generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
vpx/vp9/encoder/vp9_pickmode.c
Jingning Han ccba289f8d Fast computation path for forward transform and quantization 
This commit enables a fast path computational flow for forward
transformation. It checks the sse and variance of prediction
residuals and decides if the quantized coefficients are all
zero, dc only, or more. It then selects the corresponding coding
path in the forward transformation and quantization stage.

It is currently enabled in rtc coding mode. Will do it for rd
coding mode next.

In speed -6, the runtime for pedestrian_area 1080p at 1000 kbps
goes down from 14234 ms to 13704 ms, i.e., about 4% speed-up.
Overall coding performance for rtc set is changed by -0.18%.

Change-Id: I0452da1786d59bc8bcbe0a35fdae9f623d1d44e1
2014-06-12 11:10:54 -07:00

494 lines
18 KiB
C
Raw Blame History

/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <assert.h>
#include <limits.h>
#include <math.h>
#include <stdio.h>
#include "./vp9_rtcd.h"
#include "vpx_mem/vpx_mem.h"
#include "vp9/common/vp9_common.h"
#include "vp9/common/vp9_mvref_common.h"
#include "vp9/common/vp9_reconinter.h"
#include "vp9/common/vp9_reconintra.h"
#include "vp9/encoder/vp9_encoder.h"
#include "vp9/encoder/vp9_ratectrl.h"
#include "vp9/encoder/vp9_rdopt.h"
static void full_pixel_motion_search(VP9_COMP *cpi, MACROBLOCK *x,
BLOCK_SIZE bsize, int mi_row, int mi_col,
int_mv *tmp_mv, int *rate_mv) {
MACROBLOCKD *xd = &x->e_mbd;
MB_MODE_INFO *mbmi = &xd->mi[0]->mbmi;
struct buf_2d backup_yv12[MAX_MB_PLANE] = {{0, 0}};
int step_param;
int sadpb = x->sadperbit16;
MV mvp_full;
int ref = mbmi->ref_frame[0];
const MV ref_mv = mbmi->ref_mvs[ref][0].as_mv;
int i;
int tmp_col_min = x->mv_col_min;
int tmp_col_max = x->mv_col_max;
int tmp_row_min = x->mv_row_min;
int tmp_row_max = x->mv_row_max;
const YV12_BUFFER_CONFIG *scaled_ref_frame = vp9_get_scaled_ref_frame(cpi,
ref);
if (scaled_ref_frame) {
int i;
// Swap out the reference frame for a version that's been scaled to
// match the resolution of the current frame, allowing the existing
// motion search code to be used without additional modifications.
for (i = 0; i < MAX_MB_PLANE; i++)
backup_yv12[i] = xd->plane[i].pre[0];
vp9_setup_pre_planes(xd, 0, scaled_ref_frame, mi_row, mi_col, NULL);
}
vp9_set_mv_search_range(x, &ref_mv);
// TODO(jingning) exploiting adaptive motion search control in non-RD
// mode decision too.
step_param = 6;
for (i = LAST_FRAME; i <= LAST_FRAME && cpi->common.show_frame; ++i) {
if ((x->pred_mv_sad[ref] >> 3) > x->pred_mv_sad[i]) {
tmp_mv->as_int = INVALID_MV;
if (scaled_ref_frame) {
int i;
for (i = 0; i < MAX_MB_PLANE; i++)
xd->plane[i].pre[0] = backup_yv12[i];
}
return;
}
}
assert(x->mv_best_ref_index[ref] <= 2);
if (x->mv_best_ref_index[ref] < 2)
mvp_full = mbmi->ref_mvs[ref][x->mv_best_ref_index[ref]].as_mv;
else
mvp_full = x->pred_mv[ref];
mvp_full.col >>= 3;
mvp_full.row >>= 3;
full_pixel_search(cpi, x, bsize, &mvp_full, step_param, sadpb, &ref_mv,
&tmp_mv->as_mv, INT_MAX, 0);
x->mv_col_min = tmp_col_min;
x->mv_col_max = tmp_col_max;
x->mv_row_min = tmp_row_min;
x->mv_row_max = tmp_row_max;
if (scaled_ref_frame) {
int i;
for (i = 0; i < MAX_MB_PLANE; i++)
xd->plane[i].pre[0] = backup_yv12[i];
}
// calculate the bit cost on motion vector
mvp_full.row = tmp_mv->as_mv.row * 8;
mvp_full.col = tmp_mv->as_mv.col * 8;
*rate_mv = vp9_mv_bit_cost(&mvp_full, &ref_mv,
x->nmvjointcost, x->mvcost, MV_COST_WEIGHT);
}
static void sub_pixel_motion_search(VP9_COMP *cpi, MACROBLOCK *x,
BLOCK_SIZE bsize, int mi_row, int mi_col,
MV *tmp_mv) {
MACROBLOCKD *xd = &x->e_mbd;
MB_MODE_INFO *mbmi = &xd->mi[0]->mbmi;
struct buf_2d backup_yv12[MAX_MB_PLANE] = {{0, 0}};
int ref = mbmi->ref_frame[0];
MV ref_mv = mbmi->ref_mvs[ref][0].as_mv;
int dis;
const YV12_BUFFER_CONFIG *scaled_ref_frame = vp9_get_scaled_ref_frame(cpi,
ref);
if (scaled_ref_frame) {
int i;
// Swap out the reference frame for a version that's been scaled to
// match the resolution of the current frame, allowing the existing
// motion search code to be used without additional modifications.
