[asterisk-dev] fax/modem detection

[Matthew Rubenstein]

	I suppose that actually using the CED detection that the new dsp.c
provides does require a patch to the channel driver, or an app wrapping
it, so it can interface to a dialplan (or AGI, or Manager) API.


On Mon, 2007-12-03 at 18:51 +0300, Dmitry Andrianov wrote:
> No problem. I will create new dsp.c with CED detection for trunk in 2-3 days.
> But how are you going beta-test it when neither channel driver nor application patched to reacts to the CED tone? :)
> 
> -----Original Message-----
> From: Matthew Rubenstein [mailto:email at mattruby.com]
> Sent: Monday, December 03, 2007 6:27 PM
> To: Dmitry Andrianov
> Cc: Asterisk -Dev
> Subject: Re: [asterisk-dev] fax/modem detection
> 
>         Why not release the detection code in whichever version you can finish
> first, let us beta test it, and then use the feedback to decide whether
> to package it in the channel driver or an app, or both? Accommodating
> channels without the patch to use the app instead, maybe with a switch
> the channel driver can read to skip fax detection so an app can run late
> detection when that's appropriate.
> 
> 
> On Mon, 2007-12-03 at 04:12 -0600, asterisk-dev-request at lists.digium.com
> wrote:
> > Date: Mon, 3 Dec 2007 04:06:32 +0300
> > From: Dmitry Andrianov <dimas at dataart.com>
> > Subject: Re: [asterisk-dev] fax/modem detection
> > To: Asterisk Developers Mailing List <asterisk-dev at lists.digium.com>
> > Message-ID:
> >         <0FDEDCD05BA26E42948F997650F99CDA0396CF70 at e3.universe.dart.spb>
> > Content-Type: text/plain; charset="us-ascii"
> >
> > Asterisk DSP code does only detect CNG tones and does not detect CED
> > tones. I'm currently working on modifications of DSP code to detect
> > CED tones but it is only one part of the puzzle. The other piece is
> > how to react on tone detected by DSP code - either there should be an
> > application (I'm not sure BackgroundDetect is enough) or channel
> > drivers should jump to specific extensions when tone is detected like
> > chan_zap does. I personally like the second approach more but I still
> > haven't came to final conclusion.
> --
> 
> (C) Matthew Rubenstein
> 
-- 

(C) Matthew Rubenstein