, inelegant... */ static void interpolate18x18_switch(uint8_t * const cur, const uint8_t * const refn, const uint32_t x, const uint32_t y, const int32_t dx, const int dy, const int32_t stride, const int32_t rounding) { interpolate8x8_switch(cur, refn, x-1, y-1, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x+7, y-1, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x+9, y-1, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x-1, y+7, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x+7, y+7, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x+9, y+7, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x-1, y+9, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x+7, y+9, dx, dy, stride, rounding); interpolate8x8_switch(cur, refn, x+9, y+9, dx, dy, stride, rounding); } static void CompensateChroma( int dx, int dy, const int i, const int j, IMAGE * const Cur, const IMAGE * const Ref, uint8_t * const temp, int16_t * const coeff, const int32_t stride, const int rounding, const int rrv) { /* uv-block-based compensation */ if (!rrv) { transfer_8to16sub(coeff, Cur->u + 8 * j * stride + 8 * i, interpolate8x8_switch2(temp, Ref->u, 8 * i, 8 * j, dx, dy, stride, rounding), stride); transfer_8to16sub(coeff + 64, Cur->v + 8 * j * stride + 8 * i, interpolate8x8_switch2(temp, Ref->v, 8 * i, 8 * j, dx, dy, stride, rounding), stride); } else { uint8_t * current, * reference; current = Cur->u + 16*j*stride + 16*i; reference = temp - 16*j*stride - 16*i; interpolate18x18_switch(reference, Ref->u, 16*i, 16*j, dx, dy, stride, rounding); filter_18x18_to_8x8(coeff, current, stride); filter_diff_18x18_to_8x8(coeff, temp, stride); transfer16x16_copy(current, temp, stride); current = Cur->v + 16*j*stride + 16*i; interpolate18x18_switch(reference, Ref->v, 16*i, 16*j, dx, dy, stride, rounding); filter_18x18_to_8x8(coeff + 64, current, stride); filter_diff_18x18_to_8x8(coeff + 64, temp, stride); transfer16x16_copy(current, temp, stride); } } void MBMotionCompensation(MACROBLOCK * const mb, const uint32_t i, const uint32_t j, const IMAGE * const ref, const IMAGE * const refh, const IMAGE * const refv, const IMAGE * const refhv, const IMAGE * const refGMC, IMAGE * const cur, int16_t * dct_codes, const uint32_t width, const uint32_t height, const uint32_t edged_width, const int32_t quarterpel, const int reduced_resolution, const int32_t rounding) { int32_t dx; int32_t dy; uint8_t * const tmp = refv->u; if ( (!reduced_resolution) && (mb->mode == MODE_NOT_CODED) ) { /* quick copy for early SKIP */ /* early SKIP is only activated in P-VOPs, not in S-VOPs, so mcsel can never be 1 */ transfer16x16_copy(cur->y + 16 * (i + j * edged_width), ref->y + 16 * (i + j * edged_width), edged_width); transfer8x8_copy(cur->u + 8 * (i + j * edged_width/2), ref->u + 8 * (i + j * edged_width/2), edged_width / 2); transfer8x8_copy(cur->v + 8 * (i + j * edged_width/2), ref->v + 8 * (i + j * edged_width/2), edged_width / 2); return; } if ((mb->mode == MODE_NOT_CODED || mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q)) { /* reduced resolution + GMC: not possible */ if (mb->mcsel) { /* call normal routine once, easier than "if (mcsel)"ing all the time */ transfer_8to16sub(&dct_codes[0*64], cur->y + 16*j*edged_width + 16*i, refGMC->y + 16*j*edged_width + 16*i, edged_width); transfer_8to16sub(&dct_codes[1*64], cur->y + 16*j*edged_width + 16*i+8, refGMC->y + 16*j*edged_width + 16*i+8, edged_width); transfer_8to16sub(&dct_codes[2*64], cur->y + (16*j+8)*edged_width + 16*i, refGMC->y + (16*j+8)*edged_width + 16*i, edged_width); transfer_8to16sub(&dct_codes[3*64], cur->y + (16*j+8)*edged_width + 16*i+8, refGMC->y + (16*j+8)*edged_width + 16*i+8, edged_width); /* lumi is needed earlier for mode decision, but chroma should be done block-based, but it isn't, yet. */ transfer_8to16sub(&dct_codes[4 * 64], cur->u + 8 *j*edged_width/2 + 8*i, refGMC->u + 8 *j*edged_width/2 + 8*i, edged_width/2); transfer_8to16sub(&dct_codes[5 * 64], cur->v + 8*j* edged_width/2 + 8*i, refGMC->v + 8*j* edged_width/2 + 8*i, edged_width/2); return; } /* ordinary compensation */ dx = (quarterpel ? mb->qmvs[0].x : mb->mvs[0].x); dy = (quarterpel ? mb->qmvs[0].y : mb->mvs[0].y); if (reduced_resolution) { dx = RRV_MV_SCALEUP(dx); dy = RRV_MV_SCALEUP(dy); } compensate16x16_interpolate(&dct_codes[0 * 64], cur->y, ref->y, refh->y, refv->y, refhv->y, tmp, 16 * i, 16 * j, dx, dy, edged_width, quarterpel, reduced_resolution, rounding); dx /= (int)(1 + quarterpel); dy /= (int)(1 + quarterpel); dx = (dx >> 1) + roundtab_79[dx & 0x3]; dy = (dy >> 1) + roundtab_79[dy & 0x3]; } else { // mode == MODE_INTER4V int k, sumx = 0, sumy = 0; const VECTOR * const mvs = (quarterpel ? mb->qmvs : mb->mvs); for (k = 0; k < 4; k++) { dx = mvs[k].x; dy = mvs[k].y; sumx += dx / (1 + quarterpel); sumy += dy / (1 + quarterpel); if (reduced_resolution){ dx = RRV_MV_SCALEUP(dx); dy = RRV_MV_SCALEUP(dy); } compensate8x8_interpolate(&dct_codes[k * 64], cur->y, ref->y, refh->y, refv->y, refhv->y, tmp, 16 * i + 8*(k&1), 16 * j + 8*(k>>1), dx, dy, edged_width, quarterpel, reduced_resolution, rounding); } dx = (sumx >> 3) + roundtab_76[sumx & 0xf]; dy = (sumy >> 3) + roundtab_76[sumy & 0xf]; } CompensateChroma(dx, dy, i, j, cur, ref, tmp, &dct_codes[4 * 64], edged_width / 2, rounding, reduced_resolution); } void MBMotionCompensationBVOP(MBParam * pParam, MACROBLOCK * const mb, const uint32_t i, const uint32_t j, IMAGE * const cur, const IMAGE * const f_ref, const IMAGE * const f_refh, const IMAGE * const f_refv, const IMAGE * const f_refhv, const IMAGE * const b_ref, const IMAGE * const b_refh, const IMAGE * const b_refv, const IMAGE * const b_refhv, int16_t * dct_codes) { const uint32_t edged_width = pParam->edged_width; int32_t dx, dy, b_dx, b_dy, sumx, sumy, b_sumx, b_sumy; int