mpv/demux/demux_lavf.c

1072 lines
37 KiB
C
Raw Normal View History

/*
* Copyright (C) 2004 Michael Niedermayer <michaelni@gmx.at>
*
* This file is part of MPlayer.
*
* MPlayer 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.
*
* MPlayer 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 MPlayer; if not, write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
*/
// #include <stdio.h>
#include <stdlib.h>
// #include <unistd.h>
#include <limits.h>
Add improved relative seek mode When the new mode is active relative seeks are converted to absolute ones (current video pts + relative seek amount) and forward/backward flag before being sent to the demuxer. This mode is used if the demuxer has set the accurate_seek field in the demuxer struct and there is a video stream. At the moment the mkv and lavf demuxers enable the flag. This change is useful for later Matroska ordered chapter support (and for more general timelime editing), but also fixes problems in existing functionality. The main problem with the old mode, where relative seeks are passed directly to the demuxer, is that the user wants to seek relative to the currently displayed position but the demuxer does not know what that position is. There can be an arbitrary amount of buffering between the demuxer read position and what is displayed on the screen. In some situations this makes small seeks fail to move backward at all (especially visible at high playback speed, when audio needs to be demuxed and decoded further ahead to fill the output buffers after resampling). Some container formats that can be used with the lavf demuxer do not always have reliable timestamps that could be used for unambiguous absolute seeking. However I made the demuxer always enable the new mode because it already converted all seeks to absolute ones before sending them to libavformat, so cases without reliable absolute seeks were failing already and this should only improve the working cases.
2009-03-19 03:25:12 +00:00
#include <stdbool.h>
#include <string.h>
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
#include <assert.h>
#include <libavformat/avformat.h>
#include <libavformat/avio.h>
#include <libavutil/avutil.h>
#include <libavutil/avstring.h>
#include <libavutil/mathematics.h>
#include <libavutil/opt.h>
#include "compat/libav.h"
#include "config.h"
#include "core/options.h"
#include "core/mp_msg.h"
#include "core/av_opts.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 "core/av_common.h"
#include "core/bstr.h"
#include "stream/stream.h"
#include "aviprint.h"
#include "demux.h"
#include "stheader.h"
#include "core/m_option.h"
#include "mp_taglists.h"
#define INITIAL_PROBE_SIZE STREAM_BUFFER_SIZE
#define PROBE_BUF_SIZE (2 * 1024 * 1024)
const m_option_t lavfdopts_conf[] = {
OPT_INTRANGE("probesize", lavfdopts.probesize, 0, 32, INT_MAX),
OPT_STRING("format", lavfdopts.format, 0),
OPT_INTRANGE("analyzeduration", lavfdopts.analyzeduration, 0, 0, INT_MAX),
OPT_INTRANGE("probescore", lavfdopts.probescore, 0, 0, 100),
OPT_STRING("cryptokey", lavfdopts.cryptokey, 0),
OPT_STRING("o", lavfdopts.avopt, 0),
{NULL, NULL, 0, 0, 0, 0, NULL}
};
#define BIO_BUFFER_SIZE 32768
typedef struct lavf_priv {
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
char *filename;
AVInputFormat *avif;
AVFormatContext *avfc;
AVIOContext *pb;
uint8_t buffer[BIO_BUFFER_SIZE];
int audio_streams;
int video_streams;
int sub_streams;
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
int autoselect_sub;
int64_t last_pts;
int astreams[MAX_A_STREAMS];
int vstreams[MAX_V_STREAMS];
int sstreams[MAX_S_STREAMS];
int cur_program;
int nb_streams_last;
bool use_dts;
bool seek_by_bytes;
int bitrate;
char *mime_type;
} lavf_priv_t;
static const char *map_demuxer_mime_type[][2] = {
{"audio/aacp", "aac"},
{0}
};
static const char *find_demuxer_from_mime_type(char *mime_type)
{
for (int n = 0; map_demuxer_mime_type[n][0]; n++) {
if (strcasecmp(map_demuxer_mime_type[n][0], mime_type) == 0)
return map_demuxer_mime_type[n][1];
}
return NULL;
}
static int mp_read(void *opaque, uint8_t *buf, int size)
{
struct demuxer *demuxer = opaque;
struct stream *stream = demuxer->stream;
int ret;
ret = stream_read(stream, buf, size);
mp_msg(MSGT_HEADER, MSGL_DBG2,
"%d=mp_read(%p, %p, %d), pos: %"PRId64", eof:%d\n",
ret, stream, buf, size, stream_tell(stream), stream->eof);
return ret;
}
static int64_t mp_seek(void *opaque, int64_t pos, int whence)
{
struct demuxer *demuxer = opaque;
struct stream *stream = demuxer->stream;
int64_t current_pos;
mp_msg(MSGT_HEADER, MSGL_DBG2, "mp_seek(%p, %"PRId64", %d)\n",
stream, pos, whence);
if (whence == SEEK_CUR)
pos += stream_tell(stream);
else if (whence == SEEK_END && stream->end_pos > 0)
pos += stream->end_pos;
else if (whence == SEEK_SET)
pos += stream->start_pos;
else if (whence == AVSEEK_SIZE && stream->end_pos > 0) {
stream_update_size(stream);
return stream->end_pos - stream->start_pos;
} else
return -1;
if (pos < 0)
return -1;
current_pos = stream_tell(stream);
if (stream_seek(stream, pos) == 0) {
stream_reset(stream);
stream_seek(stream, current_pos);
return -1;
}
return pos - stream->start_pos;
}
static int64_t mp_read_seek(void *opaque, int stream_idx, int64_t ts, int flags)
{
struct demuxer *demuxer = opaque;
