to
LAST_SAMPLE_DURATION_BEHAVIOR_SET_FROM_END_OF_STREAM_BUFFER_OR_DUPLICATE_PREVIOUS
This CL also combines LAST_SAMPLE_DURATION_BEHAVIOR_SET_FROM_END_OF_STREAM_BUFFER
and LAST_SAMPLE_DURATION_BEHAVIOR_DUPLICATE_PREVIOUS.
The reason for combining the two enums is that, when the option
to use END_OF_STREAM_BUFFER is selected and if the EOS buffer is
not provided then the muxer anyways fallbacks to duplicate
duration behavior.
The last sample with 0 durations seems less useful so
change the default behavior to non-zero duration.
This will also match the behavior with MediaMuxer.
PiperOrigin-RevId: 675189932
To support for 3gpp h263 codec in Mp4Muxer currently profile and level is hardcoded and provided to h263 box. Parse profile and level from MediaFormat and use those value to write h263 box.
PiperOrigin-RevId: 675004590
The tests were flaky because CompositionPlayer registers audio sequences'
formats to the audio pipeline in first-come-first-serve order. With the change
in bc8d82355f, the audio format is deterministic.
The test is turned off in 060356ea00
PiperOrigin-RevId: 674261801
`PreloadMediaSource` allows to have a `startPositionUs` passed when `preload` is called, then in `PreloadControl.onContinueLoadingRequested`, it can be more intuitive to see the buffered duration rather than the absolute buffered position as the preload progress. Similar in `DefaultPreloadManager`, we haven't allowed the apps to set a custom start position for individual sources though, once we add this support, using the "duration from the start position" than the absolute position will be less error-prone, otherwise, it can run into a case that the position that the apps set is smaller than the start position.
PiperOrigin-RevId: 674251362
Sonic would accumulate truncation errors on float to int conversions
that caused the final output sample count to drift noticeably, by
hundreds of samples on streams of a few minutes of length. The fix now
keeps track of the truncation error and compensates for it.
Other small fixes include eliminating lossy operations (e.g. int
division) and using doubles instead of floats for resampling where
helpful.
This CL also introduces `SonicParameterizedTest`, which helps test
resampling on an arbitrary number of randomly generated parameters,
with random sample data. `SonicParameterizedTest` uses `BigDecimal`s
for calculating sample count values, as to avoid precision issues with
large sample counts.
PiperOrigin-RevId: 673852768
`CPU_FEATURES_COMPILED_ANY_ARM_NEON` is always defined when `CPU_FEATURES_ARCH_ARM` is but it can be defined as `0` or `1` ([cpu_features source](8e60d3f9be/include/cpu_features_macros.h (L237-L245))). This patch makes sure to use the value of `CPU_FEATURES_COMPILED_ANY_ARM_NEON` instead of whether it is defined or not.
PiperOrigin-RevId: 673837522
When processing edit lists in MP4 files, the media start position may be a non-keyframe. To ensure proper playback, the decoder must preroll to the preceding keyframe, as these preroll samples are essential for decoding but are not rendered.
Issue: google/ExoPlayer#1659
#cherrypick
PiperOrigin-RevId: 673457615
In case of negative PTS workaround, instead of disallowing all cases where samples are not in presentation order, we now validate that the first sample's presentation time is the smallest. This adjustment allows for correctly applying an offset to ensure all samples have a presentation time >= 0.
#cherrypick
PiperOrigin-RevId: 673434793
When converting NAL units from AnnexB to Avcc format,
one byte at a time was read. In fact many bytes were read
multiple times due to suboptimal logic.
Changed the logic to read 4 bytes at once and also to avoid
reading same bytes again.
This improved the time taken for writing a batch of 30
samples from 40ms to 20ms.
PiperOrigin-RevId: 673025781
In the current implementation due to missing "="
operator, last three bytes were not checked for
000 or 001 sequence.
In sample_no_bframes file, few samples has extra 0 at the end.
It was working fine with the bug because muxer was writing some
harmless 0 at the end.
