The previous code assumed that the `VBRI` Table of Contents (ToC)
covers all the MP3 data in the file. In a file with an invalid VBRI ToC
where this isn't the case, this results in playback silently stopping
mid-playback (and either advancing to the next item, or continuing to
count up the playback clock forever). This change considers the `bytes`
field to determine the end of the MP3 data, in addition to deriving it
from the ToC. If they disagree we log a warning and take the max value.
This is because we handle accidentally reading non-MP3 data at the end
(or hitting EoF) better than stopping reading valid MP3 data partway
through.
Issue: androidx/media#1904
#cherrypick
PiperOrigin-RevId: 700319250
This was hand-crafted with a 4-entry ToC by modifying
`bear-vbr-xing-header.mp3` in a hex editor.
The output difference from 117 samples to 116 samples is due to the
calculation in `VbriSeeker` assuming that the ToC includes the VBRI
frame itself, which I don't think is correct (fix is in a follow-up
change).
Issue: androidx/media#1904
#cherrypick
PiperOrigin-RevId: 700254516
The behaviour was changed in 1.4.0 with 0f42dd4752,
so that the buffer timestamp is compared to `outputStartTimeUs` when
deciding whether to discard a "decode only" buffer before decoding
(instead of the deprecated/removed `isDecodeOnly` property). This breaks
when the buffer timestamp is `TIME_END_OF_SOURCE` (which is
`Long.MIN_VALUE`), because `TIME_END_OF_SOURCE < outputStartTimeUs` is
always true, so the end-of-stream buffer is never passed to the decoder
and on to `TextRenderer` where it is used to
[set `inputStreamEnded = true`](40f187e4b4/libraries/exoplayer/src/main/java/androidx/media3/exoplayer/text/TextRenderer.java (L434-L436))
and so playback hangs.
#cherrypick
Issue: androidx/media#1863
PiperOrigin-RevId: 695767247
This allows users to extend and customize specific methods of the `TrackOutput` implementation while inheriting default behaviors for others.
PiperOrigin-RevId: 686454360
- Added logic to parse media duration from the `mdhd` box for accurate frame rate calculation.
- Fallbacks to track duration from `tkhd` when `mdhd` contains invalid or missing data.
- Avoids incorrect frame rate calculations in MP4 files with an edit list (`elst`) box.
- Adds frame rate calculations for partially fragmented MP4 files.
- Verified accuracy with tools like `mediainfo` and `ffprobe`.
Issue: androidx/media#1531
**Note**: The slight difference in frame rate values in dump files that aren’t MP4s with an edit list or fragmented MP4s isn’t due to differences in `tkhd` and `mdhd` duration values (which should be identical for non-edited or non-fragmented files). Rather, it’s because they are calculated using different timescales. The `mvhd` box defines a global movie timescale, which is used for the track's `tkhd` duration. Meanwhile, each track’s `mdhd` box defines its own timescale specific to its content type, which we now use for more accurate frame rate calculation.
PiperOrigin-RevId: 676046744
The `bear-cbr-no-seek-table-trailing-garbage.mp3` test file was generated by appending 150kB of `0xDEADBEEF` onto the end of `bear-cbr-variable-frame-size-no-seek-table.mp3`.
Issue: androidx/media#1563
#cherrypick
PiperOrigin-RevId: 670131828
This allows seek operations in files that were previously unseekable, particularly those with variable bitrate (VBR) or constant bitrate (CBR) with silence frames.
New samples for tests were created by adding silence frames to existing narrowband and wideband versions.
PiperOrigin-RevId: 665349241
This makes the class available to custom MP4-parsing implementations,
while also allowing it to be used by `muxer` in future.
'Box' is the term used throughout the ISO 14496-12 spec, while the
'Atom' nomenclature was used in an earlier form of the spec
(Quicktime).
This change moves it from `extractor.mp4.Atom` to `container.Mp4Box`,
to be consistent with existing MP4-specific types in the `container`
module like `Mp4TimestampData`.
