ffmpeg/libavcodec/ituh263enc.c

957 lines
32 KiB
C
Raw Normal View History

/*
* ITU H.263 bitstream encoder
* Copyright (c) 2000,2001 Fabrice Bellard
* H.263+ support.
* Copyright (c) 2001 Juan J. Sierralta P
* Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* H.263 bitstream encoder.
*/
#include "config_components.h"
#include <limits.h>
#include <string.h>
#include "libavutil/attributes.h"
#include "libavutil/thread.h"
#include "avcodec.h"
#include "codec_internal.h"
#include "mpegvideo.h"
#include "flvenc.h"
#include "mpegvideoenc.h"
#include "h263.h"
#include "h263enc.h"
#include "h263data.h"
#include "h263dsp.h"
#include "mathops.h"
#include "mpegutils.h"
#include "internal.h"
#include "put_bits.h"
/**
* Table of number of bits a motion vector component needs.
*/
static uint8_t mv_penalty[MAX_FCODE+1][MAX_DMV*2+1];
static av_cold void init_mv_penalty(void)
{
for (int f_code = 1; f_code <= MAX_FCODE; f_code++) {
for (int mv = -MAX_DMV; mv <= MAX_DMV; mv++) {
int len;
if (mv == 0) len = 1; // ff_mvtab[0][1]
else {
int val, bit_size, code;
bit_size = f_code - 1;
val = mv;
if (val < 0)
val = -val;
val--;
code = (val >> bit_size) + 1;
if (code < 33) {
len = ff_mvtab[code][1] + 1 + bit_size;
} else {
len = 12 /* ff_mvtab[32][1] */ + av_log2(code>>5) + 2 + bit_size;
}
}
mv_penalty[f_code][mv + MAX_DMV] = len;
}
}
}
#if CONFIG_H263_ENCODER
/**
* Minimal fcode that a motion vector component would need in umv.
* All entries in this table are 1.
*/
static uint8_t umv_fcode_tab[MAX_MV*2+1];
//unified encoding tables for run length encoding of coefficients
//unified in the sense that the specification specifies the encoding in several steps.
static uint8_t uni_h263_intra_aic_rl_len [64*64*2*2];
static uint8_t uni_h263_inter_rl_len [64*64*2*2];
//#define UNI_MPEG4_ENC_INDEX(last,run,level) ((last)*128 + (run)*256 + (level))
//#define UNI_MPEG4_ENC_INDEX(last,run,level) ((last)*128*64 + (run) + (level)*64)
#define UNI_MPEG4_ENC_INDEX(last,run,level) ((last)*128*64 + (run)*128 + (level))
static av_cold void init_uni_h263_rl_tab(const RLTable *rl, uint8_t *len_tab)
{
const uint16_t (*table_vlc)[2] = rl->table_vlc;
const uint8_t *table_run = rl->table_run;
const uint8_t *table_level = rl->table_level;
av_assert0(MAX_LEVEL >= 64);
av_assert0(MAX_RUN >= 63);
// Note: The LUT only covers level values for which the escape value
// is eight bits (not 8 + 5 + 6)
memset(len_tab, H263_ESCAPE_CODE_LENGTH + 1 + 6 + 8,
sizeof(uni_h263_intra_aic_rl_len));
len_tab += 64; // simplifies addressing
for (int i = 0; i < H263_RL_NB_ELEMS; ++i) {
int run = table_run[i];
int level = table_level[i];
int last = i >= H263_RL_NON_LAST_CODES;
int len = table_vlc[i][1];
len_tab[UNI_MPEG4_ENC_INDEX(last, run, level)] =
len_tab[UNI_MPEG4_ENC_INDEX(last, run, -level)] = len + 1 /* sign */;
}
for (int run = 0; run < MAX_RUN; ++run) {
len_tab[UNI_MPEG4_ENC_INDEX(0, run, 0)] =
len_tab[UNI_MPEG4_ENC_INDEX(1, run, 0)] = 0; // is this necessary?
