* Add AC-4 MIME type definition
* Add AC-4 format support in Mp4Extractor and TsExtractor
* Add AC-4 Extractor
* Add AC-4 playback support in MPEG-4, MPEG-DASH, TS and HLS
ExoPlaybackExceptions of type SOURCE are always associated with the loading
period and thus we can use the event time for the loading period in
onPlayerError. Renderer and unexpected exceptions are still associated with the
currently playing period.
Issue:#5407
PiperOrigin-RevId: 230216253
The Amlogic awesome decoder reduces the video size of interlaced videos by half
if the internal configuration isn't force reset with new maximum input size
values. The product of these new values must exceed 1920x1088 to force the
reset.
Issue:#5003
PiperOrigin-RevId: 230206675
That's the same position set in MediaPeriod.prepare (where it may be removed
in the future).
Having the position at an earlier point is necessary to fix an
issue with lazy preparation in ConcatenatingMediaSource where the prepare
position was assumed to be known but MediaPeriod.prepare hasn't been called
yet.
Issue:#5350
PiperOrigin-RevId: 229756637
The buffered position is currently based on the mimimum queued timestamp of
all AV tracks. If the tracks have unequal lengths, one track continues loading
without bounds as the "buffered position" will always stay at the shorter
track's duration.
This change adds an optional buffer flag to mark the last sample of the
stream. This is set in the Mp4Extractor only so far. ExtractorMediaSource
uses this flag to ignore AV streams in the buffered duration calculation if
they already finished loading.
Issue:#3670
PiperOrigin-RevId: 229359899
That was previously handled by the player. But since we switched to Handler
messages instead of player messages, we should do that manually.
PiperOrigin-RevId: 229341747
This is not necessary as the track selection needs to be updated with
updateSelectedTrack anyway. It's also error-prone as the selection code
calls into a protected method of a not fully initialized class.
PiperOrigin-RevId: 229331669
DataSpec.FLAG_ALLOW_CACHE_FRAGMENTATION is added to indicate to the
cache when fragmentation is allowed. This flag is set for progressive
requests only.
To avoid breaking changes, CacheDataSink defaults to ignoring the flag
(and enabling fragmentation) for now. Respecting the flag can be
enabled manually. DownloaderConstructorHelper enables respecting of
the flag.
Issue: #4253
PiperOrigin-RevId: 229176835
We currently forget whether a source is seekable at re-preparation. This was
implemented intentionally this way under the assumption that we really can't seek
until we have loaded the seek map again. However, seek operations are only
allowed after a media period is prepared. So there is no harm in remembering
whether a source is seekable.
This problem currently prevents reusing ClippingMediaSources with
ExtractorMediaSource and a non-zero start clip position.
Issue: #5351
PiperOrigin-RevId: 229169441
ExoMediaCrypto.requiresSecureDecoderComponent() is removed, and
FrameworkMediaCrypto.forceAllowInsecureDecoderComponents is made
public to allow determining whether a secure decoder is required
to be implemented in MediaCodecRenderer.
PiperOrigin-RevId: 228909771
There were some edge cases in which we'd forget to release DRM
sessions. For example if we read a format and acquired a
pendingDrmSession (in onInputFormatChanged), then immediately
read another format and overwrote pendingDrmSession, we'd
forget to release the one that's been overwritten.
This change hopefully makes release much clearer. We keep a list
of all drm sessions we're currently holding. Whenever we update
either drmSession or pendingDrmSession, we release any other
sessions that are in the list.
PiperOrigin-RevId: 228905465
The window object is used without being filled with data. This used to work
well for most cases as the same live stream is sending regular updates and the
first update is almost never used if it's not the first item in a playlist.
It causes problems when the first timeline update of a live stream is actually
used for playback (e.g. when the live stream is lazily prepared in a playlist
and played first).
PiperOrigin-RevId: 228530232
Reported in https://github.com/google/ExoPlayer/issues/3860
For failing examples see the github link above.
[Problem]
We drop matching control codes even if they are not received on
consecutive frames.
The specification says
"(4) If the first transmission of a control code pair passes parity,
it is acted upon within one video frame. If the NEXT frame contains
a perfect repeat of the same pair, the redundant code is ignored."
Keyword is the NEXT. The frames must arrive immediately after
each other.
See https://www.law.cornell.edu/cfr/text/47/79.101
[Solution]
Set an additional flag when any data is processed. Control code
duplication checks should be limited only for the first control
byte pairs processed after any control code.
