RE: [AVT] RFC3558 header-free packet format question

["Adam Li" <adamli at icsl.ucla.edu>]

Sorry to jump in the discussion a little late. I got a bit of email
build up in the last couple of days.

For Qiaobing's question, I think for EVRC the codec outputs a mixture of
frames of all different sizes even when a particular mode is selected.
My understanding is that the chance that codec outputs consecutive
frames of the same size for normal input is quite small.

Even for such rare scenarios, the one-frame-per-packet header-free
packet is as efficient as putting multiple frames in one packet, as long
as the header compression is used. 

Adam

-----Original Message-----
From: Qiaobing Xie [mailto:Qiaobing.Xie at motorola.com] 
Sent: Wednesday, September 22, 2004 8:11 PM
To: Lars-Erik Jonsson (LU/EAB)
Cc: AVT List; adamli at icsl.ucla.edu
Subject: Re: [AVT] RFC3558 header-free packet format question

Hi, Lars-Erik,

Lars-Erik Jonsson (LU/EAB) wrote:

> With frame sizes of 16, 40, 80 and 171 bits it would not be
> possible to distinguish one 80 bit frame from two 40 bit frames. 

Of cause, the sender and receiver have to agree at the session set-up to
use only a single 
rate (say, 1/2 rate only). In such a case, the frame size becomes fixed
and one can rely on 
the length of the packet to unbundle the packet with little trouble.

> In a general case, e.g. were one wants to send multiple frames
> in one packet, I believe the header-free format should not be
> the one used.

I am thinking that when the frames are of the same size, I wonder if
allowing header-free 
format to carry them is a more efficient alternative...

regards,
-Qiaobing





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