#include <math.h>
#include "common/common.h"
-#include "common/cpu.h"
#include "set.h"
#include "analyse.h"
#include "ratecontrol.h"
#include "macroblock.h"
+#include "me.h"
-#if VISUALIZE
+#if HAVE_VISUALIZE
#include "common/visualize.h"
#endif
//#define DEBUG_MB_TYPE
-#define NALU_OVERHEAD 5 // startcode + NAL type costs 5 bytes per frame
-
#define bs_write_ue bs_write_ue_big
static int x264_encoder_frame_end( x264_t *h, x264_t *thread_current,
if( f_mse <= 0.0000000001 ) /* Max 100dB */
return 100;
- return (float)(-10.0 * log( f_mse ) / log( 10.0 ));
+ return -10.0 * log10( f_mse );
+}
+
+static float x264_ssim( float ssim )
+{
+ return -10.0 * log10( 1 - ssim );
}
static void x264_frame_dump( x264_t *h )
{
FILE *f = fopen( h->param.psz_dump_yuv, "r+b" );
- int i, y;
if( !f )
return;
/* Write the frame in display order */
fseek( f, (uint64_t)h->fdec->i_frame * h->param.i_height * h->param.i_width * 3/2, SEEK_SET );
- for( i = 0; i < h->fdec->i_plane; i++ )
- for( y = 0; y < h->param.i_height >> !!i; y++ )
+ for( int i = 0; i < h->fdec->i_plane; i++ )
+ for( int y = 0; y < h->param.i_height >> !!i; y++ )
fwrite( &h->fdec->plane[i][y*h->fdec->i_stride[i]], 1, h->param.i_width >> !!i, f );
fclose( f );
}
int i_idr_pic_id, int i_frame, int i_qp )
{
x264_param_t *param = &h->param;
- int i;
- /* First we fill all field */
+ /* First we fill all fields */
sh->sps = sps;
sh->pps = pps;
sh->i_idr_pic_id = i_idr_pic_id;
/* poc stuff, fixed later */
- sh->i_poc_lsb = 0;
+ sh->i_poc = 0;
sh->i_delta_poc_bottom = 0;
sh->i_delta_poc[0] = 0;
sh->i_delta_poc[1] = 0;
sh->i_redundant_pic_cnt = 0;
- if( !h->mb.b_direct_auto_read )
+ h->mb.b_direct_auto_write = h->param.analyse.i_direct_mv_pred == X264_DIRECT_PRED_AUTO
+ && h->param.i_bframe
+ && ( h->param.rc.b_stat_write || !h->param.rc.b_stat_read );
+
+ if( !h->mb.b_direct_auto_read && sh->i_type == SLICE_TYPE_B )
{
- if( h->mb.b_direct_auto_write )
- sh->b_direct_spatial_mv_pred = ( h->stat.i_direct_score[1] > h->stat.i_direct_score[0] );
+ if( h->fref1[0]->i_poc_l0ref0 == h->fref0[0]->i_poc )
+ {
+ if( h->mb.b_direct_auto_write )
+ sh->b_direct_spatial_mv_pred = ( h->stat.i_direct_score[1] > h->stat.i_direct_score[0] );
+ else
+ sh->b_direct_spatial_mv_pred = ( param->analyse.i_direct_mv_pred == X264_DIRECT_PRED_SPATIAL );
+ }
else
- sh->b_direct_spatial_mv_pred = ( param->analyse.i_direct_mv_pred == X264_DIRECT_PRED_SPATIAL );
+ {
+ h->mb.b_direct_auto_write = 0;
+ sh->b_direct_spatial_mv_pred = 1;
+ }
}
/* else b_direct_spatial_mv_pred was read from the 2pass statsfile */
if( sh->b_ref_pic_list_reordering_l0 )
{
int pred_frame_num = i_frame;
- for( i = 0; i < h->i_ref0; i++ )
+ for( int i = 0; i < h->i_ref0; i++ )
{
int diff = h->fref0[i]->i_frame_num - pred_frame_num;
- if( diff == 0 )
- x264_log( h, X264_LOG_ERROR, "diff frame num == 0\n" );
sh->ref_pic_list_order[0][i].idc = ( diff > 0 );
- sh->ref_pic_list_order[0][i].arg = abs( diff ) - 1;
+ sh->ref_pic_list_order[0][i].arg = (abs(diff) - 1) & ((1 << sps->i_log2_max_frame_num) - 1);
pred_frame_num = h->fref0[i]->i_frame_num;
}
}
sh->b_sp_for_swidth = 0;
sh->i_qs_delta = 0;
+ int deblock_thresh = i_qp + 2 * X264_MIN(param->i_deblocking_filter_alphac0, param->i_deblocking_filter_beta);
/* If effective qp <= 15, deblocking would have no effect anyway */
- if( param->b_deblocking_filter
- && ( h->mb.b_variable_qp
- || 15 < i_qp + 2 * X264_MIN(param->i_deblocking_filter_alphac0, param->i_deblocking_filter_beta) ) )
- {
- sh->i_disable_deblocking_filter_idc = 0;
- }
+ if( param->b_deblocking_filter && (h->mb.b_variable_qp || 15 < deblock_thresh ) )
+ sh->i_disable_deblocking_filter_idc = param->b_sliced_threads ? 2 : 0;
else
- {
sh->i_disable_deblocking_filter_idc = 1;
- }
sh->i_alpha_c0_offset = param->i_deblocking_filter_alphac0 << 1;
sh->i_beta_offset = param->i_deblocking_filter_beta << 1;
}
static void x264_slice_header_write( bs_t *s, x264_slice_header_t *sh, int i_nal_ref_idc )
{
- int i;
-
if( sh->b_mbaff )
{
assert( sh->i_first_mb % (2*sh->sps->i_mb_width) == 0 );
}
if( sh->i_idr_pic_id >= 0 ) /* NAL IDR */
- {
bs_write_ue( s, sh->i_idr_pic_id );
- }
if( sh->sps->i_poc_type == 0 )
{
- bs_write( s, sh->sps->i_log2_max_poc_lsb, sh->i_poc_lsb & ((1<<sh->sps->i_log2_max_poc_lsb)-1) );
+ bs_write( s, sh->sps->i_log2_max_poc_lsb, sh->i_poc & ((1<<sh->sps->i_log2_max_poc_lsb)-1) );
if( sh->pps->b_pic_order && !sh->b_field_pic )
- {
bs_write_se( s, sh->i_delta_poc_bottom );
- }
}
else if( sh->sps->i_poc_type == 1 && !sh->sps->b_delta_pic_order_always_zero )
{
bs_write_se( s, sh->i_delta_poc[0] );
if( sh->pps->b_pic_order && !sh->b_field_pic )
- {
bs_write_se( s, sh->i_delta_poc[1] );
- }
}
if( sh->pps->b_redundant_pic_cnt )
- {
bs_write_ue( s, sh->i_redundant_pic_cnt );
- }
if( sh->i_type == SLICE_TYPE_B )
- {
bs_write1( s, sh->b_direct_spatial_mv_pred );
- }
+
if( sh->i_type == SLICE_TYPE_P || sh->i_type == SLICE_TYPE_SP || sh->i_type == SLICE_TYPE_B )
{
bs_write1( s, sh->b_num_ref_idx_override );
{
bs_write_ue( s, sh->i_num_ref_idx_l0_active - 1 );
if( sh->i_type == SLICE_TYPE_B )
- {
bs_write_ue( s, sh->i_num_ref_idx_l1_active - 1 );
- }
}
}
bs_write1( s, sh->b_ref_pic_list_reordering_l0 );
if( sh->b_ref_pic_list_reordering_l0 )
{
- for( i = 0; i < sh->i_num_ref_idx_l0_active; i++ )
+ for( int i = 0; i < sh->i_num_ref_idx_l0_active; i++ )
{
bs_write_ue( s, sh->ref_pic_list_order[0][i].idc );
bs_write_ue( s, sh->ref_pic_list_order[0][i].arg );
-
}
bs_write_ue( s, 3 );
}
bs_write1( s, sh->b_ref_pic_list_reordering_l1 );
if( sh->b_ref_pic_list_reordering_l1 )
{
- for( i = 0; i < sh->i_num_ref_idx_l1_active; i++ )
+ for( int i = 0; i < sh->i_num_ref_idx_l1_active; i++ )
{
bs_write_ue( s, sh->ref_pic_list_order[1][i].idc );
bs_write_ue( s, sh->ref_pic_list_order[1][i].arg );
/* pred_weight_table() */
bs_write_ue( s, sh->weight[0][0].i_denom );
bs_write_ue( s, sh->weight[0][1].i_denom );
- for( i = 0; i < sh->i_num_ref_idx_l0_active; i++ )
+ for( int i = 0; i < sh->i_num_ref_idx_l0_active; i++ )
{
int luma_weight_l0_flag = !!sh->weight[i][0].weightfn;
int chroma_weight_l0_flag = !!sh->weight[i][1].weightfn || !!sh->weight[i][2].weightfn;
bs_write1( s, chroma_weight_l0_flag );
if( chroma_weight_l0_flag )
{
- int j;
- for( j = 1; j < 3; j++ )
+ for( int j = 1; j < 3; j++ )
{
bs_write_se( s, sh->weight[i][j].i_scale );
bs_write_se( s, sh->weight[i][j].i_offset );
bs_write1( s, sh->i_mmco_command_count > 0 ); /* adaptive_ref_pic_marking_mode_flag */
if( sh->i_mmco_command_count > 0 )
{
- int i;
- for( i = 0; i < sh->i_mmco_command_count; i++ )
+ for( int i = 0; i < sh->i_mmco_command_count; i++ )
{
bs_write_ue( s, 1 ); /* mark short term ref as unused */
bs_write_ue( s, sh->mmco[i].i_difference_of_pic_nums - 1 );
}
if( sh->pps->b_cabac && sh->i_type != SLICE_TYPE_I )
- {
bs_write_ue( s, sh->i_cabac_init_idc );
- }
+
bs_write_se( s, sh->i_qp_delta ); /* slice qp delta */
if( sh->pps->b_deblocking_filter_control )
static int x264_bitstream_check_buffer( x264_t *h )
{
uint8_t *bs_bak = h->out.p_bitstream;
- if( ( h->param.b_cabac && (h->cabac.p_end - h->cabac.p < 2500) )
- || ( h->out.bs.p_end - h->out.bs.p < 2500 ) )
+ if( (h->param.b_cabac && (h->cabac.p_end - h->cabac.p < 2500)) ||
+ (h->out.bs.p_end - h->out.bs.p < 2500) )
{
- intptr_t delta;
- int i;
-
h->out.i_bitstream += 100000;
CHECKED_MALLOC( h->out.p_bitstream, h->out.i_bitstream );
h->mc.memcpy_aligned( h->out.p_bitstream, bs_bak, (h->out.i_bitstream - 100000) & ~15 );
- delta = h->out.p_bitstream - bs_bak;
+ intptr_t delta = h->out.p_bitstream - bs_bak;
h->out.bs.p_start += delta;
h->out.bs.p += delta;
h->cabac.p += delta;
h->cabac.p_end = h->out.p_bitstream + h->out.i_bitstream;
- for( i = 0; i <= h->out.i_nal; i++ )
+ for( int i = 0; i <= h->out.i_nal; i++ )
h->out.nal[i].p_payload += delta;
x264_free( bs_bak );
}
static int x264_validate_parameters( x264_t *h )
{
-#ifdef HAVE_MMX
+#if HAVE_MMX
+#ifdef __SSE__
if( !(x264_cpu_detect() & X264_CPU_SSE) )
{
x264_log( h, X264_LOG_ERROR, "your cpu does not support SSE1, but x264 was compiled with asm support\n");
+#else
+ if( !(x264_cpu_detect() & X264_CPU_MMXEXT) )
+ {
+ x264_log( h, X264_LOG_ERROR, "your cpu does not support MMXEXT, but x264 was compiled with asm support\n");
+#endif
x264_log( h, X264_LOG_ERROR, "to run x264, recompile without asm support (configure --disable-asm)\n");
return -1;
}
h->param.i_width, h->param.i_height );
return -1;
}
- if( h->param.i_csp != X264_CSP_I420 && h->param.i_csp != X264_CSP_YV12 )
+ int i_csp = h->param.i_csp & X264_CSP_MASK;
+ if( i_csp != X264_CSP_I420 && i_csp != X264_CSP_YV12 )
{
x264_log( h, X264_LOG_ERROR, "invalid CSP (only I420/YV12 supported)\n" );
return -1;
}
if( h->param.i_threads == X264_THREADS_AUTO )
- h->param.i_threads = x264_cpu_num_processors() * 3/2;
+ h->param.i_threads = x264_cpu_num_processors() * (h->param.b_sliced_threads?2:3)/2;
h->param.i_threads = x264_clip3( h->param.i_threads, 1, X264_THREAD_MAX );
if( h->param.i_threads > 1 )
{
