mirror of https://git.videolan.org/git/ffmpeg.git
avformat: Immersive Audio Model and Formats demuxer
Signed-off-by: James Almer <jamrial@gmail.com>
This commit is contained in:
James Almer
parent
7f9308380f
commit
4ee05182b7
|
|
@ -10,6 +10,7 @@ version <next>:
|
|||
- demuxing, decoding, filtering, encoding, and muxing in the
|
||||
ffmpeg CLI now all run in parallel
|
||||
- enable gdigrab device to grab a window using the hwnd=HANDLER syntax
|
||||
- IAMF raw demuxer
|
||||
|
||||
version 6.1:
|
||||
- libaribcaption decoder
|
||||
|
|
|
|||
|
|
@ -258,6 +258,7 @@ OBJS-$(CONFIG_EVC_MUXER) += rawenc.o
|
|||
OBJS-$(CONFIG_HLS_DEMUXER) += hls.o hls_sample_encryption.o
|
||||
OBJS-$(CONFIG_HLS_MUXER) += hlsenc.o hlsplaylist.o avc.o
|
||||
OBJS-$(CONFIG_HNM_DEMUXER) += hnm.o
|
||||
OBJS-$(CONFIG_IAMF_DEMUXER) += iamfdec.o iamf_parse.o iamf.o
|
||||
OBJS-$(CONFIG_ICO_DEMUXER) += icodec.o
|
||||
OBJS-$(CONFIG_ICO_MUXER) += icoenc.o
|
||||
OBJS-$(CONFIG_IDCIN_DEMUXER) += idcin.o
|
||||
|
|
|
|||
|
|
@ -212,6 +212,7 @@ extern const FFOutputFormat ff_hevc_muxer;
|
|||
extern const AVInputFormat ff_hls_demuxer;
|
||||
extern const FFOutputFormat ff_hls_muxer;
|
||||
extern const AVInputFormat ff_hnm_demuxer;
|
||||
extern const AVInputFormat ff_iamf_demuxer;
|
||||
extern const AVInputFormat ff_ico_demuxer;
|
||||
extern const FFOutputFormat ff_ico_muxer;
|
||||
extern const AVInputFormat ff_idcin_demuxer;
|
||||
|
|
|
|||
|
|
@ -0,0 +1,125 @@
|
|||
/*
|
||||
* Immersive Audio Model and Formats common helpers and structs
|
||||
* Copyright (c) 2023 James Almer <jamrial@gmail.com>
|
||||
*
|
||||
* 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
|
||||
*/
|
||||
|
||||
#include "libavutil/channel_layout.h"
|
||||
#include "libavutil/iamf.h"
|
||||
#include "libavutil/mem.h"
|
||||
#include "iamf.h"
|
||||
|
||||
const AVChannelLayout ff_iamf_scalable_ch_layouts[10] = {
|
||||
AV_CHANNEL_LAYOUT_MONO,
|
||||
AV_CHANNEL_LAYOUT_STEREO,
|
||||
// "Loudspeaker configuration for Sound System B"
|
||||
AV_CHANNEL_LAYOUT_5POINT1_BACK,
|
||||
// "Loudspeaker configuration for Sound System C"
|
||||
AV_CHANNEL_LAYOUT_5POINT1POINT2_BACK,
|
||||
// "Loudspeaker configuration for Sound System D"
|
||||
AV_CHANNEL_LAYOUT_5POINT1POINT4_BACK,
|
||||
// "Loudspeaker configuration for Sound System I"
|
||||
AV_CHANNEL_LAYOUT_7POINT1,
|
||||
// "Loudspeaker configuration for Sound System I" + Ltf + Rtf
|
||||
AV_CHANNEL_LAYOUT_7POINT1POINT2,
|
||||
// "Loudspeaker configuration for Sound System J"
|
||||
AV_CHANNEL_LAYOUT_7POINT1POINT4_BACK,
|
||||
// Front subset of "Loudspeaker configuration for Sound System J"
|
||||
AV_CHANNEL_LAYOUT_3POINT1POINT2,
|
||||
// Binaural
|
||||
AV_CHANNEL_LAYOUT_STEREO,
|
||||
};
|
||||
|
||||
const struct IAMFSoundSystemMap ff_iamf_sound_system_map[13] = {
|
||||
{ SOUND_SYSTEM_A_0_2_0, AV_CHANNEL_LAYOUT_STEREO },
|
||||
{ SOUND_SYSTEM_B_0_5_0, AV_CHANNEL_LAYOUT_5POINT1_BACK },
|
||||
{ SOUND_SYSTEM_C_2_5_0, AV_CHANNEL_LAYOUT_5POINT1POINT2_BACK },
|
||||
{ SOUND_SYSTEM_D_4_5_0, AV_CHANNEL_LAYOUT_5POINT1POINT4_BACK },
|
||||
{ SOUND_SYSTEM_E_4_5_1,
|
||||
{
|
||||
.nb_channels = 11,
|
||||
.order = AV_CHANNEL_ORDER_NATIVE,
|
||||
