Thanks so much for the response.
I’ m using the G722 codec for linphone. Does this
affect whether the Speex echo canceller is used and ecstate is
saved? Should I be choosing a Speex audio codec to gain these
benefits?
I don’t believe I’ve downloaded your speex
repository. Is it easy to install and ensure Linphone uses it on
the Pi? Note I don’t have the .linphone.ecstate file saved
after each call.
Hi,
The speex echo canceller is quite long to converge. It's the
reason why
we have our own repository for speex with changes to save and
restore
the echo canceller state.
To check if this is enabled in your case, look for a
.linphone.ecstate
file next to your linphonerc file after a call. If it's there
the echo
canceller should converge quite fast.
Otherwise the Linphone you are running may be using the
"standard" speex
library and there's not much to be done.
Cheers,
Ghislain
Le 02/07/2015 22:13, Mark Eastwood a ?crit :
Hi,
I?m using two RPi to
communicate as speakerphones using the Linphone SIP clients
and despite tuning the echo parameters exhaustively I still
get quite a lot of echo for the first several seconds of a
call - I?m driving the speaker quite loud and also want
maximum microphone sensitivity (Alsamixer levels set around
82 each) which isn?t ideal for reducing echo. I?m
struggling to make further improvements using the
.linphonerc file parameters and wondered if anyone has
additional suggestions for how I might reduce this initial
echo. Are there additional parameters I can play with
somewhere?
Many thanks for any help,
Mark
_______________________________________________
Linphone-developers mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/linphone-developers