-#ifndef HAVE_PTHREAD
+#if !HAVE_PTHREAD
x264_log( h, X264_LOG_WARNING, "not compiled with pthread support!\n");
h->param.i_threads = 1;
#endif
+ /* Avoid absurdly small thread slices as they can reduce performance
+ * and VBV compliance. Capped at an arbitrary 4 rows per thread. */
+ if( h->param.b_sliced_threads )
+ {
+ int max_threads = (h->param.i_height+15)/16 / 4;
+ h->param.i_threads = X264_MIN( h->param.i_threads, max_threads );
+ }
}
+ else
+ h->param.b_sliced_threads = 0;
+ h->i_thread_frames = h->param.b_sliced_threads ? 1 : h->param.i_threads;
if( h->param.b_interlaced )
{
x264_log( h, X264_LOG_WARNING, "interlace + me=esa is not implemented\n" );
h->param.analyse.i_me_method = X264_ME_UMH;
}
- if( h->param.analyse.i_direct_mv_pred > X264_DIRECT_PRED_SPATIAL )
- {
- x264_log( h, X264_LOG_WARNING, "interlace + direct=temporal is not implemented\n" );
- h->param.analyse.i_direct_mv_pred = X264_DIRECT_PRED_SPATIAL;
- }
if( h->param.analyse.i_weighted_pred > 0 )
{
x264_log( h, X264_LOG_WARNING, "interlace + weightp is not implemented\n" );
h->param.analyse.i_trellis = 0;
h->param.analyse.b_fast_pskip = 0;
h->param.analyse.i_noise_reduction = 0;
- h->param.analyse.f_psy_rd = 0;
+ h->param.analyse.b_psy = 0;
h->param.i_bframe = 0;
/* 8x8dct is not useful at all in CAVLC lossless */
if( !h->param.b_cabac )
if( h->param.rc.i_rc_method == X264_RC_CQP )
{
float qp_p = h->param.rc.i_qp_constant;
- float qp_i = qp_p - 6*log(h->param.rc.f_ip_factor)/log(2);
- float qp_b = qp_p + 6*log(h->param.rc.f_pb_factor)/log(2);
+ float qp_i = qp_p - 6*log2f( h->param.rc.f_ip_factor );
+ float qp_b = qp_p + 6*log2f( h->param.rc.f_pb_factor );
h->param.rc.i_qp_min = x264_clip3( (int)(X264_MIN3( qp_p, qp_i, qp_b )), 0, 51 );
h->param.rc.i_qp_max = x264_clip3( (int)(X264_MAX3( qp_p, qp_i, qp_b ) + .999), 0, 51 );
h->param.rc.i_aq_mode = 0;
}
h->param.rc.i_qp_max = x264_clip3( h->param.rc.i_qp_max, 0, 51 );
h->param.rc.i_qp_min = x264_clip3( h->param.rc.i_qp_min, 0, h->param.rc.i_qp_max );
+ if( h->param.rc.i_vbv_buffer_size )
+ {
+ if( h->param.rc.i_rc_method == X264_RC_CQP )
+ {
+ x264_log( h, X264_LOG_WARNING, "VBV is incompatible with constant QP, ignored.\n" );
+ h->param.rc.i_vbv_max_bitrate = 0;
+ h->param.rc.i_vbv_buffer_size = 0;
+ }
+ else if( h->param.rc.i_vbv_max_bitrate == 0 )
+ {
+ if( h->param.rc.i_rc_method == X264_RC_ABR )
+ {
+ x264_log( h, X264_LOG_WARNING, "VBV maxrate unspecified, assuming CBR\n" );
+ h->param.rc.i_vbv_max_bitrate = h->param.rc.i_bitrate;
+ }
+ else
+ {
+ x264_log( h, X264_LOG_WARNING, "VBV bufsize set but maxrate unspecified, ignored\n" );
+ h->param.rc.i_vbv_buffer_size = 0;
+ }
+ }
+ else if( h->param.rc.i_vbv_max_bitrate < h->param.rc.i_bitrate &&
+ h->param.rc.i_rc_method == X264_RC_ABR )
+ {
+ x264_log( h, X264_LOG_WARNING, "max bitrate less than average bitrate, assuming CBR\n" );
+ h->param.rc.i_vbv_max_bitrate = h->param.rc.i_bitrate;
+ }
+ }
+ else if( h->param.rc.i_vbv_max_bitrate )
+ {
+ x264_log( h, X264_LOG_WARNING, "VBV maxrate specified, but no bufsize, ignored\n" );
+ h->param.rc.i_vbv_max_bitrate = 0;
+ }
- int max_slices = (h->param.i_height+((16<<h->param.b_interlaced)-1))/(16<<h->param.b_interlaced);
- h->param.i_slice_count = x264_clip3( h->param.i_slice_count, 0, max_slices );
- h->param.i_slice_max_size = X264_MAX( h->param.i_slice_max_size, 0 );
- h->param.i_slice_max_mbs = X264_MAX( h->param.i_slice_max_mbs, 0 );
if( h->param.b_interlaced && h->param.i_slice_max_size )
{
x264_log( h, X264_LOG_WARNING, "interlaced + slice-max-size is not implemented\n" );
x264_log( h, X264_LOG_WARNING, "interlaced + slice-max-mbs is not implemented\n" );
h->param.i_slice_max_mbs = 0;
}
- if( h->param.i_slice_max_mbs || h->param.i_slice_max_size )
- h->param.i_slice_count = 0;
+ int max_slices = (h->param.i_height+((16<<h->param.b_interlaced)-1))/(16<<h->param.b_interlaced);
+ if( h->param.b_sliced_threads )
+ h->param.i_slice_count = x264_clip3( h->param.i_threads, 0, max_slices );
+ else
+ {
+ h->param.i_slice_count = x264_clip3( h->param.i_slice_count, 0, max_slices );
+ h->param.i_slice_max_size = X264_MAX( h->param.i_slice_max_size, 0 );
+ h->param.i_slice_max_mbs = X264_MAX( h->param.i_slice_max_mbs, 0 );
+ if( h->param.i_slice_max_mbs || h->param.i_slice_max_size )
+ h->param.i_slice_count = 0;
+ }
h->param.i_frame_reference = x264_clip3( h->param.i_frame_reference, 1, 16 );
if( h->param.i_keyint_max <= 0 )
h->param.i_keyint_max = 1;
if( h->param.i_scenecut_threshold < 0 )
h->param.i_scenecut_threshold = 0;
- h->param.i_keyint_min = x264_clip3( h->param.i_keyint_min, 1, h->param.i_keyint_max/2+1 );
if( !h->param.analyse.i_subpel_refine && h->param.analyse.i_direct_mv_pred > X264_DIRECT_PRED_SPATIAL )
{
x264_log( h, X264_LOG_WARNING, "subme=0 + direct=temporal is not supported\n" );
}
h->param.i_bframe = x264_clip3( h->param.i_bframe, 0, X264_BFRAME_MAX );
if( h->param.i_keyint_max == 1 )
+ {
h->param.i_bframe = 0;
+ h->param.b_intra_refresh = 0;
+ }
h->param.i_bframe_bias = x264_clip3( h->param.i_bframe_bias, -90, 100 );
if( h->param.i_bframe <= 1 )
h->param.i_bframe_pyramid = X264_B_PYRAMID_NONE;
h->param.analyse.i_direct_mv_pred = 0;
h->param.analyse.b_weighted_bipred = 0;
}
+ if( h->param.b_intra_refresh && h->param.i_bframe_pyramid == X264_B_PYRAMID_NORMAL )
+ {
+ x264_log( h, X264_LOG_WARNING, "b-pyramid normal + intra-refresh is not supported\n" );
+ h->param.i_bframe_pyramid = X264_B_PYRAMID_STRICT;
+ }
+ if( h->param.b_intra_refresh && h->param.i_frame_reference > 1 )
+ {
+ x264_log( h, X264_LOG_WARNING, "ref > 1 + intra-refresh is not supported\n" );
+ h->param.i_frame_reference = 1;
+ }
+ if( h->param.i_keyint_min == X264_KEYINT_MIN_AUTO )
+ h->param.i_keyint_min = h->param.i_keyint_max / 10;
+ h->param.i_keyint_min = x264_clip3( h->param.i_keyint_min, 1, h->param.i_keyint_max/2+1 );
h->param.rc.i_lookahead = x264_clip3( h->param.rc.i_lookahead, 0, X264_LOOKAHEAD_MAX );
{
int maxrate = X264_MAX( h->param.rc.i_vbv_max_bitrate, h->param.rc.i_bitrate );
h->param.rc.i_lookahead = X264_MIN( h->param.rc.i_lookahead, X264_MAX( h->param.i_keyint_max, bufsize*fps ) );
}
+ if( !h->param.i_timebase_num || !h->param.i_timebase_den )
+ {
+ h->param.i_timebase_num = h->param.i_fps_den;
+ h->param.i_timebase_den = h->param.i_fps_num;
+ }
+
h->param.rc.f_qcompress = x264_clip3f( h->param.rc.f_qcompress, 0.0, 1.0 );
if( !h->param.rc.i_lookahead || h->param.i_keyint_max == 1 || h->param.rc.f_qcompress == 1 )
h->param.rc.b_mb_tree = 0;
if( h->param.rc.b_stat_read )
h->param.rc.i_lookahead = 0;
-#ifdef HAVE_PTHREAD
- if( h->param.i_sync_lookahead )
- h->param.i_sync_lookahead = x264_clip3( h->param.i_sync_lookahead, h->param.i_threads + h->param.i_bframe, X264_LOOKAHEAD_MAX );
- if( h->param.rc.b_stat_read || h->param.i_threads == 1 )
+#if HAVE_PTHREAD
+ if( h->param.i_sync_lookahead < 0 )
+ h->param.i_sync_lookahead = h->param.i_bframe + 1;
+ h->param.i_sync_lookahead = X264_MIN( h->param.i_sync_lookahead, X264_LOOKAHEAD_MAX );
+ if( h->param.rc.b_stat_read || h->i_thread_frames == 1 )
h->param.i_sync_lookahead = 0;
#else
h->param.i_sync_lookahead = 0;
#endif
- h->mb.b_direct_auto_write = h->param.analyse.i_direct_mv_pred == X264_DIRECT_PRED_AUTO
- && h->param.i_bframe
- && ( h->param.rc.b_stat_write || !h->param.rc.b_stat_read );
-
h->param.i_deblocking_filter_alphac0 = x264_clip3( h->param.i_deblocking_filter_alphac0, -6, 6 );
h->param.i_deblocking_filter_beta = x264_clip3( h->param.i_deblocking_filter_beta, -6, 6 );
h->param.analyse.i_luma_deadzone[0] = x264_clip3( h->param.analyse.i_luma_deadzone[0], 0, 32 );
/* Psy trellis has a similar effect. */
if( h->mb.i_psy_trellis )
h->param.analyse.i_chroma_qp_offset -= h->param.analyse.f_psy_trellis < 0.25 ? 1 : 2;
- else
- h->mb.i_psy_trellis = 0;
h->param.analyse.i_chroma_qp_offset = x264_clip3(h->param.analyse.i_chroma_qp_offset, -12, 12);
h->param.rc.i_aq_mode = x264_clip3( h->param.rc.i_aq_mode, 0, 2 );
h->param.rc.f_aq_strength = x264_clip3f( h->param.rc.f_aq_strength, 0, 3 );
h->param.rc.i_aq_mode = 1;
h->param.rc.f_aq_strength = 0;
}
- if( h->param.rc.b_mb_tree && h->param.i_bframe_pyramid )
- {
- x264_log( h, X264_LOG_WARNING, "b-pyramid + mb-tree is not supported\n" );
- h->param.i_bframe_pyramid = X264_B_PYRAMID_NONE;
- }
h->param.analyse.i_noise_reduction = x264_clip3( h->param.analyse.i_noise_reduction, 0, 1<<16 );
if( h->param.analyse.i_subpel_refine == 10 && (h->param.analyse.i_trellis != 2 || !h->param.rc.i_aq_mode) )
h->param.analyse.i_subpel_refine = 9;
if( !h->param.analyse.i_weighted_pred && h->param.rc.b_mb_tree && h->param.analyse.b_psy && !h->param.b_interlaced )
h->param.analyse.i_weighted_pred = X264_WEIGHTP_FAKE;
- if( h->param.i_threads > 1 )
+ if( h->i_thread_frames > 1 )
{
int r = h->param.analyse.i_mv_range_thread;
int r2;
// the rest is allocated to whichever thread is far enough ahead to use it.
// reserving more space increases quality for some videos, but costs more time
// in thread synchronization.
