[*] Looking for Manitoba toll-free beta testers.

[Bill Reid]

Jim Van Meggelen wrote:
> This is a bit of a gut feeling more than a professional opinion, but the the
> echo is being artificially created (especially considering it has no loss of
> amplitude, as would normally be the case), and perhaps the 'can is deciding
> that it's a lost cause and closing the media path (or interpreting the
> entire returned signal as somehow undesirable). For sure something is making
> that decision, since the call starts with audio, anf then it goes away after
> a while.
> 

Yes, I suspect you are probably right. Since the cell phone is packet not a TDM 
circuit the cancellation algorithms are looking for echo with a much longer 
delay then on a landline. Googling would suggest that there is a whole new breed 
of echo cancellation software and hardware designed for cell phones and IP PBXs. 
They talk about catching echo up to 128ms.

-- Bill