Re: Echo Cancelation Algorithm Details and Tuning

Jozsef Vass <[email protected]> Wed, 3 Oct 2012 20:07:00 -0700
Newsgroups gmane.comp.audio.compression.speex.devel
Message-ID <CACNoOsFPn356QnRVo+=PH7AAN2B-pncmeOk3KwZbcHQ6qZVuZw@mail.gmail.com>
I suggest to take a look at echo canceller part of the webrtc project.
It is the echo cancellation developed by gips and open sourced by
google.

Jozsef

On Wed, Oct 3, 2012 at 2:14 PM, Prateek Bansal <[email protected]> wrote:
> Hi,
>
>
>
> I am just starting up on SPEEX AEC algorithm and have couple of questions
> around echo cancelation algorithm used in SPEEX.
>
>
>
> 1)     Usually Echo Cancelation Algorithm has support for number of
> components –
>
> -       Non-Linear Processor (NLP)
>
> -       Automatic Microphone Gain Control (AGC)
>
> -       Transducer Equalization (EQ)
>
> -        Dynamic Range Compression (DRC)
>
> -       Ambient Noise Reduction (ANR)
>
> -        Bi-directional Comfort Noise Generator (CNG)
>
> -       Double-talk/Coherence/Howling detection
>
> My question is whether the algorithm supports all the above functionalities
> ? Looking at speex_preprocess.h file, I find there are controls available
> for AGC, Noise Suppression, Echo Suppression during single talk and double
> talk but not clear if there are more options available.
>
> 2)     From tuning standpoint, are there references available on how we can
> tune SPEEX AEC algorithm to our needs ?
>
>
>
> Thanks in advance.
>
>
>
> Prateek
>
>
>
>
> _______________________________________________
> Speex-dev mailing list
> [email protected]
> http://lists.xiph.org/mailman/listinfo/speex-dev
>