struct stream *stream = demuxer->stream;
struct lavf_priv *priv = demuxer->priv;
AVStream *st = priv->avfc->streams[stream_idx];
double pts = (double)ts * st->time_base.num / st->time_base.den;
int ret = stream_control(stream, STREAM_CTRL_SEEK_TO_TIME, &pts);
if (ret < 0)
ret = AVERROR(ENOSYS);
return ret;
}
static void list_formats(void)
{
mp_msg(MSGT_DEMUX, MSGL_INFO, "Available lavf input formats:\n");
AVInputFormat *fmt = NULL;
while ((fmt = av_iformat_next(fmt)))
mp_msg(MSGT_DEMUX, MSGL_INFO, "%15s : %s\n", fmt->name, fmt->long_name);
}
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
static char *remove_prefix(char *s, const char **prefixes)
{
for (int n = 0; prefixes[n]; n++) {
int len = strlen(prefixes[n]);
if (strncmp(s, prefixes[n], len) == 0)
return s + len;
}
return s;
}
static const char *prefixes[] =
{"ffmpeg://", "lavf://", "avdevice://", "av://", NULL};
static int lavf_check_file(demuxer_t *demuxer)
{
struct MPOpts *opts = demuxer->opts;
struct lavfdopts *lavfdopts = &opts->lavfdopts;
AVProbeData avpd;
lavf_priv_t *priv;
int probe_data_size = 0;
int read_size = INITIAL_PROBE_SIZE;
int score;
assert(!demuxer->priv);
demuxer->priv = talloc_zero(NULL, lavf_priv_t);
priv = demuxer->priv;
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
priv->autoselect_sub = -1;
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
priv->filename = demuxer->stream->url;
if (!priv->filename) {
priv->filename = "mp:unknown";
mp_msg(MSGT_DEMUX, MSGL_WARN, "Stream url is not set!\n");
}
priv->filename = remove_prefix(priv->filename, prefixes);
char *avdevice_format = NULL;
if (demuxer->stream->type == STREAMTYPE_AVDEVICE) {
// always require filename in the form "format:filename"
char *sep = strchr(priv->filename, ':');
if (!sep) {
mp_msg(MSGT_DEMUX, MSGL_FATAL,
"Must specify filename in 'format:filename' form\n");
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
return 0;
}
avdevice_format = talloc_strndup(priv, priv->filename,
sep - priv->filename);
priv->filename = sep + 1;
}
char *format = lavfdopts->format;
if (!format)
format = demuxer->stream->lavf_type;
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
if (!format)
format = avdevice_format;
if (!format && demuxer->stream->mime_type)
format = (char *)find_demuxer_from_mime_type(demuxer->stream->mime_type);
if (format) {
if (strcmp(format, "help") == 0) {
list_formats();
return 0;
}
priv->avif = av_find_input_format(format);
if (!priv->avif) {
mp_msg(MSGT_DEMUX, MSGL_FATAL, "Unknown lavf format %s\n", format);
return 0;
}
mp_msg(MSGT_DEMUX, MSGL_INFO, "Forced lavf %s demuxer\n",
priv->avif->long_name);
goto success;
}
// AVPROBE_SCORE_MAX/4 + 1 is the "recommended" limit. Below that, the user
// is supposed to retry with larger probe sizes until a higher value is
// reached.
int min_probe = AVPROBE_SCORE_MAX/4 + 1;
if (lavfdopts->probescore)
min_probe = lavfdopts->probescore;
avpd.buf = av_mallocz(FFMAX(BIO_BUFFER_SIZE, PROBE_BUF_SIZE) +
FF_INPUT_BUFFER_PADDING_SIZE);
do {
read_size = stream_read(demuxer->stream, avpd.buf + probe_data_size,
read_size);
if (read_size < 0) {
av_free(avpd.buf);
return 0;
}
probe_data_size += read_size;
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
avpd.filename = priv->filename;
avpd.buf_size = probe_data_size;
score = 0;
priv->avif = av_probe_input_format2(&avpd, probe_data_size > 0, &score);
if (priv->avif) {
mp_msg(MSGT_HEADER, MSGL_V, "Found '%s' at score=%d size=%d.\n",
priv->avif->name, score, probe_data_size);
}
if (priv->avif && score >= min_probe)
break;
priv->avif = NULL;
read_size = FFMIN(2 * read_size, PROBE_BUF_SIZE - probe_data_size);
} while (read_size > 0 && probe_data_size < PROBE_BUF_SIZE);
av_free(avpd.buf);
if (!priv->avif) {
mp_msg(MSGT_HEADER, MSGL_V,
"No format found, try lowering probescore.\n");
return 0;
}
success:
demuxer->filetype = priv->avif->long_name;
if (!demuxer->filetype)
demuxer->filetype = priv->avif->name;
return DEMUXER_TYPE_LAVF;
}
static bool matches_avinputformat_name(struct lavf_priv *priv,
const char *name)
{
const char *avifname = priv->avif->name;
while (1) {
const char *next = strchr(avifname, ',');
if (!next)
return !strcmp(avifname, name);
int len = next - avifname;
if (len == strlen(name) && !memcmp(avifname, name, len))
return true;
avifname = next + 1;
}
}
static uint8_t char2int(char c)
{
if (c >= '0' && c <= '9') return c - '0';
if (c >= 'a' && c <= 'f') return c - 'a' + 10;
if (c >= 'A' && c <= 'F') return c - 'A' + 10;
return 0;
}
static void parse_cryptokey(AVFormatContext *avfc, const char *str)
{
int len = strlen(str) / 2;
uint8_t *key = av_mallocz(len);
int i;
avfc->keylen = len;
avfc->key = key;
for (i = 0; i < len; i++, str += 2)
*key++ = (char2int(str[0]) << 4) | char2int(str[1]);
}
static void handle_stream(demuxer_t *demuxer, AVFormatContext *avfc, int i)
{
lavf_priv_t *priv = demuxer->priv;
AVStream *st = avfc->streams[i];
AVCodecContext *codec = st->codec;
char *stream_type = NULL;
int stream_id;
AVDictionaryEntry *lang = av_dict_get(st->metadata, "language", NULL, 0);
AVDictionaryEntry *title = av_dict_get(st->metadata, "title", NULL, 0);
// Work around collisions resulting from the hacks changing codec_tag.
int lavf_codec_tag = codec->codec_tag;
// Don't use native MPEG codec tag values with our generic tag tables.