- int max_range = (h->param.i_height + X264_THREAD_HEIGHT) / h->param.i_threads - X264_THREAD_HEIGHT;
+ int max_range = (h->param.i_height + X264_THREAD_HEIGHT) / h->i_thread_frames - X264_THREAD_HEIGHT;
r = max_range / 2;
}
r = X264_MAX( r, h->param.analyse.i_me_range );
h->param.analyse.b_ssim = 0;
}
+ if( h->param.b_interlaced )
+ h->param.b_pic_struct = 1;
+
+ if( h->param.i_nal_hrd && !h->param.rc.i_vbv_buffer_size )
+ {
+ x264_log( h, X264_LOG_WARNING, "NAL HRD parameters require VBV parameters\n" );
+ h->param.i_nal_hrd = X264_NAL_HRD_NONE;
+ }
+
+ if( h->param.i_nal_hrd == X264_NAL_HRD_CBR &&
+ (h->param.rc.i_bitrate != h->param.rc.i_vbv_max_bitrate || !h->param.rc.i_vbv_max_bitrate) )
+ {
+ x264_log( h, X264_LOG_WARNING, "CBR HRD requires constant bitrate\n" );
+ h->param.i_nal_hrd = X264_NAL_HRD_VBR;
+ }
+
/* ensure the booleans are 0 or 1 so they can be used in math */
#define BOOLIFY(x) h->param.x = !!h->param.x
BOOLIFY( b_cabac );
+ BOOLIFY( b_constrained_intra );
BOOLIFY( b_deblocking_filter );
+ BOOLIFY( b_deterministic );
+ BOOLIFY( b_sliced_threads );
BOOLIFY( b_interlaced );
+ BOOLIFY( b_intra_refresh );
+ BOOLIFY( b_visualize );
+ BOOLIFY( b_aud );
+ BOOLIFY( b_repeat_headers );
+ BOOLIFY( b_annexb );
+ BOOLIFY( b_vfr_input );
+ BOOLIFY( b_pic_struct );
+ BOOLIFY( b_fake_interlaced );
BOOLIFY( analyse.b_transform_8x8 );
+ BOOLIFY( analyse.b_weighted_bipred );
BOOLIFY( analyse.b_chroma_me );
+ BOOLIFY( analyse.b_mixed_references );
BOOLIFY( analyse.b_fast_pskip );
+ BOOLIFY( analyse.b_dct_decimate );
+ BOOLIFY( analyse.b_psy );
+ BOOLIFY( analyse.b_psnr );
+ BOOLIFY( analyse.b_ssim );
BOOLIFY( rc.b_stat_write );
BOOLIFY( rc.b_stat_read );
BOOLIFY( rc.b_mb_tree );
memcpy( h->pixf.mbcmp_unaligned, satd ? h->pixf.satd : h->pixf.sad, sizeof(h->pixf.mbcmp_unaligned) );
h->pixf.intra_mbcmp_x3_16x16 = satd ? h->pixf.intra_satd_x3_16x16 : h->pixf.intra_sad_x3_16x16;
h->pixf.intra_mbcmp_x3_8x8c = satd ? h->pixf.intra_satd_x3_8x8c : h->pixf.intra_sad_x3_8x8c;
+ h->pixf.intra_mbcmp_x3_8x8 = satd ? h->pixf.intra_sa8d_x3_8x8 : h->pixf.intra_sad_x3_8x8;
h->pixf.intra_mbcmp_x3_4x4 = satd ? h->pixf.intra_satd_x3_4x4 : h->pixf.intra_sad_x3_4x4;
satd &= h->param.analyse.i_me_method == X264_ME_TESA;
memcpy( h->pixf.fpelcmp, satd ? h->pixf.satd : h->pixf.sad, sizeof(h->pixf.fpelcmp) );
/* VUI */
if( param->vui.i_sar_width > 0 && param->vui.i_sar_height > 0 )
{
- int i_w = param->vui.i_sar_width;
- int i_h = param->vui.i_sar_height;
- int old_w = h->param.vui.i_sar_width;
- int old_h = h->param.vui.i_sar_height;
+ uint32_t i_w = param->vui.i_sar_width;
+ uint32_t i_h = param->vui.i_sar_height;
+ uint32_t old_w = h->param.vui.i_sar_width;
+ uint32_t old_h = h->param.vui.i_sar_height;
x264_reduce_fraction( &i_w, &i_h );
{
x264_t *h;
char buf[1000], *p;
- int i, qp, i_slicetype_length;
+ int qp, i_slicetype_length;
CHECKED_MALLOCZERO( h, sizeof(x264_t) );
if( h->param.rc.psz_stat_in )
h->param.rc.psz_stat_in = strdup( h->param.rc.psz_stat_in );
- x264_set_aspect_ratio( h, param, 1 );
+ x264_set_aspect_ratio( h, &h->param, 1 );
x264_reduce_fraction( &h->param.i_fps_num, &h->param.i_fps_den );
+ x264_reduce_fraction( &h->param.i_timebase_num, &h->param.i_timebase_den );
/* Init x264_t */
h->i_frame = -1;
h->i_frame_num = 0;
h->i_idr_pic_id = 0;
+ uint64_t new_timebase_den = h->param.i_timebase_den;
+ if( h->param.b_dts_compress )
+ {
+ /* h->i_dts_compress_multiplier == h->frames.i_bframe_delay + 1 */
+ h->i_dts_compress_multiplier = h->param.i_bframe ? (h->param.i_bframe_pyramid ? 3 : 2) : 1;
+ if( h->i_dts_compress_multiplier != 1 )
+ {
+ new_timebase_den = h->param.i_timebase_den * h->i_dts_compress_multiplier;
+ x264_log( h, X264_LOG_DEBUG, "DTS compresion changed timebase: %u/%u -> %u/%"PRIu64"\n",
+ h->param.i_timebase_num, h->param.i_timebase_den,
+ h->param.i_timebase_num, new_timebase_den );
+ }
+ }
+ else
+ h->i_dts_compress_multiplier = 1;
+
+ if( new_timebase_den * 2 > UINT32_MAX )
+ {
+ x264_log( h, X264_LOG_ERROR, "Effective timebase denominator %"PRIu64" exceeds H.264 maximum\n", new_timebase_den );
+ goto fail;
+ }
+ h->param.i_timebase_den = new_timebase_den;
h->sps = &h->sps_array[0];
x264_sps_init( h->sps, h->param.i_sps_id, &h->param );
if( x264_cqm_init( h ) < 0 )
goto fail;
- h->mb.i_mb_count = h->sps->i_mb_width * h->sps->i_mb_height;
+ h->mb.i_mb_width = h->sps->i_mb_width;
+ h->mb.i_mb_height = h->sps->i_mb_height;
+ h->mb.i_mb_count = h->mb.i_mb_width * h->mb.i_mb_height;
/* Init frames. */
- if( h->param.i_bframe_adaptive == X264_B_ADAPT_TRELLIS )
+ if( h->param.i_bframe_adaptive == X264_B_ADAPT_TRELLIS && !h->param.rc.b_stat_read )
h->frames.i_delay = X264_MAX(h->param.i_bframe,3)*4;
else
h->frames.i_delay = h->param.i_bframe;
if( h->param.rc.b_mb_tree || h->param.rc.i_vbv_buffer_size )
h->frames.i_delay = X264_MAX( h->frames.i_delay, h->param.rc.i_lookahead );
i_slicetype_length = h->frames.i_delay;
- h->frames.i_delay += h->param.i_threads - 1;
- h->frames.i_delay = X264_MIN( h->frames.i_delay, X264_LOOKAHEAD_MAX );
+ h->frames.i_delay += h->i_thread_frames - 1;
h->frames.i_delay += h->param.i_sync_lookahead;
+ h->frames.i_delay += h->param.b_vfr_input && (h->param.rc.i_rc_method == X264_RC_ABR || h->param.rc.b_stat_write
+ || h->param.rc.i_vbv_buffer_size);
+ h->frames.i_bframe_delay = h->param.i_bframe ? (h->param.i_bframe_pyramid ? 2 : 1) : 0;
h->frames.i_max_ref0 = h->param.i_frame_reference;
- h->frames.i_max_ref1 = h->sps->vui.i_num_reorder_frames;
+ h->frames.i_max_ref1 = X264_MIN( h->sps->vui.i_num_reorder_frames, h->param.i_frame_reference );
h->frames.i_max_dpb = h->sps->vui.i_max_dec_frame_buffering;
h->frames.b_have_lowres = !h->param.rc.b_stat_read
&& ( h->param.rc.i_rc_method == X264_RC_ABR
h->frames.b_have_lowres |= h->param.rc.b_stat_read && h->param.rc.i_vbv_buffer_size > 0;
h->frames.b_have_sub8x8_esa = !!(h->param.analyse.inter & X264_ANALYSE_PSUB8x8);
- h->frames.i_last_idr = - h->param.i_keyint_max;
+ h->frames.i_last_keyframe = - h->param.i_keyint_max;
h->frames.i_input = 0;
+ h->frames.i_largest_pts = h->frames.i_second_largest_pts = -1;
CHECKED_MALLOCZERO( h->frames.unused[0], (h->frames.i_delay + 3) * sizeof(x264_frame_t *) );
/* Allocate room for max refs plus a few extra just in case. */
- CHECKED_MALLOCZERO( h->frames.unused[1], (h->param.i_threads + 20) * sizeof(x264_frame_t *) );
+ CHECKED_MALLOCZERO( h->frames.unused[1], (h->i_thread_frames + 20) * sizeof(x264_frame_t *) );
CHECKED_MALLOCZERO( h->frames.current, (h->param.i_sync_lookahead + h->param.i_bframe
- + h->param.i_threads + 3) * sizeof(x264_frame_t *) );
+ + h->i_thread_frames + 3) * sizeof(x264_frame_t *) );
if( h->param.analyse.i_weighted_pred > 0 )
- CHECKED_MALLOCZERO( h->frames.blank_unused, h->param.i_threads * 4 * sizeof(x264_frame_t *) );
+ CHECKED_MALLOCZERO( h->frames.blank_unused, h->i_thread_frames * 4 * sizeof(x264_frame_t *) );
h->i_ref0 = 0;
h->i_ref1 = 0;
-
+ h->i_cpb_delay = h->i_coded_fields = h->i_disp_fields = h->i_prev_duration = 0;
+ h->i_disp_fields_last_frame = -1;
x264_rdo_init();
/* init CPU functions */
x264_mc_init( h->param.cpu, &h->mc );
x264_quant_init( h, h->param.cpu, &h->quantf );
x264_deblock_init( h->param.cpu, &h->loopf );
+ x264_bitstream_init( h->param.cpu, &h->bsf );
x264_dct_init_weights();
mbcmp_init( h );
p = buf + sprintf( buf, "using cpu capabilities:" );
- for( i=0; x264_cpu_names[i].flags; i++ )
+ for( int i = 0; x264_cpu_names[i].flags; i++ )
{
if( !strcmp(x264_cpu_names[i].name, "SSE2")
- && param->cpu & (X264_CPU_SSE2_IS_FAST|X264_CPU_SSE2_IS_SLOW) )
+ && h->param.cpu & (X264_CPU_SSE2_IS_FAST|X264_CPU_SSE2_IS_SLOW) )
continue;
if( !strcmp(x264_cpu_names[i].name, "SSE3")
- && (param->cpu & X264_CPU_SSSE3 || !(param->cpu & X264_CPU_CACHELINE_64)) )
+ && (h->param.cpu & X264_CPU_SSSE3 || !(h->param.cpu & X264_CPU_CACHELINE_64)) )
continue;
if( !strcmp(x264_cpu_names[i].name, "SSE4.1")
- && (param->cpu & X264_CPU_SSE42) )
+ && (h->param.cpu & X264_CPU_SSE42) )
continue;
- if( (param->cpu & x264_cpu_names[i].flags) == x264_cpu_names[i].flags
+ if( (h->param.cpu & x264_cpu_names[i].flags) == x264_cpu_names[i].flags
&& (!i || x264_cpu_names[i].flags != x264_cpu_names[i-1].flags) )
p += sprintf( p, " %s", x264_cpu_names[i].name );
}
- if( !param->cpu )
+ if( !h->param.cpu )
p += sprintf( p, " none!" );
x264_log( h, X264_LOG_INFO, "%s\n", buf );
goto fail;
if( x264_analyse_init_costs( h, X264_LOOKAHEAD_QP ) )
goto fail;
+
+ /* Checks for known miscompilation issues. */
if( h->cost_mv[1][2013] != 24 )
{
x264_log( h, X264_LOG_ERROR, "MV cost test failed: x264 has been miscompiled!\n" );
goto fail;
}
+ /* Must be volatile or else GCC will optimize it out. */
+ volatile int temp = 392;
+ if( x264_clz( temp ) != 23 )
+ {
+ x264_log( h, X264_LOG_ERROR, "CLZ test failed: x264 has been miscompiled!\n" );
+#if ARCH_X86 || ARCH_X86_64
+ x264_log( h, X264_LOG_ERROR, "Are you attempting to run an SSE4a-targeted build on a CPU that\n" );
+ x264_log( h, X264_LOG_ERROR, "doesn't support it?\n" );
+#endif
+ goto fail;
+ }
+
h->out.i_nal = 0;
h->out.i_bitstream = X264_MAX( 1000000, h->param.i_width * h->param.i_height * 4
* ( h->param.rc.i_rc_method == X264_RC_ABR ? pow( 0.95, h->param.rc.i_qp_min )
h->nal_buffer_size = h->out.i_bitstream * 3/2 + 4;
h->thread[0] = h;
- h->i_thread_num = 0;
- for( i = 1; i < h->param.i_threads + !!h->param.i_sync_lookahead; i++ )
+ for( int i = 1; i < h->param.i_threads + !!h->param.i_sync_lookahead; i++ )
CHECKED_MALLOC( h->thread[i], sizeof(x264_t) );
- for( i = 0; i < h->param.i_threads; i++ )
+ if( x264_lookahead_init( h, i_slicetype_length ) )
+ goto fail;
+
+ for( int i = 0; i < h->param.i_threads; i++ )
{
+ int init_nal_count = h->param.i_slice_count + 3;
+ int allocate_threadlocal_data = !h->param.b_sliced_threads || !i;
if( i > 0 )
*h->thread[i] = *h;
- h->thread[i]->fdec = x264_frame_pop_unused( h, 1 );
- if( !h->thread[i]->fdec )
- goto fail;
+
+ if( allocate_threadlocal_data )
+ {
+ h->thread[i]->fdec = x264_frame_pop_unused( h, 1 );
+ if( !h->thread[i]->fdec )
+ goto fail;
+ }
+ else
+ h->thread[i]->fdec = h->thread[0]->fdec;
+
CHECKED_MALLOC( h->thread[i]->out.p_bitstream, h->out.i_bitstream );
- /* Start each thread with room for 8 NAL units; it'll realloc later if needed. */
- CHECKED_MALLOC( h->thread[i]->out.nal, 8*sizeof(x264_nal_t) );
- h->thread[i]->out.i_nals_allocated = 8;
- if( x264_macroblock_cache_init( h->thread[i] ) < 0 )
+ /* Start each thread with room for init_nal_count NAL units; it'll realloc later if needed. */
+ CHECKED_MALLOC( h->thread[i]->out.nal, init_nal_count*sizeof(x264_nal_t) );
+ h->thread[i]->out.i_nals_allocated = init_nal_count;
+
