I've been testing the AEC and have experienced it working on one sound card,
and not working when using different cards for capture/playback so I know
it's a real problem. And yes I know the documentation says it won't
work "regardless of what you may think." I'm the inquisitive
type; I'm looking to understand just what exactly is happening here.
One discussion in the archives: John Ridges says "The AEC can only cancel
linear changes to the audio path, and so the AEC never converges." I guess
I don't understand the "linear changes" part of that statement.
Seems to me that the "perfect" echo cancel position could basically
land "between" samples, and given that the AEC will adapt to changes
over time anyway, I guess I just don't understand why it matters if the
clocks drift. That is, why doesn't the "adaptation" just drift
too? Hoping someone could explain in layman's terms... Thanks, Mark.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/speex-dev/attachments/20091104/4c17d2fb/attachment.htm