// May contain for example value 3 for MP3, which we'd map to PCM audio.
if (matches_avinputformat_name(priv, "mpeg") ||
matches_avinputformat_name(priv, "mpegts"))
codec->codec_tag = 0;
int override_tag = mp_taglist_override(codec->codec_id);
// For some formats (like PCM) always trust CODEC_ID_* more than codec_tag
if (override_tag)
codec->codec_tag = override_tag;
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
const char *mp_codec = mp_codec_from_av_codec_id(codec->codec_id);
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
bool set_demuxer_id = matches_avinputformat_name(priv, "mpeg");
switch (codec->codec_type) {
case AVMEDIA_TYPE_AUDIO: {
WAVEFORMATEX *wf;
sh_audio_t *sh_audio;
sh_audio = new_sh_audio_aid(demuxer, i, priv->audio_streams);
if (!sh_audio)
break;
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
if (set_demuxer_id)
sh_audio->gsh->demuxer_id = st->id;
stream_type = "audio";
priv->astreams[priv->audio_streams] = 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
sh_audio->gsh->codec = mp_codec;
sh_audio->gsh->lavf_codec_tag = lavf_codec_tag;
wf = calloc(sizeof(*wf) + codec->extradata_size, 1);
// mp4a tag is used for all mp4 files no matter what they actually contain
if (codec->codec_tag == MKTAG('m', 'p', '4', 'a'))
codec->codec_tag = 0;
if (!codec->codec_tag)
codec->codec_tag = mp_taglist_audio(codec->codec_id);
if (!codec->codec_tag)
codec->codec_tag = -1;
wf->wFormatTag = codec->codec_tag;
wf->nChannels = codec->channels;
wf->nSamplesPerSec = codec->sample_rate;
wf->nAvgBytesPerSec = codec->bit_rate / 8;
wf->nBlockAlign = codec->block_align;
wf->wBitsPerSample = codec->bits_per_coded_sample;
wf->cbSize = codec->extradata_size;
if (codec->extradata_size)
memcpy(wf + 1, codec->extradata, codec->extradata_size);
sh_audio->wf = wf;
sh_audio->audio.dwSampleSize = codec->block_align;
if (codec->frame_size && codec->sample_rate) {
sh_audio->audio.dwScale = codec->frame_size;
sh_audio->audio.dwRate = codec->sample_rate;
} else {
sh_audio->audio.dwScale = codec->block_align ? codec->block_align * 8 : 8;
sh_audio->audio.dwRate = codec->bit_rate;
}
int g = av_gcd(sh_audio->audio.dwScale, sh_audio->audio.dwRate);
sh_audio->audio.dwScale /= g;
sh_audio->audio.dwRate /= g;
// printf("sca:%d rat:%d fs:%d sr:%d ba:%d\n", sh_audio->audio.dwScale, sh_audio->audio.dwRate, codec->frame_size, codec->sample_rate, codec->block_align);
sh_audio->ds = demuxer->audio;
sh_audio->format = codec->codec_tag;
sh_audio->channels = codec->channels;
sh_audio->samplerate = codec->sample_rate;
sh_audio->i_bps = codec->bit_rate / 8;
switch (codec->codec_id) {
case CODEC_ID_PCM_ALAW:
sh_audio->format = 0x6;
break;
case CODEC_ID_PCM_MULAW:
sh_audio->format = 0x7;
break;
}
if (title && title->value) {
sh_audio->gsh->title = talloc_strdup(sh_audio, title->value);
mp_msg(MSGT_IDENTIFY, MSGL_INFO, "ID_AID_%d_NAME=%s\n",
priv->audio_streams, title->value);
}
if (lang && lang->value) {
sh_audio->lang = talloc_strdup(sh_audio, lang->value);
mp_msg(MSGT_IDENTIFY, MSGL_INFO, "ID_AID_%d_LANG=%s\n",
priv->audio_streams, sh_audio->lang);
}
if (st->disposition & AV_DISPOSITION_DEFAULT)
sh_audio->gsh->default_track = 1;
if (mp_msg_test(MSGT_HEADER, MSGL_V))
print_wave_header(sh_audio->wf, MSGL_V);
st->discard = AVDISCARD_ALL;
stream_id = priv->audio_streams++;
break;
}
case AVMEDIA_TYPE_VIDEO: {
sh_video_t *sh_video;
BITMAPINFOHEADER *bih;
sh_video = new_sh_video_vid(demuxer, i, priv->video_streams);
if (!sh_video)
break;
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
if (set_demuxer_id)
sh_video->gsh->demuxer_id = st->id;
stream_type = "video";
priv->vstreams[priv->video_streams] = 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
sh_video->gsh->codec = mp_codec;
sh_video->gsh->lavf_codec_tag = lavf_codec_tag;
if (st->disposition & AV_DISPOSITION_ATTACHED_PIC)
sh_video->gsh->attached_picture = true;
bih = calloc(sizeof(*bih) + codec->extradata_size, 1);
if (codec->codec_id == CODEC_ID_RAWVIDEO) {
switch (codec->pix_fmt) {
case PIX_FMT_RGB24:
codec->codec_tag = MKTAG(24, 'B', 'G', 'R');
case PIX_FMT_BGR24:
codec->codec_tag = MKTAG(24, 'R', 'G', 'B');
}
if (!codec->codec_tag)
codec->codec_tag = avcodec_pix_fmt_to_codec_tag(codec->pix_fmt);
video, audio: use lavc decoders without codecs.conf entries Add support for using libavcodec decoders that do not have entries in codecs.conf. This is currently only used with demux_lavf, and the codec selection is based on codec_id returned by libavformat. Also modify codec-related terminal output somewhat to make it use information from libavcodec and avoid excessively long default output. The new any-lavc-codec support is implemented with codecs.conf entries that invoke vd_ffmpeg/ad_ffmpeg without directly specifying any libavcodec codec name. In this mode, the decoders now instead select the libavcodec codec based on codec_id previously set by demux_lavf (if any). These new "generic" codecs.conf entries specify "status buggy", so that they're tried after any specific entries with higher-priority status. Add new directive "anyinput" to codecs.conf syntax. This means the entry will always match regardless of fourcc. This is used for the above new codecs.conf entries (so the driver always gets to decide whether to accept the input, and will fail init() if it can't find a suitable codec in libavcodec). Remove parsing support for the obsolete codecs.conf directive "cpuflags". This directive has not had any effect and has not been used in default codecs.conf since many years ago. Shorten codec-related terminal output. When using libavcodec decoders, show the libavcodec long_name field rather than codecs.conf "info" field as the name of the codec. Stop showing the codecs.conf entry name and "vfm/afm" name by default, as these are rarely needed; they're now in verbose output only. Show "VIDEO:" line at VO initialization rather than at demuxer open. This didn't really belong in demuxer code; the new location may show more accurate values (known after decoder has been opened) and works right if video track is changed after initial demuxer open. The vd.c changes (primarily done for terminal output changes) remove round-to-even behavior from code setting dimensions based on aspect ratio. I hope nothing depended on this; at least the even values were not consistently guaranteed anyway, as the rounding code did not run if the video file did not specify a nonzero aspect value.
