mpv/video/decode/dec_video.c

497 lines
16 KiB
C
Raw Normal View History

/*
* This file is part of mpv.
*
* mpv is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* mpv 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License along
* with mpv. If not, see <http://www.gnu.org/licenses/>.
dec_video: change license to LGPL (almost) "Almost" because this might contain copyright by michael, who agreed with LGPL, but only once the core is LGPL. This is preparation for that to happen. Apart from that, the usual remarks apply. In particular, dec_video.c started out quite chaotic with no modularization, but was later basically gutted, and in general rewritten a bunch of times. Not going to give a history lesson. Special attention needs to be given to 3 patches by cehosos, who did not agree to the relicensing: 240b743ebdf: --field-dominance e32cbbf7dc3: reinit VO if aspect ratio changes 306f6243fdf: use container aspect if codec aspect unset (?) The first patch is pretty clearly still in the current code, and needs to be disabled for LGPL. The functionality of the second patch is still active, but implemented completely different, and as part of general frame parameter changes (at the time of the patch, MPlayer already reinitialized the VO on frame size and pixel format changes - all this was merged into a single check for changing image parameters). The third patch makes me a bit more uncomfortable. It appears the code was moved to dec_video.c in de68b8f23c8c, and further changed in 82f0d373, 0a0bb905, and bf13bd0d. You could claim that cehoyos' copyright still sticks. Fortunately, we implement alternative aspect detection, which is simpler and probably preferable, and which arguably contains none of the original code and logic, and thus should be fully safe. While I don't know if cehoyos' copyright actually still applies, I'm more comfortable with making the code GPL-only for now. Also change the default to use the (in future) plain LGPL code, and deprecate the one associated with the GPL code, so we can eventually remove the GPL code. But it's also possible we decide that the copyright doesn't apply, and undo the deprecation and GPL guards. I expect that users won't notice anything. If you ask me, the old aspect method was probably an accidental bug instead of intentional behavior. Although, the new aspect method was broken too, so I had to fix it.
2017-06-18 16:27:48 +00:00
*
* Almost LGPL.
*/
#include <stdio.h>
#include <stdlib.h>
#include <stdbool.h>
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
#include <assert.h>
#include <libavutil/rational.h>
#include "config.h"
#include "options/options.h"
#include "common/msg.h"
#include "osdep/timer.h"
#include "stream/stream.h"
#include "demux/demux.h"
#include "demux/packet.h"
#include "common/codecs.h"
#include "common/recorder.h"
#include "video/out/vo.h"
#include "video/csputils.h"
#include "demux/stheader.h"
#include "video/decode/vd.h"
#include "video/decode/dec_video.h"
extern const vd_functions_t mpcodecs_vd_ffmpeg;
/* Please do not add any new decoders here. If you want to implement a new
* decoder, add it to libavcodec, except for wrappers around external
* libraries and decoders requiring binary support. */
const vd_functions_t * const mpcodecs_vd_drivers[] = {
&mpcodecs_vd_ffmpeg,
/* Please do not add any new decoders here. If you want to implement a new
* decoder, add it to libavcodec, except for wrappers around external
* libraries and decoders requiring binary support. */
NULL
};
void video_reset(struct dec_video *d_video)
{
video_vd_control(d_video, VDCTRL_RESET, NULL);
d_video->first_packet_pdts = MP_NOPTS_VALUE;
d_video->start_pts = MP_NOPTS_VALUE;
d_video->decoded_pts = MP_NOPTS_VALUE;
d_video->codec_pts = MP_NOPTS_VALUE;
d_video->codec_dts = MP_NOPTS_VALUE;
d_video->has_broken_decoded_pts = 0;
d_video->last_format = d_video->fixed_format = (struct mp_image_params){0};
d_video->dropped_frames = 0;