k; const int quarterpel = pParam->m_quarterpel; const uint8_t * ptr1, * ptr2; uint8_t * const tmp = f_refv->u; const VECTOR * const fmvs = (quarterpel ? mb->qmvs : mb->mvs); const VECTOR * const bmvs = (quarterpel ? mb->b_qmvs : mb->b_mvs); switch (mb->mode) { case MODE_FORWARD: dx = fmvs->x; dy = fmvs->y; compensate16x16_interpolate(&dct_codes[0 * 64], cur->y, f_ref->y, f_refh->y, f_refv->y, f_refhv->y, tmp, 16 * i, 16 * j, dx, dy, edged_width, quarterpel, 0, 0); if (quarterpel) { dx /= 2; dy /= 2; } CompensateChroma( (dx >> 1) + roundtab_79[dx & 0x3], (dy >> 1) + roundtab_79[dy & 0x3], i, j, cur, f_ref, tmp, &dct_codes[4 * 64], edged_width / 2, 0, 0); return; case MODE_BACKWARD: b_dx = bmvs->x; b_dy = bmvs->y; compensate16x16_interpolate_ro(&dct_codes[0 * 64], cur->y, b_ref->y, b_refh->y, b_refv->y, b_refhv->y, tmp, 16 * i, 16 * j, b_dx, b_dy, edged_width, quarterpel); if (quarterpel) { b_dx /= 2; b_dy /= 2; } CompensateChroma( (b_dx >> 1) + roundtab_79[b_dx & 0x3], (b_dy >> 1) + roundtab_79[b_dy & 0x3], i, j, cur, b_ref, tmp, &dct_codes[4 * 64], edged_width / 2, 0, 0); return; case MODE_INTERPOLATE: /* _could_ use DIRECT, but would be overkill (no 4MV there) */ case MODE_DIRECT_NO4V: dx = fmvs->x; dy = fmvs->y; b_dx = bmvs->x; b_dy = bmvs->y; if (quarterpel) { if ((dx&3) | (dy&3)) { interpolate16x16_quarterpel(tmp - i * 16 - j * 16 * edged_width, (uint8_t *) f_ref->y, tmp + 32, tmp + 64, tmp + 96, 16*i, 16*j, dx, dy, edged_width, 0); ptr1 = tmp; } else ptr1 = f_ref->y + (16*j + dy/4)*edged_width + 16*i + dx/4; // fullpixel position if ((b_dx&3) | (b_dy&3)) { interpolate16x16_quarterpel(tmp - i * 16 - j * 16 * edged_width + 16, (uint8_t *) b_ref->y, tmp + 32, tmp + 64, tmp + 96, 16*i, 16*j, b_dx, b_dy, edged_width, 0); ptr2 = tmp + 16; } else ptr2 = b_ref->y + (16*j + b_dy/4)*edged_width + 16*i + b_dx/4; // fullpixel position b_dx /= 2; b_dy /= 2; dx /= 2; dy /= 2; } else { ptr1 = get_ref(f_ref->y, f_refh->y, f_refv->y, f_refhv->y, i, j, 16, dx, dy, edged_width); ptr2 = get_ref(b_ref->y, b_refh->y, b_refv->y, b_refhv->y, i, j, 16, b_dx, b_dy, edged_width); } for (k = 0; k < 4; k++) transfer_8to16sub2(&dct_codes[k * 64], cur->y + (i * 16+(k&1)*8) + (j * 16+((k>>1)*8)) * edged_width, ptr1 + (k&1)*8 + (k>>1)*8*edged_width, ptr2 + (k&1)*8 + (k>>1)*8*edged_width, edged_width); dx = (dx >> 1) + roundtab_79[dx & 0x3]; dy = (dy >> 1) + roundtab_79[dy & 0x3]; b_dx = (b_dx >> 1) + roundtab_79[b_dx & 0x3]; b_dy = (b_dy >> 1) + roundtab_79[b_dy & 0x3]; break; default: // MODE_DIRECT sumx = sumy = b_sumx = b_sumy = 0; for (k = 0; k < 4; k++) { dx = fmvs[k].x; dy = fmvs[k].y; b_dx = bmvs[k].x; b_dy = bmvs[k].y; if (quarterpel) { sumx += dx/2; sumy += dy/2; b_sumx += b_dx/2; b_sumy += b_dy/2; if ((dx&3) | (dy&3)) { interpolate8x8_quarterpel(tmp - (i * 16+(k&1)*8) - (j * 16+((k>>1)*8)) * edged_width, (uint8_t *) f_ref->y, tmp + 32, tmp + 64, tmp + 96, 16*i + (k&1)*8, 16*j + (k>>1)*8, dx, dy, edged_width, 0); ptr1 = tmp; } else ptr1 = f_ref->y + (16*j + (k>>1)*8 + dy/4)*edged_width + 16*i + (k&1)*8 + dx/4; if ((b_dx&3) | (b_dy&3)) { interpolate8x8_quarterpel(tmp - (i * 16+(k&1)*8) - (j * 16+((k>>1)*8)) * edged_width + 16, (uint8_t *) b_ref->y, tmp + 16, tmp + 32, tmp + 48, 16*i + (k&1)*8, 16*j + (k>>1)*8, b_dx, b_dy, edged_width, 0); ptr2 = tmp + 16; } else ptr2 = b_ref->y + (16*j + (k>>1)*8 + b_dy/4)*edged_width + 16*i + (k&1)*8 + b_dx/4; } else { sumx += dx; sumy += dy; b_sumx += b_dx; b_sumy += b_dy; ptr1 = get_ref(f_ref->y, f_refh->y, f_refv->y, f_refhv->y, 2*i + (k&1), 2*j + (k>>1), 8, dx, dy, edged_width); ptr2 = get_ref(b_ref->y, b_refh->y, b_refv->y, b_refhv->y, 2*i + (k&1), 2*j + (k>>1), 8, b_dx, b_dy, edged_width); } transfer_8to16sub2(&dct_codes[k * 64], cur->y + (i * 16+(k&1)*8) + (j * 16+((k>>1)*8)) * edged_width, ptr1, ptr2, edged_width); } dx = (sumx >> 3) + roundtab_76[sumx & 0xf]; dy = (sumy >> 3) + roundtab_76[sumy & 0xf]; b_dx = (b_sumx >> 3) + roundtab_76[b_sumx & 0xf]; b_dy = (b_sumy >> 3) + roundtab_76[b_sumy & 0xf]; break; } // uv block-based chroma interpolation for direct and interpolate modes transfer_8to16sub2(&dct_codes[4 * 64], cur->u + (j * 8) * edged_width / 2 + (i * 8), interpolate8x8_switch2(tmp, b_ref->u, 8 * i, 8 * j, b_dx, b_dy, edged_width / 2, 0), interpolate8x8_switch2(tmp + 8, f_ref->u, 8 * i, 8 * j, dx, dy, edged_width / 2, 0), edged_width / 2); transfer_8to16sub2(&dct_codes[5 * 64], cur->v + (j * 8) * edged_width / 2 + (i * 8), interpolate8x8_switch2(tmp, b_ref->v, 8 * i, 8 * j, b_dx, b_dy, edged_width / 2, 0), interpolate8x8_switch2(tmp + 8, f_ref->v, 8 * i, 8 * j, dx, dy, edged_width / 2, 0), edged_width / 2); } void generate_GMCparameters( const int num_wp, const int res, const WARPPOINTS *const warp, const int width, const int height, GMC_DATA *const gmc) { const int du0 = warp->duv[0].x; const int dv0 = warp->duv[0].y; const int du1 = warp->duv[1].x; const int dv1 = warp->duv[1].y; const int du2 = warp->duv[2].x; const int dv2 = warp->duv[2].y; gmc->W = width; gmc->H = height; gmc->rho = 4 - log2bin(res-1); // = {3,2,1,0} for res={2,4,8,16} gmc->alpha = log2bin(gmc->W-1); gmc->Ws = (1 << gmc->alpha); gmc->dxF = 16*gmc->Ws + RDIV( 8*gmc->Ws*du1, gmc->W ); gmc->dxG = RDIV( 8*gmc->Ws*dv1, gmc->W ); gmc->Fo = (res*du0 + 1) << (gmc->alpha+gmc->rho-1); gmc->Go = (res*dv0 + 1) << (gmc->alpha+gmc->rho-1); if (num_wp==2) { gmc->dyF = -gmc->dxG; gmc->dyG = gmc->dxF; } else if (num_wp==3) { gmc->beta = log2bin(gmc->H-1); gmc->Hs = (1 << gmc->beta); gmc->dyF = RDIV( 8*gmc->Hs*du2, gmc->H ); gmc->dyG = 16*gmc->Hs + RDIV( 8*gmc->Hs*dv2, gmc->H ); if (gmc->beta > gmc->alpha) { gmc->dxF <<= (gmc->beta - gmc->alpha); gmc->dxG <<= (gmc->beta - gmc->alpha); gmc->alpha = gmc->beta; gmc->Ws = 1<< gmc->beta; } else { gmc->dyF <<= gmc->alpha - gmc->beta; gmc->dyG <<= gmc->alpha - gmc->beta; } } gmc->cFo = gmc->dxF + gmc->dyF + (1 << (gmc->alpha+gmc->rho+1)); gmc->cFo += 16*gmc->Ws*(du0-1); gmc->cGo = gmc->dxG + gmc->dyG + (1 << (gmc->alpha+gmc->rho+1)); gmc->cGo += 16*gmc->Ws*(dv0-1); } void generate_GMCimage( const GMC_DATA *const gmc_data, // [input] precalculated data const IMAGE *const pRef, // [input] const int mb_width, const int mb_height, const int stride, const int stride2, const int fcode, // [input] some parameters... const int32_t quarterpel, // [input] for rounding avgMV const int reduced_resolution, // [input] ignored const int32_t rounding, // [input] for rounding image data MACROBLOCK *const pMBs, // [output] average motion vectors IMAGE *const pGMC) // [output] full warped image { unsigned int mj,mi; VECTOR avgMV; for (mj=0;mjW; const int H = gmc_data->H; const int rho = gmc_data->rho; const int alpha = gmc_data->alpha; const int rounder = ( 128 - (rounding<<(rho+rho)) ) << 16; const int dxF = gmc_data->dxF; const int dyF = gmc_data->dyF; const int dxG = gmc_data->dxG; const int dyG = gmc_data->dyG; uint8_t *dstY, *dstU, *dstV; int I,J; VECTOR avgMV = {0,0}; int32_t Fj, Gj; dstY = &pGMC->y[(mj*16)*stride+mi*16] + 16; Fj = gmc_data->Fo + dyF*mj*16 + dxF*mi*16; Gj = gmc_data->Go + dyG*mj*16 + dxG*mi*16; for (J=16; J>0; --J) { int32_t Fi, Gi; Fi = Fj; Fj += dyF; Gi = Gj; Gj += dyG; for (I=-16; I<0; ++I) { int32_t F, G; uint32_t ri, rj; F = ( Fi >> (alpha+rho) ) << rho; Fi += dxF; G = ( Gi >> (alpha+rho) ) << rho; Gi += dxG; avgMV.x += F; avgMV.y += G; ri = MTab[F&15]; rj = MTab[G&15]; F >>= 4; G >>= 4; if (F< -1) F=-1; else if (F>W) F=W; if (G< -1) G=-1; else if (G>H) G=H; { // MMX-like bilinear... const int offset = G*stride + F; uint32_t f0, f1; f0 = pRef->y[ offset +0 ]; f0 |= pRef->y[ offset +1 ] << 16; f1 = pRef->y[ offset+stride +0 ]; f1 |= pRef->y[ offset+stride +1 ] << 16; f0 = (ri*f0)>>16; f1 = (ri*f1) & 0x0fff0000; f0 |= f1; f0 = ( rj*f0 + rounder ) >> 24; dstY[I] = (uint8_t)f0; } } dstY += stride; } dstU = &pGMC->u[(mj*8)*stride2+mi*8] + 8; dstV = &pGMC->v[(mj*8)*stride2+mi*8] + 8; Fj = gmc_data->cFo + dyF*4 *mj*8 + dxF*4 *mi*8; Gj = gmc_data->cGo + dyG*4 *mj*8 + dxG*4 *mi*8; for (J=8; J>0; --J) { int32_t Fi, Gi; Fi = Fj; Fj += 4*dyF; Gi = Gj; Gj += 4*dyG; for (I=-8; I<0; ++I) { int32_t F, G; uint32_t ri, rj; F = ( Fi >> (alpha+rho+2) ) << rho; Fi += 4*dxF; G = ( Gi >> (alpha+rho+2) ) << rho; Gi += 4*dxG; ri = MTab[F&15]; rj = MTab[G&15]; F >>= 4; G >>= 4; if (F< -1) F=-1; else if (F>=W/2) F=W/2; if (G< -1) G=-1; else if (G>=H/2) G=H/2; { const int offset = G*stride2 + F; uint32_t f0, f1; f0 = pRef->u[ offset +0 ]; f0 |= pRef->u[ offset +1 ] << 16; f1 = pRef->u[ offset+stride2 +0 ]; f1 |= pRef->u[ offset+stride2 +1 ] << 16; f0 = (ri*f0)>>16; f1 = (ri*f1) & 0x0fff0000; f0 |= f1; f0 = ( rj*f0 + rounder ) >> 24; dstU[I] = (uint8_t)f0; f0 = pRef->v[ offset +0 ]; f0 |= pRef->v[ offset +1 ] << 16; f1 = pRef->v[ offset+stride2 +0 ]; f1 |= pRef->v[ offset+stride2 +1 ] << 16; f0 = (ri*f0)>>16; f1 = (ri*f1) & 0x0fff0000; f0 |= f1; f0 = ( rj*f0 + rounder ) >> 24; dstV[I] = (uint8_t)f0; } } dstU += stride2; dstV += stride2; } avgMV.x -= 16*((256*mi+120)<<4); // 120 = 15*16/2 avgMV.y -= 16*((256*mj+120)<<4); avgMV.x = RSHIFT( avgMV.x, (4+7-quarterpel) ); avgMV.y = RSHIFT( avgMV.y, (4+7-quarterpel) ); return avgMV; } #ifdef OLD_GRUEL_GMC void generate_GMCparameters( const int num_wp, // [input]: number of warppoints const int res, // [input]: resolution const WARPPOINTS *const warp, // [input]: warp points const int width, const int height, GMC_DATA *const gmc) // [output] precalculated parameters { /* We follow mainly two sources: The original standard, which is ugly, and the thesis from Andreas Dehnhardt, which is much nicer. Notation is: indices are written next to the variable, primes in the standard are denoted by a suffix 'p'. types are "c"=constant, "i"=input parameter, "f"=calculated, then fixed, "o"=output data, " "=other, "u" = unused, "p"=calc for every pixel type | variable name | ISO name (TeX-style) | value or range | usage ------------------------------------------------------------------------------------- c | H | H | [16 , ?] | image width (w/o edges) c | W | W | [16 , ?] | image height (w/o edges) c | i0 | i_0 | 0 | ref. point #1, X c | j0 | j_0 | 0 | ref. point #1, Y c | i1 | i_1 | W | ref. point #2, X c | j1 | j_1 | 0 | ref. point #2, Y cu | i2 | i_2 | 0 | ref. point #3, X cu | i2 | j_2 | H | ref. point #3, Y i | du0 | du[0] | [-16863,16863] | warp vector #1, Y i | dv0 | dv[0] | [-16863,16863] | warp vector #1, Y i | du1 | du[1] | [-16863,16863] | warp vector #2, Y i | dv1 | dv[1] | [-16863,16863] | warp vector #2, Y iu | du2 | du[2] | [-16863,16863] | warp vector #3, Y iu | dv2 | dv[2] | [-16863,16863] | warp vector #3, Y i | s | s | {2,4,8,16} | interpol. resolution f | sigma | - | log2(s) | X / s == X >> sigma f | r | r | =16/s | complementary res. f | rho | \rho | log2(r) | X / r == X >> rho f | i0s | i'_0 | | f | j0s | j'_0 | | f | i1s | i'_1 | | f | j1s | j'_1 | | f | i2s | i'_2 | | f | j2s | j'_2 | | f | alpha | \alpha | | 2^{alpha-1} < W <= 2^alpha f | beta | \beta | | 2^{beta-1} < H <= 2^beta f | Ws | W' | W = 2^{alpha} | scaled width f | Hs | H' | W = 2^{beta} | scaled height f | i1ss | i''_1 | "virtual sprite stuff" f | j1ss | j''_1 | "virtual sprite stuff" f | i2ss | i''_2 | "virtual sprite stuff" f | j2ss | j''_2 | "virtual sprite stuff" */ /* Some calculations are disabled because we only use 2 warppoints at the moment */ int du0 = warp->duv[0].x; int dv0 = warp->duv[0].y; int du1 = warp->duv[1].x; int dv1 = warp->duv[1].y; // int du2 = warp->duv[2].x; // int dv2 = warp->duv[2].y; gmc->num_wp = num_wp; gmc->s = res; /* scaling parameters 2,4,8 or 16 */ gmc->sigma = log2bin(res-1); /* log2bin(15)=4, log2bin(16)=5, log2bin(17)=5 */ gmc->r = 16/res; gmc->rho = 4 - gmc->sigma; /* = log2bin(r-1) */ gmc->W = width; gmc->H = height; /* fixed reference coordinates */ gmc->alpha = log2bin(gmc->W-1); gmc->Ws= 1<alpha; // gmc->beta = log2bin(gmc->H-1); // gmc->Hs= 1<beta; // printf("du0=%d dv0=%d du1=%d dv1=%d s=%d sigma=%d W=%d alpha=%d, Ws=%d, rho=%d\n",du0,dv0,du1,dv1,gmc->s,gmc->sigma,gmc->W,gmc->alpha,gmc->Ws,gmc->rho); /* i2s is only needed for num_wp >= 3, etc. */ /* the 's' values are in 1/s pel resolution */ gmc->i0s = res/2 * ( du0 ); gmc->j0s = res/2 * ( dv0 ); gmc->i1s = res/2 * (2*width + du1 + du0 ); gmc->j1s = res/2 * ( dv1 + dv0 ); // gmc->i2s = res/2 * ( du2 + du0 ); // gmc->j2s = res/2 * (2*height + dv2 + dv0 ); /* i2s and i2ss are only needed for num_wp == 3, etc. */ /* the 'ss' values are in 1/16 pel resolution */ gmc->i1ss = 16*gmc->Ws + ROUNDED_DIV(((gmc->W-gmc->Ws)*(gmc->r*gmc->i0s) + gmc->Ws*(gmc->r*gmc->i1s - 16*gmc->W)),gmc->W); gmc->j1ss = ROUNDED_DIV( ((gmc->W - gmc->Ws)*(gmc->r*gmc->j0s) + gmc->Ws*gmc->r*gmc->j1s) ,gmc->W ); // gmc->i2ss = ROUNDED_DIV( ((gmc->H - gmc->Hs)*(gmc->r*gmc->i0s) + gmc->Hs*(gmc->r*gmc->i2s)), gmc->H); // gmc->j2ss = 16*gmc->Hs + ROUNDED_DIV( ((gmc->H-gmc->Hs)*(gmc->r*gmc->j0s) + gmc->Ws*(gmc->r*gmc->j2s - 16*gmc->H)), gmc->H); return; } void generate_GMCimage( const GMC_DATA *const gmc_data, // [input] precalculated data const IMAGE *const pRef, // [input] const int mb_width, const int mb_height, const int stride, const int stride2, const int fcode, // [input] some parameters... const int32_t quarterpel, // [input] for rounding avgMV const int reduced_resolution, // [input] ignored const int32_t rounding, // [input] for rounding image data MACROBLOCK *const pMBs, // [output] average motion vectors IMAGE *const pGMC) // [output] full warped image { unsigned int mj,mi; VECTOR avgMV; for (mj=0;mjW; const int H = gmc_data->H; const int s = gmc_data->s; const int sigma = gmc_data->sigma; const int r = gmc_data->r; const int rho = gmc_data->rho; const int i0s = gmc_data->i0s; const int j0s = gmc_data->j0s; const int i1ss = gmc_data->i1ss; const int j1ss = gmc_data->j1ss; // const int i2ss = gmc_data->i2ss; // const int j2ss = gmc_data->j2ss; const int alpha = gmc_data->alpha; const int Ws = gmc_data->Ws; // const int beta = gmc_data->beta; // const int Hs = gmc_data->Hs; int I,J; VECTOR avgMV = {0,0}; for (J=16*mj;J<16*(mj+1);J++) for (I=16*mi;I<16*(mi+1);I++) { int F= i0s + ( ((-r*i0s+i1ss)*I + (r*j0s-j1ss)*J + (1<<(alpha+rho-1))) >> (alpha+rho) ); int G= j0s + ( ((-r*j0s+j1ss)*I + (-r*i0s+i1ss)*J + (1<<(alpha+rho-1))) >> (alpha+rho) ); /* this naive implementation (with lots of multiplications) isn't slower (rather faster) than working incremental. Don't ask me why... maybe the whole this is memory bound? */ const int ri= F & (s-1); // fractional