.u.mask = AV_CH_LAYOUT_5POINT1POINT4_BACK | AV_CH_BOTTOM_FRONT_CENTER,
|
||||
},
|
||||
},
|
||||
{ SOUND_SYSTEM_F_3_7_0, AV_CHANNEL_LAYOUT_7POINT2POINT3 },
|
||||
{ SOUND_SYSTEM_G_4_9_0, AV_CHANNEL_LAYOUT_9POINT1POINT4_BACK },
|
||||
{ SOUND_SYSTEM_H_9_10_3, AV_CHANNEL_LAYOUT_22POINT2 },
|
||||
{ SOUND_SYSTEM_I_0_7_0, AV_CHANNEL_LAYOUT_7POINT1 },
|
||||
{ SOUND_SYSTEM_J_4_7_0, AV_CHANNEL_LAYOUT_7POINT1POINT4_BACK },
|
||||
{ SOUND_SYSTEM_10_2_7_0, AV_CHANNEL_LAYOUT_7POINT1POINT2 },
|
||||
{ SOUND_SYSTEM_11_2_3_0, AV_CHANNEL_LAYOUT_3POINT1POINT2 },
|
||||
{ SOUND_SYSTEM_12_0_1_0, AV_CHANNEL_LAYOUT_MONO },
|
||||
};
|
||||
|
||||
void ff_iamf_free_audio_element(IAMFAudioElement **paudio_element)
|
||||
{
|
||||
IAMFAudioElement *audio_element = *paudio_element;
|
||||
|
||||
if (!audio_element)
|
||||
return;
|
||||
|
||||
for (int i = 0; i < audio_element->nb_substreams; i++)
|
||||
avcodec_parameters_free(&audio_element->substreams[i].codecpar);
|
||||
av_free(audio_element->substreams);
|
||||
av_free(audio_element->layers);
|
||||
av_iamf_audio_element_free(&audio_element->element);
|
||||
av_freep(paudio_element);
|
||||
}
|
||||
|
||||
void ff_iamf_free_mix_presentation(IAMFMixPresentation **pmix_presentation)
|
||||
{
|
||||
IAMFMixPresentation *mix_presentation = *pmix_presentation;
|
||||
|
||||
if (!mix_presentation)
|
||||
return;
|
||||
|
||||
for (int i = 0; i < mix_presentation->count_label; i++)
|
||||
av_free(mix_presentation->language_label[i]);
|
||||
av_free(mix_presentation->language_label);
|
||||
av_iamf_mix_presentation_free(&mix_presentation->mix);
|
||||
av_freep(pmix_presentation);
|
||||
}
|
||||
|
||||
void ff_iamf_uninit_context(IAMFContext *c)
|
||||
{
|
||||
if (!c)
|
||||
return;
|
||||
|
||||
for (int i = 0; i < c->nb_codec_configs; i++) {
|
||||
av_free(c->codec_configs[i]->extradata);
|
||||
av_free(c->codec_configs[i]);
|
||||
}
|
||||
av_freep(&c->codec_configs);
|
||||
c->nb_codec_configs = 0;
|
||||
|
||||
for (int i = 0; i < c->nb_audio_elements; i++)
|
||||
ff_iamf_free_audio_element(&c->audio_elements[i]);
|
||||
av_freep(&c->audio_elements);
|
||||
c->nb_audio_elements = 0;
|
||||
|
||||
for (int i = 0; i < c->nb_mix_presentations; i++)
|
||||
ff_iamf_free_mix_presentation(&c->mix_presentations[i]);
|
||||
av_freep(&c->mix_presentations);
|
||||
c->nb_mix_presentations = 0;
|
||||
|
||||
for (int i = 0; i < c->nb_param_definitions; i++)
|
||||
av_free(c->param_definitions[i]);
|
||||
av_freep(&c->param_definitions);
|
||||
c->nb_param_definitions = 0;
|
||||
}
|
||||
|
|
@ -0,0 +1,163 @@
|
|||
/*
|
||||
* Immersive Audio Model and Formats common helpers and structs
|
||||
* Copyright (c) 2023 James Almer <jamrial@gmail.com>
|
||||
*
|
||||
* 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
|
||||
*/
|
||||
|
||||
#ifndef AVFORMAT_IAMF_H
|
||||
#define AVFORMAT_IAMF_H
|
||||
|
||||
#include <stdint.h>
|
||||
|
||||
#include "libavutil/channel_layout.h"
|
||||
#include "libavutil/iamf.h"
|
||||
#include "libavcodec/codec_id.h"
|
||||
#include "libavcodec/codec_par.h"
|
||||
#include "avformat.h"
|
||||
|
||||
#define MAX_IAMF_OBU_HEADER_SIZE (1 + 8 * 3)
|
||||
|
||||
// OBU types (section 3.2).