[Test]
Sarnoff tests have equivalent CEA708 and CEA608 Streams.
This is the initialization part of mitigating issue #4253. The
remaining work is on the writing side, and is simply a case of
having startFile return File instances that are sharded into
sub-directories. We still need to decide what scheme we want
to use for doing that.
Issue: #4253
PiperOrigin-RevId: 228306327
Calls to File.length() can be O(N) where N is the number of files
in the containing folder. This is believed to be true for at least
FAT32. Repeated calls for the same file tend to be faster,
presumably due to caching in the file system, however are still
surprisingly expensive. Hence minimizing the number of calls is
preferable.
Issue: #4253
PiperOrigin-RevId: 228179921
They are not longer needed anywhere, are error-prone (because of threading
requirements), and complicate testing and using MediaSources without a player.
PiperOrigin-RevId: 227871157
We currently either use the app thread returned by the player or the thread
the commands are called on depending on whether the media source is already
prepared or not.
This change lets the application decide which callback thread to use. As a
side effect, we also don't longer need access the player instance passed to
MediaSource.prepare.
PiperOrigin-RevId: 227871111
Also configure the FFmpeg context to ignore errors as far as possible (this
appears to have an effect only for certain decoders).
Issue: #5293
PiperOrigin-RevId: 227851397
The top level requirement only tried to ensure that the entire timeline only
has one period. This is already asserted by ImaAdsLoader. AdsMediaSource
itself works fine as long as the wrapped timeline has one period only. This
is now asserted instead.
PiperOrigin-RevId: 227682141
Passing the player through MediaSource.prepare is only needed for the AdsLoader
and complicates other usages of MediaSource. Providing the player directly to
the AdsLoader is also in line with the usage pattern of PlayerView and other
components.
Also rename methods to start/stop to better reflect their usage.
PiperOrigin-RevId: 227682112
Made DownloadState top level class.
Replaced action field DownloadAction fields.
Added removing, removed and restarting states.
Renamed started state to downloading.
PiperOrigin-RevId: 227664735
ExoPlayer methods must not be called from any thread besides the specified
app thread. Therefore we shouldn't use them here. Using a regular Handler
instead is fully equivalent.
Issue:#5240
PiperOrigin-RevId: 227650489
- Increase the default cache file size to 5MB
- Recommended a minimum cache file size of 2MB to discourage
applications from specifying values small enough such that
unreasonably large numbers of cache files are generated
- Allow maxCacheFileSize=C.LENGTH_UNSET, equivalent to setting it
to MAX_VALUE. This is just for API consistency with other APIs
we have that accept LENGTH_UNSET
Issue: #4253
PiperOrigin-RevId: 227524233
CacheEvictor.onStartFile currently receives a length, which is the
maximum size of the content that might be written. The LRU cache
evictor will make sure there's sufficient space for the specified
size. If the actual size of the content is unknown, CacheDataSink
passes maxCacheFileSize.
The current behavior isn't ideal. It seems valid for a developer to
specify maxCacheFileSize=Long.MAX_VALUE if they don't want any
file fragmentation in the cache. However if they then attempt to
cache something with unknown length, LRU cache will try and make
room for content of length Long.MAX_VALUE, and end up evicting the
entire cache to do so.
This change alters the logic so a length is only passed if the
actual size of the content is known. In other cases C.LENGTH_UNSET
is now passed. The LRU evictor will not evict until after the file
is committed. Note a custom LRU evictor could still opt to evict to
ensure some application specified amount of space, if that's the
desired behavior.
PiperOrigin-RevId: 227509525
* Add AC-4 MIME type definition
* Add AC-4 format support in Mp4Extractor and TsExtractor
* Add AC-4 Extractor
* Add AC-4 playback support in MPEG-4, MPEG-DASH, TS and HLS
When a Download is created it's set to queued state but doesn't notify
listeners about this state. DownloadManager checks if it can start the
download. After this checks it notifies the listeners.
With this change Download can immediately check if it can be started and
sends correct notification.
PiperOrigin-RevId: 225967129
Problem: There is nothing in the Cache interface that enforces that
these methods are implemented by reading/writing content metadata,
however our own code (CacheDataSource) relies on this property since
it mixes use of [get|set]ContentLength and use of metadata. Using
metadata is the right approach because it minimizes the number of
index writes (e.g. it allows the redirect URI and the content length
to be committed by a single write).
Solution: Remove [get|set]ContentLength, which are redundant anyway,
and use metadata everywhere. Also expose the keys being used through
the public API; there's no particular reason why they shouldn't be.