d_video->current_state = DATA_AGAIN;
mp_image_unrefp(&d_video->current_mpi);
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
talloc_free(d_video->packet);
d_video->packet = NULL;
talloc_free(d_video->new_segment);
d_video->new_segment = NULL;
d_video->start = d_video->end = MP_NOPTS_VALUE;
}
int video_vd_control(struct dec_video *d_video, int cmd, void *arg)
{
const struct vd_functions *vd = d_video->vd_driver;
if (vd)
return vd->control(d_video, cmd, arg);
return CONTROL_UNKNOWN;
}
void video_uninit(struct dec_video *d_video)
{
if (!d_video)
return;
mp_image_unrefp(&d_video->current_mpi);
if (d_video->vd_driver) {
MP_VERBOSE(d_video, "Uninit video.\n");
d_video->vd_driver->uninit(d_video);
}
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
talloc_free(d_video->packet);
talloc_free(d_video->new_segment);
talloc_free(d_video);
}
static int init_video_codec(struct dec_video *d_video, const char *decoder)
{
if (!d_video->vd_driver->init(d_video, decoder)) {
MP_VERBOSE(d_video, "Video decoder init failed.\n");
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
return 0;
}
return 1;
}
struct mp_decoder_list *video_decoder_list(void)
{
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
struct mp_decoder_list *list = talloc_zero(NULL, struct mp_decoder_list);
for (int i = 0; mpcodecs_vd_drivers[i] != NULL; i++)
mpcodecs_vd_drivers[i]->add_decoders(list);
return list;
}
static struct mp_decoder_list *mp_select_video_decoders(struct mp_log *log,
const char *codec,
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
char *selection)
{
struct mp_decoder_list *list = video_decoder_list();
struct mp_decoder_list *new = mp_select_decoders(log, list, codec, selection);
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
talloc_free(list);
return new;
}
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
static const struct vd_functions *find_driver(const char *name)
{
for (int i = 0; mpcodecs_vd_drivers[i] != NULL; i++) {
if (strcmp(mpcodecs_vd_drivers[i]->name, name) == 0)
return mpcodecs_vd_drivers[i];
}
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
return NULL;
}
bool video_init_best_codec(struct dec_video *d_video)
{
struct MPOpts *opts = d_video->opts;
assert(!d_video->vd_driver);
video_reset(d_video);
d_video->has_broken_packet_pts = -10; // needs 10 packets to reach decision
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
struct mp_decoder_entry *decoder = NULL;
struct mp_decoder_list *list = mp_select_video_decoders(d_video->log,
d_video->codec->codec,
opts->video_decoders);
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
2013-12-21 17:46:24 +00:00
mp_print_decoders(d_video->log, MSGL_V, "Codec list:", list);
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
for (int n = 0; n < list->num_entries; n++) {
struct mp_decoder_entry *sel = &list->entries[n];
const struct vd_functions *driver = find_driver(sel->family);
if (!driver)
continue;
MP_VERBOSE(d_video, "Opening video decoder %s\n", sel->decoder);
d_video->vd_driver = driver;
if (init_video_codec(d_video, sel->decoder)) {
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
decoder = sel;
break;
}
d_video->vd_driver = NULL;
MP_WARN(d_video, "Video decoder init failed for %s\n", sel->decoder);
}
if (d_video->vd_driver) {
d_video->decoder_desc =
talloc_asprintf(d_video, "%s (%s)", decoder->decoder, decoder->desc);
MP_VERBOSE(d_video, "Selected video codec: %s\n", d_video->decoder_desc);
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
} else {
MP_ERR(d_video, "Failed to initialize a video decoder for codec '%s'.\n",
d_video->codec->codec);
}
if (d_video->header->missing_timestamps) {
MP_WARN(d_video, "This stream has no timestamps!\n");
MP_WARN(d_video, "Making up playback time using %f FPS.\n", d_video->fps);
MP_WARN(d_video, "Seeking will probably fail badly.\n");
}