|
||||
enum IAMF_OBU_Type {
|
||||
IAMF_OBU_IA_CODEC_CONFIG = 0,
|
||||
IAMF_OBU_IA_AUDIO_ELEMENT = 1,
|
||||
IAMF_OBU_IA_MIX_PRESENTATION = 2,
|
||||
IAMF_OBU_IA_PARAMETER_BLOCK = 3,
|
||||
IAMF_OBU_IA_TEMPORAL_DELIMITER = 4,
|
||||
IAMF_OBU_IA_AUDIO_FRAME = 5,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID0 = 6,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID1 = 7,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID2 = 8,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID3 = 9,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID4 = 10,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID5 = 11,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID6 = 12,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID7 = 13,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID8 = 14,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID9 = 15,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID10 = 16,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID11 = 17,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID12 = 18,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID13 = 19,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID14 = 20,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID15 = 21,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID16 = 22,
|
||||
IAMF_OBU_IA_AUDIO_FRAME_ID17 = 23,
|
||||
// 24~30 reserved.
|
||||
IAMF_OBU_IA_SEQUENCE_HEADER = 31,
|
||||
};
|
||||
|
||||
typedef struct IAMFCodecConfig {
|
||||
unsigned codec_config_id;
|
||||
enum AVCodecID codec_id;
|
||||
uint32_t codec_tag;
|
||||
unsigned nb_samples;
|
||||
int seek_preroll;
|
||||
int sample_rate;
|
||||
int extradata_size;
|
||||
uint8_t *extradata;
|
||||
} IAMFCodecConfig;
|
||||
|
||||
typedef struct IAMFLayer {
|
||||
unsigned int substream_count;
|
||||
unsigned int coupled_substream_count;
|
||||
} IAMFLayer;
|
||||
|
||||
typedef struct IAMFSubStream {
|
||||
unsigned int audio_substream_id;
|
||||
|
||||
// demux
|
||||
AVCodecParameters *codecpar;
|
||||
} IAMFSubStream;
|
||||
|
||||
typedef struct IAMFAudioElement {
|
||||
AVIAMFAudioElement *element;
|
||||
unsigned int audio_element_id;
|
||||
|
||||
IAMFSubStream *substreams;
|
||||
unsigned int nb_substreams;
|
||||
|
||||
unsigned int codec_config_id;
|
||||
|
||||
// mux
|
||||
IAMFLayer *layers;
|
||||
unsigned int nb_layers;
|
||||
} IAMFAudioElement;
|
||||
|
||||
typedef struct IAMFMixPresentation {
|
||||
AVIAMFMixPresentation *mix;
|
||||
unsigned int mix_presentation_id;
|
||||
|
||||
// demux
|
||||
unsigned int count_label;
|
||||
char **language_label;
|
||||
} IAMFMixPresentation;
|
||||
|
||||
typedef struct IAMFParamDefinition {
|
||||
const IAMFAudioElement *audio_element;
|
||||
AVIAMFParamDefinition *param;
|
||||
int mode;
|
||||
size_t param_size;
|
||||
} IAMFParamDefinition;
|
||||
|
||||
typedef struct IAMFContext {
|
||||
IAMFCodecConfig **codec_configs;
|
||||
int nb_codec_configs;
|
||||
IAMFAudioElement **audio_elements;
|
||||
int nb_audio_elements;
|
||||
IAMFMixPresentation **mix_presentations;
|
||||
int nb_mix_presentations;
|
||||
IAMFParamDefinition **param_definitions;
|
||||
int nb_param_definitions;
|
||||
} IAMFContext;
|
||||
|
||||
enum IAMF_Anchor_Element {
|
||||
IAMF_ANCHOR_ELEMENT_UNKNWONW,
|
||||
IAMF_ANCHOR_ELEMENT_DIALOGUE,
|
||||
IAMF_ANCHOR_ELEMENT_ALBUM,
|
||||
};
|
||||
|
||||
enum IAMF_Sound_System {
|
||||
SOUND_SYSTEM_A_0_2_0 = 0, // "Loudspeaker configuration for Sound System A"
|
||||
SOUND_SYSTEM_B_0_5_0 = 1, // "Loudspeaker configuration for Sound System B"
|
||||
SOUND_SYSTEM_C_2_5_0 = 2, // "Loudspeaker configuration for Sound System C"
|
||||
SOUND_SYSTEM_D_4_5_0 = 3, // "Loudspeaker configuration for Sound System D"