2012-07-24 06:01:47 +00:00
} else if (!codec->codec_tag) {
codec->codec_tag = mp_taglist_video(codec->codec_id);
video, audio: use lavc decoders without codecs.conf entries Add support for using libavcodec decoders that do not have entries in codecs.conf. This is currently only used with demux_lavf, and the codec selection is based on codec_id returned by libavformat. Also modify codec-related terminal output somewhat to make it use information from libavcodec and avoid excessively long default output. The new any-lavc-codec support is implemented with codecs.conf entries that invoke vd_ffmpeg/ad_ffmpeg without directly specifying any libavcodec codec name. In this mode, the decoders now instead select the libavcodec codec based on codec_id previously set by demux_lavf (if any). These new "generic" codecs.conf entries specify "status buggy", so that they're tried after any specific entries with higher-priority status. Add new directive "anyinput" to codecs.conf syntax. This means the entry will always match regardless of fourcc. This is used for the above new codecs.conf entries (so the driver always gets to decide whether to accept the input, and will fail init() if it can't find a suitable codec in libavcodec). Remove parsing support for the obsolete codecs.conf directive "cpuflags". This directive has not had any effect and has not been used in default codecs.conf since many years ago. Shorten codec-related terminal output. When using libavcodec decoders, show the libavcodec long_name field rather than codecs.conf "info" field as the name of the codec. Stop showing the codecs.conf entry name and "vfm/afm" name by default, as these are rarely needed; they're now in verbose output only. Show "VIDEO:" line at VO initialization rather than at demuxer open. This didn't really belong in demuxer code; the new location may show more accurate values (known after decoder has been opened) and works right if video track is changed after initial demuxer open. The vd.c changes (primarily done for terminal output changes) remove round-to-even behavior from code setting dimensions based on aspect ratio. I hope nothing depended on this; at least the even values were not consistently guaranteed anyway, as the rounding code did not run if the video file did not specify a nonzero aspect value.
2012-07-24 06:01:47 +00:00
/* 0 might mean either unset or rawvideo; if codec_id
* was not RAWVIDEO assume it's unset
*/
if (!codec->codec_tag)
codec->codec_tag = -1;
}
bih->biSize = sizeof(*bih) + codec->extradata_size;
bih->biWidth = codec->width;
bih->biHeight = codec->height;
bih->biBitCount = codec->bits_per_coded_sample;
bih->biSizeImage = bih->biWidth * bih->biHeight * bih->biBitCount / 8;
bih->biCompression = codec->codec_tag;
sh_video->bih = bih;
sh_video->disp_w = codec->width;
sh_video->disp_h = codec->height;
if (st->time_base.den) { /* if container has time_base, use that */
sh_video->video.dwRate = st->time_base.den;
sh_video->video.dwScale = st->time_base.num;
} else {
sh_video->video.dwRate = codec->time_base.den;
sh_video->video.dwScale = codec->time_base.num;
}
/* Try to make up some frame rate value, even if it's not reliable.
* FPS information is needed to support subtitle formats which base
* timing on frame numbers.
* Libavformat seems to report no "reliable" FPS value for AVI files,
* while they are typically constant enough FPS that the value this
* heuristic makes up works with subtitles in practice.
*/
double fps;
if (st->r_frame_rate.num)
fps = av_q2d(st->r_frame_rate);
else
fps = 1.0 / FFMAX(av_q2d(st->time_base),
av_q2d(st->codec->time_base) *
st->codec->ticks_per_frame);
sh_video->fps = fps;
sh_video->frametime = 1 / fps;
sh_video->format = bih->biCompression;
if (st->sample_aspect_ratio.num)
sh_video->aspect = codec->width * st->sample_aspect_ratio.num
/ (float)(codec->height * st->sample_aspect_ratio.den);
else
sh_video->aspect = codec->width * codec->sample_aspect_ratio.num
/ (float)(codec->height * codec->sample_aspect_ratio.den);
sh_video->i_bps = codec->bit_rate / 8;
if (title && title->value) {
sh_video->gsh->title = talloc_strdup(sh_video, title->value);
mp_msg(MSGT_IDENTIFY, MSGL_INFO, "ID_VID_%d_NAME=%s\n",
priv->video_streams, title->value);
}
mp_msg(MSGT_DEMUX, MSGL_DBG2, "aspect= %d*%d/(%d*%d)\n",
codec->width, codec->sample_aspect_ratio.num,
codec->height, codec->sample_aspect_ratio.den);
sh_video->ds = demuxer->video;
if (codec->extradata_size)
memcpy(sh_video->bih + 1, codec->extradata, codec->extradata_size);
if ( mp_msg_test(MSGT_HEADER, MSGL_V))
print_video_header(sh_video->bih, MSGL_V);
if (demuxer->video->id != priv->video_streams
&& demuxer->video->id != -1)
st->discard = AVDISCARD_ALL;
else {
demuxer->video->id = i;
demuxer->video->sh = demuxer->v_streams[i];
}
stream_id = priv->video_streams++;
break;
}
case AVMEDIA_TYPE_SUBTITLE: {
sh_sub_t *sh_sub;
char type;
if (codec->codec_id == CODEC_ID_TEXT ||
codec->codec_id == AV_CODEC_ID_SUBRIP)
type = 't';
else if (codec->codec_id == CODEC_ID_MOV_TEXT)
type = 'm';
else if (codec->codec_id == CODEC_ID_SSA)
type = 'a';
else if (codec->codec_id == CODEC_ID_DVD_SUBTITLE)
type = 'v';
else if (codec->codec_id == CODEC_ID_XSUB)
type = 'x';
else if (codec->codec_id == CODEC_ID_DVB_SUBTITLE)
type = 'b';
else if (codec->codec_id == CODEC_ID_DVB_TELETEXT)
type = 'd';
else if (codec->codec_id == CODEC_ID_HDMV_PGS_SUBTITLE)
type = 'p';
else
break;
sh_sub = new_sh_sub_sid(demuxer, i, priv->sub_streams);
if (!sh_sub)
break;
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
if (set_demuxer_id)
sh_sub->gsh->demuxer_id = st->id;
stream_type = "subtitle";
priv->sstreams[priv->sub_streams] = 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
sh_sub->gsh->codec = mp_codec;
sh_sub->gsh->lavf_codec_tag = lavf_codec_tag;
sh_sub->type = type;
if (codec->extradata_size) {
