[Asterisk-Dev] new jitterbuffer in 1.2?

Steve Kann stevek at stevek.com
Wed Feb 23 16:16:08 MST 2005


Mike Taht wrote:

>No, I'm not terminating to zap. The speex connection I tried was
>
>Grandstream -> Asterisk/RH9 -> IAX Internet to China -> Asterisk FC3
>-> Grandstream
>
>It's probable to me that I simply ran into a difference between speex
>on a redhat 9 system and a fedora core 3 system. I will try some other
>codecs with the PLC code enabled between these systems tonight and
>tomorrow.
>  
>

In this path, you ideally just want the jitterbuffer turned off on both 
the * boxes [this should, in theory, happen automatically], and rely on 
the VoIP _endpoints_ to handle jitterbuffering and PLC.

I don't know how well the Grandstreams do this, but they probably have 
DSP code licensed to do this, etc.

The latest jitterbuffer patch tries to disable itself in this case 
(where you're bridged to another VoIP channel), but you can turn this 
off at compile-time This is all discussed in the mantis bug, BTW.


-SteveK




More information about the asterisk-dev mailing list