}
}
#endif
static av_cold void h263_encode_init_static(void)
{
#if CONFIG_H263_ENCODER
static uint8_t rl_intra_table[2][2 * MAX_RUN + MAX_LEVEL + 3];
ff_rl_init(&ff_rl_intra_aic, rl_intra_table);
ff_h263_init_rl_inter();
init_uni_h263_rl_tab(&ff_rl_intra_aic, uni_h263_intra_aic_rl_len);
init_uni_h263_rl_tab(&ff_h263_rl_inter, uni_h263_inter_rl_len);
memset(umv_fcode_tab, 1, sizeof(umv_fcode_tab));
#endif
init_mv_penalty();
}
av_cold const uint8_t (*ff_h263_get_mv_penalty(void))[MAX_DMV*2+1]
{
static AVOnce init_static_once = AV_ONCE_INIT;
ff_thread_once(&init_static_once, h263_encode_init_static);
return mv_penalty;
}
void ff_h263_encode_motion(PutBitContext *pb, int val, int f_code)
{
if (val == 0) {
/* zero vector -- corresponds to ff_mvtab[0] */
put_bits(pb, 1, 1);
} else {
int sign, code, bits;
int bit_size = f_code - 1;
int range = 1 << bit_size;
/* modulo encoding */
val = sign_extend(val, 6 + bit_size);
sign = val>>31;
val= (val^sign)-sign;
sign&=1;
val--;
code = (val >> bit_size) + 1;
bits = val & (range - 1);
put_bits(pb, ff_mvtab[code][1] + 1, (ff_mvtab[code][0] << 1) | sign);
if (bit_size > 0) {
put_bits(pb, bit_size, bits);
}
}
}
#if CONFIG_H263_ENCODER // Snow and SVQ1 need the above
static const uint8_t wrong_run[102] = {
1, 2, 3, 5, 4, 10, 9, 8,
11, 15, 17, 16, 23, 22, 21, 20,
19, 18, 25, 24, 27, 26, 11, 7,
6, 1, 2, 13, 2, 2, 2, 2,
6, 12, 3, 9, 1, 3, 4, 3,
7, 4, 1, 1, 5, 5, 14, 6,
1, 7, 1, 8, 1, 1, 1, 1,
10, 1, 1, 5, 9, 17, 25, 24,
29, 33, 32, 41, 2, 23, 28, 31,
3, 22, 30, 4, 27, 40, 8, 26,
6, 39, 7, 38, 16, 37, 15, 10,
11, 12, 13, 14, 1, 21, 20, 18,
19, 2, 1, 34, 35, 36
};
/**
* Return the 4 bit value that specifies the given aspect ratio.
* This may be one of the standard aspect ratios or it specifies
* that the aspect will be stored explicitly later.
*/
av_const int ff_h263_aspect_to_info(AVRational aspect){
int i;
if(aspect.num==0 || aspect.den==0) aspect= (AVRational){1,1};
for(i=1; i<6; i++){
if(av_cmp_q(ff_h263_pixel_aspect[i], aspect) == 0){
return i;
}
}
return FF_ASPECT_EXTENDED;
}
static int h263_encode_picture_header(MPVMainEncContext *const m)
{
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
MPVEncContext *const s = &m->s;
int format, coded_frame_rate, coded_frame_rate_base, i, temp_ref;
int best_clock_code=1;
int best_divisor=60;
int best_error= INT_MAX;
int custom_pcf;
put_bits_assume_flushed(&s->pb);
if (s->c.codec_id == AV_CODEC_ID_H263P) {
for(i=0; i<2; i++){
int div, error;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
div= (s->c.avctx->time_base.num*1800000LL + 500LL*s->c.avctx->time_base.den) / ((1000LL+i)*s->c.avctx->time_base.den);
div= av_clip(div, 1, 127);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
error= FFABS(s->c.avctx->time_base.num*1800000LL - (1000LL+i)*s->c.avctx->time_base.den*div);
if(error < best_error){
best_error= error;
best_divisor= div;
best_clock_code= i;
}
}
}
custom_pcf = best_clock_code != 1 || best_divisor != 60;
coded_frame_rate= 1800000;
coded_frame_rate_base= (1000+best_clock_code)*best_divisor;
put_bits(&s->pb, 22, 0x20); /* PSC */
temp_ref = s->picture_number * (int64_t)coded_frame_rate * s->c.avctx->time_base.num / //FIXME use timestamp
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
(coded_frame_rate_base * (int64_t)s->c.avctx->time_base.den);
put_sbits(&s->pb, 8, temp_ref); /* TemporalReference */
put_bits(&s->pb, 1, 1); /* marker */
put_bits(&s->pb, 1, 0); /* H.263 id */
put_bits(&s->pb, 1, 0); /* split screen off */
put_bits(&s->pb, 1, 0); /* camera off */
put_bits(&s->pb, 1, 0); /* freeze picture release off */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
format = ff_match_2uint16(ff_h263_format, FF_ARRAY_ELEMS(ff_h263_format), s->c.width, s->c.height);
if (s->c.codec_id != AV_CODEC_ID_H263P) {
/* H.263v1 */
put_bits(&s->pb, 3, format);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 1, (s->c.pict_type == AV_PICTURE_TYPE_P));
/* By now UMV IS DISABLED ON H.263v1, since the restrictions
of H.263v1 UMV implies to check the predicted MV after
calculation of the current MB to see if we're on the limits */
put_bits(&s->pb, 1, 0); /* Unrestricted Motion Vector: off */
put_bits(&s->pb, 1, 0); /* SAC: off */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 1, s->c.obmc); /* Advanced Prediction */
put_bits(&s->pb, 1, 0); /* only I/P-frames, no PB-frame */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 5, s->c.qscale);
put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */
} else {
int ufep=1;
/* H.263v2 */
/* H.263 Plus PTYPE */
put_bits(&s->pb, 3, 7);
put_bits(&s->pb,3,ufep); /* Update Full Extended PTYPE */
if (format == 8)
put_bits(&s->pb,3,6); /* Custom Source Format */
else
put_bits(&s->pb, 3, format);
put_bits(&s->pb,1, custom_pcf);
put_bits(&s->pb,1, s->umvplus); /* Unrestricted Motion Vector */
put_bits(&s->pb,1,0); /* SAC: off */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb,1,s->c.obmc); /* Advanced Prediction Mode */
put_bits(&s->pb,1,s->c.h263_aic); /* Advanced Intra Coding */
put_bits(&s->pb,1,s->loop_filter); /* Deblocking Filter */
put_bits(&s->pb,1,s->h263_slice_structured); /* Slice Structured */
