This was added in 380db56928 as a
temporary crutch that is not needed anymore. The only case where this
code can be triggered is the very first frame, for which InputStream.pts
is always equal to 0.
Stop using InputStream.dts for generating missing timestamps for decoded
frames, because it contains pre-decoding timestamps and there may be
arbitrary amount of delay between input packets and output frames (e.g.
dependent on the thread count when frame threading is used). It is also
in AV_TIME_BASE (i.e. microseconds), which may introduce unnecessary
rounding issues.
New code maintains a timebase that is the inverse of the LCM of all the
samplerates seen so far, and thus can accurately represent every audio
sample. This timebase is used to generate missing timestamps after
decoding.
Changes the result of the following FATE tests
* pcm_dvd-16-5.1-96000
* lavf-smjpeg
* adpcm-ima-smjpeg
In all of these the timestamps now better correspond to actual frame
durations.
If input packets have timestamps, they will be propagated to output
frames by the decoder, so at best this block does not do anything.
There can also be an arbitrary amount of delay between packets sent to
the decoder and decoded frames (e.g. due to decoder's intrinsic delay or
frame threading), so deriving any timestamps from packet properties is
wrong.
One that is fine enough to represent all DV audio sample rates. Audio
packet durations are now sample-accurate.
This largely undoes commit 76fbb0052d. To
avoid breaking the issue fixed by that commit, resync audio timestamps
against video if they get more than one frame apart. The sample from
issue #8762 still works correctly after this commit.
Slightly changes the results of the lavf-dv seektest, due to the audio
timebase being more granular.
Current code will call avpriv_set_pts_info() for each video frame,
possibly setting a different timebase if the stream framerate changes.
This violates API conventions, as the timebase is supposed to stay
constant after stream creation.
Change the demuxer to set a single timebase that is fine enough to
handle all supported DV framerates.
The seek tests change slightly because the new timebase is more
granular.
Fixes: signed integer overflow: -2147483648 - 5 cannot be represented in type 'int'
Fixes: 58066/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HEVC_fuzzer-5312995835379712
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This makes the worst case much faster
Fixes: Timeout
Fixes: 51363/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_BONK_fuzzer-5660734784143360
Fixes: 57957/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_BONK_fuzzer-5874095467397120
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This allows the usage of codecs in builds that have a parser but no decoders
for remuxing scenarios with raw sources.
Signed-off-by: James Almer <jamrial@gmail.com>
Also remove the _minus8 part of the name to be in line with the rest of the
decoder, and fix the storage type for pps_palette_predictor_initializer,
to support hbd values.
Signed-off-by: James Almer <jamrial@gmail.com>
Anton Khirnov
Paul B Mahol
James Almer
Michael Niedermayer