Re: [MMUSIC] Displaying side by side video streams draft-jennings-mmusic-adjacent-grouping

[Cullen Jennings <[e-mail]>]

This was largely motivated from wall of screens for things like video advertising, arrays of airport monitors with TV updates, display on walls of surveillance screens and so on. I had been waiting to see where the telepresence work will go but I think that is a somewhat different problem. For example, the telepresence people would like to be able to do things like understand where camera really are relative to viewers and the field of view of the cameras so that they can display a speaker uh, "life size". That requires a far more fine graining approach to position that saying this stream goes on the left monitor and this one the right. I think there is a reasonable good document on the some of the telepresence uses cases. I used the word telepresence in this document in the case where a telepresence system was interoperating with a 2 screen video conferencing system. I just used telepresence to have a different word for the systems at each end to make it less confusing to to right but that was somewhat a mistake. I changed the latest one to try and be clear this is not taking on the telepresence problem space and removed telepresence from the document. 

However, if you got two side by side camera, and I've got two side by side screens, this can get the right camera on the right screen, so even though this does not solve all the things for a telepresence system, it seems like a tool that would work fine for solving some of the multiscreen video conferencing issues. 


On Sep 23, 2010, at 12:59 AM, Ingemar Johansson S wrote:

> Hi Cullen
> 
> I guess your input is spurred by the emerging interest in Telepresence.
> What I would believe is good is to work on a document that outlines the requirements. For instance I miss a general view on how e.g an MCU should work here. From that work I believe it will be easier to define extensions to SDP.
> Question is if this work should wait for the Telepresence WG (if it gets an OK) or if work should start already in MMUSIC  ?
> 
> FYI we earlier submitted a draft around grouping of simultaneous streams http://tools.ietf.org/id/draft-johansson-mmusic-grouping-simultaneous-00.txt
> The use case here is that an endpoint sents two represenations of the same media (e.g high-res and low-res video), an MCU then picks the media based on for instance speech activiy. This is particularly useful for the case that opposite endpoints have only one screen.
> Since the submission we did not have time for further work in this area but it is possible that we pick up this again.
> 
> /Ingemar
> 
> 
> >
> > Message: 1
> > Date: Tue, 21 Sep 2010 13:49:11 -0600
> > From: Cullen Jennings <[e-mail]>
> > Subject: [MMUSIC] Displaying side by side video streams
> >       draft-jennings-mmusic-adjacent-grouping
> > To: MMUSIC WG <[e-mail]>
> > Message-ID: <[e-mail]>
> > Content-Type: text/plain; charset=us-ascii
> >
> >
> > I just submitted a draft on using the SDP grouping framework
> > to indicate when multiple media streams should be rendered
> > side by side. It a pretty trivial draft but I'd love to get
> > some comments on it.
> >
> > http://www.ietf.org/id/draft-jennings-mmusic-adjacent-grouping-00.txt
> >
> > Thanks, Cullen
> >
> >
> >
> >
>