This is currently set from `Mp3Extractor.synchronizedHeader` which
gets overwritten every time we read a new frame. It seems safer to make
this defensive copy (and there will be at most one `XingFrame` instance
per-playback, so this is not prohibitively expensive).
PiperOrigin-RevId: 636181038
This change avoids a muxer deadlock when:
1. Sequence of items
2. First item has audio track that is shorter than video
3. Audio finishes, and muxer refuses to write more than 500ms of video
consecutively.
SequenceAssetLoader fails to progress to the second item. A muxer
deadlock is possible when the audio of the first item finishes,
audio end-of-stream is not propagated through AudioGraph, and muxer blocks
video, preventing SequenceAssetLoader to move to the next item in sequence.
By triggering silence generation early as soon as audio EOS is
encountered, we ensure SequenceAssetLoader can progress to the next item.
PiperOrigin-RevId: 636179966
`Info` header is used for CBR files, but in some cases not **every**
frame in these files is the same size. This change stops using the
single frame after the `Info` frame as the 'template' (and assuming all
subsequent frames are the same size/bitrate), and instead derives the
bitrate from fields in the `Info` header. This works for files which are
'almost' constant bitrate, like the one in Issue: androidx/media#1376 where every
frame is either 1044 or 1045 bytes except the one immediately after the
`Info` frame which is 104 bytes (32kbps), resulting in a wildly
incorrect duration calculation.
PiperOrigin-RevId: 636151605
This was previously only set on images because it was not ignored on
other media types. This parameter was made no-op for non-images in
7b2a1b4443.
PiperOrigin-RevId: 636078142
This change was originally made in 379cb3ba54.
It was then accidentally lost in when `Cea608Parser` was merged back
into `Cea608Decoder` in 25498b151b.
This was spotted when re-doing a similar lost change to `Cea708Decoder`,
reported in https://github.com/androidx/media/pull/1315.
See reasoning on e2847b3b80
about why this is the only 'lost' CEA-608 change.
PiperOrigin-RevId: 635803536
This change was originally made in 6f8249184b
It was then accidentally lost in when `Cea708Parser` was merged back
into `Cea708Decoder` in 51b4fa2cc8.
This is the only change made to the actual 'decoding' logic in
`Cea708Parser` between it being split from `Cea708Decoder` and merged
back in again, all the other changes in this period relate to the
implementation of the `SubtitleParser` interface, so don't need to be
preserved in `Cea708Decoder`:
51b4fa2cc8/libraries/extractor/src/main/java/androidx/media3/extractor/text/cea/Cea708Parser.java
`Cea608Parser` was also merged back into `Cea608Decoder` in
25498b151b
and so is vulnerable to the same risk of accidental loss of changes. To
be sure, I also checked the history of this file:
25498b151b/libraries/extractor/src/main/java/androidx/media3/extractor/text/cea/Cea608Parser.java
The only 'decoding logic' change there is 379cb3ba54,
which was also lost in 25498b151b.
I will send a separate change to resolve this.
PiperOrigin-RevId: 635796696
This shows ExoPlayer currently wrongly reports the duration of this
sample, because it assumes every frame is 32kbps (104 bytes) due to the
`PCUT` frame immediately after the `Info` frame.
A follow-up change will modify `Info` frame handling to resolve this
issue.
This sample was crafted using a hex editor to insert the additional
`PCUT` frame (the pattern of `null` and `x` is taken from the sample
file in Issue: androidx/media#1376, the header is modified to set the channel count
to 1 to match the rest of the file), and then update the frame count
and data size of the `Info` header to match.
Issue: androidx/media#1376
PiperOrigin-RevId: 635772837
Fragment shaders in OpenGL ES shader language aren't guaranteed
to support highp, required to correctly represent pixel coordinates
inside large images (e.g. 1920x1080).
This change moves coordinate mirroring for images out of fragment shader.
Fixes http://Issue: androidx/media#1331
PiperOrigin-RevId: 635732208
If MediaCodec allocates passes an image buffer with a cropped region,
SurfaceTexture.getTransformMatrix will cut off 2 pixels from each dimensions.
The resulting videos will appear a little stretched.
This patch inspects the SurfaceTexture transform matrix, and guesses what the
unscaled transform matrix should be.
