* commit 'db7968bff4851c2be79b15b2cb2ae747424d2fca':
avio: Allow custom IO users to get labels for the output bytestream
Merged-by: Matthieu Bouron <matthieu.bouron@stupeflix.com>
Docs clearly states that av_write_trailer should only be called if
avformat_write_header was successful, therefore we have to deinit if we return
failure.
Signed-off-by: Marton Balint <cus@passwd.hu>
As long as caller only writes packets using av_interleaved_write_frame
with no manual flushing, this should allow us to always have accurate
durations at the end of fragments, since there should be at least
one queued packet in each stream (except for the stream where the
current packet is being written, but if the muxer itself does the
cutting of fragments, it also has info about the next packet for that
stream).
Signed-off-by: Martin Storsjö <martin@martin.st>
This allows callers with avio write callbacks to get the bytestream
positions that correspond to keyframes, suitable for live streaming.
In the simplest form, a caller could expect that a header is written
to the bytestream during the avformat_write_header, and the data
output to the avio context during e.g. av_write_frame corresponds
exactly to the current packet passed in.
When combined with av_interleaved_write_frame, and with muxers that
do buffering (most muxers that do some sort of fragmenting or
clustering), the mapping from input data to bytestream positions
is nontrivial.
This allows callers to get directly information about what part
of the bytestream is what, without having to resort to assumptions
about the muxer behaviour.
One keyframe/fragment/block can still be split into multiple (if
they are larger than the aviocontext buffer), which would call
the callback with e.g. AVIO_DATA_MARKER_SYNC_POINT, followed by
AVIO_DATA_MARKER_UNKNOWN for the second time it is called with
the following data.
Signed-off-by: Martin Storsjö <martin@martin.st>
Currently, AVStream contains an embedded AVCodecContext instance, which
is used by demuxers to export stream parameters to the caller and by
muxers to receive stream parameters from the caller. It is also used
internally as the codec context that is passed to parsers.
In addition, it is also widely used by the callers as the decoding (when
demuxer) or encoding (when muxing) context, though this has been
officially discouraged since Libav 11.
There are multiple important problems with this approach:
- the fields in AVCodecContext are in general one of
* stream parameters
* codec options
* codec state
However, it's not clear which ones are which. It is consequently
unclear which fields are a demuxer allowed to set or a muxer allowed to
read. This leads to erratic behaviour depending on whether decoding or
encoding is being performed or not (and whether it uses the AVStream
embedded codec context).
- various synchronization issues arising from the fact that the same
context is used by several different APIs (muxers/demuxers,
parsers, bitstream filters and encoders/decoders) simultaneously, with
there being no clear rules for who can modify what and the different
processes being typically delayed with respect to each other.
- avformat_find_stream_info() making it necessary to support opening
and closing a single codec context multiple times, thus
complicating the semantics of freeing various allocated objects in the
codec context.
Those problems are resolved by replacing the AVStream embedded codec
context with a newly added AVCodecParameters instance, which stores only
the stream parameters exported by the demuxers or read by the muxers.
commit "avpacket: Deprecate av_dup_packet" broke the use
av_interleaved_write_uncoded_frame as any input uncoded frame has an
invalid packet size that will crash when av_packet_ref tries to allocate
'size' new memory. Since the packet is a temporary created within mux.c
itself it can be used directly without needing a new ref.
Signed-off-by: Matt Oliver <protogonoi@gmail.com>
This solves the problem discussed in https://ffmpeg.org/pipermail/ffmpeg-devel/2015-September/179238.html
by allowing AVCodec::write_header to be delayed until after packets have been
run through required bitstream filters in order to generate global extradata.
It also provides a mechanism by which a muxer can add a bitstream filter to a
stream automatically, rather than prompting the user to do so.
All encoders set pts and dts properly now (and have been doing that for
a while), so there is no good reason to do any timestamp guessing in the
muxer.
The newly added AVStreamInternal will be later used for storing all the
private fields currently living in AVStream.
It is well known that fabs and fabsf are at least as fast and sometimes
faster than the FFABS macro, at least on the gcc+glibc combination.
For instance, see the reference:
http://patchwork.sourceware.org/patch/6735/.
This was a patch to glibc in order to remove their usages of a macro.
The reason essentially boils down to fabs using the __builtin_fabs of
the compiler, while FFABS needs to infer to not use a branch and to
simply change the sign bit. Usually the inference works, but sometimes
it does not. This may be easily checked by looking at the asm.
This also has the added benefit of reducing macro usage, which has
problems with side-effects.
Note that avcodec is not handled here, as it is huge and
most things there are integer arithmetic anyway.
