Announcing OpenCORE 2.05 release

[Garret]

OpenCORE 2.05 has been released and is now available on the master
branch. There is a git tag "v2.05" marking the version. Below is a
summary of the changes since the previous release. The same
information is also included in the ChangeLog file in the top-level
directory of the OpenCORE code.

New Features
- Helper function to retrieve extended messages from command
responses.

- Support for eAAC+ encode in the pvAuthor engine.

- Add conversion operations to OSCL_String classes.

- RTSPT support (3GPP streaming over TCP). To test, replace "rtsp"
with "rtspt"
in the control URL (in source URL or SDP file). It is assumed that the
user has
access to a server with TCP streaming capability.

- PV Metadata Engine (ME). This engine is targeted at applications
that want
to retrieve metadata for any given clip in a generic manner.
Typically, a lot
of these applications would like to scan large batches of content.
Since ME
is mainly targeted at apps like mediascanner, where speed is the most
important
factor, thumbnail extraction is not supported. For applications that
do want
to extract thumbnails in a generic manner along with the metadata,
PV's
FrameAndMetadata Utility (FMU) is recommended instead.

- 2way Lip-Sync unit test cases.


Improvements
- A new base node implementation is introduced for the purpose of
refactoring and consolidating common logic in the nodes. The changes
will improve maintainability and make it easier to create new nodes.
Further documentation will be released in the future providing
guidelines
for authoring nodes.
- The following existing nodes were refactored to utilize the base
node
implementation mentioned above: AAC, AMR, and WAV parser nodes.
- Replace CodecSpecifier with PVMFFormatType.
- Move EventHandlerProcessing method to PVMFOMXBaseDecNode class.
- Change method type from "protected" to "private" in OMX derived
classes
(videodec node and audiodec node).
- Adding robustness to avcdecoder to verify PPS and SPS parameters.
- Display status of every test case after it finishes in 2way unit
test app.
- Add support to H.245 version 10.
- OMX node - Implement interleaved NAL sizes (to replace NAL start
codes) in
omx buffers carrying AVC frames.
- AMR-WB contributions to enable building in "C":
https://review.source.android.com/10014
https://review.source.android.com/10015
https://review.source.android.com/10016
https://review.source.android.com/10017
https://review.source.android.com/10018
- AMR-NB contributions to enable building in "C":
https://review.source.android.com/10297
https://review.source.android.com/10301
https://review.source.android.com/10302
https://review.source.android.com/10303
- Update OsclFileStats to track total time taken instead of only the
maximum time per operation.
- Rename OMX master core methods to avoid linking conflicts.
- Pass key MOUT_VIDEO_SUBFORMAT_KEY to the video MIO of FMU.
- Remove Direct Render Mode from mp4 composer library.
- AMR-NB and AMR-WB inline assembly is being re-enabled after a bug
in the QEMU assembly translation has been fixed.
- Modification of MIOs to take all configuration parameters at once.
- Move OMX input and output buffer ctrl structure away from data
buffers.
- Enable byte-stream (H264-RAW) format decoding in OMX dec node.
- PVMFInfoTrackDisable information event should be report to java
application.
- Retrieve the PLATFORM version at run time for the Android UserAgent
string.
- OMX audio and video dec nodes need to be able to handle
PortSettingsChanged
event with argument OMX_ALL.
- PVMFMediaClock latency handling improvement.
- AAC decoder creates artifacts when decoding clip.
- Modify MP4 composer node to do sample adds to mp4 composer lib in a
separate
thread.
- Add support for VIDEO_ENCODER_H264 in authordriver.
- Author major brand "3gp5" if 3gp clip has TIMED-TEXT track.
- No interleaving of mp4 files for just one track.
https://review.source.android.com/10510
- Return non-zero code when we detect a memory leak during unit tests.
- Add "3g2a", "3g2b" and "3g2c" as compatible branch in all content
authored
by PvAuthor SDK.
- Add support for OMX_COLOR_FormatYCbYCr.
- Remove unneeded libraries being linked against individual shared
libraries.
- Exclude empty directories and unncessary build files from OpenCORE.
- In OMX base node, turn the InputBufCtrlStruct from a struct into a
class
- Updates to pvmf_return_codes.pdf, pvplayer_developers_guide.pdf,
omx_decoder_test_app_guide.pdf, and omx_encoder_test_app_guide.pdf.