+ if( allocate_threadlocal_data && x264_macroblock_cache_allocate( h->thread[i] ) < 0 )
goto fail;
}
- if( x264_lookahead_init( h, i_slicetype_length ) )
- goto fail;
+ for( int i = 0; i < h->param.i_threads; i++ )
+ if( x264_macroblock_thread_allocate( h->thread[i], 0 ) < 0 )
+ goto fail;
if( x264_ratecontrol_new( h ) < 0 )
goto fail;
+ if( h->param.i_nal_hrd )
+ {
+ x264_log( h, X264_LOG_DEBUG, "HRD bitrate: %i bits/sec\n", h->sps->vui.hrd.i_bit_rate_unscaled );
+ x264_log( h, X264_LOG_DEBUG, "CPB size: %i bits\n", h->sps->vui.hrd.i_cpb_size_unscaled );
+ }
+
if( h->param.psz_dump_yuv )
{
/* create or truncate the reconstructed video file */
****************************************************************************/
int x264_encoder_reconfig( x264_t *h, x264_param_t *param )
{
- h = h->thread[h->i_thread_phase];
+ int rc_reconfig = 0;
+ h = h->thread[h->thread[0]->i_thread_phase];
x264_set_aspect_ratio( h, param, 0 );
#define COPY(var) h->param.var = param->var
COPY( i_frame_reference ); // but never uses more refs than initially specified
COPY( b_deblocking_filter );
COPY( i_deblocking_filter_alphac0 );
COPY( i_deblocking_filter_beta );
- COPY( analyse.intra );
COPY( analyse.inter );
+ COPY( analyse.intra );
COPY( analyse.i_direct_mv_pred );
/* Scratch buffer prevents me_range from being increased for esa/tesa */
if( h->param.analyse.i_me_method < X264_ME_ESA || param->analyse.i_me_range < h->param.analyse.i_me_range )
COPY( i_slice_max_size );
COPY( i_slice_max_mbs );
COPY( i_slice_count );
+
+ /* VBV can't be turned on if it wasn't on to begin with */
+ if( h->param.rc.i_vbv_max_bitrate > 0 && h->param.rc.i_vbv_buffer_size > 0 &&
+ param->rc.i_vbv_max_bitrate > 0 && param->rc.i_vbv_buffer_size > 0 )
+ {
+ COPY( rc.i_vbv_max_bitrate );
+ COPY( rc.i_vbv_buffer_size );
+ COPY( rc.i_bitrate );
+ rc_reconfig = 1;
+ }
+ if( h->param.rc.f_rf_constant != param->rc.f_rf_constant )
+ {
+ COPY( rc.f_rf_constant );
+ rc_reconfig = 1;
+ }
+ if( h->param.rc.f_rf_constant_max != param->rc.f_rf_constant_max )
+ {
+ COPY( rc.f_rf_constant_max );
+ rc_reconfig = 1;
+ }
+
#undef COPY
mbcmp_init( h );
- return x264_validate_parameters( h );
+ int ret = x264_validate_parameters( h );
+
+ /* Supported reconfiguration options (1-pass only):
+ * vbv-maxrate
+ * vbv-bufsize
+ * crf
+ * bitrate (CBR only) */
+ if( !ret && rc_reconfig )
+ x264_ratecontrol_init_reconfigurable( h, 0 );
+
+ return ret;
+}
+
+/****************************************************************************
+ * x264_encoder_parameters:
+ ****************************************************************************/
+void x264_encoder_parameters( x264_t *h, x264_param_t *param )
+{
+ memcpy( param, &h->thread[h->i_thread_phase]->param, sizeof(x264_param_t) );
}
/* internal usage */
nal->i_payload= 0;
nal->p_payload= &h->out.p_bitstream[bs_pos( &h->out.bs ) / 8];
}
-static int x264_nal_end( x264_t *h )
-{
- x264_nal_t *nal = &h->out.nal[h->out.i_nal];
- nal->i_payload = &h->out.p_bitstream[bs_pos( &h->out.bs ) / 8] - nal->p_payload;
- h->out.i_nal++;
- /* if number of allocated nals is not enough, re-allocate a larger one. */
+/* if number of allocated nals is not enough, re-allocate a larger one. */
+static int x264_nal_check_buffer( x264_t *h )
+{
if( h->out.i_nal >= h->out.i_nals_allocated )
{
x264_nal_t *new_out = x264_malloc( sizeof(x264_nal_t) * (h->out.i_nals_allocated*2) );
return 0;
}
-static int x264_encoder_encapsulate_nals( x264_t *h )
+static int x264_nal_end( x264_t *h )
{
- int nal_size = 0, i;
- for( i = 0; i < h->out.i_nal; i++ )
+ x264_nal_t *nal = &h->out.nal[h->out.i_nal];
+ nal->i_payload = &h->out.p_bitstream[bs_pos( &h->out.bs ) / 8] - nal->p_payload;
+ h->out.i_nal++;
+
+ return x264_nal_check_buffer( h );
+}
+
+static int x264_encoder_encapsulate_nals( x264_t *h, int start )
+{
+ int nal_size = 0, previous_nal_size = 0;
+
+ for( int i = 0; i < start; i++ )
+ previous_nal_size += h->out.nal[i].i_payload;
+
+ for( int i = start; i < h->out.i_nal; i++ )
nal_size += h->out.nal[i].i_payload;
/* Worst-case NAL unit escaping: reallocate the buffer if it's too small. */
uint8_t *buf = x264_malloc( nal_size * 2 + h->out.i_nal * 4 );
if( !buf )
return -1;
+ if( previous_nal_size )
+ memcpy( buf, h->nal_buffer, previous_nal_size );
x264_free( h->nal_buffer );
h->nal_buffer = buf;
}
- uint8_t *nal_buffer = h->nal_buffer;
+ uint8_t *nal_buffer = h->nal_buffer + previous_nal_size;
- for( i = 0; i < h->out.i_nal; i++ )
+ for( int i = start; i < h->out.i_nal; i++ )
{
- int size = x264_nal_encode( nal_buffer, h->param.b_annexb, &h->out.nal[i] );
+ int long_startcode = !i || h->out.nal[i].i_type == NAL_SPS || h->out.nal[i].i_type == NAL_PPS;
+ int size = x264_nal_encode( h, nal_buffer, &h->out.nal[i], long_startcode );
h->out.nal[i].i_payload = size;
h->out.nal[i].p_payload = nal_buffer;
nal_buffer += size;
}
- return nal_buffer - h->nal_buffer;
+ x264_emms();
+
+ return nal_buffer - (h->nal_buffer + previous_nal_size);
}
/****************************************************************************
bs_init( &h->out.bs, h->out.p_bitstream, h->out.i_bitstream );
/* Write SEI, SPS and PPS. */
- x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
- if( x264_sei_version_write( h, &h->out.bs ) )
- return -1;
- if( x264_nal_end( h ) )
- return -1;
/* generate sequence parameters */
x264_nal_start( h, NAL_SPS, NAL_PRIORITY_HIGHEST );
x264_pps_write( &h->out.bs, h->pps );
if( x264_nal_end( h ) )
return -1;
- bs_flush( &h->out.bs );
- frame_size = x264_encoder_encapsulate_nals( h );
+ /* identify ourselves */
+ x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
+ if( x264_sei_version_write( h, &h->out.bs ) )
+ return -1;
+ if( x264_nal_end( h ) )
+ return -1;
+
+ frame_size = x264_encoder_encapsulate_nals( h, 0 );
/* now set output*/
*pi_nal = h->out.i_nal;
* from the standard's default. */
static inline void x264_reference_check_reorder( x264_t *h )
{
- int i;
- for( i = 0; i < h->i_ref0 - 1; i++ )
+ for( int i = 0; i < h->i_ref0 - 1; i++ )
/* P and B-frames use different default orders. */
if( h->sh.i_type == SLICE_TYPE_P ? h->fref0[i]->i_frame_num < h->fref0[i+1]->i_frame_num
: h->fref0[i]->i_poc < h->fref0[i+1]->i_poc )
int x264_weighted_reference_duplicate( x264_t *h, int i_ref, const x264_weight_t *w )
{
int i = h->i_ref0;
- int j;
+ int j = 1;
x264_frame_t *newframe;
if( i <= 1 ) /* empty list, definitely can't duplicate frame */
return -1;
- /* Find a place to insert the duplicate in the reference list. */
- for( j = 0; j < i; j++ )
- if( h->fref0[i_ref]->i_frame != h->fref0[j]->i_frame )
- {
- /* found a place, after j, make sure there is not already a duplicate there */
- if( j == i-1 || ( h->fref0[j+1] && h->fref0[i_ref]->i_frame != h->fref0[j+1]->i_frame ) )
- break;
- }
-
- if( j == i ) /* No room in the reference list for the duplicate. */
- return -1;
- j++;
-
newframe = x264_frame_pop_blank_unused( h );
//FIXME: probably don't need to copy everything
static void x264_weighted_pred_init( x264_t *h )
{
- int i_ref;
- int i;
-
/* for now no analysis and set all weights to nothing */
- for( i_ref = 0; i_ref < h->i_ref0; i_ref++ )
+ for( int i_ref = 0; i_ref < h->i_ref0; i_ref++ )
h->fenc->weighted[i_ref] = h->fref0[i_ref]->filtered[0];
// FIXME: This only supports weighting of one reference frame
// and duplicates of that frame.
h->fenc->i_lines_weighted = 0;
- for( i_ref = 0; i_ref < (h->i_ref0 << h->sh.b_mbaff); i_ref++ )
- for( i = 0; i < 3; i++ )
+ for( int i_ref = 0; i_ref < (h->i_ref0 << h->sh.b_mbaff); i_ref++ )
+ for( int i = 0; i < 3; i++ )
h->sh.weight[i_ref][i].weightfn = NULL;
int denom = -1;
int weightluma = 0;
int buffer_next = 0;
- int j;
//FIXME: when chroma support is added, move this into loop
h->sh.weight[0][1].weightfn = h->sh.weight[0][2].weightfn = NULL;
h->sh.weight[0][1].i_denom = h->sh.weight[0][2].i_denom = 0;
- for( j = 0; j < h->i_ref0; j++ )
+ for( int j = 0; j < h->i_ref0; j++ )
{
if( h->fenc->weight[j][0].weightfn )
{
//scale full resolution frame
if( h->sh.weight[j][0].weightfn && h->param.i_threads == 1 )
{
- uint8_t *src = h->fref0[j]->filtered[0] - h->fref0[j]->i_stride[0]*i_padv - PADH;
- uint8_t *dst = h->fenc->weighted[j] - h->fenc->i_stride[0]*i_padv - PADH;
+ pixel *src = h->fref0[j]->filtered[0] - h->fref0[j]->i_stride[0]*i_padv - PADH;
+ pixel *dst = h->fenc->weighted[j] - h->fenc->i_stride[0]*i_padv - PADH;
int stride = h->fenc->i_stride[0];
int width = h->fenc->i_width[0] + PADH*2;
int height = h->fenc->i_lines[0] + i_padv*2;
static inline void x264_reference_build_list( x264_t *h, int i_poc )
{
- int i;
int b_ok;
/* build ref list 0/1 */
- h->i_ref0 = 0;
- h->i_ref1 = 0;
- for( i = 0; h->frames.reference[i]; i++ )
+ h->mb.pic.i_fref[0] = h->i_ref0 = 0;
+ h->mb.pic.i_fref[1] = h->i_ref1 = 0;
+ if( h->sh.i_type == SLICE_TYPE_I )
+ return;
+
+ for( int i = 0; h->frames.reference[i]; i++ )
{
if( h->frames.reference[i]->i_poc < i_poc )
- {
h->fref0[h->i_ref0++] = h->frames.reference[i];
- }
else if( h->frames.reference[i]->i_poc > i_poc )
- {
h->fref1[h->i_ref1++] = h->frames.reference[i];
- }
}
/* Order ref0 from higher to lower poc */
do
{
b_ok = 1;
- for( i = 0; i < h->i_ref0 - 1; i++ )
+ for( int i = 0; i < h->i_ref0 - 1; i++ )
{
if( h->fref0[i]->i_poc < h->fref0[i+1]->i_poc )
{
} while( !b_ok );
if( h->sh.i_mmco_remove_from_end )
- for( i = h->i_ref0-1; i >= h->i_ref0 - h->sh.i_mmco_remove_from_end; i-- )
+ for( int i = h->i_ref0-1; i >= h->i_ref0 - h->sh.i_mmco_remove_from_end; i-- )
{
int diff = h->i_frame_num - h->fref0[i]->i_frame_num;
h->sh.mmco[h->sh.i_mmco_command_count].i_poc = h->fref0[i]->i_poc;
do
{
b_ok = 1;
- for( i = 0; i < h->i_ref1 - 1; i++ )
+ for( int i = 0; i < h->i_ref1 - 1; i++ )
{
if( h->fref1[i]->i_poc > h->fref1[i+1]->i_poc )
{
/* add duplicates */
if( h->fenc->i_type == X264_TYPE_P )
{
+ int idx = -1;
if( h->param.analyse.i_weighted_pred == X264_WEIGHTP_SMART )
{
x264_weight_t w[3];
{
h->fenc->weight[0][0].i_denom = 0;
SET_WEIGHT( w[0], 1, 1, 0, -1 );
- x264_weighted_reference_duplicate( h, 0, w );
+ idx = x264_weighted_reference_duplicate( h, 0, w );
}
else
{
w[0] = h->fenc->weight[0][0];
w[0].i_offset--;
h->mc.weight_cache( h, &w[0] );
- x264_weighted_reference_duplicate( h, 0, w );
+ idx = x264_weighted_reference_duplicate( h, 0, w );
}
}
}
SET_WEIGHT( w[0], 1, 1, 0, -1 );
h->fenc->weight[0][0].i_denom = 0;
w[1].weightfn = w[2].weightfn = NULL;
- x264_weighted_reference_duplicate( h, 0, w );
+ idx = x264_weighted_reference_duplicate( h, 0, w );
}
+ h->mb.ref_blind_dupe = idx;
}
assert( h->i_ref0 + h->i_ref1 <= 16 );
h->mb.pic.i_fref[1] = h->i_ref1;
}
-static void x264_fdec_filter_row( x264_t *h, int mb_y )
+static void x264_fdec_filter_row( x264_t *h, int mb_y, int b_inloop )
{