PiperOrigin-RevId: 672994054
Avoids the dispatch of listener notifications of stream type/device volume change during construction of ExoPlayer. That was problematic because we end up blocking on `constructorFinished.blockUninterruptible()`
Issue: androidx/media#1692
PiperOrigin-RevId: 672965552
If the task executor handles an available frame (task submitted in the
SurfaceTexture listener) between the call to registerInputFrame() and
the execution of the task submitted in the method (in this CL), it
should be rejected.
PiperOrigin-RevId: 672903756
`Muxer` module needs to use this method, hence moved to common.
This CL also makes `getHevcProfileAndLevel` public because this is used
in `MediaCodecUtil`.
PiperOrigin-RevId: 671739166
Seeking was causing the player to hang in the following scenario:
1. The surfaceTexture's onFrameAvailableListener is called in
ExternalTextureManager to notify that a new frame is available.
2. This call submits a task on the GL thread.
3. A seek is performed and DefaultVideoFrameProcessor.flush() is called
before the task submitted in 2 is executed.
4. DefaultVideoFrameProcessor.flush() flushes the task executor, so that
the task submitted in 2 never gets executed.
5. Once the seek is over, the first frame is registered and rendered on
the surface texture.
6. Playback hangs because the onFrameAvailableListener is never called
for this new frame. This is because surfaceTexture.updateTexImage()
was never called on the frame that became available in 1.
This fix is making sure that the task submitted in 2 always gets executed.
Issue: androidx/media#1535
PiperOrigin-RevId: 671389215
Before this, because `Label.toString()` isn't implemented, the logged info
wasn't that useful:
```
labels=[androidx.media3.common.Label@6caac039]
```
With this change it's more useful:
```
labels=[en: english]
```
PiperOrigin-RevId: 671029474
The workaround causes issues with XML-based shared transitions, so we
can't apply it unilaterally.
Issue: androidx/media#1594
Issue: androidx/media#1237
PiperOrigin-RevId: 670960693
We have a tracking bug Issue: androidx/media#514 for supporting it, but issues like Issue: androidx/media#1542 still show that users stumble over it.
PiperOrigin-RevId: 670955189
Handling for `MediaCodec.CryptoException` was originally added only
around calls to `MediaCodec.queueSecureInputBuffer` and
`queueInputBuffer` (because these are the only methods that can throw
this exception). When asynchronous interaction with `MediaCodec` was
added in <unknown commit>, exceptions from `MediaCodec` started being stored
and bubbled out of **later** interactions with `MediaCodecAdapter`. This
means that `MediaCodecRenderer` can now see `CryptoException` thrown
from a different method, like
`MediaCodecAdapter.dequeueInputBufferIndex()`, and this ends up missing
the `catch (CryptoException)` code in `MediaCodecRenderer`. This results
in an "unexpected runtime error" stack trace like [A].
This change fixes the stack trace to:
1. Make it a "renderer exception" instead of "unexpected runtime error"
2. Include the correct DRM error code -> `@PlaybackException.ErrorCode`
mapping.
You can see the corrected stack trace below [B].