|
||||
SOUND_SYSTEM_E_4_5_1 = 4, // "Loudspeaker configuration for Sound System E"
|
||||
SOUND_SYSTEM_F_3_7_0 = 5, // "Loudspeaker configuration for Sound System F"
|
||||
SOUND_SYSTEM_G_4_9_0 = 6, // "Loudspeaker configuration for Sound System G"
|
||||
SOUND_SYSTEM_H_9_10_3 = 7, // "Loudspeaker configuration for Sound System H"
|
||||
SOUND_SYSTEM_I_0_7_0 = 8, // "Loudspeaker configuration for Sound System I"
|
||||
SOUND_SYSTEM_J_4_7_0 = 9, // "Loudspeaker configuration for Sound System J"
|
||||
SOUND_SYSTEM_10_2_7_0 = 10, // "Loudspeaker configuration for Sound System I" + Ltf + Rtf
|
||||
SOUND_SYSTEM_11_2_3_0 = 11, // Front subset of "Loudspeaker configuration for Sound System J"
|
||||
SOUND_SYSTEM_12_0_1_0 = 12, // Mono
|
||||
};
|
||||
|
||||
struct IAMFSoundSystemMap {
|
||||
enum IAMF_Sound_System id;
|
||||
AVChannelLayout layout;
|
||||
};
|
||||
|
||||
extern const AVChannelLayout ff_iamf_scalable_ch_layouts[10];
|
||||
extern const struct IAMFSoundSystemMap ff_iamf_sound_system_map[13];
|
||||
|
||||
void ff_iamf_free_audio_element(IAMFAudioElement **paudio_element);
|
||||
void ff_iamf_free_mix_presentation(IAMFMixPresentation **pmix_presentation);
|
||||
void ff_iamf_uninit_context(IAMFContext *c);
|
||||
|
||||
#endif /* AVFORMAT_IAMF_H */
|
||||
File diff suppressed because it is too large
Load Diff
|
|
@ -0,0 +1,38 @@
|
|||
/*
|
||||
* Immersive Audio Model and Formats parsing
|
||||
* Copyright (c) 2023 James Almer <jamrial@gmail.com>
|
||||
*
|
||||
* 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
|
||||
*/
|
||||
|
||||
#ifndef AVFORMAT_IAMF_PARSE_H
|
||||
#define AVFORMAT_IAMF_PARSE_H
|
||||
|
||||
#include <stdint.h>
|
||||
|
||||
#include "libavutil/iamf.h"
|
||||
#include "avio.h"
|
||||
#include "iamf.h"
|
||||
|
||||
int ff_iamf_parse_obu_header(const uint8_t *buf, int buf_size,
|
||||
unsigned *obu_size, int *start_pos, enum IAMF_OBU_Type *type,
|
||||
unsigned *skip_samples, unsigned *discard_padding);
|
||||
|
||||
int ff_iamfdec_read_descriptors(IAMFContext *c, AVIOContext *pb,
|
||||
int size, void *log_ctx);
|
||||
|
||||
#endif /* AVFORMAT_IAMF_PARSE_H */
|
||||
|
|
@ -0,0 +1,505 @@
|
|||
/*
|
||||
* Immersive Audio Model and Formats demuxer
|
||||
* Copyright (c) 2023 James Almer <jamrial@gmail.com>
|
||||
*
|
||||
* 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
|
||||
*/
|
||||
|
||||
#include "config_components.h"
|
||||
|
||||
#include "libavutil/avassert.h"
|
||||
#include "libavutil/iamf.h"
|
||||
#include "libavutil/intreadwrite.h"
|
||||
#include "libavutil/log.h"
|
||||
#include "libavcodec/mathops.h"
|
||||
#include "avformat.h"
|
||||
#include "avio_internal.h"
|
||||
#include "demux.h"
|
||||
#include "iamf.h"
|
||||
#include "iamf_parse.h"
|
||||
#include "internal.h"
|
||||
|
||||
typedef struct IAMFDemuxContext {
|
||||
IAMFContext iamf;
|
||||
|
||||
// Packet side data
|
||||
AVIAMFParamDefinition *mix;
|
||||
size_t mix_size;
|
||||
AVIAMFParamDefinition *demix;
|
||||
size_t demix_size;
|
||||
AVIAMFParamDefinition *recon;
|
||||
size_t recon_size;
|
||||
} IAMFDemuxContext;
|
||||
|
||||
static AVStream *find_stream_by_id(AVFormatContext *s, int id)
|
||||
{
|
||||
for (int i = 0; i < s->nb_streams; i++)
|
||||
if (s->streams[i]->id == id)
|
||||
return s->streams[i];
|
||||
|
||||
av_log(s, AV_LOG_ERROR, "Invalid stream id %d\n", id);
|
||||
return NULL;
|
||||
}
|
||||
|
||||
static int audio_frame_obu(AVFormatContext *s, AVPacket *pkt, int len,
|
||||
enum IAMF_OBU_Type type,
|
||||
unsigned skip_samples, unsigned discard_padding,
|
||||
int id_in_bitstream)
|
||||
{
|
||||
const IAMFDemuxContext *const c = s->priv_data;
|
||||