for (i = 0; i < MAX_MB_PLANE; i++)
backup_yv12[i] = xd->plane[i].pre[0];
vp9_setup_pre_planes(xd, 0, scaled_ref_frame, mi_row, mi_col, NULL);
}
cpi->find_fractional_mv_step(x, tmp_mv, &ref_mv,
cpi->common.allow_high_precision_mv,
x->errorperbit,
&cpi->fn_ptr[bsize],
cpi->sf.subpel_force_stop,
cpi->sf.subpel_iters_per_step,
x->nmvjointcost, x->mvcost,
&dis, &x->pred_sse[ref]);
if (scaled_ref_frame) {
int i;
for (i = 0; i < MAX_MB_PLANE; i++)
xd->plane[i].pre[0] = backup_yv12[i];
}
x->pred_mv[ref] = *tmp_mv;
}
static void model_rd_for_sb_y(VP9_COMP *cpi, BLOCK_SIZE bsize,
MACROBLOCK *x, MACROBLOCKD *xd,
int *out_rate_sum, int64_t *out_dist_sum,
unsigned int *var_y, unsigned int *sse_y) {
// Note our transform coeffs are 8 times an orthogonal transform.
// Hence quantizer step is also 8 times. To get effective quantizer
// we need to divide by 8 before sending to modeling function.
unsigned int sse;
int rate;
int64_t dist;
struct macroblock_plane *const p = &x->plane[0];
struct macroblockd_plane *const pd = &xd->plane[0];
const int quant = pd->dequant[1];
unsigned int var = cpi->fn_ptr[bsize].vf(p->src.buf, p->src.stride,
pd->dst.buf, pd->dst.stride, &sse);
*var_y = var;
*sse_y = sse;
if (sse < pd->dequant[0] * pd->dequant[0] >> 6)
x->skip_txfm = 1;
else if (var < quant * quant >> 6)
x->skip_txfm = 2;
else
x->skip_txfm = 0;
// TODO(jingning) This is a temporary solution to account for frames with
// light changes. Need to customize the rate-distortion modeling for non-RD
// mode decision.
if ((sse >> 3) > var)
sse = var;
vp9_model_rd_from_var_lapndz(var + sse, 1 << num_pels_log2_lookup[bsize],
quant >> 3, &rate, &dist);
*out_rate_sum = rate;
*out_dist_sum = dist << 3;
}
// TODO(jingning) placeholder for inter-frame non-RD mode decision.
// this needs various further optimizations. to be continued..