-----
[A] (synthesized from manually throwing a `CryptoException` from
`AsynchronousMediaCodecBufferEnqueuer#doQueueSecureInputBuffer`)
```
playerFailed [eventTime=11.56, mediaPos=10.35, window=0, period=0, errorCode=ERROR_CODE_UNSPECIFIED
androidx.media3.exoplayer.ExoPlaybackException: Unexpected runtime error
at androidx.media3.exoplayer.ExoPlayerImplInternal.handleMessage(ExoPlayerImplInternal.java:729)
at android.os.Handler.dispatchMessage(Handler.java:103)
at android.os.Looper.loopOnce(Looper.java:232)
at android.os.Looper.loop(Looper.java:317)
at android.os.HandlerThread.run(HandlerThread.java:85)
Caused by: android.media.MediaCodec$CryptoException: Test error message
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer.doQueueSecureInputBuffer(AsynchronousMediaCodecBufferEnqueuer.java:232)
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer.doHandleMessage(AsynchronousMediaCodecBufferEnqueuer.java:196)
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer.access$000(AsynchronousMediaCodecBufferEnqueuer.java:47)
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer$1.handleMessage(AsynchronousMediaCodecBufferEnqueuer.java:93)
at android.os.Handler.dispatchMessage(Handler.java:107)
at android.os.Looper.loopOnce(Looper.java:232)
at android.os.Looper.loop(Looper.java:317)
at android.os.HandlerThread.run(HandlerThread.java:85)
```
[B]
```
Playback error
androidx.media3.exoplayer.ExoPlaybackException: MediaCodecAudioRenderer error, index=1, format=Format(0, null, null, audio/mp4a-latm, mp4a.40.2, 134359, en, [-1, -1, -1.0, null], [2, 44100]), format_supported=YES
at androidx.media3.exoplayer.ExoPlayerImplInternal.handleMessage(ExoPlayerImplInternal.java:649)
at android.os.Handler.dispatchMessage(Handler.java:103)
at android.os.Looper.loopOnce(Looper.java:232)
at android.os.Looper.loop(Looper.java:317)
at android.os.HandlerThread.run(HandlerThread.java:85)
Caused by: android.media.MediaCodec$CryptoException: Test error message
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer.doQueueSecureInputBuffer(AsynchronousMediaCodecBufferEnqueuer.java:232)
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer.doHandleMessage(AsynchronousMediaCodecBufferEnqueuer.java:196)
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer.access$000(AsynchronousMediaCodecBufferEnqueuer.java:47)
at androidx.media3.exoplayer.mediacodec.AsynchronousMediaCodecBufferEnqueuer$1.handleMessage(AsynchronousMediaCodecBufferEnqueuer.java:93)
at android.os.Handler.dispatchMessage(Handler.java:107)
at android.os.Looper.loopOnce(Looper.java:232)
at android.os.Looper.loop(Looper.java:317)
at android.os.HandlerThread.run(HandlerThread.java:85)
```
PiperOrigin-RevId: 670951229
Limit input image size in Transformer to be less than 4096x4096.
For very large images, this can reduce memory usage substantially,
and stays away from `GL_MAX_TEXTURE_SIZE` - often 4096
PiperOrigin-RevId: 670555939
These tests do not test performance. Moving them out of this directory
ensures they are run on emulators and on all physical devices.
PiperOrigin-RevId: 670505992
MediaMuxer doesn't support B-frame before API 25.
The fix is added only in those test which appeared in triage failure.
It can be added to other tests as they are discovered.
PiperOrigin-RevId: 670480966
To avoid `WrongConstant` failures with methods that expect
`C.@PcmEncoding`, `OUTPUT_PCM_ENCODING` now points to the constant in
`C` instead of `AudioFormat`.
#cherrypick
PiperOrigin-RevId: 670251463
These were missed when upgrading to both SDK 19 and 21, but they now
cause failures like:
```
Caused by: java.lang.RuntimeException: Failed to parse package buildout/intermediates/apk_for_local_test/debugUnitTest/packageDebugUnitTestForUnitTest/apk-for-local-test.ap_: buildout/intermediates/apk_for_local_test/debugUnitTest/packageDebugUnitTestForUnitTest/apk-for-local-test.ap_ (at Binary XML file line #20): Requires newer sdk version #21 (current version is #19)
at org.robolectric.shadows.ShadowPackageParser.callParsePackage(ShadowPackageParser.java:61)
... 20 more
```
#cherrypick
PiperOrigin-RevId: 670241471
Make ByteBufferGlEffect public.
Build a speed test, and end to end test that verify
frames can be copied to CPU-accessible ByteBuffer
PiperOrigin-RevId: 670213343
Disable assertions and make sure playback passes.
The flake is caused by having different sequences starting with MediaItems of
different audio format, and it's undefined behaviour as to which one CompositionPlayer chooses to use.
PiperOrigin-RevId: 670195113
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 interface was used by Boxes.moov. This CL removes the interface and just uses the Track object directly.
Since Track is package-private it seems fine to use it directly.