We previously had an API for getting and setting content length, but
no API for getting and setting redirect uri.
PiperOrigin-RevId: 225383211
This adds the basic track selection capabilties (including tests).
The new capabilities are not exposed yet through the DownloadHelper implementations
and there will also be more helper methods (e.g. to select multiple audio lanuages at
once).
PiperOrigin-RevId: 224518477
We currently have two factory methods where it is completely unclear which one needs
to be overridden.
This change deprecates the old one, adds a Util method to easily map back from the new
to the old behaviour, and updates all implementations of the now deprecated method in
our code.
PiperOrigin-RevId: 224303560
[Problem]
As the alignment of each lineas are calculated individually, the
artificially introduced CENTER and RIGHT alignments can be mixed
with each other and with the default LEFT alignment. The mixed results
look worse than just using the original DEFAULT, and the readability
is radically decreased.
[Solution]
Pre-calculate the alignment for each line, and use the leftmost of them
for all lines. The intention of caption provider could be different, but
the worst case scenario is that we revert back everything to LEFT
alignment that is the only option in the CEA608 standard so we
show everything at the exact location it is set in the content.
The readability is never worse than the one given by the content
provider.
[Test]
This should influence CEA608 pop-up cations only. (Not roll up).
Test various live channels with this change.
Note: Beginning of commercials tend to have PAINT-ON mode as that
is immediately showing incoming characters, not just at the end
of the line like POP-on mode.
This instantiates the renderers and extract the capabilities. None of the known
renderes incurs any overhead during instantiation.
PiperOrigin-RevId: 224118511
- Enable GZIP for media playlist + encryption key chunk requests in
HLS, as we do during playback
- Pass around DataSpecs rather than Uris. This will be needed for if
we add manifest cacheKey support (which seems like a good idea for
completeness, if nothing else)
PiperOrigin-RevId: 224057139
We currently default to not caching data if the content length
cannot be resolved once the DataSource has been open. The
reason for this is to avoid caching progressive live streams.
By doing this we were accidentally not caching in other places
where caching is possible, such as DASH/SS/HLS segments during
playback if the server doesn't include a Content-Length header.
Also HLS encryption key chunks, which were very unlikely to be
cached during playback because we explicitly set FLAG_ALLOW_GZIP
(which normally stops content length from resolving) without
setting FLAG_ALLOW_CACHING_UNKNOWN_LENGTH.
It seems like a good idea to flip the default at this point,
and explicitly disable caching in the one case where we want
that to happen.
PiperOrigin-RevId: 223994110
This is in line with how Player.EventListener and AnalyticsListener methods are
defined and helps to only implement the callbacks needed.
PiperOrigin-RevId: 223991262
Moving most of the logic to the base DownloaderHelper helps to implement track
selection for downloading in a single place instead of multiple places.
PiperOrigin-RevId: 223964869
This option to block bandwidth already exists on the AdaptiveTrackSelection itself
but it's not currently possible to forward the total fixed track bandwidth
automatically.
PiperOrigin-RevId: 223785139
- Asus ZenFone GO (ASUS_X00AD_2)
- Sugar S9 (i9031)
- Redmi Note 3 (kate)
These devices trigger native crashes similar to
https://github.com/google/ExoPlayer/issues/4460
I'm not sure why Asus Zenfone Go (model: ZB500KL, device: ASUS_X00AD_2) was removed here
73af056da3
PiperOrigin-RevId: 223580393
A remove action will eventually just be the unique ID of the download
that should be removed. This is a step toward that goal.
PiperOrigin-RevId: 222832456
[Problem]
PAINT-ON mode is not implemented. From the compliance tests:
* RDC command has no effect except to select paint-on style.
* Next data are written directly to the display upon receipt.
* If other captioning is already on the screen, the four-row
limit is still in effect.
[Solution]
It is a rare use case, we do not support overriding characters in
existing cueBuilders as PAINT-ON would require. But several
compliance tests check if the screen is cleared when the mode
switch happens. We must keep the old captions when switching
to PAINT-ON mode
[Test]
- Live Over-the-Air content, beginning of commercials often uses
PAINT-ON mode
- Add DownloadAction.fromByteArray for symmetry with toByteArray
- Make DownloadService call fromByteArray, for symmetry with calls
to toByteArray
PiperOrigin-RevId: 222801703
We already have tests for comment blocks, and they already
pass because we discard the comment when we fail to parse it
as a cue. We should just skip it directly, however.