/* mb_y is the mb to be encoded next, not the mb to be filtered here */
int b_hpel = h->fdec->b_kept_as_ref;
- int b_deblock = !h->sh.i_disable_deblocking_filter_idc;
- int b_end = mb_y == h->sps->i_mb_height;
+ int b_deblock = h->sh.i_disable_deblocking_filter_idc != 1;
+ int b_end = mb_y == h->i_threadslice_end;
+ int b_measure_quality = 1;
int min_y = mb_y - (1 << h->sh.b_mbaff);
- int max_y = b_end ? h->sps->i_mb_height : mb_y;
+ int b_start = min_y == h->i_threadslice_start;
+ int max_y = b_end ? h->i_threadslice_end : mb_y;
b_deblock &= b_hpel || h->param.psz_dump_yuv;
+ if( h->param.b_sliced_threads && b_start && min_y && !b_inloop )
+ {
+ b_deblock = 0; /* We already deblocked on the inloop pass. */
+ b_measure_quality = 0; /* We already measured quality on the inloop pass. */
+ }
if( mb_y & h->sh.b_mbaff )
return;
- if( min_y < 0 )
+ if( min_y < h->i_threadslice_start )
return;
- if( !b_end )
- {
- int i, j;
- for( j=0; j<=h->sh.b_mbaff; j++ )
- for( i=0; i<3; i++ )
+ if( !b_end && b_inloop )
+ for( int j = 0; j <= h->sh.b_mbaff; j++ )
+ for( int i = 0; i < 3; i++ )
{
- memcpy( h->mb.intra_border_backup[j][i],
+ memcpy( h->intra_border_backup[j][i],
h->fdec->plane[i] + ((mb_y*16 >> !!i) + j - 1 - h->sh.b_mbaff) * h->fdec->i_stride[i],
- h->sps->i_mb_width*16 >> !!i );
+ (h->mb.i_mb_width*16 >> !!i) * sizeof(pixel) );
}
- }
if( b_deblock )
- {
- int y;
- for( y = min_y; y < max_y; y += (1 << h->sh.b_mbaff) )
+ for( int y = min_y; y < max_y; y += (1 << h->sh.b_mbaff) )
x264_frame_deblock_row( h, y );
- }
if( b_hpel )
{
- x264_frame_expand_border( h, h->fdec, min_y, b_end );
+ int end = mb_y == h->mb.i_mb_height;
+ x264_frame_expand_border( h, h->fdec, min_y, end );
if( h->param.analyse.i_subpel_refine )
{
- x264_frame_filter( h, h->fdec, min_y, b_end );
- x264_frame_expand_border_filtered( h, h->fdec, min_y, b_end );
+ x264_frame_filter( h, h->fdec, min_y, end );
+ x264_frame_expand_border_filtered( h, h->fdec, min_y, end );
}
}
- if( h->param.i_threads > 1 && h->fdec->b_kept_as_ref )
- {
+ if( h->i_thread_frames > 1 && h->fdec->b_kept_as_ref )
x264_frame_cond_broadcast( h->fdec, mb_y*16 + (b_end ? 10000 : -(X264_THREAD_HEIGHT << h->sh.b_mbaff)) );
- }
- min_y = X264_MAX( min_y*16-8, 0 );
- max_y = b_end ? h->param.i_height : mb_y*16-8;
+ min_y = min_y*16 - 8 * !b_start;
+ max_y = b_end ? X264_MIN( h->i_threadslice_end*16 , h->param.i_height ) : mb_y*16 - 8;
- if( h->param.analyse.b_psnr )
+ if( b_measure_quality )
{
- int i;
- for( i=0; i<3; i++ )
- h->stat.frame.i_ssd[i] +=
- x264_pixel_ssd_wxh( &h->pixf,
- h->fdec->plane[i] + (min_y>>!!i) * h->fdec->i_stride[i], h->fdec->i_stride[i],
- h->fenc->plane[i] + (min_y>>!!i) * h->fenc->i_stride[i], h->fenc->i_stride[i],
- h->param.i_width >> !!i, (max_y-min_y) >> !!i );
- }
+ if( h->param.analyse.b_psnr )
+ for( int i = 0; i < 3; i++ )
+ h->stat.frame.i_ssd[i] +=
+ x264_pixel_ssd_wxh( &h->pixf,
+ h->fdec->plane[i] + (min_y>>!!i) * h->fdec->i_stride[i], h->fdec->i_stride[i],
+ h->fenc->plane[i] + (min_y>>!!i) * h->fenc->i_stride[i], h->fenc->i_stride[i],
+ h->param.i_width >> !!i, (max_y-min_y) >> !!i );
- if( h->param.analyse.b_ssim )
- {
- x264_emms();
- /* offset by 2 pixels to avoid alignment of ssim blocks with dct blocks,
- * and overlap by 4 */
- min_y += min_y == 0 ? 2 : -6;
- h->stat.frame.f_ssim +=
- x264_pixel_ssim_wxh( &h->pixf,
- h->fdec->plane[0] + 2+min_y*h->fdec->i_stride[0], h->fdec->i_stride[0],
- h->fenc->plane[0] + 2+min_y*h->fenc->i_stride[0], h->fenc->i_stride[0],
- h->param.i_width-2, max_y-min_y, h->scratch_buffer );
+ if( h->param.analyse.b_ssim )
+ {
+ x264_emms();
+ /* offset by 2 pixels to avoid alignment of ssim blocks with dct blocks,
+ * and overlap by 4 */
+ min_y += b_start ? 2 : -6;
+ h->stat.frame.f_ssim +=
+ x264_pixel_ssim_wxh( &h->pixf,
+ h->fdec->plane[0] + 2+min_y*h->fdec->i_stride[0], h->fdec->i_stride[0],
+ h->fenc->plane[0] + 2+min_y*h->fenc->i_stride[0], h->fenc->i_stride[0],
+ h->param.i_width-2, max_y-min_y, h->scratch_buffer );
+ }
}
}
static inline int x264_reference_update( x264_t *h )
{
- int i, j;
if( !h->fdec->b_kept_as_ref )
{
- if( h->param.i_threads > 1 )
+ if( h->i_thread_frames > 1 )
{
x264_frame_push_unused( h, h->fdec );
h->fdec = x264_frame_pop_unused( h, 1 );
}
/* apply mmco from previous frame. */
- for( i = 0; i < h->sh.i_mmco_command_count; i++ )
- for( j = 0; h->frames.reference[j]; j++ )
+ for( int i = 0; i < h->sh.i_mmco_command_count; i++ )
+ for( int j = 0; h->frames.reference[j]; j++ )
if( h->frames.reference[j]->i_poc == h->sh.mmco[i].i_poc )
x264_frame_push_unused( h, x264_frame_shift( &h->frames.reference[j] ) );
static inline void x264_reference_hierarchy_reset( x264_t *h )
{
- int i, ref;
+ int ref;
int b_hasdelayframe = 0;
if( !h->param.i_bframe_pyramid )
return;
/* look for delay frames -- chain must only contain frames that are disposable */
- for( i = 0; h->frames.current[i] && IS_DISPOSABLE( h->frames.current[i]->i_type ); i++ )
- b_hasdelayframe |= h->frames.current[i]->i_dts
+ for( int i = 0; h->frames.current[i] && IS_DISPOSABLE( h->frames.current[i]->i_type ); i++ )
+ b_hasdelayframe |= h->frames.current[i]->i_coded
!= h->frames.current[i]->i_frame + h->sps->vui.i_num_reorder_frames;
if( h->param.i_bframe_pyramid != X264_B_PYRAMID_STRICT && !b_hasdelayframe )
{
x264_slice_header_init( h, &h->sh, h->sps, h->pps, -1, h->i_frame_num, i_global_qp );
- /* always set the real higher num of ref frame used */
- h->sh.b_num_ref_idx_override = 1;
h->sh.i_num_ref_idx_l0_active = h->i_ref0 <= 0 ? 1 : h->i_ref0;
h->sh.i_num_ref_idx_l1_active = h->i_ref1 <= 0 ? 1 : h->i_ref1;
+ if( h->sh.i_num_ref_idx_l0_active != h->pps->i_num_ref_idx_l0_default_active ||
+ (h->sh.i_type == SLICE_TYPE_B && h->sh.i_num_ref_idx_l1_active != h->pps->i_num_ref_idx_l1_default_active) )
+ {
+ h->sh.b_num_ref_idx_override = 1;
+ }
}
h->fdec->i_frame_num = h->sh.i_frame_num;
if( h->sps->i_poc_type == 0 )
{
- h->sh.i_poc_lsb = h->fdec->i_poc & ( (1 << h->sps->i_log2_max_poc_lsb) - 1 );
- h->sh.i_delta_poc_bottom = 0;
+ h->sh.i_poc = h->fdec->i_poc;
+ if( h->param.b_interlaced )
+ {
+ h->sh.i_delta_poc_bottom = h->param.b_tff ? 1 : -1;
+ if( h->sh.i_delta_poc_bottom == -1 )
+ h->sh.i_poc = h->fdec->i_poc + 1;
+ }
+ else
+ h->sh.i_delta_poc_bottom = 0;
}
else if( h->sps->i_poc_type == 1 )
{
{
int i_skip;
int mb_xy, i_mb_x, i_mb_y;
- int i, i_list, i_ref, i_skip_bak = 0; /* Shut up GCC. */
+ int i_skip_bak = 0; /* Shut up GCC. */
bs_t bs_bak;
x264_cabac_t cabac_bak;
uint8_t cabac_prevbyte_bak = 0; /* Shut up GCC. */
- /* Assume no more than 3 bytes of NALU escaping. */
- int slice_max_size = h->param.i_slice_max_size > 0 ? (h->param.i_slice_max_size-3-NALU_OVERHEAD)*8 : INT_MAX;
+ /* Assume no more than 3 bytes of NALU escaping.
+ * NALUs other than the first use a 3-byte startcode. */
+ int overhead_guess = (NALU_OVERHEAD - (h->param.b_annexb && h->out.i_nal)) + 3;
+ int slice_max_size = h->param.i_slice_max_size > 0 ? (h->param.i_slice_max_size-overhead_guess)*8 : INT_MAX;
int starting_bits = bs_pos(&h->out.bs);
+ int b_deblock = h->sh.i_disable_deblocking_filter_idc != 1;
+ int b_hpel = h->fdec->b_kept_as_ref;
+ b_deblock &= b_hpel || h->param.psz_dump_yuv;
+ bs_realign( &h->out.bs );
/* Slice */
x264_nal_start( h, h->i_nal_type, h->i_nal_ref_idc );
/* Slice header */
+ x264_macroblock_thread_init( h );
+
+ /* If this isn't the first slice in the threadslice, set the slice QP
+ * equal to the last QP in the previous slice for more accurate
+ * CABAC initialization. */
+ if( h->sh.i_first_mb != h->i_threadslice_start * h->mb.i_mb_width )
+ {
+ h->sh.i_qp = h->mb.i_last_qp;
+ h->sh.i_qp_delta = h->sh.i_qp - h->pps->i_pic_init_qp;
+ }
+
x264_slice_header_write( &h->out.bs, &h->sh, h->i_nal_ref_idc );
if( h->param.b_cabac )
{
h->mb.i_last_qp = h->sh.i_qp;
h->mb.i_last_dqp = 0;
- i_mb_y = h->sh.i_first_mb / h->sps->i_mb_width;
- i_mb_x = h->sh.i_first_mb % h->sps->i_mb_width;
+ i_mb_y = h->sh.i_first_mb / h->mb.i_mb_width;
+ i_mb_x = h->sh.i_first_mb % h->mb.i_mb_width;
i_skip = 0;
- while( (mb_xy = i_mb_x + i_mb_y * h->sps->i_mb_width) <= h->sh.i_last_mb )
+ while( (mb_xy = i_mb_x + i_mb_y * h->mb.i_mb_width) <= h->sh.i_last_mb )
{
int mb_spos = bs_pos(&h->out.bs) + x264_cabac_pos(&h->cabac);
if( h->param.i_slice_max_size > 0 )
}
if( i_mb_x == 0 && !h->mb.b_reencode_mb )
- x264_fdec_filter_row( h, i_mb_y );
+ x264_fdec_filter_row( h, i_mb_y, 1 );
/* load cache */
x264_macroblock_cache_load( h, i_mb_x, i_mb_y );
bs_write_ue( &h->out.bs, i_skip ); /* skip run */
i_skip = 0;
}
- x264_macroblock_write_cavlc( h, &h->out.bs );
+ x264_macroblock_write_cavlc( h );
}
}
else
h->mb.b_reencode_mb = 0;
-#if VISUALIZE
+#if HAVE_VISUALIZE
if( h->param.b_visualize )
x264_visualize_mb( h );
#endif
/* accumulate mb stats */
h->stat.frame.i_mb_count[h->mb.i_type]++;
- if( !IS_INTRA(h->mb.i_type) && !IS_SKIP(h->mb.i_type) && !IS_DIRECT(h->mb.i_type) )
+ int b_intra = IS_INTRA( h->mb.i_type );
+ if( h->param.i_log_level >= X264_LOG_INFO || h->param.rc.b_stat_write )
{
- if( h->mb.i_partition != D_8x8 )
- h->stat.frame.i_mb_partition[h->mb.i_partition] += 4;
- else
- for( i = 0; i < 4; i++ )
- h->stat.frame.i_mb_partition[h->mb.i_sub_partition[i]] ++;
- if( h->param.i_frame_reference > 1 )
- for( i_list = 0; i_list <= (h->sh.i_type == SLICE_TYPE_B); i_list++ )
- for( i = 0; i < 4; i++ )
- {
- i_ref = h->mb.cache.ref[i_list][ x264_scan8[4*i] ];
- if( i_ref >= 0 )
- h->stat.frame.i_mb_count_ref[i_list][i_ref] ++;
- }
+ if( !b_intra && !IS_SKIP( h->mb.i_type ) && !IS_DIRECT( h->mb.i_type ) )
+ {
+ if( h->mb.i_partition != D_8x8 )
+ h->stat.frame.i_mb_partition[h->mb.i_partition] += 4;
+ else
+ for( int i = 0; i < 4; i++ )
+ h->stat.frame.i_mb_partition[h->mb.i_sub_partition[i]] ++;
+ if( h->param.i_frame_reference > 1 )
+ for( int i_list = 0; i_list <= (h->sh.i_type == SLICE_TYPE_B); i_list++ )
+ for( int i = 0; i < 4; i++ )
+ {
+ int i_ref = h->mb.cache.ref[i_list][ x264_scan8[4*i] ];
+ if( i_ref >= 0 )
+ h->stat.frame.i_mb_count_ref[i_list][i_ref] ++;
+ }
+ }
}
if( h->param.i_log_level >= X264_LOG_INFO )
{
- if( h->mb.i_cbp_luma || h->mb.i_cbp_chroma )
+ if( h->mb.i_cbp_luma | h->mb.i_cbp_chroma )
{
int cbpsum = (h->mb.i_cbp_luma&1) + ((h->mb.i_cbp_luma>>1)&1)
+ ((h->mb.i_cbp_luma>>2)&1) + (h->mb.i_cbp_luma>>3);
- int b_intra = IS_INTRA(h->mb.i_type);
h->stat.frame.i_mb_cbp[!b_intra + 0] += cbpsum;
- h->stat.frame.i_mb_cbp[!b_intra + 2] += h->mb.i_cbp_chroma >= 1;
- h->stat.frame.i_mb_cbp[!b_intra + 4] += h->mb.i_cbp_chroma == 2;
+ h->stat.frame.i_mb_cbp[!b_intra + 2] += !!h->mb.i_cbp_chroma;
+ h->stat.frame.i_mb_cbp[!b_intra + 4] += h->mb.i_cbp_chroma >> 1;
}
- if( h->mb.i_cbp_luma && !IS_INTRA(h->mb.i_type) )
+ if( h->mb.i_cbp_luma && !b_intra )