core: redo how codecs are mapped, remove codecs.conf Use codec names instead of FourCCs to identify codecs. Rewrite how codecs are selected and initialized. Now each decoder exports a list of decoders (and the codec it supports) via add_decoders(). The order matters, and the first decoder for a given decoder is preferred over the other decoders. E.g. all ad_mpg123 decoders are preferred over ad_lavc, because it comes first in the mpcodecs_ad_drivers array. Likewise, decoders within ad_lavc that are enumerated first by libavcodec (using av_codec_next()) are preferred. (This is actually critical to select h264 software decoding by default instead of vdpau. libavcodec and ffmpeg/avconv use the same method to select decoders by default, so we hope this is sane.) The codec names follow libavcodec's codec names as defined by AVCodecDescriptor.name (see libavcodec/codec_desc.c). Some decoders have names different from the canonical codec name. The AVCodecDescriptor API is relatively new, so we need a compatibility layer for older libavcodec versions for codec names that are referenced internally, and which are different from the decoder name. (Add a configure check for that, because checking versions is getting way too messy.) demux/codec_tags.c is generated from the former codecs.conf (minus "special" decoders like vdpau, and excluding the mappings that are the same as the mappings libavformat's exported RIFF tables). It contains all the mappings from FourCCs to codec name. This is needed for demux_mkv, demux_mpg, demux_avi and demux_asf. demux_lavf will set the codec as determined by libavformat, while the other demuxers have to do this on their own, using the mp_set_audio/video_codec_from_tag() functions. Note that the sh_audio/video->format members don't uniquely identify the codec anymore, and sh->codec takes over this role. Replace the --ac/--vc/--afm/--vfm with new --vd/--ad options, which provide cover the functionality of the removed switched. Note: there's no CODECS_FLAG_FLIP flag anymore. This means some obscure container/video combinations (e.g. the sample Film_200_zygo_pro.mov) are played flipped. ffplay/avplay doesn't handle this properly either, so we don't care and blame ffmeg/libav instead.
2013-02-09 14:15:19 +00:00
talloc_free(list);
return !!d_video->vd_driver;
}
static bool is_valid_peak(float sig_peak)
{
return !sig_peak || (sig_peak >= 1 && sig_peak <= 100);
}
static void fix_image_params(struct dec_video *d_video,
struct mp_image_params *params)
{
struct MPOpts *opts = d_video->opts;
struct mp_image_params p = *params;
struct mp_codec_params *c = d_video->codec;
MP_VERBOSE(d_video, "Decoder format: %s\n", mp_image_params_to_str(params));
d_video->dec_format = *params;
// While mp_image_params normally always have to have d_w/d_h set, the
// decoder signals unknown bitstream aspect ratio with both set to 0.
bool use_container = true;
if (opts->aspect_method == 1 && p.p_w > 0 && p.p_h > 0) {
MP_VERBOSE(d_video, "Using bitstream aspect ratio.\n");
use_container = false;
}
if (use_container && c->par_w > 0 && c->par_h) {
MP_VERBOSE(d_video, "Using container aspect ratio.\n");
p.p_w = c->par_w;
p.p_h = c->par_h;
}
if (opts->movie_aspect >= 0) {
MP_VERBOSE(d_video, "Forcing user-set aspect ratio.\n");
if (opts->movie_aspect == 0) {
p.p_w = p.p_h = 1;
} else {
AVRational a = av_d2q(opts->movie_aspect, INT_MAX);
mp_image_params_set_dsize(&p, a.num, a.den);
}
}
// Assume square pixels if no aspect ratio is set at all.
if (p.p_w <= 0 || p.p_h <= 0)
p.p_w = p.p_h = 1;
if (opts->video_rotate < 0) {
p.rotate = 0;
} else {
p.rotate = (p.rotate + opts->video_rotate) % 360;
}
p.stereo_out = opts->video_stereo_mode;
mp_colorspace_merge(&p.color, &c->color);
// Sanitize the HDR peak. Sadly necessary
if (!is_valid_peak(p.color.sig_peak)) {
MP_WARN(d_video, "Invalid HDR peak in stream: %f\n", p.color.sig_peak);
p.color.sig_peak = 0.0;
}
// Guess missing colorspace fields from metadata. This guarantees all
// fields are at least set to legal values afterwards.