PiperOrigin-RevId: 663274752
AMR samples with identical data but different names, previously used to generate uniquely named dump files, have been deleted. Instead, `AssertionConfig` is now used to set the dump file prefix, ensuring files are generated with unique names.
PiperOrigin-RevId: 662883541
This is groundwork to moving `Atom` to the `container` library, which
we want to do before making it public (so it can be used by `muxer` in
future).
PiperOrigin-RevId: 662453520
If the length of the `ExtractorInput` is not known then the
`subtitleData` field is re-sized by 1kB each time
(`SubtitleExtractor.DEFAULT_BUFFER_SIZE`), so the end of the array is
often not populated. This change ensures that `length` is propagated to
`SubtitleParser`, so that implementations don't try and parse the
garbage/zero bytes at the end of the array.
Discovered while investigating Issue: androidx/media#1516
#cherrypick
PiperOrigin-RevId: 661195634
This is caused when the requested "output start time" is equal to or
larger than the last event time in a `Subtitle` object.
This resolves the error in Issue: androidx/media#1516, but subtitles are still not
renderered (probably because the timestamps aren't what we expect
somewhere, but I need to investigate this part further).
#cherrypick
PiperOrigin-RevId: 660462720
This is needed to correctly handle files with trailing non-MP3 data
(which is indicated by the length in the `Info` frame being shorter than
the overall length of the file).
The test file was generated by appending 150kB of `DEADBEEF` onto the
end of `test-cbr-info-header.mp3`, and the test asserts that the
extracted samples are identical.
Issue: androidx/media#1480
#cherrypick
PiperOrigin-RevId: 658727595
The integration with `SubtitleTranscodingExtractorOutput` has been
moved inside the relevant `Extractor` implementations instead.
PiperOrigin-RevId: 651213564
Stereo view information is stored in the 3D reference displays information SEI and the optional vexu box. Parsing of the SEI and vexu box is added, and based on the parsed info, proper mapping of primary/secondary view to left/right eye is determined.
PiperOrigin-RevId: 651002190
Changes to FragmentedMp4Extractor to parse additional sample dependency
information and mark output samples as "no other samples depend on this".
Only applies to H.264 tracks.
Controlled by new fMP4 flag: FLAG_READ_WITHIN_GOP_SAMPLE_DEPENDENCIES
PiperOrigin-RevId: 650538377
Changes to Mp4Extractor to parse additional sample dependency information
and mark output samples as "no other sample depend on this".
Only applies to H.264 tracks.
Controlled by new mp4 flag: FLAG_READ_WITHIN_GOP_SAMPLE_DEPENDENCIES
PiperOrigin-RevId: 649640184
The percentage should be interpreted as relative to the size of a parent
node.
This change makes this inheritance work correctly for percentages in
both the parent and child. It does not fix the case of a non-percentage
parent size with a percentage child size.
PiperOrigin-RevId: 649631055
Parse LHEVCDecoderConfigurationRecord with the ‘lhvC’ type and set the corresponding sample mime type to video/mv-hevc. With no MV-HEVC decoder available, fallback to single-layer HEVC decoding.
PiperOrigin-RevId: 649119173
This is required before we can move CEA-6/708 parsing from the rendering
side of the sample queue to the extraction side.
This re-ordering is needed for video encodings with different decoder
and presentation orders, because the CEA-6/708 data is attached to each
frame and needs to be processed in presentation order instead of decode
order. This change re-orders frames within a group-of-pictures, but also
takes advantage of `maxNumReorderFrames/Pics` values to cap the size of
the re-ordering queue, allowing caption data to be released 'earlier'
than the end of a GoP.
Annex D of the CEA-708 spec (which also applies for CEA-608 embedded in
SEI messages), makes the need to re-order from decode to presentation
order clear.
PiperOrigin-RevId: 648648002
This value is used in a follow-up change to re-order SEI messages
containing CEA-6/708 data from decode order to presentation order.
PiperOrigin-RevId: 643296338
