ChangeLog.txt

//------------------------------------------------------------------------------------------------------------------
09/09/2018

- Update project to Visual Studio 2017 Community


//------------------------------------------------------------------------------------------------------------------
04/07/2015

- Update project to Visual Studio 2013 Community

- Update MFTDxva2Decoder 1.1.0.0 (beta)

* the decoder is now able to decode mpeg1/2 format with dxva2

- Update MFNodePlayer 1.11.3.3 (beta)

* add code to use MFTDxva2Decoder with file drop

- Update MFSrMpeg2Splitter 1.3.1.0 (beta)

* add seek functionality (this introduces a bug : sometimes when loading file)


//------------------------------------------------------------------------------------------------------------------
15/11/2014

- Update MFTCudaDecoder 1.4.0.0 (beta)

* Handle video timestamp

- Update MFSrMpeg2Splitter 1.2.1.0 (beta)

* Handle decoded time stamp

- Update MFTMpeg12Decoder 1.1.0.0 (beta)

* Few change for later check audio/video synch (according to MFTCudaDecoder)


//------------------------------------------------------------------------------------------------------------------
01/11/2014

- Update MFNodePlayer 1.10.3.3 (beta)

* Bug in CMFNodeForm::OnDrop : allocate MAX_PATH * sizeof(WCHAR), not just the name file size
* Set Autolock in CMFNodePlayer::Invoke for m_state (design has to be changed...)
* Add a define to test cuda decoder with drag and drop

- Update MFTCudaDecoder 1.3.0.0 (beta)

* B frames are handle correctly
* Code has been refactored
* note:
. Cuda never decodes the first I frame (don't know why)
. Now there is a latency before presenting the first frame (20 frames), perhaps we can improve this


//------------------------------------------------------------------------------------------------------------------
17/10/2014

- Update MFNodePlayer 1.9.2.3 (beta)

* Add free session to load a file by drag and drop
* No media session blocking on loading failure (create media source and topology before media session)
* Change buttons text to "Select".
* VideoRendererSession : add option to select decoder (cuda or not)

- Update MFSrMpeg2Splitter 1.1.1.0 (beta)

* update MFByteStream.cpp to be used by both topoedit and MFNodePlayer

- Update MFTCudaDecoder 1.2.0.0 (beta)

* Change project properties to Cuda SDK 6.0
* Add output media type YV12, to be used by MFNodePlayer (video renderer session)


//------------------------------------------------------------------------------------------------------------------
10/05/2014

- Update MFNodePlayer 1.9.2.2 (beta)

* Refactor MFNodeForm
* Improve MediaSession event handling
* No more memory leaks, on loading failure

- Update MFSkDxva2Renderer 1.1.0.0 (beta)

* To be registered as MFT_CATEGORY_OTHER

- Update MFSkVideoRenderer 1.3.0.1 (beta)

* To be registered as MFT_CATEGORY_OTHER

- Update MFSkImageWriter 1.1.0.0 (beta)

* To be registered as MFT_CATEGORY_OTHER

- Update MFSkJpegHttpStreamer 1.1.1.0 (beta)

* To be registered as MFT_CATEGORY_OTHER


//------------------------------------------------------------------------------------------------------------------
05/05/2014

- Adding MFSkDxva2Renderer 1.0.0.0 (beta)

- Update MFSrFlvSplitter 1.2.1.1 (beta)

* The small bug in CreateVideoSample wasn't a bug...

- Update MFSkVideoRenderer 1.2.0.1 (beta)

- Update MFTCudaDecoder 1.1.0.0 (beta)

- Update MFTMpeg2Decoder 1.2.0.3 (beta)

- Update MFNodePlayer 1.8.1.2 (beta)

* Adding Dxva2 session

- Move MFSafeQueue.h in common

- Rename Mpeg2Lib in LibMpeg2, to be compiled before MFTMpeg2Decoder


//------------------------------------------------------------------------------------------------------------------
30/04/2014

- Update MFNodePlayer 1.7.1.2 (beta)

* Adding audio/video capture session (show how aggregate capture source)

- Update MFSrFlvSplitter 1.2.0.1 (beta)

* A small bug in CreateVideoSample


//------------------------------------------------------------------------------------------------------------------
28/04/2014

- Update MFNodePlayer 1.6.1.2 (beta)

* Adding sequencer source sample

- Update MFSkVideoRenderer 1.1.0.0 (beta)

* fx file is loaded as resource.

- Update MFSrFlvSplitter 1.1.0.0 (beta)

* Adding audio/video stream selection


//------------------------------------------------------------------------------------------------------------------
21/04/2014

- Adding MFSkVideoRenderer 1.0.0.0 (beta)

- Adding Mp3Decoding

- Update MFNodePlayer 1.5.1.2 (beta)

* Adding Renderer Session to use MFSkVideoRenderer
* Adding controls to handle shader parameters from the MFSkVideoRenderer


//------------------------------------------------------------------------------------------------------------------
30/03/2014

- Adding MFSrA52Decoder 1.0.0.0 (alpha)

- Adding MFTA52Decoder 1.0.0.0 (alpha)

- Update MFNodePlayer 1.4.1.1 (beta)

* Adding Flv Session and an editbox to select the flv file

- Update MFSrFlvSplitter 1.1.0.0 (beta)