{
h->stat.frame.i_mb_count_8x8dct[0] ++;
h->stat.frame.i_mb_count_8x8dct[1] += h->mb.b_transform_8x8;
}
- if( IS_INTRA(h->mb.i_type) && h->mb.i_type != I_PCM )
+ if( b_intra && h->mb.i_type != I_PCM )
{
if( h->mb.i_type == I_16x16 )
h->stat.frame.i_mb_pred_mode[0][h->mb.i_intra16x16_pred_mode]++;
else if( h->mb.i_type == I_8x8 )
- for( i = 0; i < 16; i += 4 )
+ for( int i = 0; i < 16; i += 4 )
h->stat.frame.i_mb_pred_mode[1][h->mb.cache.intra4x4_pred_mode[x264_scan8[i]]]++;
else //if( h->mb.i_type == I_4x4 )
- for( i = 0; i < 16; i++ )
+ for( int i = 0; i < 16; i++ )
h->stat.frame.i_mb_pred_mode[2][h->mb.cache.intra4x4_pred_mode[x264_scan8[i]]]++;
+ h->stat.frame.i_mb_pred_mode[3][x264_mb_pred_mode8x8c_fix[h->mb.i_chroma_pred_mode]]++;
}
}
+ /* calculate deblock strength values (actual deblocking is done per-row along with hpel) */
+ if( b_deblock )
+ {
+ int mvy_limit = 4 >> h->sh.b_mbaff;
+ uint8_t (*bs)[4][4] = h->deblock_strength[h->mb.i_mb_y&h->sh.b_mbaff][h->mb.i_mb_x];
+ x264_macroblock_cache_load_deblock( h );
+ if( IS_INTRA( h->mb.type[h->mb.i_mb_xy] ) )
+ memset( bs, 3, 2*4*4*sizeof(uint8_t) );
+ else
+ h->loopf.deblock_strength( h->mb.cache.non_zero_count, h->mb.cache.ref, h->mb.cache.mv,
+ bs, mvy_limit, h->sh.i_type == SLICE_TYPE_B );
+ }
+
x264_ratecontrol_mb( h, mb_size );
if( h->sh.b_mbaff )
{
i_mb_x += i_mb_y & 1;
- i_mb_y ^= i_mb_x < h->sps->i_mb_width;
+ i_mb_y ^= i_mb_x < h->mb.i_mb_width;
}
else
i_mb_x++;
- if( i_mb_x == h->sps->i_mb_width )
+ if( i_mb_x == h->mb.i_mb_width )
{
i_mb_y++;
i_mb_x = 0;
if( x264_nal_end( h ) )
return -1;
- if( h->sh.i_last_mb == h->mb.i_mb_count-1 )
+ if( h->sh.i_last_mb == (h->i_threadslice_end * h->mb.i_mb_width - 1) )
{
h->stat.frame.i_misc_bits = bs_pos( &h->out.bs )
+ (h->out.i_nal*NALU_OVERHEAD * 8)
- h->stat.frame.i_tex_bits
- h->stat.frame.i_mv_bits;
- x264_fdec_filter_row( h, h->sps->i_mb_height );
+ x264_fdec_filter_row( h, h->i_threadslice_end, 1 );
}
return 0;
static void x264_thread_sync_context( x264_t *dst, x264_t *src )
{
- x264_frame_t **f;
if( dst == src )
return;
// reference counting
- for( f = src->frames.reference; *f; f++ )
+ for( x264_frame_t **f = src->frames.reference; *f; f++ )
(*f)->i_reference_count++;
- for( f = dst->frames.reference; *f; f++ )
+ for( x264_frame_t **f = dst->frames.reference; *f; f++ )
x264_frame_push_unused( src, *f );
src->fdec->i_reference_count++;
x264_frame_push_unused( src, dst->fdec );
static void *x264_slices_write( x264_t *h )
{
int i_slice_num = 0;
+ int last_thread_mb = h->sh.i_last_mb;
if( h->param.i_sync_lookahead )
x264_lower_thread_priority( 10 );
-#ifdef HAVE_MMX
+#if HAVE_MMX
/* Misalign mask has to be set separately for each thread. */
if( h->param.cpu&X264_CPU_SSE_MISALIGN )
x264_cpu_mask_misalign_sse();
#endif
-#if VISUALIZE
+#if HAVE_VISUALIZE
if( h->param.b_visualize )
if( x264_visualize_init( h ) )
return (void *)-1;
/* init stats */
memset( &h->stat.frame, 0, sizeof(h->stat.frame) );
h->mb.b_reencode_mb = 0;
- while( h->sh.i_first_mb < h->mb.i_mb_count )
+ while( h->sh.i_first_mb <= last_thread_mb )
{
- h->sh.i_last_mb = h->mb.i_mb_count - 1;
+ h->sh.i_last_mb = last_thread_mb;
if( h->param.i_slice_max_mbs )
h->sh.i_last_mb = h->sh.i_first_mb + h->param.i_slice_max_mbs - 1;
- else if( h->param.i_slice_count )
+ else if( h->param.i_slice_count && !h->param.b_sliced_threads )
{
- x264_emms();
+ int height = h->mb.i_mb_height >> h->param.b_interlaced;
+ int width = h->mb.i_mb_width << h->param.b_interlaced;
i_slice_num++;
- double height = h->sps->i_mb_height >> h->param.b_interlaced;
- int width = h->sps->i_mb_width << h->param.b_interlaced;
- h->sh.i_last_mb = (int)(height * i_slice_num / h->param.i_slice_count + 0.5) * width - 1;
+ h->sh.i_last_mb = (height * i_slice_num + h->param.i_slice_count/2) / h->param.i_slice_count * width - 1;
}
- h->sh.i_last_mb = X264_MIN( h->sh.i_last_mb, h->mb.i_mb_count - 1 );
+ h->sh.i_last_mb = X264_MIN( h->sh.i_last_mb, last_thread_mb );
if( x264_stack_align( x264_slice_write, h ) )
return (void *)-1;
h->sh.i_first_mb = h->sh.i_last_mb + 1;
}
-#if VISUALIZE
+#if HAVE_VISUALIZE
if( h->param.b_visualize )
{
x264_visualize_show( h );
return (void *)0;
}
+static int x264_threaded_slices_write( x264_t *h )
+{
+ void *ret = NULL;
+#if HAVE_MMX
+ if( h->param.cpu&X264_CPU_SSE_MISALIGN )
+ x264_cpu_mask_misalign_sse();
+#endif
+ /* set first/last mb and sync contexts */
+ for( int i = 0; i < h->param.i_threads; i++ )
+ {
+ x264_t *t = h->thread[i];
+ if( i )
+ {
+ t->param = h->param;
+ memcpy( &t->i_frame, &h->i_frame, offsetof(x264_t, rc) - offsetof(x264_t, i_frame) );
+ }
+ int height = h->mb.i_mb_height >> h->param.b_interlaced;
+ t->i_threadslice_start = ((height * i + h->param.i_slice_count/2) / h->param.i_threads) << h->param.b_interlaced;
+ t->i_threadslice_end = ((height * (i+1) + h->param.i_slice_count/2) / h->param.i_threads) << h->param.b_interlaced;
+ t->sh.i_first_mb = t->i_threadslice_start * h->mb.i_mb_width;
+ t->sh.i_last_mb = t->i_threadslice_end * h->mb.i_mb_width - 1;
+ }
+
+ x264_stack_align( x264_analyse_weight_frame, h, h->mb.i_mb_height*16 + 16 );
+
+ x264_threads_distribute_ratecontrol( h );
+
+ /* dispatch */
+ for( int i = 0; i < h->param.i_threads; i++ )
+ {
+ if( x264_pthread_create( &h->thread[i]->thread_handle, NULL, (void*)x264_slices_write, (void*)h->thread[i] ) )
+ return -1;
+ h->thread[i]->b_thread_active = 1;
+ }
+ for( int i = 0; i < h->param.i_threads; i++ )
+ {
+ x264_pthread_join( h->thread[i]->thread_handle, &ret );
+ h->thread[i]->b_thread_active = 0;
+ if( (intptr_t)ret )
+ return (intptr_t)ret;
+ }
+
+ /* Go back and fix up the hpel on the borders between slices. */
+ for( int i = 1; i < h->param.i_threads; i++ )
+ {
+ x264_fdec_filter_row( h->thread[i], h->thread[i]->i_threadslice_start + 1, 0 );
+ if( h->sh.b_mbaff )
+ x264_fdec_filter_row( h->thread[i], h->thread[i]->i_threadslice_start + 2, 0 );
+ }
+
+ x264_threads_merge_ratecontrol( h );
+
+ for( int i = 1; i < h->param.i_threads; i++ )
+ {
+ x264_t *t = h->thread[i];
+ for( int j = 0; j < t->out.i_nal; j++ )
+ {
+ h->out.nal[h->out.i_nal] = t->out.nal[j];
+ h->out.i_nal++;
+ x264_nal_check_buffer( h );
+ }
+ /* All entries in stat.frame are ints except for ssd/ssim. */
+ for( int j = 0; j < (offsetof(x264_t,stat.frame.i_ssd) - offsetof(x264_t,stat.frame.i_mv_bits)) / sizeof(int); j++ )
+ ((int*)&h->stat.frame)[j] += ((int*)&t->stat.frame)[j];
+ for( int j = 0; j < 3; j++ )
+ h->stat.frame.i_ssd[j] += t->stat.frame.i_ssd[j];
+ h->stat.frame.f_ssim += t->stat.frame.f_ssim;
+ }
+
+ return 0;
+}
+
+void x264_encoder_intra_refresh( x264_t *h )
+{
+ h = h->thread[h->i_thread_phase];
+ h->b_queued_intra_refresh = 1;
+}
+
/****************************************************************************
* x264_encoder_encode:
* XXX: i_poc : is the poc of the current given picture
x264_picture_t *pic_out )
{
x264_t *thread_current, *thread_prev, *thread_oldest;
- int i_nal_type;
- int i_nal_ref_idc;
-
- int i_global_qp;
+ int i_nal_type, i_nal_ref_idc, i_global_qp;
+ int overhead = NALU_OVERHEAD;
- if( h->param.i_threads > 1)
+ if( h->i_thread_frames > 1 )
{
thread_prev = h->thread[ h->i_thread_phase ];
- h->i_thread_phase = (h->i_thread_phase + 1) % h->param.i_threads;
+ h->i_thread_phase = (h->i_thread_phase + 1) % h->i_thread_frames;
thread_current = h->thread[ h->i_thread_phase ];
- thread_oldest = h->thread[ (h->i_thread_phase + 1) % h->param.i_threads ];
+ thread_oldest = h->thread[ (h->i_thread_phase + 1) % h->i_thread_frames ];
x264_thread_sync_context( thread_current, thread_prev );
x264_thread_sync_ratecontrol( thread_current, thread_prev, thread_oldest );
h = thread_current;
-// fprintf(stderr, "current: %p prev: %p oldest: %p \n", thread_current, thread_prev, thread_oldest);
}
else
{
if( x264_frame_copy_picture( h, fenc, pic_in ) < 0 )
return -1;
- if( h->param.i_width != 16 * h->sps->i_mb_width ||
- h->param.i_height != 16 * h->sps->i_mb_height )
+ if( h->param.i_width != 16 * h->mb.i_mb_width ||
+ h->param.i_height != 16 * h->mb.i_mb_height )
x264_frame_expand_border_mod16( h, fenc );
fenc->i_frame = h->frames.i_input++;
- if( h->frames.b_have_lowres )
+ if( h->frames.i_bframe_delay && fenc->i_frame == h->frames.i_bframe_delay )
+ h->frames.i_bframe_delay_time = fenc->i_pts;
+
+ if( h->param.b_vfr_input && fenc->i_pts <= h->frames.i_largest_pts )
+ x264_log( h, X264_LOG_WARNING, "non-strictly-monotonic PTS\n" );
+
+ h->frames.i_second_largest_pts = h->frames.i_largest_pts;
+ h->frames.i_largest_pts = fenc->i_pts;
+
+ if( (fenc->i_pic_struct < PIC_STRUCT_AUTO) || (fenc->i_pic_struct > PIC_STRUCT_TRIPLE) )
+ fenc->i_pic_struct = PIC_STRUCT_AUTO;
+
+ if( fenc->i_pic_struct == PIC_STRUCT_AUTO )
{
- if( h->param.analyse.i_weighted_pred )
- x264_weight_plane_analyse( h, fenc );
- x264_frame_init_lowres( h, fenc );
+ int b_interlaced = fenc->param ? fenc->param->b_interlaced : h->param.b_interlaced;
+ if( b_interlaced )
+ {
+ int b_tff = fenc->param ? fenc->param->b_tff : h->param.b_tff;
+ fenc->i_pic_struct = b_tff ? PIC_STRUCT_TOP_BOTTOM : PIC_STRUCT_BOTTOM_TOP;
+ }
+ else
+ fenc->i_pic_struct = PIC_STRUCT_PROGRESSIVE;
}
if( h->param.rc.b_mb_tree && h->param.rc.b_stat_read )
{
- if( x264_macroblock_tree_read( h, fenc ) )
+ if( x264_macroblock_tree_read( h, fenc, pic_in->prop.quant_offsets ) )
return -1;
}
- else if( h->param.rc.i_aq_mode )
- x264_adaptive_quant_frame( h, fenc );
+ else
+ x264_adaptive_quant_frame( h, fenc, pic_in->prop.quant_offsets );
+
+ if( pic_in->prop.quant_offsets_free )
+ pic_in->prop.quant_offsets_free( pic_in->prop.quant_offsets );
+
+ if( h->frames.b_have_lowres )
+ x264_frame_init_lowres( h, fenc );
/* 2: Place the frame into the queue for its slice type decision */
x264_lookahead_put_frame( h, fenc );
- if( h->frames.i_input <= h->frames.i_delay + 1 - h->param.i_threads )
+ if( h->frames.i_input <= h->frames.i_delay + 1 - h->i_thread_frames )
{
/* Nothing yet to encode, waiting for filling of buffers */
pic_out->i_type = X264_TYPE_AUTO;
/* ------------------- Get frame to be encoded ------------------------- */
/* 4: get picture to encode */
h->fenc = x264_frame_shift( h->frames.current );
+ if( h->i_frame == h->i_thread_frames - 1 )
+ h->i_reordered_pts_delay = h->fenc->i_reordered_pts;
if( h->fenc->param )
{
x264_encoder_reconfig( h, h->fenc->param );
h->fenc->param->param_free( h->fenc->param );
}
- if( h->fenc->i_type == X264_TYPE_IDR )
+ if( h->fenc->b_keyframe )
{
- h->frames.i_last_idr = h->fenc->i_frame;
- h->i_frame_num = 0;
+ h->frames.i_last_keyframe = h->fenc->i_frame;
+ if( h->fenc->i_type == X264_TYPE_IDR )
+ h->i_frame_num = 0;
}
- h->sh.i_mmco_command_count = 0;
+ h->sh.i_mmco_command_count =