put_bits(&s->pb,1,0); /* Reference Picture Selection: off */
put_bits(&s->pb,1,0); /* Independent Segment Decoding: off */
put_bits(&s->pb,1,s->alt_inter_vlc); /* Alternative Inter VLC */
put_bits(&s->pb,1,s->modified_quant); /* Modified Quantization: */
put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
put_bits(&s->pb,3,0); /* Reserved */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 3, s->c.pict_type == AV_PICTURE_TYPE_P);
put_bits(&s->pb,1,0); /* Reference Picture Resampling: off */
put_bits(&s->pb,1,0); /* Reduced-Resolution Update: off */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb,1,s->c.no_rounding); /* Rounding Type */
put_bits(&s->pb,2,0); /* Reserved */
put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
/* This should be here if PLUSPTYPE */
put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */
if (format == 8) {
/* Custom Picture Format (CPFMT) */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
unsigned aspect_ratio_info = ff_h263_aspect_to_info(s->c.avctx->sample_aspect_ratio);
put_bits(&s->pb,4, aspect_ratio_info);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb,9,(s->c.width >> 2) - 1);
put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb,9,(s->c.height >> 2));
if (aspect_ratio_info == FF_ASPECT_EXTENDED){
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 8, s->c.avctx->sample_aspect_ratio.num);
put_bits(&s->pb, 8, s->c.avctx->sample_aspect_ratio.den);
}
}
if (custom_pcf) {
if(ufep){
put_bits(&s->pb, 1, best_clock_code);
put_bits(&s->pb, 7, best_divisor);
}
put_sbits(&s->pb, 2, temp_ref>>8);
}
/* Unlimited Unrestricted Motion Vectors Indicator (UUI) */
if (s->umvplus)
// put_bits(&s->pb,1,1); /* Limited according tables of Annex D */
//FIXME check actual requested range
put_bits(&s->pb,2,1); /* unlimited */
if (s->h263_slice_structured)
put_bits(&s->pb,2,0); /* no weird submodes */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 5, s->c.qscale);
}
put_bits(&s->pb, 1, 0); /* no PEI */
if (s->h263_slice_structured) {
put_bits(&s->pb, 1, 1);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
av_assert1(s->c.mb_x == 0 && s->c.mb_y == 0);
ff_h263_encode_mba(s);
put_bits(&s->pb, 1, 1);
}
return 0;
}
void ff_h263_mpeg4_reset_dc(MPVEncContext *s)
{
int16_t *dc = s->c.dc_val;
// The "- 1" is for the top-left entry
const int l_xy = s->c.block_index[2];
for (int i = l_xy - 2 * s->c.b8_stride - 1; i < l_xy; i += 2)
AV_WN32A(dc + i, 1024 << 16 | 1024);
const int u_xy = s->c.block_index[4];
const int v_xy = s->c.block_index[5];
int16_t *dc2 = dc + v_xy - u_xy;
for (int i = u_xy - s->c.mb_stride - 1; i < u_xy; ++i)
dc[i] = dc2[i] = 1024;
}
/**
* Encode a group of blocks header.
*/
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
void ff_h263_encode_gob_header(MPVEncContext *const s, int mb_line)
{
put_bits(&s->pb, 17, 1); /* GBSC */
if (s->h263_slice_structured) {
put_bits(&s->pb, 1, 1);
ff_h263_encode_mba(s);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if(s->c.mb_num > 1583)
put_bits(&s->pb, 1, 1);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 5, s->c.qscale); /* GQUANT */
put_bits(&s->pb, 1, 1);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 2, s->c.pict_type == AV_PICTURE_TYPE_I); /* GFID */
}else{
int gob_number = mb_line / s->gob_index;
put_bits(&s->pb, 5, gob_number); /* GN */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
put_bits(&s->pb, 2, s->c.pict_type == AV_PICTURE_TYPE_I); /* GFID */
put_bits(&s->pb, 5, s->c.qscale); /* GQUANT */
}
}
/**
* modify qscale so that encoding is actually possible in H.263 (limit difference to -2..2)
*/
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
void ff_clean_h263_qscales(MPVEncContext *const s)
{
int8_t * const qscale_table = s->c.cur_pic.qscale_table;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
for (int i = 1; i < s->c.mb_num; i++) {
if (qscale_table[ s->c.mb_index2xy[i] ] - qscale_table[ s->c.mb_index2xy[i-1] ] > 2)
qscale_table[ s->c.mb_index2xy[i] ] = qscale_table[ s->c.mb_index2xy[i-1] ] + 2;
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
for(int i = s->c.mb_num - 2; i >= 0; i--) {
if (qscale_table[ s->c.mb_index2xy[i] ] - qscale_table[ s->c.mb_index2xy[i+1] ] > 2)
qscale_table[ s->c.mb_index2xy[i] ] = qscale_table[ s->c.mb_index2xy[i+1] ] + 2;
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.codec_id != AV_CODEC_ID_H263P) {
for (int i = 1; i < s->c.mb_num; i++) {
int mb_xy = s->c.mb_index2xy[i];
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (qscale_table[mb_xy] != qscale_table[s->c.mb_index2xy[i - 1]] &&
(s->mb_type[mb_xy] & CANDIDATE_MB_TYPE_INTER4V)) {
s->mb_type[mb_xy]|= CANDIDATE_MB_TYPE_INTER;
}
}
}
}
static const int dquant_code[5]= {1,0,9,2,3};
static void flv2_encode_ac_esc(PutBitContext *pb, int slevel, int level,
int run, int last)
{
unsigned code;
int bits;
if (level < 64) { // 7-bit level
bits = 1 + 1 + 6 + 7;
code = (0 << (1 + 6 + 7)) |
(last << (6 + 7)) |
(run << 7) |
(slevel & 0x7f);
} else {
/* 11-bit level */
bits = 1 + 1 + 6 + 11;
code = (1 << (1 + 6 + 11)) |
(last << (6 + 11)) |
(run << 11) |
(slevel & 0x7ff);
}
put_bits(pb, bits, code);
}
/**
* Encode an 8x8 block.