int64_t vp9_pick_inter_mode(VP9_COMP *cpi, MACROBLOCK *x,
const TileInfo *const tile,
int mi_row, int mi_col,
int *returnrate,
int64_t *returndistortion,
BLOCK_SIZE bsize) {
MACROBLOCKD *xd = &x->e_mbd;
MB_MODE_INFO *mbmi = &xd->mi[0]->mbmi;
struct macroblock_plane *const p = &x->plane[0];
struct macroblockd_plane *const pd = &xd->plane[0];
PREDICTION_MODE this_mode, best_mode = ZEROMV;
MV_REFERENCE_FRAME ref_frame, best_ref_frame = LAST_FRAME;
INTERP_FILTER best_pred_filter = EIGHTTAP;
int_mv frame_mv[MB_MODE_COUNT][MAX_REF_FRAMES];
struct buf_2d yv12_mb[4][MAX_MB_PLANE];
static const int flag_list[4] = { 0, VP9_LAST_FLAG, VP9_GOLD_FLAG,
VP9_ALT_FLAG };
int64_t best_rd = INT64_MAX;
int64_t this_rd = INT64_MAX;
int skip_txfm = 0;
int rate = INT_MAX;
int64_t dist = INT64_MAX;
// var_y and sse_y are saved to be used in skipping checking
unsigned int var_y = UINT_MAX;
unsigned int sse_y = UINT_MAX;
VP9_COMMON *cm = &cpi->common;
int intra_cost_penalty = 20 * vp9_dc_quant(cm->base_qindex, cm->y_dc_delta_q);
const int64_t inter_mode_thresh = RDCOST(x->rdmult, x->rddiv,
intra_cost_penalty, 0);
const int64_t intra_mode_cost = 50;
unsigned char segment_id = mbmi->segment_id;
const int *const rd_threshes = cpi->rd.threshes[segment_id][bsize];
const int *const rd_thresh_freq_fact = cpi->rd.thresh_freq_fact[bsize];
// Mode index conversion form THR_MODES to PREDICTION_MODE for a ref frame.
int mode_idx[MB_MODE_COUNT] = {0};
INTERP_FILTER filter_ref = SWITCHABLE;
int bsl = mi_width_log2_lookup[bsize];
const int pred_filter_search = (((mi_row + mi_col) >> bsl) +
get_chessboard_index(cm)) % 2;
x->skip_encode = cpi->sf.skip_encode_frame && x->q_index < QIDX_SKIP_THRESH;
x->skip = 0;
// initialize mode decisions
*returnrate = INT_MAX;
*returndistortion = INT64_MAX;
vpx_memset(mbmi, 0, sizeof(MB_MODE_INFO));
mbmi->sb_type = bsize;
mbmi->ref_frame[0] = NONE;
mbmi->ref_frame[1] = NONE;
mbmi->tx_size = MIN(max_txsize_lookup[bsize],
tx_mode_to_biggest_tx_size[cpi->common.tx_mode]);
mbmi->interp_filter = cpi->common.interp_filter == SWITCHABLE ?
EIGHTTAP : cpi->common.interp_filter;
mbmi->skip = 0;
mbmi->segment_id = segment_id;
for (ref_frame = LAST_FRAME; ref_frame <= LAST_FRAME ; ++ref_frame) {
x->pred_mv_sad[ref_frame] = INT_MAX;
if (cpi->ref_frame_flags & flag_list[ref_frame]) {
vp9_setup_buffer_inter(cpi, x, tile,
ref_frame, bsize, mi_row, mi_col,
frame_mv[NEARESTMV], frame_mv[NEARMV], yv12_mb);
}
frame_mv[NEWMV][ref_frame].as_int = INVALID_MV;
frame_mv[ZEROMV][ref_frame].as_int = 0;
}
if (xd->up_available)
filter_ref = xd->mi[-xd->mi_stride]->mbmi.interp_filter;
else if (xd->left_available)
filter_ref = xd->mi[-1]->mbmi.interp_filter;
for (ref_frame = LAST_FRAME; ref_frame <= LAST_FRAME ; ++ref_frame) {
if (!(cpi->ref_frame_flags & flag_list[ref_frame]))
continue;
// Select prediction reference frames.
xd->plane[0].pre[0] = yv12_mb[ref_frame][0];
clamp_mv2(&frame_mv[NEARESTMV][ref_frame].as_mv, xd);
clamp_mv2(&frame_mv[NEARMV][ref_frame].as_mv, xd);
mbmi->ref_frame[0] = ref_frame;
// Set conversion index for LAST_FRAME.