Bugs Fixed
- Repositioning takes a long time for YouTube videos after a few
repositions.
- Crash in CPVH223Multiplex::Stop().
- PlayerEngine cannot finish a playback session after repeated pause
and resume near EndOfStream.
- Allow file path to be passed to the pvPlayer SDK on Android.
- Add a new constructor for the AVCSampleEntry to be used when parsing
the
encv atom
https://review.source.android.com/9683
- MP3 DurationCalcAO logic needs to moved to a place during Prepare
and not
Init.
- Change PV_GET_ROW definition in M4V decoder.
- Player Engine unit test TC 55 and 56 has memory leaks
- Crash in authordriver if InitializeForThread fails.
- Crash in playerdriver if InitializeForThread fails.
- Potential memory leak in OMX encoder node.
- Memory leaks in AAC file parser node & lib
- FindFirst returns wrong element type when find a directory in
android
- Intermittent timeouts on 2way unit tests
- Incorrect DeBlocking logic
- AVC Clip Specific Issue: (Play till EOS) OR (Reposition to near EOS)
The
video gets stuck at near EOS
- iTunes genre is not parsed correctly
- Youtube : AV sync is lost after a seek when playing youtube contents
- Crash after recording one minute of Video when duration is specified
- Tracknumber metadata key string should be consistent through out all
nodes
- Possible race condition in singleton lock in omx_init
- CPV2WayDatapath::GetPortFormatType() selects least preferred type
- AAC OMX decoder component needs to process whole audio config buffer
- Modify MP3Parser to make seek to 0 a special case and do not use
XING header
for calculating file offset
- AAC Decoder should not try decode any AAC+ clip whose final sampling
freq
is > 48 KHz. Should default to AAC in these cases.
- Update OMX encoder node to calculate the correct output buffer size
in case
the encoder has multiple roles.
- OMX Encoder MPEG4 unit test fix
- Setting AMR_NB bitrate to GSM_AMR_4_75 will fail prepare.
- AuthorEngine should return Success incase Stop is called in
PVAE_STATE_INITIALIZED state.
- OMX M4V encoding drops the first I-frame
- VIDEO_INTERLEAVE_BUFFER_SIZE is too short for some buffers when
encoding
D1 resolutions
- Occasional audio glitch in MP3 parser during repositioning
- OMX tests for AMR give seg fault on android
- pvPlayer sends incorrect HTTP Range header
http://code.google.com/p/android/issues/detail?id=3031
- PDL/PS - Fix the Mp4 Parser node to pass the correct timestamp to PE
node
in RequestResumeNotification during Underflow condition
- H.245 TerminalCapabilitySet (TCS) and MasterSlaveDetermination (MSD)
requests must be sent before any other H.245 messages.
- Logic to set iKeepDroppingMsgsUntilMarkerBit to false in AMR RTSP
streaming
- 2way tests hanging when run in Android emulator
- Sound Recorder: Incorrect Duration of clip sometimes observed when
sdcard is
full
- Memory leak in PV2WayMIO
- Playerengine needs to be updated to return success even if there are
no keys
for GetMetadataValue
- Correct the Android RTSP UserAgent string
- Author Engine : Don't access the queue element without checking its
queue
size
- If SSRC is not provided in setup response, then pvplayer cleint
doesn't send
firewall packets to the server
- GetMetadataValues returns invalid number of entries
- Modify AUT to parse authored files and validate the files as much as
possible
- Eliminate media layer node in case of RTSP streaming
- AAC timestamps provided to OMX component may be wrong
- Clean up warnings and some negotiate parameter changes from omx dec
& enc
test apps
- Fix valgrind issue reported with android simulator in omx base node
- Add more supported rate control types in pvmp4h263encextension.h
- Memory leak in player engine test case 807
- Addressed compiler warnings in PVME
- nBufferAlignment of OMX_PARAM_PORTDEFINITIONTYPE is initialized to
negative
value when playing clip on android using qcom decoders
- MP3 parser crash
- pvplayer_engine_test is detecting 'rtspt_test.sdp' as a url instead
a local
file
- Count of command line arguments being read in PVME Test App is
incorrect
- Player engine test case 1307 fails when run with a mp4 file
containing 3GPP ti
med text track
- PVMFInfoDataReady event being sent to the engine after the unit test
calls
stop
- PlayerEngine doesn't free reference to
PVMFDataSourceNodeRegistryInitInterface
- Combine the AMR if statement in pv_omxmastercore.cpp
- Player Engine unit test case 154 fails on execution
- Author Engine unit test cases 101-105 fail
- Add new return code PVMFErrContentInvalidForProgressivePlayback for
a
NOT_PROGRESSIVE_STREAMABLE content
- Add USE (AL2, AL3) to 2way test files
- H223 2way test not being run
- Avoid calling CCYUV422toYUV420::New() for each thumbnail
- Author engine unit test case 101 fails
- Ignore encoding flaw, which generates aspec_ratio of value 0 in file
codecs_v2\video\m4v_h263\dec\src
- Memory leaks in pvmf_mio_fileinput.cpp and performance improvement
in mp4
composer node
- Do not cancel Init() immediately in case of 3gpp streaming
- Transferring MP3 with invalid ID2 tags crashes mediaScanner

