[SIPForum-discussion] payload type for G729A and differentflavorsof G729

Herve Jourdain herve.jourdain at mstarsemi.com
Thu Jan 10 06:52:16 UTC 2008


Hi,

 

Unfortunately, yes. On some dual-mode phones, while doing interoperability
testing at Telecom Italia.

They used a "G.729a" MIME type, which is not standard, and negotiated
dynamically. I have the feeling this might have been done on purpose,
though, for "opening less" the environment to devices.

 

About the G.729A, the RFC 3551 is clear on that topic, G.729 or G.729A are
considered the same, as the payloads are bit-compatible - G.729A being an
algorithmic variant of G.729 providing a little less quality but using much
less CPU.

 

So you should use "G729" as the MIME type, and use 18 as PT.

 

Regards,

 

Herve

 

  _____  

From: discussion-bounces at sipforum.org
[mailto:discussion-bounces at sipforum.org] On Behalf Of Jeff Wright
Sent: jeudi 10 janvier 2008 07:02
To: udayabhanu parida; discussion at sipforum.org
Subject: Re: [SIPForum-discussion] payload type for G729A and
differentflavorsof G729

 

All implementations I've seen in the past use static PT 18.  Anyone else
have any other experiences?

Jeffrey Wright
System Test Engineering Manager
Aztek Networks



-----Original Message-----
From: discussion-bounces at sipforum.org on behalf of udayabhanu parida
Sent: Wed 1/9/2008 10:39 PM
To: discussion at sipforum.org
Subject: [SIPForum-discussion] payload type for G729A and different
flavorsof G729

Hi,

I have a confusion in whether to use static payload type i.e. 18 or a
dynamic payload type for G729A while transmitting data over RTP.
The rfc 3551 doesn't specify the PT for G729A. It seems that for G729 or
G729A both the payload types are applicable.
A static payload type of 18 as well as a dynamic payload type can be used.

Please clarify what should be used for the different flavors of G729 codec.

Thanks,
Uday

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sipforum.org/pipermail/discussion/attachments/20080110/20647ad5/attachment-0002.html>


More information about the discussion mailing list