if (ref_frame == LAST_FRAME) {
mode_idx[NEARESTMV] = THR_NEARESTMV; // LAST_FRAME, NEARESTMV
mode_idx[NEARMV] = THR_NEARMV; // LAST_FRAME, NEARMV
mode_idx[ZEROMV] = THR_ZEROMV; // LAST_FRAME, ZEROMV
mode_idx[NEWMV] = THR_NEWMV; // LAST_FRAME, NEWMV
}
for (this_mode = NEARESTMV; this_mode <= NEWMV; ++this_mode) {
int rate_mv = 0;
if (!(cpi->sf.inter_mode_mask[bsize] & (1 << this_mode)))
continue;
if (rd_less_than_thresh(best_rd, rd_threshes[mode_idx[this_mode]],
rd_thresh_freq_fact[this_mode]))
continue;
if (this_mode == NEWMV) {
int rate_mode = 0;
if (this_rd < (int64_t)(1 << num_pels_log2_lookup[bsize]))
continue;
full_pixel_motion_search(cpi, x, bsize, mi_row, mi_col,
&frame_mv[NEWMV][ref_frame], &rate_mv);
if (frame_mv[NEWMV][ref_frame].as_int == INVALID_MV)
continue;
rate_mode = cpi->inter_mode_cost[mbmi->mode_context[ref_frame]]
[INTER_OFFSET(this_mode)];
if (RDCOST(x->rdmult, x->rddiv, rate_mv + rate_mode, 0) > best_rd)
continue;
sub_pixel_motion_search(cpi, x, bsize, mi_row, mi_col,
&frame_mv[NEWMV][ref_frame].as_mv);
}
if (this_mode != NEARESTMV)
if (frame_mv[this_mode][ref_frame].as_int ==
frame_mv[NEARESTMV][ref_frame].as_int)
continue;
mbmi->mode = this_mode;
mbmi->mv[0].as_int = frame_mv[this_mode][ref_frame].as_int;
// Search for the best prediction filter type, when the resulting
// motion vector is at sub-pixel accuracy level for luma component, i.e.,
// the last three bits are all zeros.
if ((this_mode == NEWMV || filter_ref == SWITCHABLE) &&
pred_filter_search &&
((mbmi->mv[0].as_mv.row & 0x07) != 0 ||
(mbmi->mv[0].as_mv.col & 0x07) != 0)) {
int pf_rate[3];
int64_t pf_dist[3];
unsigned int pf_var[3];
unsigned int pf_sse[3];
int64_t best_cost = INT64_MAX;
INTERP_FILTER best_filter = SWITCHABLE, filter;
for (filter = EIGHTTAP; filter <= EIGHTTAP_SHARP; ++filter) {
int64_t cost;
mbmi->interp_filter = filter;
vp9_build_inter_predictors_sby(xd, mi_row, mi_col, bsize);
model_rd_for_sb_y(cpi, bsize, x, xd, &pf_rate[filter],
&pf_dist[filter], &pf_var[filter], &pf_sse[filter]);
cost = RDCOST(x->rdmult, x->rddiv,
vp9_get_switchable_rate(cpi) + pf_rate[filter],
pf_dist[filter]);
if (cost < best_cost) {
best_filter = filter;
best_cost = cost;
skip_txfm = x->skip_txfm;
}
}
mbmi->interp_filter = best_filter;
rate = pf_rate[mbmi->interp_filter];
dist = pf_dist[mbmi->interp_filter];
var_y = pf_var[mbmi->interp_filter];
sse_y = pf_sse[mbmi->interp_filter];
x->skip_txfm = skip_txfm;
} else {
mbmi->interp_filter = (filter_ref == SWITCHABLE) ? EIGHTTAP: filter_ref;
vp9_build_inter_predictors_sby(xd, mi_row, mi_col, bsize);
model_rd_for_sb_y(cpi, bsize, x, xd, &rate, &dist, &var_y, &sse_y);
}
rate += rate_mv;
rate += cpi->inter_mode_cost[mbmi->mode_context[ref_frame]]
[INTER_OFFSET(this_mode)];
this_rd = RDCOST(x->rdmult, x->rddiv, rate, dist);
// Skipping checking: test to see if this block can be reconstructed by
// prediction only.
if (!x->in_active_map) {
x->skip = 1;
} else if (cpi->allow_encode_breakout && x->encode_breakout) {
const BLOCK_SIZE uv_size = get_plane_block_size(bsize, &xd->plane[1]);
unsigned int var = var_y, sse = sse_y;
// Skipping threshold for ac.
unsigned int thresh_ac;
// Skipping threshold for dc.
unsigned int thresh_dc;
// Set a maximum for threshold to avoid big PSNR loss in low bit rate
// case. Use extreme low threshold for static frames to limit skipping.