* @param block the 8x8 block
* @param n block index (0-3 are luma, 4-5 are chroma)
*/
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
static void h263_encode_block(MPVEncContext *const s, int16_t block[], int n)
{
int level, run, last, i, j, last_index, last_non_zero, sign, slevel, code;
const RLTable *rl;
rl = &ff_h263_rl_inter;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.mb_intra && !s->c.h263_aic) {
/* DC coef */
level = block[0];
/* 255 cannot be represented, so we clamp */
if (level > 254) {
level = 254;
block[0] = 254;
}
/* 0 cannot be represented also */
else if (level < 1) {
level = 1;
block[0] = 1;
}
if (level == 128) //FIXME check rv10
put_bits(&s->pb, 8, 0xff);
else
put_bits(&s->pb, 8, level);
i = 1;
} else {
i = 0;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.h263_aic && s->c.mb_intra)
rl = &ff_rl_intra_aic;
if (s->alt_inter_vlc && !s->c.mb_intra) {
int aic_vlc_bits=0;
int inter_vlc_bits=0;
int wrong_pos=-1;
int aic_code;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
last_index = s->c.block_last_index[n];
last_non_zero = i - 1;
for (; i <= last_index; i++) {
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
j = s->c.intra_scantable.permutated[i];
level = block[j];
if (level) {
run = i - last_non_zero - 1;
last = (i == last_index);
if(level<0) level= -level;
code = get_rl_index(rl, last, run, level);
aic_code = get_rl_index(&ff_rl_intra_aic, last, run, level);
inter_vlc_bits += rl->table_vlc[code][1]+1;
aic_vlc_bits += ff_rl_intra_aic.table_vlc[aic_code][1]+1;
if (code == rl->n) {
inter_vlc_bits += 1+6+8-1;
}
if (aic_code == ff_rl_intra_aic.n) {
aic_vlc_bits += 1+6+8-1;
wrong_pos += run + 1;
}else
wrong_pos += wrong_run[aic_code];
last_non_zero = i;
}
}
i = 0;
if(aic_vlc_bits < inter_vlc_bits && wrong_pos > 63)
rl = &ff_rl_intra_aic;
}
}
/* AC coefs */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
last_index = s->c.block_last_index[n];
last_non_zero = i - 1;
for (; i <= last_index; i++) {
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
j = s->c.intra_scantable.permutated[i];
level = block[j];
if (level) {
run = i - last_non_zero - 1;
last = (i == last_index);
sign = 0;
slevel = level;
if (level < 0) {
sign = 1;
level = -level;
}
code = get_rl_index(rl, last, run, level);
put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
if (code == rl->n) {
if (!CONFIG_FLV_ENCODER || s->c.codec_id != AV_CODEC_ID_FLV1) {
put_bits(&s->pb, 1, last);
put_bits(&s->pb, 6, run);
av_assert2(slevel != 0);
if (level < 128) {
put_sbits(&s->pb, 8, slevel);
} else {
put_bits(&s->pb, 8, 128);
put_sbits(&s->pb, 5, slevel);
put_sbits(&s->pb, 6, slevel>>5);
}
} else {
flv2_encode_ac_esc(&s->pb, slevel, level, run, last);
}
} else {
put_bits(&s->pb, 1, sign);
}
last_non_zero = i;
}
}
}
/* Encode MV differences on H.263+ with Unrestricted MV mode */
static void h263p_encode_umotion(PutBitContext *pb, int val)
{
if ( val == 0)
put_bits(pb, 1, 1);
else {
unsigned code = (val < 0) << 1;
unsigned aval = val < 0 ? -val : val;
unsigned n_bits = 2;
while (aval != 1) { // The leading digit is implicitly coded via length
unsigned tmp = (aval & 1) << 1 | 1;
aval >>= 1;
code |= tmp << n_bits;
n_bits += 2;
}
put_bits(pb, n_bits + 1, code);
}
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
static int h263_pred_dc(MPVEncContext *const s, int n, int16_t **dc_val_ptr)
{
const int wrap = s->c.block_wrap[n];
const int xy = s->c.block_index[n];
int16_t *const dc_val = s->c.dc_val + xy;
int pred_dc;
/* find prediction */
/* B C
* A X
*/
int a = dc_val[-1];
int c = dc_val[-wrap];
/* just DC prediction */
if (a != 1024 && c != 1024)
pred_dc = (a + c) >> 1;
else if (a != 1024)
pred_dc = a;
else
pred_dc = c;
/* we assume pred is positive */
*dc_val_ptr = dc_val;
return pred_dc;
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
static void h263_encode_mb(MPVEncContext *const s,
int16_t block[][64],
int motion_x, int motion_y)
{
int cbpc, cbpy, i, cbp, pred_x, pred_y;
int16_t pred_dc;
int16_t rec_intradc[6];
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
const int interleaved_stats = s->c.avctx->flags & AV_CODEC_FLAG_PASS1;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (!s->c.mb_intra) {
/* compute cbp */
