[vlc-devel] [PATCH] Merge VoD module into RTP stream output module

Rémi Denis-Courmont remi at remlab.net
Sat Oct 30 21:12:47 CEST 2010


Le samedi 30 octobre 2010 21:54:23 Pierre Ynard, vous avez écrit :
> > Le samedi 30 octobre 2010 09:00:29 Pierre Ynard, vous avez écrit :
> > > > Why do you need to a specific PTS?
> > > 
> > > The RTSP headers already give the initial RTP timestamps to the
> > > client, before the instance and the rtp sout are started. So I need
> > > to use a common reference, and I need to pass it somewhere.
> > 
> > Might as well create a value that can be recomputed from the infos
> > that the vod stream already has. The use of mdate() was purely
> > arbitrary.
> 
> Yeah, I guess the timestamp can be derived from the p_media address
> and/or the session identifier. Any security issue with that?

What does security have to do with this? We use a pseudo-random start value to 
maximize the chance that it's completely different from an earlier session. 
This is important for the sequence number, not so important for the timestamp.

> > > We could also implement streaming several payload types in the same
> > > RTP session,
> > 
> > Yes, but that's rarely useful. First, the design of RTP implicitly
> > limits this to payload types with equal timestamp frequency. Otherwise
> > delay jitter computations and synchronization would go crazy. Second,
> > I have to expect many receivers will fail miserably, including current
> > VLC.
> 
> RFC 3550 §5.2 suggests using different SSRC for each payload type in the
> same RTP session to avoid these problems.

That's for multiple sender using the same RTP session, i.e. conferencing. 
That's not for multiple (synchronized) elementary streams from the same 
sender.

I expect most RTP receiver only support one SSRC at a time. In fact, VLC can 
do more, but it defaults to 1.

-- 
Rémi Denis-Courmont
http://www.remlab.net/
http://fi.linkedin.com/in/remidenis



More information about the vlc-devel mailing list