AVStream *st;
|
||||
int ret, audio_substream_id;
|
||||
|
||||
if (id_in_bitstream) {
|
||||
unsigned explicit_audio_substream_id;
|
||||
int64_t pos = avio_tell(s->pb);
|
||||
explicit_audio_substream_id = ffio_read_leb(s->pb);
|
||||
len -= avio_tell(s->pb) - pos;
|
||||
audio_substream_id = explicit_audio_substream_id;
|
||||
} else
|
||||
audio_substream_id = type - IAMF_OBU_IA_AUDIO_FRAME_ID0;
|
||||
|
||||
st = find_stream_by_id(s, audio_substream_id);
|
||||
if (!st)
|
||||
return AVERROR_INVALIDDATA;
|
||||
|
||||
ret = av_get_packet(s->pb, pkt, len);
|
||||
if (ret < 0)
|
||||
return ret;
|
||||
if (ret != len)
|
||||
return AVERROR_INVALIDDATA;
|
||||
|
||||
if (skip_samples || discard_padding) {
|
||||
uint8_t *side_data = av_packet_new_side_data(pkt, AV_PKT_DATA_SKIP_SAMPLES, 10);
|
||||
if (!side_data)
|
||||
return AVERROR(ENOMEM);
|
||||
AV_WL32(side_data, skip_samples);
|
||||
AV_WL32(side_data + 4, discard_padding);
|
||||
}
|
||||
if (c->mix) {
|
||||
uint8_t *side_data = av_packet_new_side_data(pkt, AV_PKT_DATA_IAMF_MIX_GAIN_PARAM, c->mix_size);
|
||||
if (!side_data)
|
||||
return AVERROR(ENOMEM);
|
||||
memcpy(side_data, c->mix, c->mix_size);
|
||||
}
|
||||
if (c->demix) {
|
||||
uint8_t *side_data = av_packet_new_side_data(pkt, AV_PKT_DATA_IAMF_DEMIXING_INFO_PARAM, c->demix_size);
|
||||
if (!side_data)
|
||||
return AVERROR(ENOMEM);
|
||||
memcpy(side_data, c->demix, c->demix_size);
|
||||
}
|
||||
if (c->recon) {
|
||||
uint8_t *side_data = av_packet_new_side_data(pkt, AV_PKT_DATA_IAMF_RECON_GAIN_INFO_PARAM, c->recon_size);
|
||||
if (!side_data)
|
||||
return AVERROR(ENOMEM);
|
||||
memcpy(side_data, c->recon, c->recon_size);
|
||||
}
|
||||
|
||||
pkt->stream_index = st->index;
|
||||
return 0;
|
||||
}
|
||||
|
||||
static const IAMFParamDefinition *get_param_definition(AVFormatContext *s, unsigned int parameter_id)
|
||||
{
|
||||
const IAMFDemuxContext *const c = s->priv_data;
|
||||
const IAMFContext *const iamf = &c->iamf;
|
||||
const IAMFParamDefinition *param_definition = NULL;
|
||||
|
||||
for (int i = 0; i < iamf->nb_param_definitions; i++)
|
||||
if (iamf->param_definitions[i]->param->parameter_id == parameter_id) {
|
||||
param_definition = iamf->param_definitions[i];
|
||||
break;
|
||||
}
|
||||
|
||||
return param_definition;
|
||||
}
|
||||
|
||||
static int parameter_block_obu(AVFormatContext *s, int len)
|
||||
{
|
||||
IAMFDemuxContext *const c = s->priv_data;
|
||||
const IAMFParamDefinition *param_definition;
|
||||
const AVIAMFParamDefinition *param;
|
||||
AVIAMFParamDefinition *out_param = NULL;
|
||||
FFIOContext b;
|
||||
AVIOContext *pb;
|
||||
uint8_t *buf;
|
||||
unsigned int duration, constant_subblock_duration;
|
||||
unsigned int nb_subblocks;
|
||||
unsigned int parameter_id;
|
||||
size_t out_param_size;
|
||||
int ret;
|
||||
|
||||
buf = av_malloc(len);
|
||||
if (!buf)
|
||||
return AVERROR(ENOMEM);
|
||||
|
||||
ret = avio_read(s->pb, buf, len);
|
||||
if (ret != len) {
|
||||
if (ret >= 0)
|
||||
ret = AVERROR_INVALIDDATA;
|
||||
goto fail;
|
||||
}
|
||||
|
||||
ffio_init_context(&b, buf, len, 0, NULL, NULL, NULL, NULL);
|
||||
pb = &b.pub;
|
||||
|
||||
parameter_id = ffio_read_leb(pb);
|
||||
param_definition = get_param_definition(s, parameter_id);
|
||||
if (!param_definition) {
|
||||
av_log(s, AV_LOG_VERBOSE, "Non existant parameter_id %d referenced in a parameter block. Ignoring\n",
|
||||
parameter_id);
|
||||
ret = 0;
|
||||
goto fail;
|
||||
}
|
||||
|
||||
param = param_definition->param;
|
||||
if (!param_definition->mode) {
|
||||
duration = ffio_read_leb(pb);
|
||||
if (!duration) {
|
||||
ret = AVERROR_INVALIDDATA;
|
||||
goto fail;
|
||||
}
|
||||