cbp= get_p_cbp(s, block, motion_x, motion_y);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if ((cbp | motion_x | motion_y | s->dquant | (s->c.mv_type - MV_TYPE_16X16)) == 0) {
/* skip macroblock */
put_bits(&s->pb, 1, 1);
if(interleaved_stats){
s->misc_bits++;
s->last_bits++;
}
return;
}
put_bits(&s->pb, 1, 0); /* mb coded */
cbpc = cbp & 3;
cbpy = cbp >> 2;
if (!s->alt_inter_vlc || cbpc!=3)
cbpy ^= 0xF;
if(s->dquant) cbpc+= 8;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if(s->c.mv_type==MV_TYPE_16X16){
put_bits(&s->pb,
ff_h263_inter_MCBPC_bits[cbpc],
ff_h263_inter_MCBPC_code[cbpc]);
put_bits(&s->pb, ff_h263_cbpy_tab[cbpy][1], ff_h263_cbpy_tab[cbpy][0]);
if(s->dquant)
put_bits(&s->pb, 2, dquant_code[s->dquant+2]);
if(interleaved_stats){
s->misc_bits+= get_bits_diff(s);
}
/* motion vectors: 16x16 mode */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
ff_h263_pred_motion(&s->c, 0, 0, &pred_x, &pred_y);
if (!s->umvplus) {
ff_h263_encode_motion_vector(s, motion_x - pred_x,
motion_y - pred_y, 1);
}
else {
h263p_encode_umotion(&s->pb, motion_x - pred_x);
h263p_encode_umotion(&s->pb, motion_y - pred_y);
if (((motion_x - pred_x) == 1) && ((motion_y - pred_y) == 1))
/* To prevent Start Code emulation */
put_bits(&s->pb,1,1);
}
}else{
put_bits(&s->pb,
ff_h263_inter_MCBPC_bits[cbpc+16],
ff_h263_inter_MCBPC_code[cbpc+16]);
put_bits(&s->pb, ff_h263_cbpy_tab[cbpy][1], ff_h263_cbpy_tab[cbpy][0]);
if(s->dquant)
put_bits(&s->pb, 2, dquant_code[s->dquant+2]);
if(interleaved_stats){
s->misc_bits+= get_bits_diff(s);
}
for(i=0; i<4; i++){
/* motion vectors: 8x8 mode*/
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
ff_h263_pred_motion(&s->c, i, 0, &pred_x, &pred_y);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
motion_x = s->c.cur_pic.motion_val[0][s->c.block_index[i]][0];
motion_y = s->c.cur_pic.motion_val[0][s->c.block_index[i]][1];
if (!s->umvplus) {
ff_h263_encode_motion_vector(s, motion_x - pred_x,
motion_y - pred_y, 1);
}
else {
h263p_encode_umotion(&s->pb, motion_x - pred_x);
h263p_encode_umotion(&s->pb, motion_y - pred_y);
if (((motion_x - pred_x) == 1) && ((motion_y - pred_y) == 1))
/* To prevent Start Code emulation */
put_bits(&s->pb,1,1);
}
}
}
if(interleaved_stats){
s->mv_bits+= get_bits_diff(s);
}
} else {
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
av_assert2(s->c.mb_intra);
cbp = 0;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.h263_aic) {
/* Predict DC */
for(i=0; i<6; i++) {
int16_t level = block[i][0];
int16_t *dc_ptr;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
int scale = i < 4 ? s->c.y_dc_scale : s->c.c_dc_scale;
pred_dc = h263_pred_dc(s, i, &dc_ptr);
level -= pred_dc;
/* Quant */
if (level >= 0)
level = (level + (scale>>1))/scale;
else
level = (level - (scale>>1))/scale;
if (!s->modified_quant) {
if (level < -127)
level = -127;
else if (level > 127)
level = 127;
}
block[i][0] = level;
/* Reconstruction */
rec_intradc[i] = scale*level + pred_dc;
/* Oddify */
rec_intradc[i] |= 1;
//if ((rec_intradc[i] % 2) == 0)
// rec_intradc[i]++;
/* Clipping */
if (rec_intradc[i] < 0)
rec_intradc[i] = 0;
else if (rec_intradc[i] > 2047)
rec_intradc[i] = 2047;
/* Update AC/DC tables */
*dc_ptr = rec_intradc[i];
/* AIC can change CBP */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.block_last_index[i] > 0 ||
(s->c.block_last_index[i] == 0 && level !=0))
cbp |= 1 << (5 - i);
}
}else{
for(i=0; i<6; i++) {
/* compute cbp */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.block_last_index[i] >= 1)
cbp |= 1 << (5 - i);
}
}
cbpc = cbp & 3;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.pict_type == AV_PICTURE_TYPE_I) {
if(s->dquant) cbpc+=4;
put_bits(&s->pb,
ff_h263_intra_MCBPC_bits[cbpc],
ff_h263_intra_MCBPC_code[cbpc]);
} else {
if(s->dquant) cbpc+=8;
put_bits(&s->pb, 1, 0); /* mb coded */
put_bits(&s->pb,
ff_h263_inter_MCBPC_bits[cbpc + 4],
ff_h263_inter_MCBPC_code[cbpc + 4]);
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.h263_aic) {
/* XXX: currently, we do not try to use ac prediction */
put_bits(&s->pb, 1, 0); /* no AC prediction */
}
cbpy = cbp >> 2;
put_bits(&s->pb, ff_h263_cbpy_tab[cbpy][1], ff_h263_cbpy_tab[cbpy][0]);
if(s->dquant)
put_bits(&s->pb, 2, dquant_code[s->dquant+2]);
if(interleaved_stats){