mp_image_params_guess_csp(&p);
d_video->last_format = *params;
d_video->fixed_format = p;
}
static bool send_packet(struct dec_video *d_video, struct demux_packet *packet)
{
double pkt_pts = packet ? packet->pts : MP_NOPTS_VALUE;
double pkt_dts = packet ? packet->dts : MP_NOPTS_VALUE;
if (pkt_pts == MP_NOPTS_VALUE)
d_video->has_broken_packet_pts = 1;
bool dts_replaced = false;
if (packet && packet->dts == MP_NOPTS_VALUE && !d_video->codec->avi_dts) {
packet->dts = packet->pts;
dts_replaced = true;
}
double pkt_pdts = pkt_pts == MP_NOPTS_VALUE ? pkt_dts : pkt_pts;
if (pkt_pdts != MP_NOPTS_VALUE && d_video->first_packet_pdts == MP_NOPTS_VALUE)
d_video->first_packet_pdts = pkt_pdts;
MP_STATS(d_video, "start decode video");
bool res = d_video->vd_driver->send_packet(d_video, packet);
MP_STATS(d_video, "end decode video");
// Stream recording can't deal with almost surely wrong fake DTS.
if (dts_replaced)
packet->dts = MP_NOPTS_VALUE;
return res;
}
static bool receive_frame(struct dec_video *d_video, struct mp_image **out_image)
{
struct MPOpts *opts = d_video->opts;
struct mp_image *mpi = NULL;
assert(!*out_image);
MP_STATS(d_video, "start decode video");
bool progress = d_video->vd_driver->receive_frame(d_video, &mpi);
MP_STATS(d_video, "end decode video");
// Error, EOF, discarded frame, dropped frame, or initial codec delay.
if (!mpi)
return progress;
// Note: the PTS is reordered, but the DTS is not. Both should be monotonic.
double pts = mpi->pts;
double dts = mpi->dts;
if (pts != MP_NOPTS_VALUE) {
if (pts < d_video->codec_pts)
d_video->num_codec_pts_problems++;
d_video->codec_pts = mpi->pts;
}
if (dts != MP_NOPTS_VALUE) {
if (dts <= d_video->codec_dts)
d_video->num_codec_dts_problems++;
d_video->codec_dts = mpi->dts;
}
if (d_video->has_broken_packet_pts < 0)
d_video->has_broken_packet_pts++;
if (d_video->num_codec_pts_problems)
d_video->has_broken_packet_pts = 1;
// If PTS is unset, or non-monotonic, fall back to DTS.
if ((d_video->num_codec_pts_problems > d_video->num_codec_dts_problems ||
pts == MP_NOPTS_VALUE) && dts != MP_NOPTS_VALUE)
pts = dts;
if (!opts->correct_pts || pts == MP_NOPTS_VALUE) {
if (opts->correct_pts && !d_video->header->missing_timestamps) {
if (d_video->has_broken_decoded_pts <= 1) {
MP_WARN(d_video, "No video PTS! Making something up.\n");
if (d_video->has_broken_decoded_pts == 1)
MP_WARN(d_video, "Ignoring further missing PTS warnings.\n");
d_video->has_broken_decoded_pts++;
}
}
double frame_time = 1.0f / (d_video->fps > 0 ? d_video->fps : 25);
double base = d_video->first_packet_pdts;
pts = d_video->decoded_pts;
if (pts == MP_NOPTS_VALUE) {
pts = base == MP_NOPTS_VALUE ? 0 : base;
} else {
pts += frame_time;
}
}
if (!mp_image_params_equal(&d_video->last_format, &mpi->params))
fix_image_params(d_video, &mpi->params);
mpi->params = d_video->fixed_format;
mpi->pts = pts;
d_video->decoded_pts = pts;
video: approximate AVI timestamps via DTS handling Until now (and in mplayer traditionally), avi timestamps were handled with a timestamp FIFO. AVI timestamps are essentially just strictly increasing frame numbers and are not reordered like normal timestamps. Limiting the FIFO is required because frames can be dropped. To make it worse, frame dropping can't be distinguished from the decoder not returning output due to increasing the buffering required for B-frames. ("Measuring" the buffering at playback start seems like an interesting idea, but won't work as the buffering could be increased mid-playback.) Another problem are skipped frames (packets with data, but which do not contain a video frame). Besides dropped and skipped frames, there is the problem that we can't always know the delay. External decoders like MMAL are not going to tell us. (And later perhaps others, like direct VideoToolbox usage.) In general, this works not-well enough that I prefer the solution of passing through AVI timestamps as DTS. This is slightly incorrect, because most decoders treat DTS as mpeg-style timestamps, which already include a b-frame delay, and thus will be shifted by a few frames. This means there will be a problem with A/V sync in some situations. Note that the FFmpeg AVI demuxer shifts timestamps by an additional amount (which increases after the first seek!?!?), which makes the situation worse. It works well with VfW-muxed Matroska files, though. On RPI, the first X timestamps are broken until the MMAL decoder "locks on".