The drawback with interface is that, with every new field addition in the
Track class, we need to update the interface as well (if we need to access that field for moov box).
PiperOrigin-RevId: 669295399
Adds a class that represents an image rectangle
in OpenGL coordinate convention.
android.graphics.Rect puts (0, 0) as the top-left corner:
it enforces `Rect.top <= Rect.bottom` and this matches
`android.graphics.Bitmap` coordinates: docs https://developer.android.com/reference/android/graphics/Rect
This is different from OpenGL coordinates where (0, 0) is
at the bottom-left corner. I.e. GlRect.bottom <= GlRect.top: docs https://registry.khronos.org/OpenGL-Refpages/es3.0/html/glReadPixels.xhtml
The reason for this change is to allow a public API GlRect
getScaledRegion() which selects a region of pixels of a GL texture
to be copied to CPU memory.
PiperOrigin-RevId: 669231826
The CL aims to
1. Shorten unnecessary lengthy chatty comments.
2. Remove dead TODOs.
3. nit fixes for comment style consistency.
4. Remove usage of "we" in the comments.
5. Media3 muxer does not need to mention the behaviour of framework muxer
unless its required for some purpose, so remove them.
PiperOrigin-RevId: 668985875
The existing logic was not working sometimes because:
1. The repeated scheduling in releaseAllFramesFromMediaCodec was
starving the thread on which the SurfaceTexture frameAvailableListener
was called.
2. The case where a pending frame arrives on the surface after flush
finishes executing was not handled.
The consequence of both problems is that availableFrameCount ended up
being > pendingFrames.size().
PiperOrigin-RevId: 668916256
These tests do not test performance. Moving them out of this directory
ensures they are run on emulators and on more than 1 physical device.
PiperOrigin-RevId: 668859017
DASH periods don't have to start at the beginning of a segment. In
these cases, they should report an initial discontinuity to let the
player know it needs to expect preroll data (e.g. to flush renderers)
This information is only available in the ChunkSampleStream after
loading the initialization data, so we need to check the sample
streams and tell them to only report discontinuities at the very
beginning of playback. All other position resets are triggered by
the player itself and don't need this method.
Issue: androidx/media#1440
PiperOrigin-RevId: 668831563
In the flaky test `ExoPlayerTest.loading_withLargeAllocationCausingOom_playsRemainingMediaAndThenThrows`, it is indeterministic that when the message `MSG_IO_EXCEPTION` from the loader thread will arrive on the playback looper. If it arrives after `MSG_PERIOD_PREPARED`, then the period can continue loading and get the three samples written to the `SampleQueue` before the intentional OOM surfacing to the `ExoPlayerImplInternal`, otherwise, the OOM will be detected by `ExoPlayerImplInternal` very early and the player disallows to load three samples, which will cause the assertion to fail.
As we are expecting the three samples to play until the playback fails, we should assume that the `Loader` encounters the OOM after those samples loaded, thus we need to put this trigger a bit later until the `SampleStreamItem`s are handled by the `FakeSampleStream`. This could be checked by the return value of `FakeMediaPeriod.continueLoading` (super class implementation). However, `FakeMediaPeriod.continueLoading` originally always returns `true`, which is not aligned with the javadoc of `MediaPeriod.continueLoading`:
"return `true` if progress was made, meaning that `getNextLoadPositionUs()` will return a different value than prior to the call, `false` otherwise."
then we should also modify that logic.
PiperOrigin-RevId: 668438316
This callback allows listeners to track when individual renderers
allow or prevent playback from being ready. For example, this is useful
to figure out which renderer blocked the playback the longest.
PiperOrigin-RevId: 667970933
Before this CL, the video sink was stuck if a flush was executed:
- after VideoSink.onInputStreamChanged, which is setting
pendingInputStreamBufferPresentationTimeUs and
- before the previous stream was fully rendered.
This is because pendingInputStreamBufferPresentationTimeUs was not reset
to TIME_UNSET when flushing the sink, so that the sink was still waiting
for the last frame of the previous stream to be rendered in
handleInputFrame/Bitmap.
PiperOrigin-RevId: 667924517
