RE: [AVT] RE: [Sip] RE: [Sip-implementors] media description in SDP is for sender or receiver?

["Ling, Fan" <fanl at qualcomm.com>]

Magnus,
As you pointed out, inband transport of VOL headers may be one solution.
We also have explored this possibility. However, due to possible loss of
RTP packets, this may cause corrupted images for one whole GOP, which
can be several seconds long. 
It looks like the current standards don't have a satisfactory solution
to our problem. We are now looking for proprietary methods.
Thanks for your wonderful comments and advices.
Fan

-----Original Message-----
From: Magnus Westerlund [mailto:magnus.westerlund at ericsson.com] 
Sent: Monday, November 07, 2005 5:45 AM
To: Ling, Fan
Cc: sip at ietf.org; sip-implementors at cs.columbia.edu; someshss at yahoo.com;
IETF AVT WG
Subject: Re: [AVT] RE: [Sip] RE: [Sip-implementors] media description in
SDP is for sender or receiver?

Hi Ling,

See inline.

Ling, Fan wrote:
> Magnus,
> You are dead on on the issues we are concerned about. Initially, our
> understanding is the same as yours. However, when we proceed from
> one-to-one video telephony application to multiparty video conference
> application, the same understanding brings in some trouble.
> Let's assume "MPEG4 simple profile level 1" is the negotiated codec. 
> In one-to-one video telephony, both ends can use different video
sending
> parameters, that is, "config=" or VOL parameters, both ends decoder
just
> configure itself to the incoming video bitstream's setting. In
> multiparty conference, this means any one terminal has to configure
its
> multiple decoders according to different incoming bitstreams. This
makes
> the system a lot more complicated.

I don't see that you can avoid this. You anyway need to save the per 
stream decoder state, thus all active participants will result in state 
on the receiver side.

However isn't the real issue in multi-party that in many of the setup 
scenarios you are not able to describe in the answer more than a single 
configuration. Thus you are forced to use the inband transport that is 
not as robust.

> That's why we are coming back to the standards, to see whether there
is
> a rule that dictates that not only the codec type needs to be the
same,
> the video bitstream parameters (VOL) also needs to be the same for
both
> encoder and decoder.

To my knowledge this is not possible. However input from video coding 
experts are solicited.

Cheers

Magnus Westerlund

Multimedia Technologies, Ericsson Research EAB/TVA/A
----------------------------------------------------------------------
Ericsson AB                | Phone +46 8 4048287
Torshamsgatan 23           | Fax   +46 8 7575550
S-164 80 Stockholm, Sweden | mailto: magnus.westerlund at ericsson.com


_______________________________________________
Audio/Video Transport Working Group
avt at ietf.org
https://www1.ietf.org/mailman/listinfo/avt