constant_subblock_duration = ffio_read_leb(pb);
|
||||
if (constant_subblock_duration == 0)
|
||||
nb_subblocks = ffio_read_leb(pb);
|
||||
else
|
||||
nb_subblocks = duration / constant_subblock_duration;
|
||||
} else {
|
||||
duration = param->duration;
|
||||
constant_subblock_duration = param->constant_subblock_duration;
|
||||
nb_subblocks = param->nb_subblocks;
|
||||
}
|
||||
|
||||
out_param = av_iamf_param_definition_alloc(param->type, nb_subblocks, &out_param_size);
|
||||
if (!out_param) {
|
||||
ret = AVERROR(ENOMEM);
|
||||
goto fail;
|
||||
}
|
||||
|
||||
out_param->parameter_id = param->parameter_id;
|
||||
out_param->type = param->type;
|
||||
out_param->parameter_rate = param->parameter_rate;
|
||||
out_param->duration = duration;
|
||||
out_param->constant_subblock_duration = constant_subblock_duration;
|
||||
out_param->nb_subblocks = nb_subblocks;
|
||||
|
||||
for (int i = 0; i < nb_subblocks; i++) {
|
||||
void *subblock = av_iamf_param_definition_get_subblock(out_param, i);
|
||||
unsigned int subblock_duration = constant_subblock_duration;
|
||||
|
||||
if (!param_definition->mode && !constant_subblock_duration)
|
||||
subblock_duration = ffio_read_leb(pb);
|
||||
|
||||
switch (param->type) {
|
||||
case AV_IAMF_PARAMETER_DEFINITION_MIX_GAIN: {
|
||||
AVIAMFMixGain *mix = subblock;
|
||||
|
||||
mix->animation_type = ffio_read_leb(pb);
|
||||
if (mix->animation_type > AV_IAMF_ANIMATION_TYPE_BEZIER) {
|
||||
ret = 0;
|
||||
av_free(out_param);
|
||||
goto fail;
|
||||
}
|
||||
|
||||
mix->start_point_value = av_make_q(sign_extend(avio_rb16(pb), 16), 1 << 8);
|
||||
if (mix->animation_type >= AV_IAMF_ANIMATION_TYPE_LINEAR)
|
||||
mix->end_point_value = av_make_q(sign_extend(avio_rb16(pb), 16), 1 << 8);
|
||||
if (mix->animation_type == AV_IAMF_ANIMATION_TYPE_BEZIER) {
|
||||
mix->control_point_value = av_make_q(sign_extend(avio_rb16(pb), 16), 1 << 8);
|
||||
mix->control_point_relative_time = av_make_q(avio_r8(pb), 1 << 8);
|
||||
}
|
||||
mix->subblock_duration = subblock_duration;
|
||||
break;
|
||||
}
|
||||
case AV_IAMF_PARAMETER_DEFINITION_DEMIXING: {
|
||||
AVIAMFDemixingInfo *demix = subblock;
|
||||
|
||||
demix->dmixp_mode = avio_r8(pb) >> 5;
|
||||
demix->subblock_duration = subblock_duration;
|
||||
break;
|
||||
}
|
||||
case AV_IAMF_PARAMETER_DEFINITION_RECON_GAIN: {
|
||||
AVIAMFReconGain *recon = subblock;
|
||||
const IAMFAudioElement *audio_element = param_definition->audio_element;
|
||||
const AVIAMFAudioElement *element = audio_element->element;
|
||||
|
||||
av_assert0(audio_element && element);
|
||||
for (int i = 0; i < element->nb_layers; i++) {
|
||||
const AVIAMFLayer *layer = element->layers[i];
|
||||
if (layer->flags & AV_IAMF_LAYER_FLAG_RECON_GAIN) {
|
||||
unsigned int recon_gain_flags = ffio_read_leb(pb);
|
||||
unsigned int bitcount = 7 + 5 * !!(recon_gain_flags & 0x80);
|
||||
recon_gain_flags = (recon_gain_flags & 0x7F) | ((recon_gain_flags & 0xFF00) >> 1);
|
||||
for (int j = 0; j < bitcount; j++) {
|
||||
if (recon_gain_flags & (1 << j))
|
||||
recon->recon_gain[i][j] = avio_r8(pb);
|
||||
}
|
||||
}
|
||||
}
|
||||
recon->subblock_duration = subblock_duration;
|
||||
break;
|
||||
}
|
||||
default:
|
||||
av_assert0(0);
|
||||
}
|
||||
}
|
||||
|
||||
len -= avio_tell(pb);
|
||||
if (len) {
|
||||
int level = (s->error_recognition & AV_EF_EXPLODE) ? AV_LOG_ERROR : AV_LOG_WARNING;
|
||||
av_log(s, level, "Underread in parameter_block_obu. %d bytes left at the end\n", len);
|
||||
}
|
||||
|
||||
switch (param->type) {
|
||||
case AV_IAMF_PARAMETER_DEFINITION_MIX_GAIN:
|
||||
av_free(c->mix);
|
||||
c->mix = out_param;
|
||||
c->mix_size = out_param_size;
|
||||
break;
|
||||