* Disable one header check, see ParseHeader in FlvParser.cpp

- Update MFSrMpeg2Splitter 1.1.0.0 (beta)

* Adding vob file extension


//------------------------------------------------------------------------------------------------------------------
30/03/2014

- Adding MFSrMpeg2Splitter 1.0.0.0 (beta)

- Adding MFSkImageWriter 1.0.0.0 (beta)

- Adding MFTCudaDecoder 1.0.0.0 (beta) (compiled with Cuda SDK v5.5)

- Adding MFTDxva2Decoder 1.0.0.0 (alpha) (project stopped)

- Adding MFSrFlvSplitter 1.0.0.0 (beta)

- Adding MFTVp6Decoder 1.0.0.0 (beta)

- Update all StdAfx.h according to MFInclude.h

- MFSrMpeg12Decoder 1.1.1.0 (beta)

* Return failure if layer is 3

- Update common project files

* adding MFTraceSample.h
* adding CLSID in MFNodeGuid.h
* Change MFBuffer.h to be more robust
* adding CMFWriteFile in MFFile.h
* Change MFInclude.h to handle more common include files
* Adding MFDxva2String in MFTrace.h
* Adding ClockStateString in MFState.h
* Adding MFReadParam.h
* Adding MFByteStream.h


//------------------------------------------------------------------------------------------------------------------
16/01/2014

- Update MFNodePlayer 1.3.1.1 (beta)

* Adding Screen Session and HttpStreamer Session (Jpeg)
* Adding Check for registered dll
* Adding MFNodePlayer_Topology.cpp


//------------------------------------------------------------------------------------------------------------------
14/01/2014

- Update MFNodePlayer 1.1.1.1 (beta)

* Adding MF_TRACE_PIPELINE_EVENT
* Handle cursor
* If using stop then play, the MediaSession did not restart

- Update MFTVideoMpeg2 1.1.0.3 (beta)

* Reset MEDIASUBTYPE_MPEG1Payload to keep compatibility with Mpeg1Source from SDK


//------------------------------------------------------------------------------------------------------------------
01/12/2013

- Update MFSkJpegHttpStreamer 1.0.1.0 (beta)

* Fix a big problem with reference counting. The design was to reuse the sink after ShutDown.
It was a bad idea due to circular reference counting (between Sink and StreamSink).
When ShutDown is called, MediaFoundation is designing to delete the sink after. Not to reuse it.


//------------------------------------------------------------------------------------------------------------------
18/11/2013

- Adding MFSkJpegHttpStreamer 1.0.0.0 (beta)

- Update MFTJpegEncoder 1.1.0.0 (beta) :

Change SetEndPosition in CJpegStream::Write.
Adding m_bHaveImage

- Update MFSVersion in MFSrScreenCapture and Trace

- Update MFSVersion in MFSrMpeg12Decoder

- Update MFInclude/MFLogging/MFMacro/MFNodeGuid

- Reset the Common folder in the solution and all the ToDo.txt


//------------------------------------------------------------------------------------------------------------------
18/11/2013

- Adding MFTJpegEncoder 1.0.0.0 (beta)

* I did not test it. I have to write the sink that will stream the jpeg data. Coming soon

- Rename a lot of project. See MFNodeGuid.h

- Update MFTMpeg2Decoder 1.1.0.2 (beta)

* When frame size is not multiple of 16, we get wrong image. It's because the decoder use 16 bytes alignment.
Adding GetWidth from the decoder, and Mpeg2CopyFrameYV12Stride to setup YUV correctly.

- Update MFTMpegAudioDecoder 1.0.0.1 (beta)

* Activate m_bValidTime to use splitter source time, if valid. When not use, the audio was not sync

- Update MFBuffer.h

* in Reserve, change GetBufferSize by m_dwTotalSize. We must check the total size, not the current
buffer size.
* add some check for DWORD overflow

- Update some common files to handle more things


//------------------------------------------------------------------------------------------------------------------
26/07/2013

- Adding LibMpegAudio

* original source code from lipmp2dec and mpeg123 rewritten in C++
* decode mp1 and mp2 audio format
* does not handle free format

- Adding MFSrMpegAudio 1.0.0.0 (beta)

- Adding MFTMpegAudioDecoder 1.0.0.0 (beta)

- Adding MFAsyncCallback.h and MFBuffer.h


//------------------------------------------------------------------------------------------------------------------
20/05/2013

- Common project files are now in only one diretory
- Enhanced Logging mediatype/guid/clsid/etc...
- New file for MFNode Guid

- Adding MFTAudioMixer 1.0.0.0 (beta)

- Adding MFSrScreenCapture 1.0.0.0 (beta)

- Update MFTVideoMpeg2 1.1.0.1 (beta)

* to be compliant with new common project files

- Update MFKinectSource 1.1.0.0 (beta)

* now relative to Kinect SDK V1.7
* to be compliant with new common project files
* adding GNU General Public License header (oups...)


//------------------------------------------------------------------------------------------------------------------
28/10/2012

- Adding MFKinectSource 1.0.0.0 (beta)

- Update MFTVideoMpeg2 1.0.0.1 (beta)

* remove BaseObject in Mpeg2Decoder.h. Not needed because Mpeg2Decoder is an internal class of MFTMpeg2Decoder
* add (std::nothrow) to all new allocation class
* Configuration for release mode