[Asterisk-Users] Echo canceller data-points

[Steve Davies]

oops. attachments are blocked :) I'll email it directly to anyone who
provides an email address.

Regards,
Steve

On 3/16/06, Steve Davies <davies147 at gmail.com> wrote:
> Here is the patch file which I use (I manually removed some other
> parts of the patch, so I hope it is okay!) - It should be sufficient
> to get you going.
>
> cd into the zaptel-1.0.9.2 source directory, and
> patch -p1 <zap-patch.txt
>
> Cheers,
> Steve
>
> On 3/15/06, Colin Anderson <ColinA at landmarkmasterbuilder.com> wrote:
> > Is it onerous to backport or is it a case of fiddling around with the
> > makefile? Care to post a backported tar?
> >
> > -----Original Message-----
> > From: Steve Davies [mailto:davies147 at gmail.com]
> > Sent: Wednesday, March 15, 2006 2:47 PM
> > To: asterisk-users at lists.digium.com
> > Subject: [Asterisk-Users] Echo canceller data-points
> >
> >
> > In case this is useful to someone...
> >
> > Initially running * 1.0.7 and the default canceller, about 1 in 20 E1
> > PRI calls still had echo, sometimes quite bad.
> >
> > Updated to * 1.0.9, and backported KB1 from 1.2 HEAD to this build -
> > No noticable change over the previous version, but we ran with it
> > anyway as the small changes in the EC code looked sensible.
> >
> > Recently we backported MG2 from Zaptel 1.2.4 into our * 1.0.9 build,
> > and noticed a significant improvement.
> >
> > I thought I would try the 1.2 trunk/HEAD version of MG2 with the extra
> > knobs and whistles, but found 2 problems. This version trains even a
> > normally clean line in about 10 seconds, rather than instantly, and
> > its CPU usage is through the roof compared to the 1.2.4 version of the
> > code. (FYI I got very similar resuls at all intermediate SVN versions
> > of the MG2 canceller between 1.2.4 and HEAD)
> >
> > My advice: Go with the 1.2.4 MG2 echo canceller, perhaps if you have
> > plenty of spare CPU the newer code will become useful, but I could not
> > cancel even 20 PRI channels using a 1GHz processor on the latest code
> > - I got clicks, buzzing and eventually a dead PRI. With the 1.2.4
> > branch I had 40% CPU free when cancelling 30 channels.
> >
> > Hope this helps.
> >
> > Cheers,
> > Steve
>
>
>