s->misc_bits+= get_bits_diff(s);
}
}
for(i=0; i<6; i++) {
/* encode each block */
h263_encode_block(s, block[i], i);
/* Update INTRADC for decoding */
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.h263_aic && s->c.mb_intra)
block[i][0] = rec_intradc[i];
}
if(interleaved_stats){
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (!s->c.mb_intra) {
s->p_tex_bits+= get_bits_diff(s);
}else{
s->i_tex_bits+= get_bits_diff(s);
s->i_count++;
}
}
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
void ff_h263_update_mb(MPVEncContext *const s)
{
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
const int mb_xy = s->c.mb_y * s->c.mb_stride + s->c.mb_x;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.cur_pic.mbskip_table)
s->c.cur_pic.mbskip_table[mb_xy] = s->c.mb_skipped;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.mv_type == MV_TYPE_8X8)
s->c.cur_pic.mb_type[mb_xy] = MB_TYPE_FORWARD_MV | MB_TYPE_8x8;
else if(s->c.mb_intra)
s->c.cur_pic.mb_type[mb_xy] = MB_TYPE_INTRA;
else
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
s->c.cur_pic.mb_type[mb_xy] = MB_TYPE_FORWARD_MV | MB_TYPE_16x16;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
ff_h263_update_motion_val(&s->c);
}
av_cold void ff_h263_encode_init(MPVMainEncContext *const m)
{
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
MPVEncContext *const s = &m->s;
s->me.mv_penalty = ff_h263_get_mv_penalty(); // FIXME exact table for MSMPEG4 & H.263+
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
ff_h263dsp_init(&s->c.h263dsp);
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.codec_id == AV_CODEC_ID_MPEG4)
return;
s->intra_ac_vlc_length =s->inter_ac_vlc_length = uni_h263_inter_rl_len;
s->intra_ac_vlc_last_length=s->inter_ac_vlc_last_length= uni_h263_inter_rl_len + 128*64;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if (s->c.h263_aic) {
s->intra_ac_vlc_length = uni_h263_intra_aic_rl_len;
s->intra_ac_vlc_last_length= uni_h263_intra_aic_rl_len + 128*64;
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
s->c.y_dc_scale_table =
s->c.c_dc_scale_table = ff_aic_dc_scale_table;
}
s->ac_esc_length= 7+1+6+8;
if (s->modified_quant)
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
s->c.chroma_qscale_table = ff_h263_chroma_qscale_table;
// Only used for H.263 and H.263+
s->gob_index = H263_GOB_HEIGHT(s->c.height);
// use fcodes >1 only for MPEG-4 & H.263 & H.263+ FIXME
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
switch(s->c.codec_id){
case AV_CODEC_ID_H263P:
if (s->umvplus)
m->fcode_tab = umv_fcode_tab + MAX_MV;
if (s->modified_quant) {
s->min_qcoeff= -2047;
s->max_qcoeff= 2047;
}else{
s->min_qcoeff= -127;
s->max_qcoeff= 127;
}
break;
// Note for MPEG-4 & H.263 the dc-scale table will be set per frame as needed later
#if CONFIG_FLV_ENCODER
case AV_CODEC_ID_FLV1:
m->encode_picture_header = ff_flv_encode_picture_header;
/* format = 1; 11-bit codes */
s->min_qcoeff = -1023;
s->max_qcoeff = 1023;
break;
#endif
default: //nothing needed - default table already set in mpegvideo.c
s->min_qcoeff= -127;
s->max_qcoeff= 127;
}
// H.263, H.263+; will be overwritten for MSMPEG-4 later
if (!m->encode_picture_header)
m->encode_picture_header = h263_encode_picture_header;
if (!s->encode_mb)
s->encode_mb = h263_encode_mb;
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
void ff_h263_encode_mba(MPVEncContext *const s)
{
int i, mb_pos;
for(i=0; i<6; i++){
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
if(s->c.mb_num-1 <= ff_mba_max[i]) break;
}
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
mb_pos= s->c.mb_x + s->c.mb_width*s->c.mb_y;
put_bits(&s->pb, ff_mba_length[i], mb_pos);
}
#define OFFSET(x) offsetof(MpegEncContext, x)
#define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
static const AVOption h263_options[] = {
{ "obmc", "use overlapped block motion compensation.", OFFSET(obmc), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE },
