[Asterisk-Dev] Re: DTMF when transferring calls - not handled well?

[Dan Evans]

Asterisk has always "normalized" RFC2833 DTMF input.  Initially, it sent 
1 "DOWN" ("ON") followed by 3 "UP"s ("OFF").  I believe the 3 
terminating packets are recommended in the RFC.  But having only 1 DOWN 
meant that your DTMF would disappear if that one packet was dropped. 
The UP's would just look like events associated with the previous keystroke.

The normaization was changed to 3 DOWN and 3 UP in patch 0003339 (sic), 
to improve robustness.

Ideally, when acting as a proxy, * should just copy the packets through 
(resequencing, etc) instead of normalizing them, so that they still 
approximate the timing of the original source.  There is most often more 
than enough redundancy in the original source, so normalization tends to 
remove it.

But please don't make a change that removes the redundancy.

Dan

Bryan Field-Elliot wrote:
> Either way -- someone please post some working code! At this point I 
> strongly suspect (but cannot totally prove) that the rtp.c 
> ast_rtp_senddigit() function does not work as advertised, for more than 
> one manufacturer's SIP ATA.
> 
> Just eyeballing the code - granted I don't know the RTP specification 
> very well (or RFC2833) - but why is it sending 6 packets per digit (3 
> "ON" followed by 3 "OFF")? Does the spec call for redundancy in the 
> packets, or is this function trying to do some funky future-timing 
> without actually having to wait?
> 
> 
> 
> 
> On Mon, 2005-05-23 at 15:14 +0000, Tony Mountifield wrote:
> 
>>In article <4291D17E.6080506 at arminco.com <mailto:4291D17E.6080506 at arminco.com>>,
>>Vahan Yerkanian <vahan at arminco.com <mailto:vahan at arminco.com>> wrote:
>>> 
>>> I think posting the patch to the list would be benificial too, at least 
>>> for archival purposes.
>>
>>It's better to post such patches to bugs.digium.com, and then possibly
>>inform the list of the bug report's URL.
>>
>>Cheers
>>Tony
>>
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Asterisk-Dev mailing list
> Asterisk-Dev at lists.digium.com
> http://lists.digium.com/mailman/listinfo/asterisk-dev
> To UNSUBSCRIBE or update options visit:
>    http://lists.digium.com/mailman/listinfo/asterisk-dev