2016-02-11 15:01:11 +00:00
// Compensate for incorrectly using mpeg-style DTS for avi timestamps.
if (d_video->codec->avi_dts && opts->correct_pts &&
mpi->pts != MP_NOPTS_VALUE && d_video->fps > 0)
{
video: approximate AVI timestamps via DTS handling Until now (and in mplayer traditionally), avi timestamps were handled with a timestamp FIFO. AVI timestamps are essentially just strictly increasing frame numbers and are not reordered like normal timestamps. Limiting the FIFO is required because frames can be dropped. To make it worse, frame dropping can't be distinguished from the decoder not returning output due to increasing the buffering required for B-frames. ("Measuring" the buffering at playback start seems like an interesting idea, but won't work as the buffering could be increased mid-playback.) Another problem are skipped frames (packets with data, but which do not contain a video frame). Besides dropped and skipped frames, there is the problem that we can't always know the delay. External decoders like MMAL are not going to tell us. (And later perhaps others, like direct VideoToolbox usage.) In general, this works not-well enough that I prefer the solution of passing through AVI timestamps as DTS. This is slightly incorrect, because most decoders treat DTS as mpeg-style timestamps, which already include a b-frame delay, and thus will be shifted by a few frames. This means there will be a problem with A/V sync in some situations. Note that the FFmpeg AVI demuxer shifts timestamps by an additional amount (which increases after the first seek!?!?), which makes the situation worse. It works well with VfW-muxed Matroska files, though. On RPI, the first X timestamps are broken until the MMAL decoder "locks on".
2016-02-11 15:01:11 +00:00
int delay = -1;
video_vd_control(d_video, VDCTRL_GET_BFRAMES, &delay);
mpi->pts -= MPMAX(delay, 0) / d_video->fps;
}
*out_image = mpi;
return true;
}
void video_reset_params(struct dec_video *d_video)
{
d_video->last_format = (struct mp_image_params){0};
}
void video_get_dec_params(struct dec_video *d_video, struct mp_image_params *p)
{
*p = d_video->dec_format;
}
void video_set_framedrop(struct dec_video *d_video, bool enabled)
{
d_video->framedrop_enabled = enabled;
}
// Frames before the start timestamp can be dropped. (Used for hr-seek.)
void video_set_start(struct dec_video *d_video, double start_pts)
{
d_video->start_pts = start_pts;
}
void video_work(struct dec_video *d_video)
{
if (d_video->current_mpi || !d_video->vd_driver)
return;
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
if (!d_video->packet && !d_video->new_segment &&
demux_read_packet_async(d_video->header, &d_video->packet) == 0)
{
d_video->current_state = DATA_WAIT;
return;
}
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
if (d_video->packet && d_video->packet->new_segment) {
assert(!d_video->new_segment);
d_video->new_segment = d_video->packet;
d_video->packet = NULL;
}
double start_pts = d_video->start_pts;
if (d_video->start != MP_NOPTS_VALUE && (start_pts == MP_NOPTS_VALUE ||
d_video->start > start_pts))
start_pts = d_video->start;
int framedrop_type = d_video->framedrop_enabled ? 1 : 0;
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
if (start_pts != MP_NOPTS_VALUE && d_video->packet &&
d_video->packet->pts < start_pts - .005 &&
!d_video->has_broken_packet_pts)
{
framedrop_type = 2;
}
d_video->vd_driver->control(d_video, VDCTRL_SET_FRAMEDROP, &framedrop_type);
if (send_packet(d_video, d_video->packet)) {
if (d_video->recorder_sink)
mp_recorder_feed_packet(d_video->recorder_sink, d_video->packet);
talloc_free(d_video->packet);
d_video->packet = NULL;
}
bool progress = receive_frame(d_video, &d_video->current_mpi);
d_video->current_state = DATA_OK;
if (!progress) {
d_video->current_state = DATA_EOF;
} else if (!d_video->current_mpi) {
if (framedrop_type == 1)
d_video->dropped_frames += 1;
d_video->current_state = DATA_AGAIN;
}
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
bool segment_ended = d_video->current_state == DATA_EOF;
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
if (d_video->current_mpi && d_video->current_mpi->pts != MP_NOPTS_VALUE) {
double vpts = d_video->current_mpi->pts;
segment_ended = d_video->end != MP_NOPTS_VALUE && vpts >= d_video->end;
if ((d_video->start != MP_NOPTS_VALUE && vpts < d_video->start)
|| segment_ended)
{
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
talloc_free(d_video->current_mpi);
d_video->current_mpi = NULL;
}
}
// If there's a new segment, start it as soon as we're drained/finished.