Tested with FATE.
Reviewed-by: Clément Bœsch <u@pkh.me>
Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
* commit '948f3c19a8bd069768ca411212aaf8c1ed96b10d':
lavc: Make AVPacket.duration int64, and deprecate convergence_duration
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
* commit '01bcc2d5c23fa757d163530abb396fd02f1be7c8':
lavc: Drop deprecated destruct_packet related functions
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
Move field to internal part of AVStream and struct to internal.h
Reviewed-by: Andreas Cadhalpun <andreas.cadhalpun@googlemail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes Ticket3514
See: ETSI EN 300 743 V1.3.1 (2006-11)
"In summary, all of the segments of a single display set shall be carried in one (or more) PES packets that have the same
PTS value."
with PTS = DTS and remuxing of such a stream it is to be expected that sometimes
multiple packets would have the same DTS
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit '9257692ac15eff7b07540c1f61cebde0d8823fbd':
lavf: Only initialize s->offset once when using avoid_negative_ts make_zero
Conflicts:
libavformat/mux.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
When given a stream starting at dts=0, it would previously consider
s->offset as uninitialized and set an offset when the second packet
was written, ending up writing two packets with dts=0. By initializing
this field to AV_NOPTS_VALUE, we make sure that we only initialize it
once, on the first packet.
Signed-off-by: Martin Storsjö <martin@martin.st>
* commit '897d5c3a4296f3da80b8699d1487328ca2de8e55':
lavf: Print a warning if failed to avoid negative timestamps when requested
Conflicts:
libavformat/mux.c
See: ec6a5fc6cc
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This is the same logic as is invoked on AVFMT_TS_NEGATIVE,
but which can be enabled manually, or can be enabled
in muxers which only need it in certain conditions.
Also allow using the same mechanism to force streams to start
at 0.
Signed-off-by: Martin Storsjö <martin@martin.st>
* commit '91e8d2eb1f7bf3af949008b106ec1ca037b88b0e':
lavf: use the format context strict_std_compliance instead of the codec one
Conflicts:
libavformat/mux.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '2d6e58497e76836604364b037df9b00ba3d75b69':
lavf: switch to AVCodecContext.framerate for demuxing
Conflicts:
libavformat/utils.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
It is basically a wrapper around av_get_audio_frame_duration(), with a
fallback to AVCodecContext.frame_size. However, that field is set only
when the stream codec context is actually used for encoding or decoding,
which is discouraged.
For muxing, it is generally the responsibility of the caller to set the
packet duration.
For demuxing, if the duration is not stored at the container level, it
should be set by the parser.
Therefore, removing the frame_size fallback should not break any
important case.
(cherry picked from commit 30e50c5027)
Conflicts:
libavformat/utils.c
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This makes problematic unconditional flushes in mpegts redundant
And is thus part of a fix for ticket 2748
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
libvpx adds very significant delay, which appears normal and we must
buffer all other streams no matter what to interleave them correctly
Fixes Ticket3440
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit '194be1f43ea391eb986732707435176e579265aa':
lavf: switch to AVStream.time_base as the hint for the muxer timebase
Conflicts:
doc/APIchanges
libavformat/filmstripenc.c
libavformat/movenc.c
libavformat/mxfenc.c
libavformat/oggenc.c
libavformat/swf.h
libavformat/version.h
tests/ref/lavf/mkv
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Previously, AVStream.codec.time_base was used for that purpose, which
was quite confusing for the callers. This change also opens the path for
removing AVStream.codec.
The change in the lavf-mkv test is due to the native timebase (1/1000)
being used instead of the default one (1/90000), so the packets are now
sent to the crc muxer in the same order in which they are demuxed
(previously some of them got reordered because of inexact timestamp
conversion).
* commit 'a312f71090ee620ee252f2034aef6b13e2dafe9c':
lavf: deprecate now unused AVStream.pts
Conflicts:
libavformat/mux.c
libavformat/version.h
mostly not merged as the code is needed for a/vsync drop handling
and what the code does is what is needed, it could maybe be moved
elsewhere or factored somehow but simply removing it would be droping
these features.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'c9281a01b78cc3f09e36300a0ca3f5824d1c74cf':
lavf: drop the zero-sized packets hack
Conflicts:
libavformat/mux.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Use it instead of checking CODEC_FLAG_BITEXACT in the first stream's
codec context.
Using codec options inside lavf is fragile and can easily break when the
muxing codec context is not the encoding context.
This is continuation of commit 330d547e
Nested struct is set in two places.
Previous commit set nested struct only in one case.
