Sign in

username:

password:



Not a member?

Search echocancel



Search tips

Subscribe to echocancel



echocancel by Keywords

AEC | Double Talk | ERLE | G.168 | LEC | NLMS

Discussion Groups

Discussion Groups | Echo Cancellation | RE: acoustic echo cancellation performance?

Technical discussions about echo cancellation and adaptive filtering in general.

  

Post a new Thread

RE: acoustic echo cancellation performance? - Shaw, David G David - Sep 12 16:07:00 2003



From experience you should not count on echo cancellation
like you expect from the LEC numbers you may be familiar with.
Measure the microphone noise floor and make sure that your
residual error is not dominated by it.
It is VERY, VERY hard to reach 30dB from just the AEC.
Issues: dynamic range (immediate echo vs local talker levels),
dynamic echo path, noise floor, lack of tail length, noise
floor of the sub-band filters (if used) etc.

Regards,
Dave

-----Original Message-----
From: andrew_xiang [mailto:]
Sent: Friday, September 12, 2003 10:58 AM
To:
Subject: [echocancel] acoustic echo cacellation performance? Anyone implemented a system that get you 30dB of reduction for AEC?

In my system, I only get 4X for voice signal. I will give it a try
with white noise. But seems 30X is far away. But I don't expect users
to utter white noise when they speak.

My calculation:

4X=SUM(abs(microphone))/SUM(abs(error))

4X leaves me a lot of residual.

Are there other tricks in increasing the reduction?

thanks
Andrew _____________________________________






(You need to be a member of echocancel -- send a blank email to echocancel-subscribe@yahoogroups.com )

RE: acoustic echo cancellation performance? - Prabhu T N V R K - Sep 25 6:19:00 2003

Hello all,

How to take care if the echo path has a gain. (no more
erl)

Thanks,
Prabhu.

--- "Shaw, David G (David)" <> wrote:
> From experience you should not count on echo
> cancellation
> like you expect from the LEC numbers you may be
> familiar with.
> Measure the microphone noise floor and make sure
> that your
> residual error is not dominated by it.
> It is VERY, VERY hard to reach 30dB from just the
> AEC.
> Issues: dynamic range (immediate echo vs local
> talker levels),
> dynamic echo path, noise floor, lack of tail length,
> noise
> floor of the sub-band filters (if used) etc.
>
> Regards,
> Dave
>
> -----Original Message-----
> From: andrew_xiang [mailto:]
> Sent: Friday, September 12, 2003 10:58 AM
> To:
> Subject: [echocancel] acoustic echo cacellation
> performance? > Anyone implemented a system that get you 30dB of
> reduction for AEC?
>
> In my system, I only get 4X for voice signal. I will
> give it a try
> with white noise. But seems 30X is far away. But I
> don't expect users
> to utter white noise when they speak.
>
> My calculation:
>
> 4X=SUM(abs(microphone))/SUM(abs(error))
>
> 4X leaves me a lot of residual.
>
> Are there other tricks in increasing the reduction?
>
> thanks
> Andrew > _____________________________________
__________________________________



(You need to be a member of echocancel -- send a blank email to echocancel-subscribe@yahoogroups.com )