sh_sub->extradata = malloc(codec->extradata_size);
memcpy(sh_sub->extradata, codec->extradata, codec->extradata_size);
sh_sub->extradata_len = codec->extradata_size;
}
if (title && title->value) {
sh_sub->gsh->title = talloc_strdup(sh_sub, title->value);
mp_msg(MSGT_IDENTIFY, MSGL_INFO, "ID_SID_%d_NAME=%s\n",
priv->sub_streams, title->value);
}
if (lang && lang->value) {
sh_sub->lang = talloc_strdup(sh_sub, lang->value);
mp_msg(MSGT_IDENTIFY, MSGL_INFO, "ID_SID_%d_LANG=%s\n",
priv->sub_streams, sh_sub->lang);
}
if (st->disposition & AV_DISPOSITION_DEFAULT)
sh_sub->gsh->default_track = 1;
stream_id = priv->sub_streams++;
break;
}
case AVMEDIA_TYPE_ATTACHMENT: {
AVDictionaryEntry *ftag = av_dict_get(st->metadata, "filename",
NULL, 0);
char *filename = ftag ? ftag->value : NULL;
if (st->codec->codec_id == CODEC_ID_TTF)
demuxer_add_attachment(demuxer, bstr0(filename),
bstr0("application/x-truetype-font"),
(struct bstr){codec->extradata,
codec->extradata_size});
break;
}
default:
st->discard = AVDISCARD_ALL;
}
if (stream_type) {
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
mp_msg(MSGT_DEMUX, MSGL_V, "[lavf] stream %d: %s, -%cid %d",
i, stream_type, *stream_type, stream_id);
if (lang && lang->value && *stream_type != 'v')
mp_msg(MSGT_DEMUX, MSGL_V, ", -%clang %s",
*stream_type, lang->value);
if (title && title->value)
mp_msg(MSGT_DEMUX, MSGL_V, ", %s", title->value);
mp_msg(MSGT_DEMUX, MSGL_V, "\n");
}
}
static demuxer_t *demux_open_lavf(demuxer_t *demuxer)
{
2008-04-16 04:11:12 +00:00
struct MPOpts *opts = demuxer->opts;
struct lavfdopts *lavfdopts = &opts->lavfdopts;
AVFormatContext *avfc;
AVDictionaryEntry *t = NULL;
lavf_priv_t *priv = demuxer->priv;
int i;
// do not allow forcing the demuxer
if (!priv->avif)
return NULL;
stream_seek(demuxer->stream, 0);
avfc = avformat_alloc_context();
if (lavfdopts->cryptokey)
parse_cryptokey(avfc, lavfdopts->cryptokey);
if (matches_avinputformat_name(priv, "avi")) {
/* for avi libavformat returns the avi timestamps in .dts,
* some made-up stuff that's not really pts in .pts */
priv->use_dts = true;
demuxer->timestamp_type = TIMESTAMP_TYPE_SORT;
} else {
if (opts->user_correct_pts != 0)
avfc->flags |= AVFMT_FLAG_GENPTS;
}
if (index_mode == 0)
avfc->flags |= AVFMT_FLAG_IGNIDX;
if (lavfdopts->probesize) {
if (av_opt_set_int(avfc, "probesize", lavfdopts->probesize, 0) < 0)
mp_msg(MSGT_HEADER, MSGL_ERR,
"demux_lavf, couldn't set option probesize to %u\n",
lavfdopts->probesize);
}
if (lavfdopts->analyzeduration) {
if (av_opt_set_int(avfc, "analyzeduration",
lavfdopts->analyzeduration * AV_TIME_BASE, 0) < 0)
mp_msg(MSGT_HEADER, MSGL_ERR, "demux_lavf, couldn't set option "
"analyzeduration to %u\n", lavfdopts->analyzeduration);
}
if (lavfdopts->avopt) {
if (parse_avopts(avfc, lavfdopts->avopt) < 0) {
mp_msg(MSGT_HEADER, MSGL_ERR,
"Your options /%s/ look like gibberish to me pal\n",
lavfdopts->avopt);
return NULL;
}
}
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
if (!(priv->avif->flags & AVFMT_NOFILE) &&
demuxer->stream->type != STREAMTYPE_AVDEVICE)
{
priv->pb = avio_alloc_context(priv->buffer, BIO_BUFFER_SIZE, 0,
demuxer, mp_read, NULL, mp_seek);
priv->pb->read_seek = mp_read_seek;
priv->pb->seekable = demuxer->stream->end_pos
&& (demuxer->stream->flags & MP_STREAM_SEEK) == MP_STREAM_SEEK
? AVIO_SEEKABLE_NORMAL : 0;
avfc->pb = priv->pb;
}
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
if (avformat_open_input(&avfc, priv->filename, priv->avif, NULL) < 0) {
mp_msg(MSGT_HEADER, MSGL_ERR,
"LAVF_header: avformat_open_input() failed\n");
return NULL;
}
priv->avfc = avfc;
if (avformat_find_stream_info(avfc, NULL) < 0) {
mp_msg(MSGT_HEADER, MSGL_ERR,
"LAVF_header: av_find_stream_info() failed\n");
return NULL;
}
/* Add metadata. */
while ((t = av_dict_get(avfc->metadata, "", t,
AV_DICT_IGNORE_SUFFIX)))
demux_info_add(demuxer, t->key, t->value);
for (i = 0; i < avfc->nb_chapters; i++) {
AVChapter *c = avfc->chapters[i];
uint64_t start = av_rescale_q(c->start, c->time_base,
(AVRational){1, 1000000000});
uint64_t end = av_rescale_q(c->end, c->time_base,
(AVRational){1, 1000000000});
t = av_dict_get(c->metadata, "title", NULL, 0);
demuxer_add_chapter(demuxer, t ? bstr0(t->value) : bstr0(NULL),
start, end);
}
for (i = 0; i < avfc->nb_streams; i++)
handle_stream(demuxer, avfc, i);
priv->nb_streams_last = avfc->nb_streams;
if (avfc->nb_programs) {
int p;
for (p = 0; p < avfc->nb_programs; p++) {
AVProgram *program = avfc->programs[p];
t = av_dict_get(program->metadata, "title", NULL, 0);
mp_msg(MSGT_HEADER, MSGL_INFO, "LAVF: Program %d %s\n",
program->id, t ? t->value : "");
mp_msg(MSGT_IDENTIFY, MSGL_V, "PROGRAM_ID=%d\n", program->id);
}
}
mp_msg(MSGT_HEADER, MSGL_V, "LAVF: %d audio and %d video streams found\n",
priv->audio_streams, priv->video_streams);
mp_msg(MSGT_HEADER, MSGL_V, "LAVF: build %d\n", LIBAVFORMAT_BUILD);
demuxer->audio->id = -2; // wait for higher-level code to select track
if (!priv->video_streams) {
demuxer->video->id = -2; // audio-only / sub-only
}
// disabled because unreliable per-stream bitrate values returned
// by libavformat trigger this heuristic incorrectly and break things
#if 0
/* libavformat sets bitrate for mpeg based on pts at start and end
* of file, which fails for files with pts resets. So calculate our
* own bitrate estimate. */
if (priv->avif->flags & AVFMT_TS_DISCONT) {
for (int i = 0; i < avfc->nb_streams; i++)
priv->bitrate += avfc->streams[i]->codec->bit_rate;
/* pts-based is more accurate if there are no resets; try to make
* a somewhat reasonable guess */
if (!avfc->duration || avfc->duration == AV_NOPTS_VALUE
|| priv->bitrate && (avfc->bit_rate < priv->bitrate / 2
|| avfc->bit_rate > priv->bitrate * 2))
priv->seek_by_bytes = true;
if (!priv->bitrate)
priv->bitrate = 1440000;
}
#endif
demuxer->accurate_seek = !priv->seek_by_bytes;
