summaryrefslogtreecommitdiff
path: root/dxr3demuxdevice.c
AgeCommit message (Collapse)Author
2010-01-09there is no need to store the type of a bufferChristian Gmeiner
The type is known by the used buffer.
2010-01-09make more use of AccessorsChristian Gmeiner
2010-01-08code style changes... nothing moreChristian Gmeiner
2010-01-08kick m_dxr3Device member from cDxr3DemuxDevice and use singleton insteadChristian Gmeiner
2010-01-07check for max sizes supported by syncbufferChristian Gmeiner
2009-12-10make DemuxPes much simplerChristian Gmeiner
2009-12-09make cDxr3AudioDecoder::Init() privateChristian Gmeiner
There is no need to call Init() as it gets called by is self, if a different audio header is found.
2009-12-09add a way to enable/disable audio via iAudioChristian Gmeiner
As all audio realated stuff uses iAudio I added a way to control if write gets called or not. This depends if it was enabled or not. Also I did changes to the demuxer to call the new method instead of the old stub. I am not happy to have that chain: m_pAudioThread->audio()->setEnabled(false) Need some time to redesign this part.
2009-12-09digital audio works if enabled by vdr startupChristian Gmeiner
2009-12-09make call to cDxr3AudioDecoder::DecodeAc3Dts simplerChristian Gmeiner
2009-12-09isAc3Dts() checks are not neededChristian Gmeiner
2009-12-09a MPEG-1 or MPEG-2 audio stream contains only MPEG audio data, no need for ↵Christian Gmeiner
ac3dts check
2009-12-09change structure of iAudio classChristian Gmeiner
This change introduces some bigger changes in how analog/digital audio will be handled. VDR notifices the plugin via cDxr3Device::SetDigitalAudioDevice, if if the we need to use analog or digital audio. With digital audio I mean digital PCM. Based on the informations get via the pes packet ac3dts will be switched on or off. So the svdrp commands SDO, SAO and SAC3 do not make sense at the moment.
2009-12-08remove IsAudioModeAC3() method - instead make use of iAudioChristian Gmeiner
2009-11-19remove method cDxr3Interface::ReOpenAudio, as it is a pure stubChristian Gmeiner
2009-11-18Revert "make it simpler"Christian Gmeiner
This fixes hopefully some problems with playback. This reverts commit 34176db79c1d6bca5d4f76fc34f80e00565091e4.
2009-11-09some small cleanupsChristian Gmeiner
2009-10-23kick PlayVideoFrame(const uint8_t* pBuf, int length, int times) method as ↵Christian Gmeiner
not needed anymore
2009-10-23play cDxr3PesFrame directlyChristian Gmeiner
2009-10-16current implementation of StillPicture in cDxr3DemuxDevice is wrongChristian Gmeiner
We call 3 times DemuxPes - wow. The big problem is that a still picture can not only contain a pes packet. Also I want to refactor out as much as possible from the demuxdevice as i dont like it and want to remove it in the long term.
2009-10-16remove cDxr3Interface::SingleStep methodChristian Gmeiner
Driver does _NOT_ support SingleStep playmode, so there is no need to call it. The big problem is that the driver does not return an error by this syscall. Have send a pull request to fix this issue in the driver.
2009-08-28make it simplerChristian Gmeiner
- get rid of SetPlayMode() - add method to do flushing of video buffer - dont spam log with spu buffer sizes
2009-08-19merge methods SetHorizontalSize and SetVerticalSize to method setDimensionChristian Gmeiner
2009-06-15Update copyright boilerplates from http://www.gnu.org/licenses/lgpl-2.1.txt ↵Ville Skyttä
and http://www.gnu.org/licenses/gpl-2.0.txt
2009-06-05update emacs code style partChristian Gmeiner
2009-06-05make gcc happier by fixing some warningsChristian Gmeiner
2009-06-02rework singleton classChristian Gmeiner
2009-05-02cleanup header includesChristian Gmeiner
2009-04-24create cDxr3PesFrame on the heapChristian Gmeiner
This commit is a small step into the right direction - use one datastructe from parsing to playback and this datastructe will be cDxr3PesFrame. One of the next things to do is to change the syncbuffer to store cDxr3PesFrames :)
2009-04-20this is a small preperation for later patchesChristian Gmeiner
Reduce the number of method arguments and simply use a pointer to a cDxr3PesFrame.
2009-04-20get rid of offset in cDxr3PesFrameChristian Gmeiner
2009-04-19get rid of m_staticVerticalSize and m_staticHorizontalSize in pes frameChristian Gmeiner
2009-04-18get rid of cDxr3SafeArray and all the exception stuffChristian Gmeiner
2009-04-18get rid of IsValid from cDxr3PesFrameChristian Gmeiner
2009-04-18rename cDxr3PesFrame::ExtractNextFrame to cDxr3PesFrame::parseChristian Gmeiner
2009-04-18get rid of an assumption of older vdr versionChristian Gmeiner
If we have a lookt at comments about Play* mehtods in device.h, we can read the following line: ' ///< Data points to exactly one complete PES packet of the given Length.' As a result of this, there is no need to parse in the data pointer for more then one pes packet.
2009-04-18fix coding style before starting rewriteChristian Gmeiner
2009-04-15rename es to payload in pesframe and coChristian Gmeiner
2009-04-01simplify cDxr3DemuxDevice::setAudioChristian Gmeiner
2009-02-26remove unused code in the demux deviceChristian Gmeiner
The big goal for the next days will be to throw away the whole demuxdevice and add a very small and easy to understand demuxer. The current demuxdevice has a very bad coding style and is quite hard to follow whats going on.
2009-02-21first work to get alsa audio outputChristian Gmeiner
Introduce a iAudio class - will be later base class for oss and alas output -, which handles volume and channel handling.
2008-12-29Convert to UTF-8.scop
2007-04-05Apply patches/vdr-dxr3-ac3.patch, start countdown to 0.2.7.scop
2007-01-28Use cCondWait::SleepMs() instead of usleep().scop
2006-02-11Some commented out code updates (no-op).scop
2005-08-02Update FSF's address.scop
2005-06-22Use VDR's logging facilities, no more dxr3plugin.log.scop
2005-05-28Improve recovery and avoid lockups with bad streams (Jon Burgess).scop
2005-05-28Clean up non-working dead code for VDR < 1.3.11.scop
2005-04-19Mass indentation/whitespace cleanup.scop