I'm (slowly) learning some of the subtle differences between the different types of streams. Basically, there's the sync issue that about 3/4 of the utilies out there do as a single, static offset. If the A/V sync changes throughout the stream, sync dies.... that's why many/most people don't have issues with most recordings. Once in awhile, however, they hiccup and screw up. I've got a clip that plays fine from mplayer as a direct capture. With dvb-mplex, mplex, or any strip A/V and reassemble, the sync is broken. With replex, it corrects the PTS offsets as it remuxes.Yes. I simply want a way to do demux, then reassemble the two streams into a single file that I could then directly use DVDAuthor on to make what would then be burned straight to DVD. I am uninterested in having the raw capture file split into audio and video into two separate files. I want a finished product that will not have that insufferable desync problem anymore. btw...Why is it that the raw capture file gives DVDAuthor heartburn? I fail to understand why burning the capture directly to DVD is such a problem. I know my woeful lack of knowledge of the nuts and bolts of digital A/V is showing here.
one... hopefully one with cutting.[...]
The most likely reason DVDAuthor doesn't like them directly is that it doesn't have DVD NAV packets. There can be lots of other streams in a single MPEG2 stream. One thing that's *required* on a DVD MPEG2 is a DVD NAV packet every so often. Also, DVDs like to have a single presentation unit with lots of PES packets. A RAW ivtv capture has *lots* of presentation units, with about 20-50 packets each. Consider these different muxings of the same two A and V files:
'mplex -f 8' produces (with tcscan -i out.mpg):
tcscan -i test.vobfound first packet header at stream offset 0x0
found unknown stream [0xbb]
found navigation pack [0xbf]
found ISO/IEC 13818-2 or 11172-2 MPEG video stream [0xe0]
sequence: 704x480 4:3, 29.97 fps, 5000 kbps, VBV 224 kB , Custom Intra Matrix , Custom Non-Intra Matrix
found ISO/IEC 13818-3 or 11172-3 MPEG audio track 0 [0xc0]
found padding stream [0xbe]
found program end code [0xb9]
------------- presentation unit [0] ---------------
stream id [0xbb] 2915
stream id [0xbe] 2924
stream id [0xbf] 5830
stream id [0xc0] 17336
stream id [0xe0] 361878
390883 packetized elementary stream(s) PES packets found
presentation unit PU [0] contains 2915 MPEG video sequence(s)
Average Bitrate is 579. Min Bitrate is 5000, max is 5000 (CBR)
'mplex -f 3' produces (with tcscan -i out.mpg): <snip> ------------- presentation unit [381] --------------- stream id [0xc0] 2 stream id [0xe0] 49 51 packetized elementary stream(s) PES packets found presentation unit PU [381] contains 3 MPEG video sequence(s) Average Bitrate is 8333. Min Bitrate is 5000, max is 5000 (CBR) --------------------------------------------------- found ISO/IEC 13818-2 or 11172-2 MPEG video stream [0xe0] found ISO/IEC 13818-3 or 11172-3 MPEG audio track 0 [0xc0] ------------- presentation unit [382] --------------- stream id [0xc0] 1 stream id [0xe0] 32 33 packetized elementary stream(s) PES packets found presentation unit PU [382] contains 3 MPEG video sequence(s) Average Bitrate is 8333. Min Bitrate is 5000, max is 5000 (CBR) <snip>
'mplex -f 9' produces (with tcscan -i out.mpg):
tcscan -i test.mpgfound first packet header at stream offset 0x0
found ISO/IEC 13818-2 or 11172-2 MPEG video stream [0xe0]
sequence: 704x480 4:3, 29.97 fps, 5000 kbps, VBV 224 kB , Custom Intra Matrix , Custom Non-Intra Matrix
found ISO/IEC 13818-3 or 11172-3 MPEG audio track 0 [0xc0]
found padding stream [0xbe]
found program end code [0xb9]
------------- presentation unit [0] ---------------
stream id [0xbe] 2924
stream id [0xc0] 17336
stream id [0xe0] 361878
382138 packetized elementary stream(s) PES packets found
presentation unit PU [0] contains 2915 MPEG video sequence(s)
Average Bitrate is 579. Min Bitrate is 5000, max is 5000 (CBR)
The -f 3 is how the output of an ivtv capture looks. The -f 9 is a "DVD without the VOBU sectors"... note the lack of 0xbb and 0xbf packets in the stream like the -f 8 type has.
Basically, a raw capture from ivtv will require a replex to insert the empty DVD sectors, and hopefully adjust any PTS offsets. If only it would cut frames as well, it'd be the ticket.
-Cory
************************************************************************* * Cory Papenfuss * * Electrical Engineering candidate Ph.D. graduate student * * Virginia Polytechnic Institute and State University * *************************************************************************
------------------------------------------------------- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag-&-drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl _______________________________________________ ivtv-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ivtv-devel