h->sh.i_mmco_remove_from_end = 0;
h->b_ref_reorder[0] =
h->b_ref_reorder[1] = 0;
}
h->fdec->i_poc =
- h->fenc->i_poc = 2 * (h->fenc->i_frame - h->frames.i_last_idr);
+ h->fenc->i_poc = 2 * (h->fenc->i_frame - h->frames.i_last_keyframe);
h->fdec->i_type = h->fenc->i_type;
h->fdec->i_frame = h->fenc->i_frame;
h->fenc->b_kept_as_ref =
/* ---------------------- Write the bitstream -------------------------- */
/* Init bitstream context */
- h->out.i_nal = 0;
- bs_init( &h->out.bs, h->out.p_bitstream, h->out.i_bitstream );
+ if( h->param.b_sliced_threads )
+ {
+ for( int i = 0; i < h->param.i_threads; i++ )
+ {
+ bs_init( &h->thread[i]->out.bs, h->thread[i]->out.p_bitstream, h->thread[i]->out.i_bitstream );
+ h->thread[i]->out.i_nal = 0;
+ }
+ }
+ else
+ {
+ bs_init( &h->out.bs, h->out.p_bitstream, h->out.i_bitstream );
+ h->out.i_nal = 0;
+ }
if( h->param.b_aud )
{
bs_rbsp_trailing( &h->out.bs );
if( x264_nal_end( h ) )
return -1;
+ overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD - (h->param.b_annexb && h->out.i_nal-1);
}
h->i_nal_type = i_nal_type;
h->i_nal_ref_idc = i_nal_ref_idc;
- int overhead = NALU_OVERHEAD;
+ if( h->param.b_intra_refresh )
+ {
+ if( IS_X264_TYPE_I( h->fenc->i_type ) )
+ {
+ h->fdec->i_frames_since_pir = 0;
+ h->b_queued_intra_refresh = 0;
+ /* PIR is currently only supported with ref == 1, so any intra frame effectively refreshes
+ * the whole frame and counts as an intra refresh. */
+ h->fdec->f_pir_position = h->mb.i_mb_width;
+ }
+ else if( h->fenc->i_type == X264_TYPE_P )
+ {
+ int pocdiff = (h->fdec->i_poc - h->fref0[0]->i_poc)/2;
+ float increment = X264_MAX( ((float)h->mb.i_mb_width-1) / h->param.i_keyint_max, 1 );
+ h->fdec->f_pir_position = h->fref0[0]->f_pir_position;
+ h->fdec->i_frames_since_pir = h->fref0[0]->i_frames_since_pir + pocdiff;
+ if( h->fdec->i_frames_since_pir >= h->param.i_keyint_max ||
+ (h->b_queued_intra_refresh && h->fdec->f_pir_position + 0.5 >= h->mb.i_mb_width) )
+ {
+ h->fdec->f_pir_position = 0;
+ h->fdec->i_frames_since_pir = 0;
+ h->b_queued_intra_refresh = 0;
+ h->fenc->b_keyframe = 1;
+ }
+ h->fdec->i_pir_start_col = h->fdec->f_pir_position+0.5;
+ h->fdec->f_pir_position += increment * pocdiff;
+ h->fdec->i_pir_end_col = h->fdec->f_pir_position+0.5;
+ }
+ }
- /* Write SPS and PPS */
- if( i_nal_type == NAL_SLICE_IDR && h->param.b_repeat_headers )
+ if( h->fenc->b_keyframe )
{
- if( h->fenc->i_frame == 0 )
+ /* Write SPS and PPS */
+ if( h->param.b_repeat_headers )
+ {
+ /* generate sequence parameters */
+ x264_nal_start( h, NAL_SPS, NAL_PRIORITY_HIGHEST );
+ x264_sps_write( &h->out.bs, h->sps );
+ if( x264_nal_end( h ) )
+ return -1;
+ overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD;
+
+ /* generate picture parameters */
+ x264_nal_start( h, NAL_PPS, NAL_PRIORITY_HIGHEST );
+ x264_pps_write( &h->out.bs, h->pps );
+ if( x264_nal_end( h ) )
+ return -1;
+ overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD;
+ }
+
+ /* buffering period sei is written in x264_encoder_frame_end */
+
+ if( h->param.b_repeat_headers && h->fenc->i_frame == 0 )
{
/* identify ourself */
x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
return -1;
if( x264_nal_end( h ) )
return -1;
- overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD;
+ overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD - (h->param.b_annexb && h->out.i_nal-1);
}
- /* generate sequence parameters */
- x264_nal_start( h, NAL_SPS, NAL_PRIORITY_HIGHEST );
- x264_sps_write( &h->out.bs, h->sps );
- if( x264_nal_end( h ) )
- return -1;
- overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD;
+ if( h->fenc->i_type != X264_TYPE_IDR )
+ {
+ int time_to_recovery = X264_MIN( h->mb.i_mb_width - 1, h->param.i_keyint_max ) + h->param.i_bframe;
+ x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
+ x264_sei_recovery_point_write( h, &h->out.bs, time_to_recovery );
+ x264_nal_end( h );
+ overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD - (h->param.b_annexb && h->out.i_nal-1);
+ }
+ }
- /* generate picture parameters */
- x264_nal_start( h, NAL_PPS, NAL_PRIORITY_HIGHEST );
- x264_pps_write( &h->out.bs, h->pps );
+ /* generate sei pic timing */
+ if( h->sps->vui.b_pic_struct_present || h->sps->vui.b_nal_hrd_parameters_present )
+ {
+ x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
+ x264_sei_pic_timing_write( h, &h->out.bs );
if( x264_nal_end( h ) )
return -1;
- overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD;
+ overhead += h->out.nal[h->out.i_nal-1].i_payload + NALU_OVERHEAD - (h->param.b_annexb && h->out.i_nal-1);
}
/* Init the rate control */
x264_reference_check_reorder( h );
}
+ if( h->i_ref0 )
+ h->fdec->i_poc_l0ref0 = h->fref0[0]->i_poc;
+
if( h->sh.i_type == SLICE_TYPE_B )
x264_macroblock_bipred_init( h );
h->i_frame_num++;
/* Write frame */
- if( h->param.i_threads > 1 )
+ h->i_threadslice_start = 0;
+ h->i_threadslice_end = h->mb.i_mb_height;
+ if( h->i_thread_frames > 1 )
{
if( x264_pthread_create( &h->thread_handle, NULL, (void*)x264_slices_write, h ) )
return -1;
h->b_thread_active = 1;
}
+ else if( h->param.b_sliced_threads )
+ {
+ if( x264_threaded_slices_write( h ) )
+ return -1;
+ }
else
if( (intptr_t)x264_slices_write( h ) )
return -1;
x264_nal_t **pp_nal, int *pi_nal,
x264_picture_t *pic_out )
{
- int i, j, i_list, frame_size;
char psz_message[80];
if( h->b_thread_active )
{
void *ret = NULL;
x264_pthread_join( h->thread_handle, &ret );
+ h->b_thread_active = 0;
if( (intptr_t)ret )
return (intptr_t)ret;
- h->b_thread_active = 0;
}
if( !h->out.i_nal )
{
return 0;
}
- x264_frame_push_unused( thread_current, h->fenc );
-
- /* End bitstream, set output */
- *pi_nal = h->out.i_nal;
- *pp_nal = h->out.nal;
+ x264_emms();
+ /* generate sei buffering period and insert it into place */
+ if( h->fenc->b_keyframe && h->sps->vui.b_nal_hrd_parameters_present )
+ {
+ h->initial_cpb_removal_delay = x264_hrd_fullness( h );
- frame_size = x264_encoder_encapsulate_nals( h );
+ x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
+ x264_sei_buffering_period_write( h, &h->out.bs );
+ if( x264_nal_end( h ) )
+ return -1;
+ /* buffering period sei must follow AUD, SPS and PPS and precede all other SEIs */
+ int idx = 0;
+ while( h->out.nal[idx].i_type == NAL_AUD ||
+ h->out.nal[idx].i_type == NAL_SPS ||
+ h->out.nal[idx].i_type == NAL_PPS )
+ idx++;
+ x264_nal_t nal_tmp = h->out.nal[h->out.i_nal-1];
+ memmove( &h->out.nal[idx+1], &h->out.nal[idx], (h->out.i_nal-idx-1)*sizeof(x264_nal_t) );
+ h->out.nal[idx] = nal_tmp;
+ }
- h->out.i_nal = 0;
+ int frame_size = x264_encoder_encapsulate_nals( h, 0 );
/* Set output picture properties */
- if( h->sh.i_type == SLICE_TYPE_I )
- pic_out->i_type = h->i_nal_type == NAL_SLICE_IDR ? X264_TYPE_IDR : X264_TYPE_I;
- else if( h->sh.i_type == SLICE_TYPE_P )
- pic_out->i_type = X264_TYPE_P;
+ pic_out->i_type = h->fenc->i_type;
+
+ pic_out->b_keyframe = h->fenc->b_keyframe;
+
+ pic_out->i_pts = h->fenc->i_pts *= h->i_dts_compress_multiplier;
+ if( h->frames.i_bframe_delay )
+ {
+ int64_t *prev_reordered_pts = thread_current->frames.i_prev_reordered_pts;
+ if( h->i_frame <= h->frames.i_bframe_delay )
+ {
+ if( h->i_dts_compress_multiplier == 1 )
+ pic_out->i_dts = h->fenc->i_reordered_pts - h->frames.i_bframe_delay_time;
+ else
+ {
+ /* DTS compression */
+ if( h->i_frame == 1 )
+ thread_current->frames.i_init_delta = h->fenc->i_reordered_pts * h->i_dts_compress_multiplier;
+ pic_out->i_dts = h->i_frame * thread_current->frames.i_init_delta / h->i_dts_compress_multiplier;
+ }
+ }
+ else
+ pic_out->i_dts = prev_reordered_pts[ (h->i_frame - h->frames.i_bframe_delay) % h->frames.i_bframe_delay ];
+ prev_reordered_pts[ h->i_frame % h->frames.i_bframe_delay ] = h->fenc->i_reordered_pts * h->i_dts_compress_multiplier;
+ }
else
- pic_out->i_type = X264_TYPE_B;
- pic_out->i_pts = h->fenc->i_pts;
+ pic_out->i_dts = h->fenc->i_reordered_pts;
+ if( pic_out->i_pts < pic_out->i_dts )
+ x264_log( h, X264_LOG_WARNING, "invalid DTS: PTS is less than DTS\n" );
pic_out->img.i_plane = h->fdec->i_plane;
- for(i = 0; i < 3; i++)
+ for( int i = 0; i < 3; i++ )
{
pic_out->img.i_stride[i] = h->fdec->i_stride[i];
pic_out->img.plane[i] = h->fdec->plane[i];
}
+ x264_frame_push_unused( thread_current, h->fenc );
+
/* ---------------------- Update encoder state ------------------------- */
/* update rc */
- x264_emms();
- if( x264_ratecontrol_end( h, frame_size * 8 ) < 0 )
+ int filler = 0;
+ if( x264_ratecontrol_end( h, frame_size * 8, &filler ) < 0 )
return -1;
- x264_noise_reduction_update( thread_current );
+ pic_out->hrd_timing = h->fenc->hrd_timing;
+
+ while( filler > 0 )
+ {
+ int f, overhead;
+ overhead = (FILLER_OVERHEAD - h->param.b_annexb);
+ if( h->param.i_slice_max_size && filler > h->param.i_slice_max_size )
+ {
+ int next_size = filler - h->param.i_slice_max_size;
+ int overflow = X264_MAX( overhead - next_size, 0 );
+ f = h->param.i_slice_max_size - overhead - overflow;
+ }
+ else
+ f = X264_MAX( 0, filler - overhead );
+
+ x264_nal_start( h, NAL_FILLER, NAL_PRIORITY_DISPOSABLE );
+ x264_filler_write( h, &h->out.bs, f );
+ if( x264_nal_end( h ) )
+ return -1;
+ int total_size = x264_encoder_encapsulate_nals( h, h->out.i_nal-1 );
+ frame_size += total_size;
+ filler -= total_size;
+ }
+
+ /* End bitstream, set output */
+ *pi_nal = h->out.i_nal;
+ *pp_nal = h->out.nal;
+
+ h->out.i_nal = 0;
+
+ x264_noise_reduction_update( h );
/* ---------------------- Compute/Print statistics --------------------- */
x264_thread_sync_stat( h, h->thread[0] );
h->stat.i_frame_size[h->sh.i_type] += frame_size;
h->stat.f_frame_qp[h->sh.i_type] += h->fdec->f_qp_avg_aq;
- for( i = 0; i < X264_MBTYPE_MAX; i++ )
+ for( int i = 0; i < X264_MBTYPE_MAX; i++ )
h->stat.i_mb_count[h->sh.i_type][i] += h->stat.frame.i_mb_count[i];
- for( i = 0; i < X264_PARTTYPE_MAX; i++ )
+ for( int i = 0; i < X264_PARTTYPE_MAX; i++ )
h->stat.i_mb_partition[h->sh.i_type][i] += h->stat.frame.i_mb_partition[i];
- for( i = 0; i < 2; i++ )
+ for( int i = 0; i < 2; i++ )
h->stat.i_mb_count_8x8dct[i] += h->stat.frame.i_mb_count_8x8dct[i];
- for( i = 0; i < 6; i++ )
+ for( int i = 0; i < 6; i++ )
h->stat.i_mb_cbp[i] += h->stat.frame.i_mb_cbp[i];
- for( i = 0; i < 3; i++ )
- for( j = 0; j < 13; j++ )
+ for( int i = 0; i < 4; i++ )
+ for( int j = 0; j < 13; j++ )
h->stat.i_mb_pred_mode[i][j] += h->stat.frame.i_mb_pred_mode[i][j];
if( h->sh.i_type != SLICE_TYPE_I )
- for( i_list = 0; i_list < 2; i_list++ )
- for( i = 0; i < 32; i++ )
+ for( int i_list = 0; i_list < 2; i_list++ )
+ for( int i = 0; i < 32; i++ )
h->stat.i_mb_count_ref[h->sh.i_type][i_list][i] += h->stat.frame.i_mb_count_ref[i_list][i];
if( h->sh.i_type == SLICE_TYPE_P )
{