Add improved relative seek mode When the new mode is active relative seeks are converted to absolute ones (current video pts + relative seek amount) and forward/backward flag before being sent to the demuxer. This mode is used if the demuxer has set the accurate_seek field in the demuxer struct and there is a video stream. At the moment the mkv and lavf demuxers enable the flag. This change is useful for later Matroska ordered chapter support (and for more general timelime editing), but also fixes problems in existing functionality. The main problem with the old mode, where relative seeks are passed directly to the demuxer, is that the user wants to seek relative to the currently displayed position but the demuxer does not know what that position is. There can be an arbitrary amount of buffering between the demuxer read position and what is displayed on the screen. In some situations this makes small seeks fail to move backward at all (especially visible at high playback speed, when audio needs to be demuxed and decoded further ahead to fill the output buffers after resampling). Some container formats that can be used with the lavf demuxer do not always have reliable timestamps that could be used for unambiguous absolute seeking. However I made the demuxer always enable the new mode because it already converted all seeks to absolute ones before sending them to libavformat, so cases without reliable absolute seeks were failing already and this should only improve the working cases.
2009-03-19 03:25:12 +00:00
return demuxer;
}
static int destroy_avpacket(void *pkt)
{
av_free_packet(pkt);
return 0;
}
static int demux_lavf_fill_buffer(demuxer_t *demux, demux_stream_t *dsds)
{
lavf_priv_t *priv = demux->priv;
demux_packet_t *dp;
demux_stream_t *ds;
int id;
mp_msg(MSGT_DEMUX, MSGL_DBG2, "demux_lavf_fill_buffer()\n");
demux->filepos = stream_tell(demux->stream);
AVPacket *pkt = talloc(NULL, AVPacket);
if (av_read_frame(priv->avfc, pkt) < 0) {
talloc_free(pkt);
return 0;
}
talloc_set_destructor(pkt, destroy_avpacket);
// handle any new streams that might have been added
for (id = priv->nb_streams_last; id < priv->avfc->nb_streams; id++)
handle_stream(demux, priv->avfc, id);
priv->nb_streams_last = priv->avfc->nb_streams;
id = pkt->stream_index;
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
assert(id >= 0 && id < MAX_S_STREAMS);
if (demux->s_streams[id] && demux->sub->id == -1 &&
demux->s_streams[id]->gsh->demuxer_id == priv->autoselect_sub)
{
priv->autoselect_sub = -1;
demux->sub->id = id;
}
if (id == demux->audio->id) {
// audio
ds = demux->audio;
} else if (id == demux->video->id) {
// video
ds = demux->video;
} else if (id == demux->sub->id) {
// subtitle
ds = demux->sub;
} else {
talloc_free(pkt);
return 1;
}
// If the packet has pointers to temporary fields that could be
// overwritten/freed by next av_read_frame(), copy them to persistent
// allocations so we can safely queue the packet for any length of time.
if (av_dup_packet(pkt) < 0)
abort();
dp = new_demux_packet_fromdata(pkt->data, pkt->size);
dp->avpacket = pkt;
AVStream *st = priv->avfc->streams[id];
int64_t ts = priv->use_dts ? pkt->dts : pkt->pts;
if (ts == AV_NOPTS_VALUE && (st->disposition & AV_DISPOSITION_ATTACHED_PIC))
ts = 0;
if (ts != AV_NOPTS_VALUE) {
dp->pts = ts * av_q2d(priv->avfc->streams[id]->time_base);
priv->last_pts = dp->pts * AV_TIME_BASE;
// always set duration for subtitles, even if AV_PKT_FLAG_KEY isn't set,
// otherwise they will stay on screen to long if e.g. ASS is demuxed
// from mkv
if ((ds == demux->sub || (pkt->flags & AV_PKT_FLAG_KEY)) &&
pkt->convergence_duration > 0)
dp->duration = pkt->convergence_duration *
av_q2d(priv->avfc->streams[id]->time_base);
}
dp->pos = demux->filepos;
dp->keyframe = pkt->flags & AV_PKT_FLAG_KEY;
// append packet to DS stream:
ds_add_packet(ds, dp);
return 1;
}
static void demux_seek_lavf(demuxer_t *demuxer, float rel_seek_secs,
float audio_delay, int flags)
{
lavf_priv_t *priv = demuxer->priv;
int avsflags = 0;
mp_msg(MSGT_DEMUX, MSGL_DBG2, "demux_seek_lavf(%p, %f, %f, %d)\n",
demuxer, rel_seek_secs, audio_delay, flags);
if (priv->seek_by_bytes) {
int64_t pos = demuxer->filepos;
rel_seek_secs *= priv->bitrate / 8;
pos += rel_seek_secs;
av_seek_frame(priv->avfc, -1, pos, AVSEEK_FLAG_BYTE);
return;
}
if (flags & SEEK_ABSOLUTE)
priv->last_pts = 0;
else if (rel_seek_secs < 0)
avsflags = AVSEEK_FLAG_BACKWARD;
Add improved relative seek mode When the new mode is active relative seeks are converted to absolute ones (current video pts + relative seek amount) and forward/backward flag before being sent to the demuxer. This mode is used if the demuxer has set the accurate_seek field in the demuxer struct and there is a video stream. At the moment the mkv and lavf demuxers enable the flag. This change is useful for later Matroska ordered chapter support (and for more general timelime editing), but also fixes problems in existing functionality. The main problem with the old mode, where relative seeks are passed directly to the demuxer, is that the user wants to seek relative to the currently displayed position but the demuxer does not know what that position is. There can be an arbitrary amount of buffering between the demuxer read position and what is displayed on the screen. In some situations this makes small seeks fail to move backward at all (especially visible at high playback speed, when audio needs to be demuxed and decoded further ahead to fill the output buffers after resampling). Some container formats that can be used with the lavf demuxer do not always have reliable timestamps that could be used for unambiguous absolute seeking. However I made the demuxer always enable the new mode because it already converted all seeks to absolute ones before sending them to libavformat, so cases without reliable absolute seeks were failing already and this should only improve the working cases.
