andy vt's tools & blog

automating lazy

This Blog


andy vt's blog

QSTranscode -
- Minor GUI changes
- Populate decoder parameters using data from ffmpeg context if the MSDK can't get it from extradata
- Don't feed PTS into the MSDK if we can tell that the timestamps are messed up
- Set PTS = DTS when muxing if PTS is invalid from MSDK
- Make sure the first DTS is valid before using it as an offset
- Change basis of rendered framerate
- Expose CAVLC Off/On
- Add some descriptive error messages for initialization errors
- Retrive and write GFX driver version to output
- Reorder include/lib folders for cleaner inclusion
- Retrive and write IPG model to output
- Only add CodingOptions2 if API version is >= 1.6 (fix encoder init issue with SNB)
- Write MSDK version before initializing components
- Only use AR to calculate width/height
- Don't use ticks_per_frame to calculate input timestamps
- Calculate audio/video start PTS even when not seeking forward so it can be used to adjust A/V sync
- Seek to 0 (unless -ss is used to override) for all files
- Respect that MSDK versions < 1.6 do not use DTS
- Try to adapt to situations where the input timestamps are crap
- Test AV_CODEC_ID_H264 video streams to determine if they are AVC1 to assign the correct parser
- Add a feature to seek to a common start point to attempt to deal with files with A/V that doesn't start at the same time. Use -seekstart to enable.
- Pick the best video stream manually
- Start moving logging from printf to av_log
- Add bfr arg
- Fix ICQ & LA_ICQ rate control methods
- Don't trust the codec context timebase to make sense. Use the avg/r timebase if it doesn't when calculating input timestamps.
- Skip all audio/subtitle processing in benchmark mode
- Fix an issue with -lads validation
- Add -genpts, -rtphint, -empty_moov, -frag_keyframe, -separate_moof options
- Use first audio track PTS to generate PTS when muxing
- Rev to latest Ffmpeg (2.2.3?)
- Better work around files with crazy FPS information
- Set surface frame flags based on decoded frame information when decoding with Ffmpeg
- Sanity check selected audio track
- Use rendered framerate to set input timestamps
- Expose -genpts in GUI
- Fix WMV3 header construction
- Fix issue where unparsed audio codecs wouldn't be decoded properly (e.g. WMAPro)


Published Jun 15 2014, 09:18 AM by babgvant
Filed under:


No Comments
@2008 andy vt
Powered by Community Server (Non-Commercial Edition), by Telligent Systems