h->stat.i_consecutive_bframes[h->fdec->i_frame - h->fref0[0]->i_frame - 1]++;
if( h->param.analyse.i_weighted_pred == X264_WEIGHTP_SMART )
- {
- for( i = 0; i < 3; i++ )
- for( j = 0; j < h->i_ref0; j++ )
+ for( int i = 0; i < 3; i++ )
+ for( int j = 0; j < h->i_ref0; j++ )
if( h->sh.weight[0][i].i_denom != 0 )
{
h->stat.i_wpred[i]++;
break;
}
- }
}
if( h->sh.i_type == SLICE_TYPE_B )
{
{
//FIXME somewhat arbitrary time constants
if( h->stat.i_direct_score[0] + h->stat.i_direct_score[1] > h->mb.i_mb_count )
- {
- for( i = 0; i < 2; i++ )
+ for( int i = 0; i < 2; i++ )
h->stat.i_direct_score[i] = h->stat.i_direct_score[i] * 9/10;
- }
- for( i = 0; i < 2; i++ )
+ for( int i = 0; i < 2; i++ )
h->stat.i_direct_score[i] += h->stat.frame.i_direct_score[i];
}
}
{
static const char mb_chars[] = { 'i', 'i', 'I', 'C', 'P', '8', 'S',
'D', '<', 'X', 'B', 'X', '>', 'B', 'B', 'B', 'B', '8', 'S' };
- int mb_xy;
- for( mb_xy = 0; mb_xy < h->sps->i_mb_width * h->sps->i_mb_height; mb_xy++ )
+ for( int mb_xy = 0; mb_xy < h->mb.i_mb_width * h->mb.i_mb_height; mb_xy++ )
{
if( h->mb.type[mb_xy] < X264_MBTYPE_MAX && h->mb.type[mb_xy] >= 0 )
fprintf( stderr, "%c ", mb_chars[ h->mb.type[mb_xy] ] );
else
fprintf( stderr, "? " );
- if( (mb_xy+1) % h->sps->i_mb_width == 0 )
+ if( (mb_xy+1) % h->mb.i_mb_width == 0 )
fprintf( stderr, "\n" );
}
}
/* Remove duplicates, must be done near the end as breaks h->fref0 array
* by freeing some of its pointers. */
- for( i = 0; i < h->i_ref0; i++ )
+ for( int i = 0; i < h->i_ref0; i++ )
if( h->fref0[i] && h->fref0[i]->b_duplicate )
{
x264_frame_push_blank_unused( h, h->fref0[i] );
int64_t i_yuv_size = 3 * h->param.i_width * h->param.i_height / 2;
int64_t i_mb_count_size[2][7] = {{0}};
char buf[200];
- int i, j, i_list, i_type;
int b_print_pcm = h->stat.i_mb_count[SLICE_TYPE_I][I_PCM]
|| h->stat.i_mb_count[SLICE_TYPE_P][I_PCM]
|| h->stat.i_mb_count[SLICE_TYPE_B][I_PCM];
x264_lookahead_delete( h );
- for( i=0; i<h->param.i_threads; i++ )
+ if( h->param.i_threads > 1 )
{
// don't strictly have to wait for the other threads, but it's simpler than canceling them
- if( h->thread[i]->b_thread_active )
+ for( int i = 0; i < h->param.i_threads; i++ )
+ if( h->thread[i]->b_thread_active )
+ x264_pthread_join( h->thread[i]->thread_handle, NULL );
+ if( h->i_thread_frames > 1 )
{
- x264_pthread_join( h->thread[i]->thread_handle, NULL );
- assert( h->thread[i]->fenc->i_reference_count == 1 );
- x264_frame_delete( h->thread[i]->fenc );
- }
- }
-
- if( h->param.i_threads > 1 )
- {
- x264_t *thread_prev;
+ for( int i = 0; i < h->i_thread_frames; i++ )
+ if( h->thread[i]->b_thread_active )
+ {
+ assert( h->thread[i]->fenc->i_reference_count == 1 );
+ x264_frame_delete( h->thread[i]->fenc );
+ }
- thread_prev = h->thread[h->i_thread_phase];
- x264_thread_sync_ratecontrol( h, thread_prev, h );
- x264_thread_sync_ratecontrol( thread_prev, thread_prev, h );
- h->i_frame = thread_prev->i_frame + 1 - h->param.i_threads;
+ x264_t *thread_prev = h->thread[h->i_thread_phase];
+ x264_thread_sync_ratecontrol( h, thread_prev, h );
+ x264_thread_sync_ratecontrol( thread_prev, thread_prev, h );
+ h->i_frame = thread_prev->i_frame + 1 - h->i_thread_frames;
+ }
}
h->i_frame++;
/* Slices used and PSNR */
- for( i=0; i<5; i++ )
+ for( int i = 0; i < 5; i++ )
{
- static const int slice_order[] = { SLICE_TYPE_I, SLICE_TYPE_SI, SLICE_TYPE_P, SLICE_TYPE_SP, SLICE_TYPE_B };
- static const char *slice_name[] = { "P", "B", "I", "SP", "SI" };
+ static const uint8_t slice_order[] = { SLICE_TYPE_I, SLICE_TYPE_SI, SLICE_TYPE_P, SLICE_TYPE_SP, SLICE_TYPE_B };
+ static const char * const slice_name[] = { "P", "B", "I", "SP", "SI" };
int i_slice = slice_order[i];
if( h->stat.i_frame_count[i_slice] > 0 )
char *p = buf;
int den = 0;
// weight by number of frames (including the P-frame) that are in a sequence of N B-frames
- for( i=0; i<=h->param.i_bframe; i++ )
+ for( int i = 0; i <= h->param.i_bframe; i++ )
den += (i+1) * h->stat.i_consecutive_bframes[i];
- for( i=0; i<=h->param.i_bframe; i++ )
+ for( int i = 0; i <= h->param.i_bframe; i++ )
p += sprintf( p, " %4.1f%%", 100. * (i+1) * h->stat.i_consecutive_bframes[i] / den );
x264_log( h, X264_LOG_INFO, "consecutive B-frames:%s\n", buf );
}
- for( i_type = 0; i_type < 2; i_type++ )
- for( i = 0; i < X264_PARTTYPE_MAX; i++ )
+ for( int i_type = 0; i_type < 2; i_type++ )
+ for( int i = 0; i < X264_PARTTYPE_MAX; i++ )
{
if( i == D_DIRECT_8x8 ) continue; /* direct is counted as its own type */
i_mb_count_size[i_type][x264_mb_partition_pixel_table[i]] += h->stat.i_mb_partition[i_type][i];
int64_t *i_mb_size = i_mb_count_size[SLICE_TYPE_B];
int64_t list_count[3] = {0}; /* 0 == L0, 1 == L1, 2 == BI */
x264_print_intra( i_mb_count, i_count, b_print_pcm, buf );
- for( i = 0; i < X264_PARTTYPE_MAX; i++ )
- for( j = 0; j < 2; j++ )
+ for( int i = 0; i < X264_PARTTYPE_MAX; i++ )
+ for( int j = 0; j < 2; j++ )
{
int l0 = x264_mb_type_list_table[i][0][j];
int l1 = x264_mb_type_list_table[i][1][j];
h->stat.i_frame_count[SLICE_TYPE_B];
int64_t i_mb_count = i_count * h->mb.i_mb_count;
float fps = (float) h->param.i_fps_num / h->param.i_fps_den;
- float f_bitrate = fps * SUM3(h->stat.i_frame_size) / i_count / 125;
+ float f_bitrate;
+ /* duration algorithm fails with one frame */
+ if( !h->param.b_vfr_input || i_count == 1 )
+ f_bitrate = fps * SUM3(h->stat.i_frame_size) / i_count / 125;
+ else
+ {
+ float duration = (float)(2 * h->frames.i_largest_pts - h->frames.i_second_largest_pts) * h->param.i_timebase_num / h->param.i_timebase_den;
+ f_bitrate = SUM3(h->stat.i_frame_size) / duration / 125;
+ }
if( h->pps->b_transform_8x8_mode )
{
x264_log( h, X264_LOG_INFO, "8x8 transform intra:%.1f%%%s\n", 100. * i_i8x8 / i_intra, buf );
}
- if( h->param.analyse.i_direct_mv_pred == X264_DIRECT_PRED_AUTO
+ if( (h->param.analyse.i_direct_mv_pred == X264_DIRECT_PRED_AUTO ||
+ (h->stat.i_direct_frames[0] && h->stat.i_direct_frames[1]))
&& h->stat.i_frame_count[SLICE_TYPE_B] )
{
x264_log( h, X264_LOG_INFO, "direct mvs spatial:%.1f%% temporal:%.1f%%\n",
h->stat.i_mb_cbp[2] * 100.0 / (i_all_intra ),
h->stat.i_mb_cbp[4] * 100.0 / (i_all_intra ), buf );
- int64_t fixed_pred_modes[3][9] = {{0}};
- int64_t sum_pred_modes[3] = {0};
- for( i = 0; i <= I_PRED_16x16_DC_128; i++ )
+ int64_t fixed_pred_modes[4][9] = {{0}};
+ int64_t sum_pred_modes[4] = {0};
+ for( int i = 0; i <= I_PRED_16x16_DC_128; i++ )
{
fixed_pred_modes[0][x264_mb_pred_mode16x16_fix[i]] += h->stat.i_mb_pred_mode[0][i];
sum_pred_modes[0] += h->stat.i_mb_pred_mode[0][i];
fixed_pred_modes[0][1] * 100.0 / sum_pred_modes[0],
fixed_pred_modes[0][2] * 100.0 / sum_pred_modes[0],
fixed_pred_modes[0][3] * 100.0 / sum_pred_modes[0] );
- for( i = 1; i <= 2; i++ )
+ for( int i = 1; i <= 2; i++ )
{
- for( j = 0; j <= I_PRED_8x8_DC_128; j++ )
+ for( int j = 0; j <= I_PRED_8x8_DC_128; j++ )
{
fixed_pred_modes[i][x264_mb_pred_mode4x4_fix(j)] += h->stat.i_mb_pred_mode[i][j];
sum_pred_modes[i] += h->stat.i_mb_pred_mode[i][j];
fixed_pred_modes[i][7] * 100.0 / sum_pred_modes[i],
fixed_pred_modes[i][8] * 100.0 / sum_pred_modes[i] );
}
+ for( int i = 0; i <= I_PRED_CHROMA_DC_128; i++ )
+ {
+ fixed_pred_modes[3][x264_mb_pred_mode8x8c_fix[i]] += h->stat.i_mb_pred_mode[3][i];
+ sum_pred_modes[3] += h->stat.i_mb_pred_mode[3][i];
+ }
+ if( sum_pred_modes[3] )
+ x264_log( h, X264_LOG_INFO, "i8c dc,h,v,p: %2.0f%% %2.0f%% %2.0f%% %2.0f%%\n",
+ fixed_pred_modes[3][0] * 100.0 / sum_pred_modes[3],
+ fixed_pred_modes[3][1] * 100.0 / sum_pred_modes[3],
+ fixed_pred_modes[3][2] * 100.0 / sum_pred_modes[3],
+ fixed_pred_modes[3][3] * 100.0 / sum_pred_modes[3] );
if( h->param.analyse.i_weighted_pred == X264_WEIGHTP_SMART && h->stat.i_frame_count[SLICE_TYPE_P] > 0 )
x264_log( h, X264_LOG_INFO, "Weighted P-Frames: Y:%.1f%%\n",
h->stat.i_wpred[0] * 100.0 / h->stat.i_frame_count[SLICE_TYPE_P] );
- for( i_list = 0; i_list < 2; i_list++ )
- {
- int i_slice;
- for( i_slice = 0; i_slice < 2; i_slice++ )
+ for( int i_list = 0; i_list < 2; i_list++ )
+ for( int i_slice = 0; i_slice < 2; i_slice++ )
{
char *p = buf;
int64_t i_den = 0;
int i_max = 0;
- for( i = 0; i < 32; i++ )
+ for( int i = 0; i < 32; i++ )
if( h->stat.i_mb_count_ref[i_slice][i_list][i] )
{
i_den += h->stat.i_mb_count_ref[i_slice][i_list][i];
}
if( i_max == 0 )
continue;
- for( i = 0; i <= i_max; i++ )
+ for( int i = 0; i <= i_max; i++ )
p += sprintf( p, " %4.1f%%", 100. * h->stat.i_mb_count_ref[i_slice][i_list][i] / i_den );
x264_log( h, X264_LOG_INFO, "ref %c L%d:%s\n", "PB"[i_slice], i_list, buf );
}
- }
if( h->param.analyse.b_ssim )
{
- x264_log( h, X264_LOG_INFO,
- "SSIM Mean Y:%.7f\n",
- SUM3( h->stat.f_ssim_mean_y ) / i_count );
+ float ssim = SUM3( h->stat.f_ssim_mean_y ) / i_count;
+ x264_log( h, X264_LOG_INFO, "SSIM Mean Y:%.7f (%6.3fdb)\n", ssim, x264_ssim( ssim ) );
}
if( h->param.analyse.b_psnr )
{
x264_free( h->nal_buffer );
x264_analyse_free_costs( h );
- if( h->param.i_threads > 1)
+ if( h->i_thread_frames > 1)
h = h->thread[h->i_thread_phase];
/* frames */
h = h->thread[0];
- for( i = h->param.i_threads - 1; i >= 0; i-- )
+ for( int i = h->param.i_threads - 1; i >= 0; i-- )
{
x264_frame_t **frame;
- for( frame = h->thread[i]->frames.reference; *frame; frame++ )
+ if( !h->param.b_sliced_threads || i == 0 )
{
+ for( frame = h->thread[i]->frames.reference; *frame; frame++ )
+ {
+ assert( (*frame)->i_reference_count > 0 );
+ (*frame)->i_reference_count--;
+ if( (*frame)->i_reference_count == 0 )
+ x264_frame_delete( *frame );
+ }
+ frame = &h->thread[i]->fdec;
assert( (*frame)->i_reference_count > 0 );
(*frame)->i_reference_count--;
if( (*frame)->i_reference_count == 0 )
x264_frame_delete( *frame );
+ x264_macroblock_cache_free( h->thread[i] );
}
- frame = &h->thread[i]->fdec;
- assert( (*frame)->i_reference_count > 0 );
- (*frame)->i_reference_count--;
- if( (*frame)->i_reference_count == 0 )
- x264_frame_delete( *frame );
-
- x264_macroblock_cache_end( h->thread[i] );
+ x264_macroblock_thread_free( h->thread[i], 0 );
x264_free( h->thread[i]->out.p_bitstream );
x264_free( h->thread[i]->out.nal);
x264_free( h->thread[i] );
int x264_encoder_delayed_frames( x264_t *h )
{
int delayed_frames = 0;
- int i;
- for( i=0; i<h->param.i_threads; i++ )
- delayed_frames += h->thread[i]->b_thread_active;
- h = h->thread[h->i_thread_phase];
- for( i=0; h->frames.current[i]; i++ )
+ if( h->i_thread_frames > 1 )
+ {
+ for( int i = 0; i < h->i_thread_frames; i++ )
+ delayed_frames += h->thread[i]->b_thread_active;
+ h = h->thread[h->i_thread_phase];
+ }
+ for( int i = 0; h->frames.current[i]; i++ )
delayed_frames++;
x264_pthread_mutex_lock( &h->lookahead->ofbuf.mutex );
x264_pthread_mutex_lock( &h->lookahead->ifbuf.mutex );
projects / x264 / blobdiff