[Brian 1143]

Does this release include fix for issue 3319 (http://code.google.com/p/
android/issues/detail?id=3319)?
Thanks.

On 8月15日, 上午4時38分, Garret <pic...@pv.com> wrote:
> OpenCORE 2.05 has been released and is now available on the master
> branch.  There is a git tag "v2.05" marking the version.  Below is a
> summary of the changes since the previous release.  The same
> information is also included in the ChangeLog file in the top-level
> directory of the OpenCORE code.
>
> New Features
> - Helper function to retrieve extended messages from command
> responses.
>
> - Support for eAAC+ encode in the pvAuthor engine.
>
> - Add conversion operations to OSCL_String classes.
>
> - RTSPT support (3GPP streaming over TCP). To test, replace "rtsp"
> with "rtspt"

> in the control URL (in source URL orSDPfile). It is assumed that the

> - AMR-WB contributions to enable building in "C":https://review.source.android.com/10014https://review.source.android.com/10015https://review.source.android.com/10016https://review.source.android.com/10017https://review.source.android.com/10018
> - AMR-NB contributions to enable building in "C":https://review.source.android.com/10297https://review.source.android.com/10301https://review.source.android.com/10302https://review.source.android.com/10303

> - No interleaving of mp4 files for just one track.https://review.source.android.com/10510

> - Return non-zero code when we detect a memory leak during unit tests.
> - Add "3g2a", "3g2b" and "3g2c" as compatible branch in all content
> authored
> by PvAuthor SDK.
> - Add support for OMX_COLOR_FormatYCbYCr.
> - Remove unneeded libraries being linked against individual shared
> libraries.
> - Exclude empty directories and unncessary build files from OpenCORE.
> - In OMX base node, turn the InputBufCtrlStruct from a struct into a
> class
> - Updates to pvmf_return_codes.pdf, pvplayer_developers_guide.pdf,
> omx_decoder_test_app_guide.pdf, and omx_encoder_test_app_guide.pdf.
>
> Bugs Fixed
> - Repositioning takes a long time for YouTube videos after a few
> repositions.
> - Crash in CPVH223Multiplex::Stop().
> - PlayerEngine cannot finish a playback session after repeated pause
> and resume near EndOfStream.
> - Allow file path to be passed to the pvPlayer SDK on Android.
> - Add a new constructor for the AVCSampleEntry to be used when parsing
> the

> encv atomhttps://review.source.android.com/9683

> - pvPlayer sends incorrect HTTP Range headerhttp://code.google.com/p/android/issues/detail?id=3031

[RaviY]

No.

> > - AMR-WB contributions to enable building in "C":https://review.source.android.com/10014https://review.source.android....
> > - AMR-NB contributions to enable building in "C":https://review.source.android.com/10297https://review.source.android....

> ...
>
> read more »

[San Mehat]

On Fri, Aug 21, 2009 at 11:25 AM, RaviY <yen...@pv.com> wrote:

No.

<cues the trobones>
waaah-waaaaaaah.
 

--
----------
San Mehat
Staff Software Engineer
Google Inc.
o: 650-253-7422
c: 408-382-1249
s...@google.com

[RaviY]

On popular demand, I will try to get that in sometime next week.

-Ravi

On Aug 21, 2:32 pm, San Mehat <s...@google.com> wrote:

> ...
>
> read more »