2009-03-19 03:25:12 +00:00
if (flags & SEEK_FORWARD)
avsflags = 0;
else if (flags & SEEK_BACKWARD)
avsflags = AVSEEK_FLAG_BACKWARD;
if (flags & SEEK_FACTOR) {
if (priv->avfc->duration == 0 || priv->avfc->duration == AV_NOPTS_VALUE)
return;
priv->last_pts += rel_seek_secs * priv->avfc->duration;
} else
priv->last_pts += rel_seek_secs * AV_TIME_BASE;
if (!priv->avfc->iformat->read_seek2) {
// Normal seeking.
av_seek_frame(priv->avfc, -1, priv->last_pts, avsflags);
} else {
// av_seek_frame() won't work. Use "new" seeking API. We don't use this
// API by default, because there are some major issues.
// Set max_ts==ts, so that demuxing starts from an earlier position in
// the worst case.
avformat_seek_file(priv->avfc, -1, INT64_MIN,
priv->last_pts, priv->last_pts, avsflags);
}
}
static int demux_lavf_control(demuxer_t *demuxer, int cmd, void *arg)
{
lavf_priv_t *priv = demuxer->priv;
switch (cmd) {
case DEMUXER_CTRL_CORRECT_PTS:
return DEMUXER_CTRL_OK;
case DEMUXER_CTRL_GET_TIME_LENGTH:
if (priv->seek_by_bytes) {
/* Our bitrate estimate may be better than would be used in
* otherwise similar fallback code at higher level */
if (demuxer->movi_end <= 0)
return DEMUXER_CTRL_DONTKNOW;
*(double *)arg = (demuxer->movi_end - demuxer->movi_start) * 8 /
priv->bitrate;
return DEMUXER_CTRL_GUESS;
}
if (priv->avfc->duration == 0 || priv->avfc->duration == AV_NOPTS_VALUE)
return DEMUXER_CTRL_DONTKNOW;
*((double *)arg) = (double)priv->avfc->duration / AV_TIME_BASE;
return DEMUXER_CTRL_OK;
case DEMUXER_CTRL_GET_PERCENT_POS:
if (priv->seek_by_bytes)
return DEMUXER_CTRL_DONTKNOW; // let it use the fallback code
if (priv->avfc->duration == 0 || priv->avfc->duration == AV_NOPTS_VALUE)
return DEMUXER_CTRL_DONTKNOW;
*((int *)arg) = (int)((priv->last_pts - priv->avfc->start_time) * 100 /
priv->avfc->duration);
return DEMUXER_CTRL_OK;
case DEMUXER_CTRL_SWITCH_AUDIO:
case DEMUXER_CTRL_SWITCH_VIDEO:
{
int id = *((int *)arg);
int newid = -2;
int i, curridx = -1;
int nstreams, *pstreams;
demux_stream_t *ds;
if (cmd == DEMUXER_CTRL_SWITCH_VIDEO) {
ds = demuxer->video;
nstreams = priv->video_streams;
pstreams = priv->vstreams;
} else {
ds = demuxer->audio;
nstreams = priv->audio_streams;
pstreams = priv->astreams;
}
for (i = 0; i < nstreams; i++) {
if (pstreams[i] == ds->id) { //current stream id
curridx = i;
break;
}
}
if (id == -1) { // next track
i = (curridx + 2) % (nstreams + 1) - 1;
if (i >= 0)
newid = pstreams[i];
} else if (id >= 0 && id < nstreams) { // select track by id
i = id;
newid = pstreams[i];
} else // no sound
i = -1;
if (i == curridx) {
*(int *) arg = curridx < 0 ? -2 : curridx;
return DEMUXER_CTRL_OK;
} else {
ds_free_packs(ds);
if (ds->id >= 0)
priv->avfc->streams[ds->id]->discard = AVDISCARD_ALL;
ds->id = newid;
*(int *) arg = i < 0 ? -2 : i;
if (newid >= 0)
priv->avfc->streams[newid]->discard = AVDISCARD_NONE;
return DEMUXER_CTRL_OK;
}
}
core: fix DVD subtitle selection Add all subtitle tracks as reported by libdvdread at playback start. Display language for subtitle and audio tracks. This commit restores these features to the state when demux_mpg was default for DVD playback, and makes them work with demux_lavf and the recent changes to subtitle selection in the frontend. demux_mpg, which was the default demuxer for DVD playback, reordered the subtitle streams according to the "logical" subtitle track number, which conforms to the track layout reported by libdvdread, and is what stream_dvd expects for the STREAM_CTRL_GET_LANG call. demux_lavf, on the other hand, adds the streams in the order it encounters them in the MPEG stream. It seems this order is essentially random, and can't be mapped easily to what stream_dvd expects. Solve this by making demux_lavf hand out the MPEG stream IDs (using the demuxer_id field). The MPEG IDs are mapped by mplayer.c by special casing DVD playback (map_id_from/to_demuxer() functions). This mapping is essentially the same what demux_mpg did. Making demux_lavf reorder the streams is out of the question, because its stream handling is already messy enough. (Note that demux_lavf doesn't export stream IDs for other formats, because most time libavformat demuxers do not set AVStream.id, and we don't know which demuxers do. But we know that MPEG is safe.) Another major complication is that subtitle tracks are added lazily, as soon as the demuxer encounters the first subtitle packet for a given subtitle stream. Add the streams in advance. If a yet non-existent stream is selected, demux_lavf must be made to auto-select that subtitle stream as soon as it is added. Otherwise, the first subtitle packet would be lost. This is done by DEMUXER_CTRL_PRESELECT_SUBTITLE. demux_mpg didn't need this: the frontend code could just set ds->id to the desired stream number. But demux_lavf's stream IDs don't map directly to the stream number as used by libdvdread, which is why this hack is needed.