avcodec/mpegvideoenc: Add MPVEncContext Many of the fields of MpegEncContext (which is also used by decoders) are actually only used by encoders. Therefore this commit adds a new encoder-only structure and moves all of the encoder-only fields to it except for those which require more explicit synchronisation between the main slice context and the other slice contexts. This synchronisation is currently mainly provided by ff_update_thread_context() which simply copies most of the main slice context over the other slice contexts. Fields which are moved to the new MPVEncContext no longer participate in this (which is desired, because it is horrible and for the fields b) below wasteful) which means that some fields can only be moved when explicit synchronisation code is added in later commits. More explicitly, this commit moves the following fields: a) Fields not copied by ff_update_duplicate_context(): dct_error_sum and dct_count; the former does not need synchronisation, the latter is synchronised in merge_context_after_encode(). b) Fields which do not change after initialisation (these fields could also be put into MPVMainEncContext at the cost of an indirection to access them): lambda_table, adaptive_quant, {luma,chroma}_elim_threshold, new_pic, fdsp, mpvencdsp, pdsp, {p,b_forw,b_back,b_bidir_forw,b_bidir_back,b_direct,b_field}_mv_table, [pb]_field_select_table, mb_{type,var,mean}, mc_mb_var, {min,max}_qcoeff, {inter,intra}_quant_bias, ac_esc_length, the *_vlc_length fields, the q_{intra,inter,chroma_intra}_matrix{,16}, dct_offset, mb_info, mjpeg_ctx, rtp_mode, rtp_payload_size, encode_mb, all function pointers, mpv_flags, quantizer_noise_shaping, frame_reconstruction_bitfield, error_rate and intra_penalty. c) Fields which are already (re)set explicitly: The PutBitContexts pb, tex_pb, pb2; dquant, skipdct, encoding_error, the statistics fields {mv,i_tex,p_tex,misc,last}_bits and i_count; last_mv_dir, esc_pos (reset when writing the header). d) Fields which are only used by encoders not supporting slice threading for which synchronisation doesn't matter: esc3_level_length and the remaining mb_info fields. e) coded_score: This field is only really used when FF_MPV_FLAG_CBP_RD is set (which implies trellis) and even then it is only used for non-intra blocks. For these blocks dct_quantize_trellis_c() either sets coded_score[n] or returns a last_non_zero value of -1 in which case coded_score will be reset in encode_mb_internal(). Therefore no old values are ever used. The MotionEstContext has not been moved yet. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2025-03-19 08:11:01 +01:00
{ "mb_info", "emit macroblock info for RFC 2190 packetization, the parameter value is the maximum payload size", FF_MPV_OFFSET(mb_info), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, VE },
FF_MPV_COMMON_OPTS
FF_MPV_COMMON_MOTION_EST_OPTS
{ NULL },
};
static const AVClass h263_class = {
.class_name = "H.263 encoder",
.item_name = av_default_item_name,
.option = h263_options,
.version = LIBAVUTIL_VERSION_INT,
};
const FFCodec ff_h263_encoder = {
.p.name = "h263",
CODEC_LONG_NAME("H.263 / H.263-1996"),
.p.type = AVMEDIA_TYPE_VIDEO,
.p.id = AV_CODEC_ID_H263,
CODEC_PIXFMTS(AV_PIX_FMT_YUV420P),
avcodec/internal: add FFCodec.color_ranges I went through all codecs and put them into five basic categories: 1. JPEG range only 2. MPEG range only 3. Explicitly tagged 4. Broken (codec supports both but encoder ignores tags) 5. N/A (headerless or pseudo-formats) Filters in category 5 remain untouched. The rest gain an explicit assignment of their supported color ranges, with codecs in category 4 being set to MPEG-only for safety. It might be considered redundant to distinguish between 0 (category 5) and MPEG+JPEG (category 3), but in doing so we effectively communicate that we can guarantee that these tags will be encoded, which is distinct from the situation where there are some codecs that simply don't have tagging or implied semantics (e.g. rawvideo). A full list of codecs follows: JPEG range only: - amv - roqvideo MPEG range only: - asv1, asv2 - avui - cfhd - cljr - dnxhd - dvvideo - ffv1 - flv - h261, h263, h263p - {h263,vp8}_v4l2m2m - huffyuv, ffvhuff - jpeg2000 - libopenjpeg - libtheora - libwebp, libwebp_anim - libx262 - libxavs, libxavs2 - libxvid - mpeg1video, mpeg2video - mpeg2_qsv - mpeg2_vaapi - mpeg4, msmpeg4, msmpeg4v2, wmv1, wmv2 - mpeg4_omx - prores, prores_aw, prores_ks - rv10, rv20 - snow - speedhq - svq1 - tiff - utvideo Explicitly tagged (MPEG/JPEG): - {av1,h264,hevc}_nvenc - {av1,h264,hevc}_vaapi - {av1,h264,hevc,vp8,vp9,mpeg4}_mediacodec - {av1,h264,hevc,vp9}_qsv - h264_amf - {h264,hevc,prores}_videotoolbox - libaom-av1 - libkvazaar - libopenh264 - librav1e - libsvtav1 - libvpx, libvpx-vp9 - libx264 - libx265 - ljpeg - mjpeg - vc2 Broken (encoder ignores tags): - {av1,hevc}_amf - {h264,hevc,mpeg4}_v4l2m2m - h264_omx - libxeve - magicyuv - {vp8,vp9,mjpeg}_vaapi N/A: - ayuv, yuv4, y41p, v308, v210, v410, v408 (headerless) - pgmyuv (headerless) - rawvideo, bitpacked (headerless) - vnull, wrapped_avframe (pseudocodecs)