Signed-off-by: Lukasz Marek <lukasz.m.luki2@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This commit allows to benefit from implementing child_next
callback for muxers' AVClasses.
Without that, options cannot be set in nested structs.
Signed-off-by: Lukasz Marek <lukasz.m.luki2@gmail.com>
* commit '394fb56c29eee7f4f8f0334d8b5d30d3c54ac703':
lavf: always unref the packet passed to av_interleaved_write_frame() on error
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Currently ff_interleave_packet_per_dts() waits until it gets a frame for
each stream before outputting packets in interleaved order.
Sparse streams (i.e. streams with much fewer packets than the other
streams, like subtitles or audio with DTX) tend to add up latency and in
specific cases end up allocating a large amount of memory.
Emit the top packet from the packet_buffer if it has a time delta
larger than a specified threshold.
Original report of the issue and initial proposed solution by
mus.svz@gmail.com.
Bug-id: 31
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Since we don't write lavf's string when bitexact is requested, this will
prevent the tag from being copied from the source stream.
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit '596e5d4783ca951258a7c580951fd161f1785ec1':
lavf: Add a flag to enable/disable per-packet flushing
Conflicts:
libavformat/avformat.h
libavformat/mux.c
libavformat/version.h
This adds a 2nd API to set per packet flushing
If the user application indicates through either a non default then this non default takes
precedence over the other still default value
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This is enabled by default and can be disabled with
"-fflags -flush_packets".
Inspired by a patch from Nicolas George <nicolas.george@normalesup.org>.
Signed-off-by: Martin Storsjö <martin@martin.st>
* commit '7950e519bb094897f957b9a9531cc60ba46cbc91':
Disable deprecation warnings for cases where a replacement is available
Conflicts:
libavcodec/avpacket.c
libavcodec/pthread.c
libavcodec/utils.c
libavdevice/v4l2.c
libavfilter/avfiltergraph.c
libavfilter/buffersrc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
lavf: don't abort if both encoder and muxer aspect ratios are not set
Conflicts:
libavformat/mux.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This patch adds support for encoding VP8 files with alpha. The alpha channel
is encoded separately and the output is placed in AVPacket's side_data. The
muxer then muxes it into the BlockAdditional element of the matroska container.
More details on spec here: http://goo.gl/wCP1y
Signed-off-by: Vignesh Venkatasubramanian <vigneshv@google.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'b3ea76624ad1baab0b6bcc13f3f856be2f958110':
vf_aspect: use the name 's' for the pointer to the private context
Remove commented-out debug #define cruft
Conflicts:
libavcodec/4xm.c
libavcodec/dvdsubdec.c
libavcodec/ituh263dec.c
libavcodec/mpeg12.c
libavfilter/avfilter.c
libavfilter/vf_aspect.c
libavfilter/vf_fieldorder.c
libavformat/rtmpproto.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'e926b5ceb1962833f0c884a328382bc2eca67aff':
avformat: Drop unnecessary ff_ name prefixes from static functions
Conflicts:
libavformat/audiointerleave.c
libavformat/mux.c
libavformat/mxfenc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'c2cb01d418dd18e1cf997c038d37378d773121be':
lavf: introduce AVFMT_TS_NEGATIVE
Conflicts:
libavformat/avformat.h
libavformat/mux.c
tests/ref/lavf/asf
tests/ref/lavf/mkv
tests/ref/lavf/mpg
tests/ref/lavf/ts
tests/ref/seek/lavf-asf
tests/ref/seek/lavf-mkv
tests/ref/seek/lavf-mpg
tests/ref/seek/lavf-ts
This commit does not change the default ts behaviour, such
change will, if its done, be done separately.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Most formats do not support negative timestamps, shift them to avoid
unexpected behaviour and a number of bad crashes.
CC:libav-stable@libav.org
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
The code did not account properly for packets that where added to
the end of the packet list. Also flags for such packets where not
set correctly leading to incorrect chunked interleaving.
Reported-by: bcoudurier
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
a small value was rounded to 0 and then treated special as if
chunked_duration was 0. This led to a inconsistency that further led
to wrong interleaving
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Sometimes the muxer modifies the packet, like for instance lavf/mp3enc
changing pkt->destruct in order to keep a copy. These changes must be
kept, even though the muxer behaviour is questionable. Regression since
0072116.
Fixes#2124.
After demuxing, data and side are merged. Before decoding, they are
split. Encoder will perform with data and side split. This means that a
muxer can receive split data (after encoding) but also merged data (if
called directly after demuxing). This commit makes sure data and side
are split for the muxer.
otherwise a unexpected timebase could be choosen
that is one that is thousand times more precisse than requested
which can have sideeffects.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
