Other possible Echo Cause

Al Johnson openmoko at mazikeen.demon.co.uk
Mon Dec 15 12:28:30 CET 2008

On Monday 15 December 2008, Paul wrote:
> I have been thinking about the possible causes for the echo bug.  I
> have a feeling that it's actually feedback and not an echo.  I think
> what's happening is the sound is being picked up by the mic and being
> sent to the remote phone for an extra round of fun.

I hadn't realised there was any doubt that this was the mechanism. 

> Does anyone know how I would start testing this?  In particular,  I
> would like to try the following.
> 1. Disable the mic totally and raise the handset volume.  I know the
> handset volume is set by Control.4 values in gsmhandset.state file.  I
> would make some phone calls and have the remote party talk to see if
> an echo occurs.  If not echo occurs, well then we know the mic is part
> of the cause.

If you mute either the mic or the earpiece there is no echo. If you use the 
wired headset there is no echo. The level of echo is affected by both the mic 
gain and the earpiece level. Using the speakerphone gives much more echo, 
which is expected because the volume is higher and the speaker is close to 
the mic.

> I also know that our telephone systems do not transmit the whole
> audible frequency range.  Is there a way to clip the audio from the
> mic for phone calls?  I think this would translate to
> reduce/elimination of buzz.

It's possible, but the buzz is within the voice frequency range. If we could 
characterise the buzz sufficiently it may be possible to remove it by 
applying a suitable digital filter in the CPU, but a hardware fix is 
certainly preferable.

More information about the support mailing list