The default sampling rate chosen for an AudioTrack or AudioRecord
constructed with a Builder not specifying a sampling rate is not
the primary output sampling rate but depends on the device actually
selected.
Bug: 29541155
Change-Id: Ifcab65418b816b7321b6e9542b9e4bb95f748f95
Support additional corner cases in level ordering:
- MPEG4Level1 only implies support for MPEG4Level0
- HEVC high tier levels are only supported by other high tier levels
H263 unusual level definitions:
- levels 45/50+ define minimum requirements that codecs can extend
- levels 10-45 define minimum size
- restrict to QCIF and CIF only if codecs only supports levels 10-40
(or 45 in profiles 0 and 2)
MPEG4 max width/height/frame rate:
- Only specified for MPEG4Level0/0b
- Using arbitrary 2:1 aspect ratio and 60fps for other levels
This is sure to include typical sizes for levels
Fix level limits:
- MPEG2LevelML max 25fps for D1 PAL
- H263Level20 max 30fps, but only 15fps for CIF
- H263Level45 max 15fps
- H263Level45+ min alignment is 4
- MPEG4Level0b max 15fps
- add support for VP9 HDR profiles
- fix HEVC overflow for Level62 calculations
Bug: 28671284
Change-Id: Ie75ae1d19f20d5c6eb40807d03c3b487854eb9a0
Explain buffer size optimization.
This CL has no code changes, just documentation changes.
Bug: 28802853
Change-Id: I91942f00a3e9691e4fc05b24656b8bf4cb538bc2
Signed-off-by: Phil Burk <philburk@google.com>
Improve documentation for error codes returned by
AudioTrack.write() and AudioRecord.read() methods.
Fix native to JAVA error code conversion in JNI.
Bug: 28906466
Change-Id: I4d48b1d428834b7a39a14e2d81b6c164696817a8
Requiring uniqueness of ID across all track types seems a little too
much demand for developers and they do want to use the same ID on
different track types. Also, this doesn't seem terribly useful.
Hence remove the check.
Bug: 28436618
Change-Id: I9040310c5ce31bcff522021d83f7508506cbd9d3
A workaround change for supporting support lib 23.x in N release.
This change should be reverted in some point before O release.
Bug: 28572934
Change-Id: I7e6afc0293688e425736bc93da0ea10233c07e9a
Update documentation for VOICE_UPLINK, VOICE_DOWNKINK and
VOICE_CALL MediaRecorder audio sources to indicate those
sources are protected by the signature|privileged permission
CAPTURE_AUDIO_OUTPUT
This was missing in M SDK doc.
Bug: 28081587.
Change-Id: I24f6d9978069115a0f2c41e48038b935e4395124
Registering with the maximum potential memory made by visible by an
ImageReader/Writer can cause the VM to try to heavily garbage-collect
the rest of the application. This can have significant impact on camera
applications, since they often have large ImageReader/Writer queues.
Bug: 28454727
Change-Id: I2ba43635f93da66655be024165a15631b3b421d3