const unsigned int max_thresh = 36000;
// The encode_breakout input
const unsigned int min_thresh =
MIN(((unsigned int)x->encode_breakout << 4), max_thresh);
// Calculate threshold according to dequant value.
thresh_ac = (xd->plane[0].dequant[1] * xd->plane[0].dequant[1]) / 9;
thresh_ac = clamp(thresh_ac, min_thresh, max_thresh);
// Adjust ac threshold according to partition size.
thresh_ac >>= 8 - (b_width_log2_lookup[bsize] +
b_height_log2_lookup[bsize]);
thresh_dc = (xd->plane[0].dequant[0] * xd->plane[0].dequant[0] >> 6);
// Y skipping condition checking for ac and dc.
if (var <= thresh_ac && (sse - var) <= thresh_dc) {
unsigned int sse_u, sse_v;
unsigned int var_u, var_v;
// Skip u v prediction for less calculation, that won't affect
// result much.
var_u = cpi->fn_ptr[uv_size].vf(x->plane[1].src.buf,
x->plane[1].src.stride,
xd->plane[1].dst.buf,
xd->plane[1].dst.stride, &sse_u);
// U skipping condition checking
if ((var_u * 4 <= thresh_ac) && (sse_u - var_u <= thresh_dc)) {
var_v = cpi->fn_ptr[uv_size].vf(x->plane[2].src.buf,
x->plane[2].src.stride,
xd->plane[2].dst.buf,
xd->plane[2].dst.stride, &sse_v);
// V skipping condition checking
if ((var_v * 4 <= thresh_ac) && (sse_v - var_v <= thresh_dc)) {
x->skip = 1;
// The cost of skip bit needs to be added.
rate = rate_mv;
rate += cpi->inter_mode_cost[mbmi->mode_context[ref_frame]]
[INTER_OFFSET(this_mode)];
// More on this part of rate
// rate += vp9_cost_bit(vp9_get_skip_prob(cm, xd), 1);
// Scaling factor for SSE from spatial domain to frequency
// domain is 16. Adjust distortion accordingly.
// TODO(yunqingwang): In this function, only y-plane dist is
// calculated.
dist = (sse << 4); // + ((sse_u + sse_v) << 4);
this_rd = RDCOST(x->rdmult, x->rddiv, rate, dist);
// *disable_skip = 1;
}
}
}
}
if (this_rd < best_rd || x->skip) {
best_rd = this_rd;
*returnrate = rate;
*returndistortion = dist;
best_mode = this_mode;
best_pred_filter = mbmi->interp_filter;
best_ref_frame = ref_frame;
skip_txfm = x->skip_txfm;
}
if (x->skip)
break;
}
}
mbmi->mode = best_mode;
mbmi->interp_filter = best_pred_filter;
mbmi->ref_frame[0] = best_ref_frame;
mbmi->mv[0].as_int = frame_mv[best_mode][best_ref_frame].as_int;
xd->mi[0]->bmi[0].as_mv[0].as_int = mbmi->mv[0].as_int;
x->skip_txfm = skip_txfm;
// Perform intra prediction search, if the best SAD is above a certain
// threshold.
if (!x->skip && best_rd > inter_mode_thresh &&
bsize < cpi->sf.max_intra_bsize) {
for (this_mode = DC_PRED; this_mode <= DC_PRED; ++this_mode) {
vp9_predict_intra_block(xd, 0, b_width_log2(bsize),
mbmi->tx_size, this_mode,
&p->src.buf[0], p->src.stride,
&pd->dst.buf[0], pd->dst.stride, 0, 0, 0);
model_rd_for_sb_y(cpi, bsize, x, xd, &rate, &dist, &var_y, &sse_y);
rate += cpi->mbmode_cost[this_mode];
rate += intra_cost_penalty;
this_rd = RDCOST(x->rdmult, x->rddiv, rate, dist);
if (this_rd + intra_mode_cost < best_rd) {
best_rd = this_rd;
*returnrate = rate;
*returndistortion = dist;
mbmi->mode = this_mode;
mbmi->ref_frame[0] = INTRA_FRAME;
mbmi->uv_mode = this_mode;
mbmi->mv[0].as_int = INVALID_MV;
} else {
x->skip_txfm = skip_txfm;
}
}
}
return INT64_MAX;
}
Reference in New Issue View Git Blame Copy Permalink