[asterisk-dev] New codec support in chan_skinny

Dan Austin Dan_Austin at Phoenix.com
Tue Sep 26 12:42:45 MST 2006


I found that.  The major differences I see between chan_sip
and chan_skinny is how preferences are honored.

I did get chan_skinny to work with G729 by disallowing 
ulaw and alaw, but that should not be required.
If this is the allow directive:
allow = g729,ulaw then a new call should try to use g729

It looked like ast_codec_choose would be better than
ast_best_codec, as it appeared to honor the device/line
preferences, but I still get ulaw when g729 has a higher
preference.

I'll keep digging.

Dan
 

-----Original Message-----
From: asterisk-dev-bounces at lists.digium.com
[mailto:asterisk-dev-bounces at lists.digium.com] On Behalf Of Jason Parker
Sent: Tuesday, September 26, 2006 12:21 PM
To: Asterisk Developers Mailing List
Subject: Re: [asterisk-dev] New codec support in chan_skinny

----- Dan Austin <Dan_Austin at Phoenix.com> wrote:
> I'm testing the new codec support that qwell commited to
> chan_skinny this morning, and it looks good for the most
> part except for g729.
> 
> Any time I add g729 to an allow list, it is ignored.  If
> I only allow g729 it is still enabling the ulaw and alaw
> codecs based on the phone offering them.
> 
> Packetization is working on ulaw/alaw, but I cannot trigger
> a G729 so far.  Any thoughts before I dig into the codec
> negotiation code in chan_skinny?
> 
> Dan

I did test G.729, and it worked for me..

It should be defaulting to ulaw and alaw being allowed.  If you want to
use something besides those, you'll need to disallow them
(ast_best_codec will choose them over anything else, iirc).  I do
believe this is the same way any of the other channel drivers work -
where it has default codecs that are allowed.

-- 
Jason Parker
Digium

_______________________________________________
--Bandwidth and Colocation provided by Easynews.com --

asterisk-dev mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-dev


More information about the asterisk-dev mailing list