2023-10-11 16:09:33 +02:00
.color_ranges = AVCOL_RANGE_MPEG,
.p.priv_class = &h263_class,
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE,
.caps_internal = FF_CODEC_CAP_INIT_CLEANUP,
.priv_data_size = sizeof(MPVMainEncContext),
.init = ff_mpv_encode_init,
FF_CODEC_ENCODE_CB(ff_mpv_encode_picture),
.close = ff_mpv_encode_end,
};
static const AVOption h263p_options[] = {
{ "umv", "Use unlimited motion vectors.", FF_MPV_OFFSET(umvplus), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE },
{ "aiv", "Use alternative inter VLC.", FF_MPV_OFFSET(alt_inter_vlc), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE },
{ "obmc", "use overlapped block motion compensation.", OFFSET(obmc), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE },
{ "structured_slices", "Write slice start position at every GOB header instead of just GOB number.", FF_MPV_OFFSET(h263_slice_structured), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE},
FF_MPV_COMMON_OPTS
FF_MPV_COMMON_MOTION_EST_OPTS
{ NULL },
};
static const AVClass h263p_class = {
.class_name = "H.263p encoder",
.item_name = av_default_item_name,
.option = h263p_options,
.version = LIBAVUTIL_VERSION_INT,
};
const FFCodec ff_h263p_encoder = {
.p.name = "h263p",
CODEC_LONG_NAME("H.263+ / H.263-1998 / H.263 version 2"),
.p.type = AVMEDIA_TYPE_VIDEO,
.p.id = AV_CODEC_ID_H263P,
CODEC_PIXFMTS(AV_PIX_FMT_YUV420P),
avcodec/internal: add FFCodec.color_ranges I went through all codecs and put them into five basic categories: 1. JPEG range only 2. MPEG range only 3. Explicitly tagged 4. Broken (codec supports both but encoder ignores tags) 5. N/A (headerless or pseudo-formats) Filters in category 5 remain untouched. The rest gain an explicit assignment of their supported color ranges, with codecs in category 4 being set to MPEG-only for safety. It might be considered redundant to distinguish between 0 (category 5) and MPEG+JPEG (category 3), but in doing so we effectively communicate that we can guarantee that these tags will be encoded, which is distinct from the situation where there are some codecs that simply don't have tagging or implied semantics (e.g. rawvideo). A full list of codecs follows: JPEG range only: - amv - roqvideo MPEG range only: - asv1, asv2 - avui - cfhd - cljr - dnxhd - dvvideo - ffv1 - flv - h261, h263, h263p - {h263,vp8}_v4l2m2m - huffyuv, ffvhuff - jpeg2000 - libopenjpeg - libtheora - libwebp, libwebp_anim - libx262 - libxavs, libxavs2 - libxvid - mpeg1video, mpeg2video - mpeg2_qsv - mpeg2_vaapi - mpeg4, msmpeg4, msmpeg4v2, wmv1, wmv2 - mpeg4_omx - prores, prores_aw, prores_ks - rv10, rv20 - snow - speedhq - svq1 - tiff - utvideo Explicitly tagged (MPEG/JPEG): - {av1,h264,hevc}_nvenc - {av1,h264,hevc}_vaapi - {av1,h264,hevc,vp8,vp9,mpeg4}_mediacodec - {av1,h264,hevc,vp9}_qsv - h264_amf - {h264,hevc,prores}_videotoolbox - libaom-av1 - libkvazaar - libopenh264 - librav1e - libsvtav1 - libvpx, libvpx-vp9 - libx264 - libx265 - ljpeg - mjpeg - vc2 Broken (encoder ignores tags): - {av1,hevc}_amf - {h264,hevc,mpeg4}_v4l2m2m - h264_omx - libxeve - magicyuv - {vp8,vp9,mjpeg}_vaapi N/A: - ayuv, yuv4, y41p, v308, v210, v410, v408 (headerless) - pgmyuv (headerless) - rawvideo, bitpacked (headerless) - vnull, wrapped_avframe (pseudocodecs)
2023-10-11 16:09:33 +02:00
.color_ranges = AVCOL_RANGE_MPEG,
.p.priv_class = &h263p_class,
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_SLICE_THREADS |
AV_CODEC_CAP_ENCODER_REORDERED_OPAQUE,
.caps_internal = FF_CODEC_CAP_INIT_CLEANUP,
.priv_data_size = sizeof(MPVMainEncContext),
.init = ff_mpv_encode_init,
FF_CODEC_ENCODE_CB(ff_mpv_encode_picture),
.close = ff_mpv_encode_end,
};
#endif