[Asterisk-Users] Re: Echo SIP-T100P-PRI

Rich Adamson radamson at routers.com
Thu Aug 19 15:33:12 MST 2004


> Mike Schwartz wrote:
> >>>I'm experience echo on outgoing calls:
> >>>  Snom 200 ----> Asterisk ----> T100P ----> PRI ----> called party
> >>>
> >>>I am getting echo on the Snom 200 phone. The called party does not
> >>>hear the echo. 
> > 
> > 
> > 
> >>Rule of thumb (i.e. a good starting point): if you hear the echo it is
> >>coming from somewhere else, your equipment mearly makes it noticable. In
> >>this case the echo is probably caused by an analog conversion at the
> >>remote party. This is quite common but for calls with a very short latency
> >>in the call path this is experienced as a reverb and not an echo.
> >>
> > 
> > 
> > Reverb might be a good description. 
> > 
> > If this is the case--that my CLEC is the source of the echo/reverb when they do 
> > a digital to analog conversion, is there anything I can do about it, except try 
> > to cancel the echo? Can I ask them to change any settings? They have a 5E 
> > switch. My collo is right next door--I just cross connect to get my PRI. 
> > 
> > The echo I am experiencing is on local calls.
> > 
> > My * server is a dual-processor XEON (2.4 Ghz) Supermicro server. Its 
> > relatively fast. I will reduce the number of taps, and I was also thinking of 
> > decreasing the sample size on the phone from 20ms to 10ms.
> 
> It seems that echo is not related to processor speed but type of 
> hardware inside. Somebody post, one month ago, to have echo with a 
> Supermicro server then moved to a slower machine, a DL380, and the echo 
> disappear.

When that discussion was going on a few weeks ago, the echo issue
seemed to have been narrowed down to two possibiliites; 1) interrupt
service latency, or, 2) PCI bus latencies. Processor speed does not
seem to be a driving factor as noted above.

I've not heard anyone (as yet) come up with the tools or process for
actually identifying the root-cause. Would be nice for those of us
that aren't programmers.






More information about the asterisk-users mailing list