| Summary: | ASTERISK-02054: SIP generated inband DTMF too short | ||
| Reporter: | glendale2x (glendale2x) | Labels: | |
| Date Opened: | 2004-07-18 13:50:25 | Date Closed: | 2011-06-07 14:05:15 |
| Priority: | Major | Regression? | No |
| Status: | Closed/Complete | Components: | Core/General |
| Versions: | Frequency of Occurrence | ||
| Related Issues: | |||
| Environment: | Attachments: | ||
| Description: | The DTMF generated by SIP info/rfc devices through asterisk to inband (like BroadVoice) can be too short to be reliable in some applications. An option in the sip.conf file to change the DTMF length and delay between multiple digits would be ideal. | ||
| Comments: | By: Mark Spencer (markster) 2004-07-18 14:41:35 Please try updated CVS first. By: glendale2x (glendale2x) 2004-07-18 16:58:55 I don't see an option anywhere for DTMF tone length in latest CVS. The SIP inband DTMF length is way too short, even for an asterisk IVR to recognize; it can't understand its own DTMF. I'm testing this by dialing a Zap channel (X100P) from BroadVoice. The Zap channel answers into the IVR, but it won't recognize the inband DTMF from BroadVoice. It needs to be increased, like I can adjust the DTMF in digits.h with the zaptel channels. I've tried grepping through the source of asterisk for a place to make the DTMF longer, but I can't seem to find anything. By: Mark Spencer (markster) 2004-07-18 17:06:11 I'm not suggesting that there is an option, I'm suggesting you try the code and see if the audio is longer appropriate. The tones are done in channel.c in the send dtmf stuff. By: glendale2x (glendale2x) 2004-07-19 01:37:34 It's still unreliable in latest CVS; several times during testing it would misread the actual digits pressed. (200 became 202, for example.) The only workaround I've found is to increase the tone length. By: Brian West (bkw918) 2004-07-19 18:25:53 What phones and gateways? Because I never have this problem. By: cshaw59 (cshaw59) 2004-07-21 18:19:15 U sure it's not the phone's fault? I have a GrandStream that's dyslexic and thinks that 2 is 5 sometimes. I've checked and it's not *'s fault, the GS ACTUALLY sends 5 instead of 2... By: Olle Johansson (oej) 2004-08-08 16:15:47 Any updates to this bug report? Still a problem with current cvs? By: glendale2x (glendale2x) 2004-08-09 17:35:33 The problem doesn't seem to be present in the RC1 tarball. I haven't tried the latest CVS. By: Brian West (bkw918) 2004-08-11 16:52:29 Fixed | ||