case AV_IAMF_PARAMETER_DEFINITION_DEMIXING:
|
||||
av_free(c->demix);
|
||||
c->demix = out_param;
|
||||
c->demix_size = out_param_size;
|
||||
break;
|
||||
case AV_IAMF_PARAMETER_DEFINITION_RECON_GAIN:
|
||||
av_free(c->recon);
|
||||
c->recon = out_param;
|
||||
c->recon_size = out_param_size;
|
||||
break;
|
||||
default:
|
||||
av_assert0(0);
|
||||
}
|
||||
|
||||
ret = 0;
|
||||
fail:
|
||||
if (ret < 0)
|
||||
av_free(out_param);
|
||||
av_free(buf);
|
||||
|
||||
return ret;
|
||||
}
|
||||
|
||||
static int iamf_read_packet(AVFormatContext *s, AVPacket *pkt)
|
||||
{
|
||||
IAMFDemuxContext *const c = s->priv_data;
|
||||
uint8_t header[MAX_IAMF_OBU_HEADER_SIZE + AV_INPUT_BUFFER_PADDING_SIZE];
|
||||
unsigned obu_size;
|
||||
int ret;
|
||||
|
||||
while (1) {
|
||||
enum IAMF_OBU_Type type;
|
||||
unsigned skip_samples, discard_padding;
|
||||
int len, size, start_pos;
|
||||
|
||||
if ((ret = ffio_ensure_seekback(s->pb, MAX_IAMF_OBU_HEADER_SIZE)) < 0)
|
||||
return ret;
|
||||
size = avio_read(s->pb, header, MAX_IAMF_OBU_HEADER_SIZE);
|
||||
if (size < 0)
|
||||
return size;
|
||||
|
||||
len = ff_iamf_parse_obu_header(header, size, &obu_size, &start_pos, &type,
|
||||
&skip_samples, &discard_padding);
|
||||
if (len < 0) {
|
||||
av_log(s, AV_LOG_ERROR, "Failed to read obu\n");
|
||||
return len;
|
||||
}
|
||||
avio_seek(s->pb, -(size - start_pos), SEEK_CUR);
|
||||
|
||||
if (type >= IAMF_OBU_IA_AUDIO_FRAME && type <= IAMF_OBU_IA_AUDIO_FRAME_ID17)
|
||||
return audio_frame_obu(s, pkt, obu_size, type,
|
||||
skip_samples, discard_padding,
|
||||
type == IAMF_OBU_IA_AUDIO_FRAME);
|
||||
else if (type == IAMF_OBU_IA_PARAMETER_BLOCK) {
|
||||
ret = parameter_block_obu(s, obu_size);
|
||||
if (ret < 0)
|
||||
return ret;
|
||||
} else if (type == IAMF_OBU_IA_TEMPORAL_DELIMITER) {
|
||||
av_freep(&c->mix);
|
||||
c->mix_size = 0;
|
||||
av_freep(&c->demix);
|
||||
c->demix_size = 0;
|
||||
av_freep(&c->recon);
|
||||
c->recon_size = 0;
|
||||
} else {
|
||||
int64_t offset = avio_skip(s->pb, obu_size);
|
||||
if (offset < 0) {
|
||||
ret = offset;
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return ret;
|
||||
}
|
||||
|
||||
//return < 0 if we need more data
|
||||
static int get_score(const uint8_t *buf, int buf_size, enum IAMF_OBU_Type type, int *seq)
|
||||
{
|
||||
if (type == IAMF_OBU_IA_SEQUENCE_HEADER) {
|
||||
if (buf_size < 4 || AV_RB32(buf) != MKBETAG('i','a','m','f'))
|
||||
return 0;
|
||||
*seq = 1;
|
||||
return -1;
|
||||
}
|
||||
if (type >= IAMF_OBU_IA_CODEC_CONFIG && type <= IAMF_OBU_IA_TEMPORAL_DELIMITER)
|
||||
return *seq ? -1 : 0;
|
||||
if (type >= IAMF_OBU_IA_AUDIO_FRAME && type <= IAMF_OBU_IA_AUDIO_FRAME_ID17)
|
||||
return *seq ? AVPROBE_SCORE_EXTENSION + 1 : 0;
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int iamf_probe(const AVProbeData *p)
|
||||
{
|
||||
unsigned obu_size;
|
||||
enum IAMF_OBU_Type type;
|
||||
int seq = 0, cnt = 0, start_pos;
|
||||
int ret;
|
||||
|
||||
while (1) {
|
||||
int size = ff_iamf_parse_obu_header(p->buf + cnt, p->buf_size - cnt,
|
||||
&obu_size, &start_pos, &type,
|
||||
NULL, NULL);
|
||||
if (size < 0)
|
||||
return 0;
|
||||
|
||||
ret = get_score(p->buf + cnt + start_pos,
|
||||
p->buf_size - cnt - start_pos,
|
||||
type, &seq);
|
||||
if (ret >= 0)
|
||||
return ret;
|
||||
|
||||
cnt += FFMIN(size, p->buf_size - cnt);
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int iamf_read_header(AVFormatContext *s)
|
||||
{
|
||||
IAMFDemuxContext *const c = s->priv_data;
|
||||
IAMFContext *const iamf = &c->iamf;
|
||||
int ret;
|
||||
|
||||
ret = ff_iamfdec_read_descriptors(iamf, s->pb, INT_MAX, s);
|
||||
if (ret < 0)
|
||||
return ret;
|
||||
|
||||