2012-08-30 14:43:31 +00:00
case DEMUXER_CTRL_AUTOSELECT_SUBTITLE:
{
demuxer->sub->id = -1;
priv->autoselect_sub = *((int *)arg);
return DEMUXER_CTRL_OK;
}
case DEMUXER_CTRL_IDENTIFY_PROGRAM:
{
demux_program_t *prog = arg;
AVProgram *program;
int p, i;
int start;
prog->vid = prog->aid = prog->sid = -2;
if (priv->avfc->nb_programs < 1)
return DEMUXER_CTRL_DONTKNOW;
if (prog->progid == -1) {
p = 0;
while (p < priv->avfc->nb_programs && priv->avfc->programs[p]->id != priv->cur_program)
p++;
p = (p + 1) % priv->avfc->nb_programs;
} else {
for (i = 0; i < priv->avfc->nb_programs; i++)
if (priv->avfc->programs[i]->id == prog->progid)
break;
if (i == priv->avfc->nb_programs)
return DEMUXER_CTRL_DONTKNOW;
p = i;
}
start = p;
redo:
program = priv->avfc->programs[p];
for (i = 0; i < program->nb_stream_indexes; i++) {
switch (priv->avfc->streams[program->stream_index[i]]->codec->codec_type) {
case AVMEDIA_TYPE_VIDEO:
if (prog->vid == -2)
prog->vid = program->stream_index[i];
break;
case AVMEDIA_TYPE_AUDIO:
if (prog->aid == -2)
prog->aid = program->stream_index[i];
break;
case AVMEDIA_TYPE_SUBTITLE:
if (prog->sid == -2)
prog->sid = program->stream_index[i];
break;
}
}
if (prog->aid >= 0 && prog->aid < MAX_A_STREAMS &&
demuxer->a_streams[prog->aid]) {
sh_audio_t *sh = demuxer->a_streams[prog->aid];
prog->aid = sh->aid;
} else
prog->aid = -2;
if (prog->vid >= 0 && prog->vid < MAX_V_STREAMS &&
demuxer->v_streams[prog->vid]) {
sh_video_t *sh = demuxer->v_streams[prog->vid];
prog->vid = sh->vid;
} else
prog->vid = -2;
if (prog->progid == -1 && prog->vid == -2 && prog->aid == -2) {
p = (p + 1) % priv->avfc->nb_programs;
if (p == start)
return DEMUXER_CTRL_DONTKNOW;
goto redo;
}
priv->cur_program = prog->progid = program->id;
return DEMUXER_CTRL_OK;
}
case DEMUXER_CTRL_RESYNC:
/* NOTE:
*
* We actually want to call ff_read_frame_flush() here, but it is
* internal.
*
* This function call seems to do the same for now.
*
* Once ff_read_frame_flush() is exported in some way, change this to
* call the new API instead of relying on av_seek_frame() to do this
* for us.
*/
avio_flush(priv->avfc->pb);
av_seek_frame(priv->avfc, 0, avio_tell(priv->avfc->pb),
AVSEEK_FLAG_BYTE);
avio_flush(priv->avfc->pb);
return DEMUXER_CTRL_OK;
default:
return DEMUXER_CTRL_NOTIMPL;
}
}
static void demux_close_lavf(demuxer_t *demuxer)
{
lavf_priv_t *priv = demuxer->priv;
if (priv) {
if (priv->avfc) {
av_freep(&priv->avfc->key);
avformat_close_input(&priv->avfc);
}
av_freep(&priv->pb);
demux_lavf: add support for libavdevice libavdevice supports various "special" video and audio inputs, such as screen-capture or libavfilter filter graphs. libavdevice inputs are implemented as demuxers. They don't use the custom stream callbacks (in AVFormatContext.pb). Instead, input parameters are passed as filename. This means the mpv stream layer has to be disabled. Do this by adding the pseudo stream handler avdevice://, whose only purpose is passing the filename to demux_lavf, without actually doing anything. Change the logic how the filename is passed to libavformat. Remove handling of the filename from demux_open_lavf() and move it to lavf_check_file(). (This also fixes a possible bug when skipping the "lavf://" prefix.) libavdevice now can be invoked by specifying demuxer and args as in: mpv avdevice://demuxer:args The args are passed as filename to libavformat. When using libavdevice demuxers, their actual meaning is highly implementation specific. They don't refer to actual filenames. Note: libavdevice is disabled by default. There is one problem: libavdevice pulls in libavfilter, which in turn causes symbol clashes with mpv internals. The problem is that libavfilter includes a mplayer filter bridge, which is used to interface with a set of nearly unmodified mplayer filters copied into libavfilter. This filter bridge uses the same symbol names as mplayer/mpv's filter chain, which results in symbol clashes at link-time. This can be prevented by building ffmpeg with --disable-filter=mp, but unfortunately this is not the default. This means linking to libavdevice (which in turn forces linking with libavfilter by default) must be disabled. We try doing this by compiling a test file that defines one of the clashing symbols (vf_mpi_clear). To enable libavdevice input, ffmpeg should be built with the options: --disable-filter=mp and mpv with: --enable-libavdevice Originally, I tried to auto-detect it. But the resulting complications in configure did't seem worth the trouble.
2012-11-30 17:41:04 +00:00
talloc_free(priv);
demuxer->priv = NULL;
}
}
const demuxer_desc_t demuxer_desc_lavf = {
"libavformat demuxer",
"lavf",
"libavformat",
"Michael Niedermayer",
"supports many formats, requires libavformat",
DEMUXER_TYPE_LAVF,
1,
lavf_check_file,
demux_lavf_fill_buffer,
demux_open_lavf,
demux_close_lavf,
demux_seek_lavf,
demux_lavf_control
};