Re: [MMUSIC] draft-jennings-mmusic-adjacent-grouping

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

> -----Original Message-----
> From: Paul Kyzivat (pkyzivat)
> Sent: Thursday, March 24, 2011 4:01 AM
> To: Cullen Jennings (fluffy); Ali C. Begen (abegen)
> Cc: IETF MMUSIC WG
> Subject: Re: draft-jennings-mmusic-adjacent-grouping
> 
> (Just catching up on my reading...)
> 
> One observation about this draft:
> 
> The grouping framework permits multiple groups to be formed with the
> same semantics. (E.g. you could define two audio/video groups, each to
> be lip synchronized.)
> 
> Analogously, one could presumably use the ADJ grouping semantic to form
> two independent groups of video streams.
> 
> OTOH, you might want to form two different groupings of m-lines that
> both map onto the same grid of devices. (For instance, a grouping of
> video m-lines and a corresponding grouping of audio m-lines.)

Yes. I just replied to Christer's email, he brought up the same thing. We realized this open issue after the revision.
 
> But the a=media-grid-dims attribute doesn't provide any way to identify
> which group it applies to if there is more than one.

So it is a session-level attribute but this still does not solve the audio-video overlay problem (suppose I have 3x2 screens - 6 video streams and 6x2 audio (stereo) streams).

If such an overlay is a desired feature, a general solution needs to be identified.
 
> At the very least, the issue of what to do when there are multiple
> adjacency groupings ought to be discussed.

-acbegen
 
> 	Thanks,
> 	Paul