Behind experimentalAdjustSurfaceTextureTransformationMatrix flag
PiperOrigin-RevId: 635721267
Gainmaps don't currently have an equals override, only reference equality is checked by Objects.equals(gainmap1, gainmap2). Create an equals method for gainmaps with the fields we care about to ensure we don't incur false positives in our equality checks.
PiperOrigin-RevId: 635510451
This is a fix for the fix in 319854d624. The original fix did
not reset the firstPeriodId to avoid any id clashes with future
updates. This however only works under the assumption that the
next manifest load at the next call to prepare() is exactly the
same as the current manifest. This is not true unless the call
happens very quickly (and may fail even then). Instead we should
keep the existing manifest directly as a reference so we can use
it to find the number of removed periods when we get a new manifest
at the next call to prepare().
Issue: androidx/media#1329
PiperOrigin-RevId: 634853524
This should be no-op overall and helps to disentangle the clock sync
update from the state of the manifest.
We currently check oldPeriodCount==0 to trigger the clock sync load,
which only works because the manifest happens to be null whenever
we need a new clock sync. We can decouple these concepts by directly
checking whether we have an existing elapsedRealtimeOffsetMs.
This also requires to set this value explicitly at the point where we
consider it set to the local device clock fallback when the timing
element load fails.
PiperOrigin-RevId: 634844921
This case is most likely to happen when re-preparing a multi-period
live stream after an error. The live timeline can easily move on to
new periods in the meantime, creating this type of update.
The behavior before this change has two bugs:
- The player resolves the new start position to a subsequent period
that existed in the old timeline, or ends playback if that cannot
be found. The more useful behavior is to restart playback in the
same live item if it still exists.
- MaskingMediaSource creates a pending MaskingMediaPeriod using the
old timeline and then attempts to create the real period from the
updated source. This fails because MediaSource.createPeriod is
called with a periodUid that does no longer exist at this point.
We already have logic to not override the start position and need
to extend this to also not prepare the real source.
Issue: androidx/media#1329
PiperOrigin-RevId: 634833030
For each item, AudioGraphInput now pads the input audio track with silence
to the duration given in onMediaItemChanged.
Possibly resolves Issue: androidx/media#921 .
PiperOrigin-RevId: 634753721
The two affected tests where playing until a specific
position to enable the player to read ahead. The method
pauses at exactly the target position, but then has
temporarily undetermined behavior because the playback
thread uses player.getClock().onThreadBlocked() that lets
the playback thread make progress in parallel to the test
thread. The tests were flaky because they sometimes made
so much progress that they ended playback before we could
query the updated renderer state.
This can be fixed by using
run(player).untilBackgroundThreadCondition instead, which
is guaranteed to be fully deterministic, but may not be able
to stop at exactly the desired position (which we don't
really need anyway for this test).
PiperOrigin-RevId: 634699752
The method is only allowed to be called on prepared items.
This check was currently missing and also causing the
corresponding test to be flaky in ExoPlayerTest.
PiperOrigin-RevId: 634694077
Before this change, if a playback error is thrown the test fails with a
timeout and no additional info:
```
java.util.concurrent.TimeoutException
at androidx.media3.exoplayer.e2etest.WebvttPlaybackTest.stallPlayerUntilCondition(WebvttPlaybackTest.java:361)
```
After this change, the test failure includes a much more useful stack
trace, e.g. from 0352db9a37:
```
Caused by: java.lang.IllegalStateException: Legacy decoding is disabled, can't handle text/vtt samples (expected application/x-media3-cues).
at androidx.media3.common.util.Assertions.checkState(Assertions.java:100)
at androidx.media3.exoplayer.text.TextRenderer.assertLegacyDecodingEnabledIfRequired(TextRenderer.java:587)
at androidx.media3.exoplayer.text.TextRenderer.onStreamChanged(TextRenderer.java:210)
```
PiperOrigin-RevId: 634672138
ExoPlayer sometimes calls AudioSink.playToEndOfStream before configuring
the sink. Before this CL, the composition player was failing if this
happened.
PiperOrigin-RevId: 634306592
Fix "subtitles rendered with border and no fill color" problem:
https://github.com/moneytoo/Player/issues/413
because '0' means it's an index to the palette table, not an RGBA value of 0
To override this change, and go back to parsing during rendering,
apps must make two method calls:
1. `MediaSource.Factory.experimentalParseSubtitlesDuringExtraction(false)`
2. `TextRenderer.experimentalSetLegacyDecodingEnabled(true)`
PiperOrigin-RevId: 634262798
