> -----Original Message----- > From: Andrew Kohlsmith [mailto:[EMAIL PROTECTED] > Sent: February 19, 2006 6:45 AM > To: [email protected] > Subject: Re: [on-asterisk] Persistent DTMF Issue > > On Saturday 18 February 2006 23:52, Jim Van Meggelen wrote: > > DTMF and SIP simply do not seem to get along. > > ?? I have *never* had DTMF issues with SIP. Are you > following the golden rule? "If ye have the gold to be able > to spend the bandwidth for ulaw, ye can use inband dtmf! > Otherwise, use ye RFC2833 lest ye wish to pull thy hair out > in frustration!"
Tried that. The problem with InBand is that Comedian Mail does not support it (i.e., set it on your phones and you can't login to voicemail). The problem with RFC2833 is that the duration of the digit is generally too short, and thus the tones sent out are of such a duration that they are not accurately interpreted by the far end. The Polycom phones do not let you control this as much as one would like (or, rather, it doesn't seem to make any difference). Specifically, the duration of the tone should be directly related to how long the user holds down the button. Instead, it waits until the button is released, and then sends a tone of extremely short duration - regardless of how long the button was held down for. I have not managed to stay awake through all of RFC-2833, so I cannot say with authority, but I suspect that the mechanism for determining what device defines the duration of the generated tone is either not specified, or it is not adhered to. So, even though I try and ensure that the tones generated are at least 250ms, all I can convince the Sipura to produce is a chirp. This is not a chirp-from-a-chickadee-celebrating-the-first-beautiful-day-of-spring, but more the kind of chirp a chipmunk might make if it lost a game of chicken with a bus. This problem is compunded when you involve a local loop that is marginal (i.e. greater than, say, 4.5dB of attenuation). This is something that is probably never tested in the lab, but is in fact quite common in reality. > > For whatever reasons, DTMF over IP is a pain. > > Strange. I'd have to say of all the VOIP problems I've ever > run up against in my tenure I have never had to worry about > DTMF. I've never had to worry about 1-way audio, either, but > I think that's because I simply don't try to push SIP through > NAT. :-) You are running PRI mostly? That'll help keep things clean. I think it's when an analogue link is involved somewhere that things can dicey. Jim. -- Jim Van Meggelen [EMAIL PROTECTED] http://www.oreillynet.com/pub/au/2177 "A child is the ultimate startup, and I have three. This makes me rich." Guy Kawasaki -- -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.375 / Virus Database: 267.15.11/264 - Release Date: 17/02/2006