This allows more efficient range removals and is consistent with addition,
which supports adding multiple elements in a single operation.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222396310
This allows to account for bandwidth used by fixed track selections.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222383736
BandwidthProvider allows bandwidth allocation logic to be customized.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222293581
1. Pull up all subclasses of DownloadAction into DownloadAction
2. Add DownloaderFactory for Downloader instantiation, and DefaultDownloaderFactory
to replace the instantiation logic being removed from the DownloadAction
subclasses.
This change will upgrade existing action files gracefully (i.e. it does not
lose compatibility with the existing offline implementation, other than some
minor breaking changes to the API).
TODOs:
1. Move test methods from the XDownloadActionTest classes into DownloadActionTest.
This will be done in a subsequent CL. There's a lot of consolidation that can
be done here, including de-duplicating some of the test code added in this CL.
2. Look at merging DownloaderConstructorHelper into DefaultDownloaderFactory.
3. Use customCacheKey in DASH/HLS/SS Downloaders, for completeness.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222258983
createTrackSelections decides whether to create an adaptive or a fixed track seletion to create.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222231011
Forwarding the listeners to the notification update is meant to ensure we
only notify the listeners which were registered at the time the event happened
However, we currently just copy the reference to the actual list instead of
doing a deep copy of the listeners.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222227735
This ensures that all callbacks use the same notification mechanism.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222226913
The update is actually unrelated and putting it all in one large update
object complicates adding the same update mechanism for other values.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222226825
WindowedTrackBitrateEstimator, uses future and past chunks in a window
of time to estimate track bitrate values.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=222065409
Currently no path may be chosen if
minBufferUs == maxBufferUs == bufferedDurationUs.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=221639199
This is needed because the allocator can't be reused for example.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=221638233
- Increase the search window size to fix TS seeking for problematic
media we've had provided to us.
- As per my comments on the issue, we should look at doing more here
to better fix the problem. This will solve the worst of the
immediate problem, however.
- The memory usage is non-trivial, particularly with the increased
search window size. I've made the allocations only live whilst
determining duration and seeking to address this. I've done the same
for PS just for consistency.
Issue: #5097
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=221449988
The remaining work is to split Window.isDynamic so that it's
possible to represent a window that wont be appended to, but
may still be removed from.
Issue: #4780
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=221439220
This is a precursor for fixing the ref'd issue. These classes are
well tested, so the tests passing should give you reasonable
confidence I didn't break anything :).
Issue: #5097
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=221435824
LegacyVideoTrackSelection is the previous implementation (unchanged).
VideoTrackSelection is rewritten to extend AdaptiveTrackSelection.
An experiment is added to allow use of the new VideoTrackSelection
implementation. It's envisaged that this transition will be more or
less a no-op change. The experiment flag is only in case of bugs
during roll-out, and will be removed with LegacyVideoTrackSelection
upon successful transition.
VideoTrackSelection also implements discard-on-trick-play-exit, which
was a TODO in the previous implementation.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=221153827
When in REPEAT_MODE_ONE, it's unlikely apps want next/previous methods
on the player to keep them in the same window. Music apps in particular
tend to implement next/previous functionality as though repeat mode were
off when in this mode (i.e. current song loops forever during playback,
but next/previous navigation still navigates to next/previous items).
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=220469655
This fixes an issue where disabling audio focus handling
while audio focus is held would not release audio focus.
A new test was added for this situation.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=220316866
*** Reason for rollback ***
Rolling forward again as [] should fix issue that prompted the rollback
*** Original change description ***
Automated g4 rollback of changelist 219130576.
*** Original change description ***
Re-enable codec re-use
Issue: #2826
***
***
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=220124362
It's no longer guaranteed that onOutputFormatChanged will always be
called when the renderer is re-enabled, since the codec may be kept
and its output format may not change. Hence we need to avoid resetting
state in onDisabled that we need when the renderer is enabled again.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=219821519
Currently, MediaCodecRenderer's render loop does not have any time limit. We
always try to drain and feed as much buffers as possible. This may lead to a
pattern of feeding phase takes too much time from previous loop, which makes
the next draining phase drops all buffers, making these buffers available for
feeding all at once, and the pattern keeps repeating.
This CL adds an experimental parameter for setting a time limit. If the time
limit is exceeded, the feeding process stops even if more input could be fed.
The default behavior is unchanged
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=219820386
This prevents leaving an inconsistent state after a EOF exception.
Issue:#5039
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=219785275