for (int i = 0; i < iamf->nb_audio_elements; i++) {
|
||||
IAMFAudioElement *audio_element = iamf->audio_elements[i];
|
||||
AVStreamGroup *stg = avformat_stream_group_create(s, AV_STREAM_GROUP_PARAMS_IAMF_AUDIO_ELEMENT, NULL);
|
||||
|
||||
if (!stg)
|
||||
return AVERROR(ENOMEM);
|
||||
|
||||
stg->id = audio_element->audio_element_id;
|
||||
stg->params.iamf_audio_element = audio_element->element;
|
||||
|
||||
for (int j = 0; j < audio_element->nb_substreams; j++) {
|
||||
IAMFSubStream *substream = &audio_element->substreams[j];
|
||||
AVStream *st = avformat_new_stream(s, NULL);
|
||||
|
||||
if (!st)
|
||||
return AVERROR(ENOMEM);
|
||||
|
||||
ret = avformat_stream_group_add_stream(stg, st);
|
||||
if (ret < 0)
|
||||
return ret;
|
||||
|
||||
ret = avcodec_parameters_copy(st->codecpar, substream->codecpar);
|
||||
if (ret < 0)
|
||||
return ret;
|
||||
|
||||
st->id = substream->audio_substream_id;
|
||||
avpriv_set_pts_info(st, 64, 1, st->codecpar->sample_rate);
|
||||
}
|
||||
}
|
||||
|
||||
for (int i = 0; i < iamf->nb_mix_presentations; i++) {
|
||||
IAMFMixPresentation *mix_presentation = iamf->mix_presentations[i];
|
||||
AVStreamGroup *stg = avformat_stream_group_create(s, AV_STREAM_GROUP_PARAMS_IAMF_MIX_PRESENTATION, NULL);
|
||||
const AVIAMFMixPresentation *mix = mix_presentation->mix;
|
||||
|
||||
if (!stg)
|
||||
return AVERROR(ENOMEM);
|
||||
|
||||
stg->id = mix_presentation->mix_presentation_id;
|
||||
stg->params.iamf_mix_presentation = mix_presentation->mix;
|
||||
|
||||
for (int j = 0; j < mix->nb_submixes; j++) {
|
||||
AVIAMFSubmix *sub_mix = mix->submixes[j];
|
||||
|
||||
for (int k = 0; k < sub_mix->nb_elements; k++) {
|
||||
AVIAMFSubmixElement *submix_element = sub_mix->elements[k];
|
||||
AVStreamGroup *audio_element = NULL;
|
||||
|
||||
for (int l = 0; l < s->nb_stream_groups; l++)
|
||||
if (s->stream_groups[l]->type == AV_STREAM_GROUP_PARAMS_IAMF_AUDIO_ELEMENT &&
|
||||
s->stream_groups[l]->id == submix_element->audio_element_id) {
|
||||
audio_element = s->stream_groups[l];
|
||||
break;
|
||||
}
|
||||
av_assert0(audio_element);
|
||||
|
||||
for (int l = 0; l < audio_element->nb_streams; l++) {
|
||||
ret = avformat_stream_group_add_stream(stg, audio_element->streams[l]);
|
||||
if (ret < 0 && ret != AVERROR(EEXIST))
|
||||
return ret;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int iamf_read_close(AVFormatContext *s)
|
||||
{
|
||||
IAMFDemuxContext *const c = s->priv_data;
|
||||
IAMFContext *const iamf = &c->iamf;
|
||||
|
||||
for (int i = 0; i < iamf->nb_audio_elements; i++) {
|
||||
IAMFAudioElement *audio_element = iamf->audio_elements[i];
|
||||
audio_element->element = NULL;
|
||||
}
|
||||
for (int i = 0; i < iamf->nb_mix_presentations; i++) {
|
||||
IAMFMixPresentation *mix_presentation = iamf->mix_presentations[i];
|
||||
mix_presentation->mix = NULL;
|
||||
}
|
||||
|
||||
ff_iamf_uninit_context(&c->iamf);
|
||||
|
||||
av_freep(&c->mix);
|
||||
c->mix_size = 0;
|
||||
av_freep(&c->demix);
|
||||
c->demix_size = 0;
|
||||
av_freep(&c->recon);
|
||||
c->recon_size = 0;
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
const AVInputFormat ff_iamf_demuxer = {
|
||||
.name = "iamf",
|
||||
.long_name = NULL_IF_CONFIG_SMALL("Raw Immersive Audio Model and Formats"),
|
||||
.priv_data_size = sizeof(IAMFDemuxContext),
|
||||
.flags_internal = FF_FMT_INIT_CLEANUP,
|
||||
.read_probe = iamf_probe,
|
||||
.read_header = iamf_read_header,
|
||||
.read_packet = iamf_read_packet,
|
||||
.read_close = iamf_read_close,
|
||||
.extensions = "iamf",
|
||||
.flags = AVFMT_GENERIC_INDEX | AVFMT_NO_BYTE_SEEK | AVFMT_NOTIMESTAMPS | AVFMT_SHOW_IDS,
|
||||
};
|
||||
Loading…
Reference in New Issue