part of pelwise MV X const int rj= G & (s-1); // fractional part of pelwise MV Y int Y00,Y01,Y10,Y11; /* unclipped values are used for avgMV */ avgMV.x += F-(I<>= sigma; G >>= sigma; /* clip values to be in range. Since we have edges, clip to 1 less than lower boundary this way positions F+1/G+1 are still right */ if (F< -1) F=-1; else if (F>W) F=W; /* W or W-1 doesn't matter, so save 1 subtract ;-) */ if (G< -1) G=-1; else if (G>H) G=H; /* dito */ Y00 = pRef->y[ G*stride + F ]; // Lumi values Y01 = pRef->y[ G*stride + F+1 ]; Y10 = pRef->y[ G*stride + F+stride ]; Y11 = pRef->y[ G*stride + F+stride+1 ]; /* bilinear interpolation */ Y00 = ((s-ri)*Y00 + ri*Y01); Y10 = ((s-ri)*Y10 + ri*Y11); Y00 = ((s-rj)*Y00 + rj*Y10 + s*s/2 - rounding ) >> (sigma+sigma); pGMC->y[J*stride+I] = (uint8_t)Y00; /* output 1 Y-pixel */ } /* doing chroma _here_ is even more stupid and slow, because won't be used until Compensation and most likely not even then (only if the block really _is_ GMC) */ for (J=8*mj;J<8*(mj+1);J++) /* this plays the role of j_c,i_c in the standard */ for (I=8*mi;I<8*(mi+1);I++) /* For I_c we have to use I_c = 4*i_c+1 ! */ { /* same positions for both chroma components, U=Cb and V=Cr */ int Fc=((-r*i0s+i1ss)*(4*I+1) + (r*j0s-j1ss)*(4*J+1) +2*Ws*r*i0s -16*Ws +(1<<(alpha+rho+1)))>>(alpha+rho+2); int Gc=((-r*j0s+j1ss)*(4*I+1) +(-r*i0s+i1ss)*(4*J+1) +2*Ws*r*j0s -16*Ws +(1<<(alpha+rho+1))) >>(alpha+rho+2); const int ri= Fc & (s-1); // fractional part of pelwise MV X const int rj= Gc & (s-1); // fractional part of pelwise MV Y int C00,C01,C10,C11; Fc >>= sigma; Gc >>= sigma; if (Fc< -1) Fc=-1; else if (Fc>=W/2) Fc=W/2; /* W or W-1 doesn't matter, so save 1 subtraction ;-) */ if (Gc< -1) Gc=-1; else if (Gc>=H/2) Gc=H/2; /* dito */ /* now calculate U data */ C00 = pRef->u[ Gc*stride2 + Fc ]; // chroma-value Cb C01 = pRef->u[ Gc*stride2 + Fc+1 ]; C10 = pRef->u[ (Gc+1)*stride2 + Fc ]; C11 = pRef->u[ (Gc+1)*stride2 + Fc+1 ]; /* bilinear interpolation */ C00 = ((s-ri)*C00 + ri*C01); C10 = ((s-ri)*C10 + ri*C11); C00 = ((s-rj)*C00 + rj*C10 + s*s/2 - rounding ) >> (sigma+sigma); pGMC->u[J*stride2+I] = (uint8_t)C00; /* output 1 U-pixel */ /* now calculate V data */ C00 = pRef->v[ Gc*stride2 + Fc ]; // chroma-value Cr C01 = pRef->v[ Gc*stride2 + Fc+1 ]; C10 = pRef->v[ (Gc+1)*stride2 + Fc ]; C11 = pRef->v[ (Gc+1)*stride2 + Fc+1 ]; /* bilinear interpolation */ C00 = ((s-ri)*C00 + ri*C01); C10 = ((s-ri)*C10 + ri*C11); C00 = ((s-rj)*C00 + rj*C10 + s*s/2 - rounding ) >> (sigma+sigma); pGMC->v[J*stride2+I] = (uint8_t)C00; /* output 1 V-pixel */ } /* The average vector is rounded from 1/s-pel to 1/2 or 1/4 using the '//' operator*/ avgMV.x = RSHIFT( avgMV.x, (sigma+7-quarterpel) ); avgMV.y = RSHIFT( avgMV.y, (sigma+7-quarterpel) ); /* ^^^^ this is the way MS Reference Software does it */ return avgMV; /* clipping to fcode area is done outside! */ } #endif