>Thank you very much for your reply. 

Np

>> The timecode needs to be at the beginning of the packet. The demuxer
>> (mpegts) will, when opening the stream "guess" the packet size. Based on
>> that, it will resync itself every packet and find the true start of the
>> packet (usually by skipping those 4 bytes)

>Does this means, that after a sync byte (0x47) starts the timecode byte?
And yes how does it looks like? I can't get infos from the iso13818-1
document (maybe my version is to old).

Before. 4 byte timestamp header, then packet (starting with sync). You can't
break the elem. packet.

See also 

http://www.blu-raydisc.com/Assets/Downloadablefile/2b_bdrom_audiovisualappli
cation_0305-12955-15269.pdf

page 16

and

http://www.blu-raydisc.com/Assets/Downloadablefile/BD-RE_Part3_V2.1_WhitePap
er_080406-15271.pdf 

(somewhere 1/2 way)

As far as I can see, the timestamps are clock based on the recording system
used. Not sure on that (I never used them, I maintain my own queues in a
decoder).

FYI: the relavant code is @ read_packet() in mpegts.c (around line 1454)

FYI2: Also note that the AVC demuxer is broken in ffmpeg in ts streams (for
about 2 years now). I have a patch for that if you are interested.

Erik

>regards,
>Sven
_______________________________________________
libav-user mailing list
[email protected]
https://lists.mplayerhq.hu/mailman/listinfo/libav-user

_______________________________________________
libav-user mailing list
[email protected]
https://lists.mplayerhq.hu/mailman/listinfo/libav-user

Reply via email to