Hi Ken, I added some code that wrapped a simple finite-state-machine around the existing tone detection code in chan_usbradio.c some time back. With this code, DTMF tone passed thru the repeater until the FSM "accept" state was reached--after a couple taps, as I recall. This worked well....In the meantime, I discovered that if I turned OFF the RADIO_RELAX compiler flag (from "make menuconfig" in the asterisk directory), I got even better results. And, I subsequently removed the FSM code... With RADIO_RELAX unset, I've had a few cases where DTMF from weak/noisey signals wouldn't decode. But, this hasn't been bad enough for me to re-enable RADIO_RELAX. ...Hope this helps! 73, David kb4fxc On Thu, 25 Feb 2010, Ken wrote:

Is there a parameter that can be tweeked to reduce susceptibility of Asterisk to decoding voice as tones?? We have only one user who is bad at this, so it may not be fixable at the repeater end, but I think it would be interesting to try.? The parameter?would be the one that determines how long a tone pair must persist before it is considered valid.? I suppose this would be in the core Asterisk part of the program. ? I understand that there is a trade off with "real" tones going out over the air. This would probably be more of an issue with a system that has autopatch.... ? Ken Jamrogowicz