[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MMUSIC] WGLC on draft-ietf-mmusic-rfc4756bis-03

Hi Thomas,

> -----Original Message-----
> From: Thomas Schierl [mailto:schierl at hhi.fhg.de]
> Sent: Monday, October 19, 2009 8:16 AM
> To: Jean-Francois Mule
> Cc: mmusic at ietf.org; Joerg Ott; draft-ietf-mmusic-rfc4756bis at tools.ietf.org
> Subject: Re: [MMUSIC] WGLC on draft-ietf-mmusic-rfc4756bis-03
> 
> Dear Ali,
> 
> I am not sure if the SDP signaling is covering all cases. What is about
> the case of having a SVC or any other layered media, where one session
> SB contains the base layer and another session SE contains the
> enhancement layer. Let us assume session SB is protected by  repair flow
> RB. Furthermore, session SB and SE are jointly protected by repair flow
> RB+E. Additionally, RB+E is additive to RB. The only way expressing this
> with the current draft seems to be:
> 
>         a=group:FEC-XR SB RB
>         a=group:FEC-XR SB SE RB RB+E

If RB+E and RB were not additive, we would have:

a=group:FEC-XR SB RB
a=group:FEC-XR SB SE RB+E

But, now you are saying that they are additive, I believe we should have:
 
a=group:FEC-XR SB RB RB+E
a=group:FEC-XR SB SE RB+E

Because RB+E is additive to RB, that means it is protecting SB as well. Here I assumed RB exclusively protected SB. If this assumption does not hold, see below.

> I would expect that the grouping can signal that RB can be used
> separately for SB, but RB and RB+E can be also jointly used for the
> combination of SB and SE. Would the above grouping be in line with the
> current draft? I wonder if this may be somehow contrary to definition in

Now, this sounds different than the first scenario you described above, IMO. And, this leads to the original SDP you gave:

a=group:FEC-XR SB RB
a=group:FEC-XR SB SE RB RB+E

> section 4.1:
> 
> "Repair flows that are in different FEC groups are non-
>    additive.".

Why would it be contrary? Here, RB and RB+E are in the same line and that is enough to state their additivity. But, in this scenario, their joint decoding helps the group of SB+SE. The above SDP says if the source consists of SB only, then just use RB only since RB+E would not help with decoding.
 
> If the signaling in the example above is in line with the intention of
> the draft, I would make the following modifications:
> 
> - You may add an "...exclusively in different FEC groups..." to the
> sentence of section 4.1 above.

We can add this to make it stronger. But, I believe if two repair flows are additive in a group, it does not mean that they will also be additive in any other group where one of the source flows they protect is listed.

Agree?

> - Further, I would add the following text to section 4.1 to make it clearer:
> 
> "In order to express multiple relations between source and repair flows,
> source and repair flows may appear in more than one FEC group."

I believe this was clear, but it is a good suggestion.
 
> I think the same should be valid for the FEC-SSRC grouping.

Sure.

Thanks for the review. Let me know if you agree with the above.

Cheers, acbegen.

 
> Best regards,
> Thomas
> 
> 
> --
> Thomas Schierl
> --------------
> Fraunhofer HHI
> 
> 
> 
> Jean-Francois Mule wrote:
> > This is to start a 2-week working group last call on draft
> > draft-ietf-mmusic-rfc4756bis-03,
> > http://tools.ietf.org/html/draft-ietf-mmusic-rfc4756bis-03.  The comment
> > period will end on Wednesday October 28.
> >
> > Please send your comments to the mmusic list and the authors.
> >
> > Thank you,
> > Jean-Francois.
> >
> >
> > _______________________________________________
> > mmusic mailing list
> > mmusic at ietf.org
> > https://www.ietf.org/mailman/listinfo/mmusic
> >
> >
> >
> 
> 
>