Re: [AVTCORE] Duplication semantics drafts

["Ali C. Begen (abegen)" <[e-mail]>]

> -----Original Message-----
> From: Magnus Westerlund [mailto:[e-mail]]
> Sent: Tuesday, October 11, 2011 5:42 AM
> To: Ali C. Begen (abegen)
> Cc: mmusic (E-mail); [e-mail]
> Subject: Re: [AVTCORE] Duplication semantics drafts
> 
> On 2011-10-10 20:17, Ali C. Begen (abegen) wrote:
> > I revised the two drafts per the feedback in Quebec.
> >
> > If there are further comments, we would like to hear about them. As
> > of now, there are no open issues from our standpoint except possibly
> > defining a new RTCP XR report, which can be done later. Thus, we
> > would like these two drafts to be considered and adopted by the
> > MMUSIC wg.
> >
> > https://datatracker.ietf.org/doc/draft-begen-mmusic-redundancy-grouping/
> >
> > https://datatracker.ietf.org/doc/draft-begen-mmusic-temporal-interleaving/
> >
> > CC'ing AVT(core) since the topic is relevant but please post your
> > comments to mmusic where the drafts belong to.
> >
> 
> Hi,
> 
> (as individual)
> 
> I have looked at the redundancy-grouping draft and still think it fails
> as RTP specifications for how duplication should be done in RTP, either
> using SSRC multiplexing with two different SSRCs in the same RTP session
> or for the alternative how having two different RTP sessions. This first
> appears to be sufficiently specified, where my only comment is the
> dependency on signalling to determine that this grouping is in place. I
> do propose that an RTP/RTCP based complement is derived. I would
> personally suggest that my own proposal for a new SDES item SRCNAME is
> used but that should clearly be discussed.

Right, the ssrc muxing case is straightforward and easy to implement. Everything works fine without the need for defining new stuff except the grouping semantics. I am not sure why you think there is a problem in the signaling.

If duplication will be done across sessions (your second case), then it gets complicated. And as I understand your claim is regular CNAME falls short to define this relation. Is that correct? And your proposal is SRCNAME, right?
 
> For the case where multiple RTP sessions are used, it is still not clear
> how a receiver determine which SSRC in each RTP session that are
> related. And I would like to point out that CNAME is not sufficient.
> 
> I still think we should consider splitting this into one AVTCORE
> document discussing the RTP aspects of duplication, and another MMUSIC
> document defining the grouping semantics.

This sounds like a good plan to me.
-acbegen
 
> Cheers
> 
> Magnus Westerlund
> 
> ----------------------------------------------------------------------
> Multimedia Technologies, Ericsson Research EAB/TVM
> ----------------------------------------------------------------------
> Ericsson AB                | Phone  [phone]
> Färögatan 6                | Mobile [phone]
> SE-164 80 Stockholm, Sweden| mailto: [e-mail]
> ----------------------------------------------------------------------