if (segment_ended && d_video->new_segment) {
struct demux_packet *new_segment = d_video->new_segment;
d_video->new_segment = NULL;
if (d_video->codec == new_segment->codec) {
video_reset(d_video);
} else {
d_video->codec = new_segment->codec;
d_video->vd_driver->uninit(d_video);
d_video->vd_driver = NULL;
video_init_best_codec(d_video);
}
Rewrite ordered chapters and timeline stuff This uses a different method to piece segments together. The old approach basically changes to a new file (with a new start offset) any time a segment ends. This meant waiting for audio/video end on segment end, and then changing to the new segment all at once. It had a very weird impact on the playback core, and some things (like truly gapless segment transitions, or frame backstepping) just didn't work. The new approach adds the demux_timeline pseudo-demuxer, which presents an uniform packet stream from the many segments. This is pretty similar to how ordered chapters are implemented everywhere else. It also reminds of the FFmpeg concat pseudo-demuxer. The "pure" version of this approach doesn't work though. Segments can actually have different codec configurations (different extradata), and subtitles are most likely broken too. (Subtitles have multiple corner cases which break the pure stream-concatenation approach completely.) To counter this, we do two things: - Reinit the decoder with each segment. We go as far as allowing concatenating files with completely different codecs for the sake of EDL (which also uses the timeline infrastructure). A "lighter" approach would try to make use of decoder mechanism to update e.g. the extradata, but that seems fragile. - Clip decoded data to segment boundaries. This is equivalent to normal playback core mechanisms like hr-seek, but now the playback core doesn't need to care about these things. These two mechanisms are equivalent to what happened in the old implementation, except they don't happen in the playback core anymore. In other words, the playback core is completely relieved from timeline implementation details. (Which honestly is exactly what I'm trying to do here. I don't think ordered chapter behavior deserves improvement, even if it's bad - but I want to get it out from the playback core.) There is code duplication between audio and video decoder common code. This is awful and could be shareable - but this will happen later. Note that the audio path has some code to clip audio frames for the purpose of codec preroll/gapless handling, but it's not shared as sharing it would cause more pain than it would help.
2016-02-15 20:04:07 +00:00
d_video->start = new_segment->start;
d_video->end = new_segment->end;
new_segment->new_segment = false;
d_video->packet = new_segment;
d_video->current_state = DATA_AGAIN;
}
}
// Fetch an image decoded with video_work(). Returns one of:
// DATA_OK: *out_mpi is set to a new image
// DATA_WAIT: waiting for demuxer; will receive a wakeup signal
// DATA_EOF: end of file, no more frames to be expected
// DATA_AGAIN: dropped frame or something similar
int video_get_frame(struct dec_video *d_video, struct mp_image **out_mpi)
{
*out_mpi = NULL;
if (d_video->current_mpi) {
*out_mpi = d_video->current_mpi;
d_video->current_mpi = NULL;
return DATA_OK;
}
if (d_video->current_state == DATA_OK)
return DATA_AGAIN;
return d_video->current_state;
}