Merge commit 'b35988329ac1a574898db9226039e37e352c57a8'
* commit 'b35988329ac1a574898db9226039e37e352c57a8':
Allow application to set two more encoding paramters: video profile and level
Merge commit '877045aa853aaf6e58ca3ac7bfa2bde9165b55ae'
* commit '877045aa853aaf6e58ca3ac7bfa2bde9165b55ae':
An AHandler can now find its associated ALooper.
Merge commit 'c4be155a540695c42bcd6589604f86d300f4548f'
* commit 'c4be155a540695c42bcd6589604f86d300f4548f':
Add an option to ALooper::start that allows it to call back into java or not.
Current implementation looks at the timestamps of all incoming frames in
CameraSource::dataCallbackTimestamp().
It drops all frames until enough time has elapsed to get the next time lapse frame.
When enough time has passed to capture the next time lapse frame, the frame is no longer dropped
and the timestamp of this frame is modified to be one frame time (1/framerate) ahead of the
last encoded frame's time stamp.
Change-Id: I82b9d5e96113dffa6901aac3b8a8ef999ffc1d0b
Also:
- Allowed start() call when encoder already starts and stop() call when encoder has not started yet
- Handled default value for audio/video sources/encoders and file output format
Change-Id: I03b2f7d3cf570baa0fd011a8c0ad200f2f2a5da1
- An alternative would be to define a common base class that
both MediaRecorderClient and MediaPlayerClient can derive.
But since the common code, onTransact() and notify() uses
some Binder code, having a common base class may not gain
us too much in terms of code reuse.
Change-Id: Ibc06720278ad173fceacff3d267b7060856c6316
- Track either the number of A/V frames authored, or the time elapsed
- Track the completion of the authoring
- Add multiple camera support for authoring by accepting a camera id parameter
- Set file type based on the OUTPUT_FORMAT requested
Change-Id: I0f9d31b3b7a8fa43eb53f572410fb0ebd4fa0bb7
Effect API:
- Use different definitions for audio device, channels, formats... in AudioSystem and EffectApi:
Removed media/AudioCommon.h file created for initial version of EffectApi
- Indicate audio session and output ID to effect library when calling EffectCreate(). Session ID can be useful to optimize
the implementation of effect chains in the same audio session. Output ID can be used for effects implemented in audio hardware.
- Renamed EffectQueryNext() function to EffectQueryEffect() and changed operating mode:
now an index is passed for the queried effect instead of implicitly querying the next one.
- Added CPU load and memory usage indication in effects descriptor
- Added flags and commands to indicate changes in audio mode (ring tone, in call...) to effect engine
- Added flag to indicate hardware accelerated effect implementation.
- Renamed EffectFactoryApi.h to EffectsFactoryApi.h for consistency with EffectsFactory.c/h
Effect libraries:
- Reflected changes in Effect API
- Several fixes in reverb implementation
- Added build option TEST_EFFECT_LIBRARIES in makefile to prepare integration of actual effect library.
- Replaced pointer by integer identifier for library handle returned by effects factory
Audio effect framework:
- Added support for audio session -1 in preparation of output stage effects configuration.
- Reflected changes in Effect API
- Removed volume ramp up/down when effect is inserted/removed: this has to be taken care of by effect engines.
- Added some overflow verification on indexes used for deferred parameter updates via shared memory
- Added hardcoded CPU and memory limit check when creating a new effect instance
Change-Id: I43fee5182ee201384ea3479af6d0acb95092901d
- estimate the moov box size for mp4 file writer based on the file
size/duration limit and target bit rate.
- can switch to use 64 bit file offset at runtime
rebased
Change-Id: Ibbe1f57e91ab2605820d5d96e8048d11e5559c53
- Mainly correcting the location of stbl box which should be a child of minf box.
This resolved the issue where the mis-muxed encoded file could not be played by QT/VLC.
- Enabled the the recorded tracks by setting the flags to 0x07 by default
- Allows for encoding either 32-bit or 64-bit offsets. By default encoding
32-bit offsets to reduce the metadata overhead
- Fixed a edts box issue where an empty elst box was used at the end
Change-Id: I570621a26714a81dc9400271aa5d3a07b483172f
Audio sessions are used to associate audio effects to particular instances (or groups) of MediaPlayers or AudioTracks.
Change-Id: Ib94eec43241cfcb416590f435ddce7ab39a07640
We don't need to do interleave when the total number of tracks to be recorded is one.
Metadata-wise, we only need to have one chunk in chunk offset table, and a
single entry in the stsc table.
Change-Id: I46f0e4b3860620311e7a91b68a9067acaa137bb2
Added AudioEffect C++ class. AudioEffect is the base class for effect specific implementations,
OpenSL ES effect interfaces and audio effect JNI.
Added the AudioEffect JNI and AudioEffect JAVA class. AudioEffect is the base class
to implement more specific JAVA classes to control audio effects from JAVA applications.
Change-Id: If300a1b708f2e6605891261e67bfb4f8330a4624
1. Avoid copying the input recording frames to the encoder via OMX interface
for TI video encoder
This is a missing change for part one which help reduces the CPU load.
2. Release output buffers as early as possible. This is a little bit helpful, but not critical.
TODO:
We should save the underlying pointers allocated by the OMX component before we replace them
and restore them before we call OMX_FreeBuffer()!
Change-Id: Ib3a88978f4c3b1153808872eaa7ac4c265a811ff
