So if I have a file, its content is like this: This is important: in MPEG2 transport streams, audio and video data is often interleaved in such a way that the video data for PTS t is near to audio data for PTS t+about 1s (in other words, the channels transmit the video data earlier than the corresponding audio data, probably because the decoder latency of video is bigger). This is actually not the usual gapless playback. It should not try to determine which tracks all subsequent parts contain, but should treat them as one big file.
(Currently I do a rough remux first and then remux the mkv In my usecase it should not even finish one media before starting the next one, but treat them as one thing (just like the concat protocol). And for precise remuxing with MKVToolNix I need precise timing infos.
My intended usecase would fall under this very small set of cases: I have a DVR that writes to a FAT32 disc and therefore has to split the file (MPEG2 transport stream) into smaller parts. The ffmpeg concat protocol is extremely limited as it is, it would only work in a very small set of cases, so its probably not a good idea to base any such features on top of it.ĭoes this mean that LAV Filters can't be made to use the concat protocol (or expose it to other applications or however it would be called) and that therefore there is nothing for a player like MPC-HC to use anyway? DirectSound: AMD HDMI Output (AMD High D.īelow the MediaInfo without audio details.Ĭomplete name : E:\\\BDMV\PLAYLIST\00802.mplsįormat profile : Stereo / settings, CABAC : Yes I have tested the one or two differences before posting.
0 kommentar(er)
