Summary:ASTERISK-17237: inband DTMF cannot be detected and trigger service execute when A and B both use u-law (the same codec)
Reporter:aaa (shihchuan)Labels:
Date Opened:2011-01-13 05:04:49.000-0600Date Closed:2011-07-19 13:00:18
Versions:Frequency of
Environment:Attachments:( 0) A.log
( 1) B.log
Description:1. A and B both use the same codec (u-law, a-law).
When A enter *96(transfer feature key), transfer service don't be executed but B hear dtmf tone.

2. If A and B use different codec.(A use u-law, B use a-law)
When A enter *96(transfer feature key), transfer sevice is executed.
Comments:By: David Woolley (davidw) 2011-01-13 05:35:49.000-0600

I think it almost certain that the channel technology affects this.  In particular, it may be erroneously doing an external bridge.  Assuming you are using SIP, I think you need to follow the rules for reporting SIP problems.  You also need to provide the dialplan, just in case the real problem is that the feature code is being erroneously detected when transfers are disabled.

By: Leif Madsen (lmadsen) 2011-01-13 11:13:37.000-0600

I bet you don't have the 't' or 'T' flags enabled in Dial() which would make this a configuration problem.

By: aaa (shihchuan) 2011-01-14 00:26:28.000-0600

Yes, I did have put 'tT' in Dial() in these 2 cases.

The problem is that,
when using the SAME codec in a conversation, the inband dtmf is not detected.
I set debug level to 5 but no any debug messages about dtmf is displayed when using the same codec.

However, if using DIFFERENT codec, debug message:'Got DTMF begin on channel (SIP/2112-00000027)' will be shown, which is correct.

By: David Woolley (davidw) 2011-01-14 05:06:10.000-0600

You need to provide the SIP(?) debug output, so one can tell whether the problem is that the call is being erroneously externally bridged, and more generally, the nature of the bridge being created.

More generally, it is generally better to include the actual debug output in a report, rather than a interpretation that it shows nothing.

By: Nickilo ( ThinkroSystem) (nickilo) 2011-01-16 05:43:54.000-0600

I can confirm this bug exist. But at the moment i have two asterisk with the same binary, the same configuration files. On can reproduce the problem, the other not.
I'm gonna try to understand the problem, give you a way to reproduce the problem and the debug output.

I confirm also that using différent codecs "solves" the problem.

By: Nickilo ( ThinkroSystem) (nickilo) 2011-01-17 17:10:26.000-0600

Hi, i just uploaded Bug_18611_case_1.zip

You can find explanations and logs inside the zip file.

I hope it can help, if you need anything else, just ask.

Thank you !

By: Leif Madsen (lmadsen) 2011-01-19 10:46:22.000-0600

Please re-upload your data as text files attached to the issue and not as an archive file.

By: Nickilo ( ThinkroSystem) (nickilo) 2011-01-19 10:58:40.000-0600

Done Imadsen, i've just uploaded README wich describe the test i have made.
A.log and B.log contains logs files.

By: Leif Madsen (lmadsen) 2011-01-19 14:49:27.000-0600

You're still missing the dialplan to be used to reproduce this issue.

By: Nickilo ( ThinkroSystem) (nickilo) 2011-01-21 17:53:01.000-0600

The dialplan is uploaded as a separate file

By: Nickilo ( ThinkroSystem) (nickilo) 2011-01-27 15:39:56.000-0600

I found that difference between the A and the B case is dtmf configuration of the sip client.

If using pure in-band, asterisk is no more working.

The bug still exist in version

By: David Woolley (davidw) 2011-01-28 05:11:59.000-0600

That sounds like a configuration error involving dtmfmode, rather than a bug.

By: Nickilo ( ThinkroSystem) (nickilo) 2011-01-28 17:43:57.000-0600

In sip.conf i have dtmfmode = inband and no dtmf parameter in the phone configuration.

By: Remi Quezada (remiq) 2011-02-11 09:59:45.000-0600

I am also experiencing this problem.  I am able to reproduce this in 1.6.1 and 1.6.2.

By: Kinsey Moore (kmoore) 2011-06-29 17:23:32.472-0500

I have posted a fix for this issue here: https://reviewboard.asterisk.org/r/1302/

By: Kinsey Moore (kmoore) 2011-07-19 13:00:18.566-0500

fixed in 1.8, 1.10, and trunk