[asterisk-users] DTMF issues in 1.4.19 with missing digits

Fri May 2 15:14:54 CDT 2008

Hello, all!

Trying to figure out an issue with DTMF recognition with 1.4.19. This  
is somewhat similar to the issue described here:  
http://bugs.digium.com/view.php?id=11740, but it might be a different  
issue altogether.

I have 1.4.19 running with TE212P on a US PRI.

I'm sending digits 823229999. Sometimes the digits are making it all  
in the asterisk, and sometimes some are missing.

In the case when the digits are all caught, my DTMF log enteries are  
something like this:

<snip>
[May  2 14:48:56] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '8'  
received on Zap/1-1, duration 0 ms
[May  2 14:48:56] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '8' on Zap/1-1
[May  2 14:48:56] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '8' on Zap/1-1
[May  2 14:48:57] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '2'  
received on Zap/1-1, duration 0 ms
[May  2 14:48:57] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '2' on Zap/1-1
[May  2 14:48:57] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '2' on Zap/1-1
[May  2 14:48:57] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '3'  
received on Zap/1-1, duration 0 ms
[May  2 14:48:57] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '3' on Zap/1-1
[May  2 14:48:57] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '3' on Zap/1-1
[May  2 14:48:58] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '2'  
received on Zap/1-1, duration 0 ms
[May  2 14:48:58] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '2' on Zap/1-1
[May  2 14:48:58] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '2' on Zap/1-1
[May  2 14:48:58] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '2'  
received on Zap/1-1, duration 0 ms
[May  2 14:48:58] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '2' on Zap/1-1
[May  2 14:48:58] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '2' on Zap/1-1
[May  2 14:48:59] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:48:59] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:48:59] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
[May  2 14:49:00] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:49:00] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:49:00] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
[May  2 14:49:00] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:49:00] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:49:00] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
[May  2 14:49:01] DTMF[28649]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:49:01] DTMF[28649]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:49:01] DTMF[28649]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
</snip>

In the case when digits are not fully recognized (one is missing), I get this:
<snip>
[May  2 14:36:16] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '8'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:16] DTMF[28461]: channel.c:2128 __ast_read: DTMF begin  
emulation of '8' with duration 100 queued on Zap/1-1
[May  2 14:36:16] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '2'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:16] DTMF[28461]: channel.c:2098 __ast_read: DTMF end '2'  
put into dtmf queue on Zap/1-1
[May  2 14:36:16] DTMF[28461]: channel.c:2237 __ast_read: DTMF end  
emulation of '8' queued on Zap/1-1
[May  2 14:36:16] DTMF[28461]: channel.c:1961 __ast_read: DTMF begin  
emulation of '2' with duration 100 queued on Zap/1-1
[May  2 14:36:16] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '3'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:16] DTMF[28461]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '3' on Zap/1-1
[May  2 14:36:16] DTMF[28461]: channel.c:2155 __ast_read: DTMF end  
passthrough '3' on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2237 __ast_read: DTMF end  
emulation of '2' queued on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '2'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:17] DTMF[28461]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '2' on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2155 __ast_read: DTMF end  
passthrough '2' on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:17] DTMF[28461]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:17] DTMF[28461]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:36:17] DTMF[28461]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
[May  2 14:36:18] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:18] DTMF[28461]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:36:18] DTMF[28461]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1
[May  2 14:36:18] DTMF[28461]: channel.c:2092 __ast_read: DTMF end '9'  
received on Zap/1-1, duration 0 ms
[May  2 14:36:18] DTMF[28461]: channel.c:2144 __ast_read: DTMF end  
accepted without begin '9' on Zap/1-1
[May  2 14:36:18] DTMF[28461]: channel.c:2155 __ast_read: DTMF end  
passthrough '9' on Zap/1-1

</snip>

Looks like the 2 after the 3 is not getting recognized, and the 2  
after the 8 is being queued up and then dequeued later. I don't know  
why we need to defer DTMF here, but either way the issue is that DTMF  
digits are missing in the end.

The behavior is a bit erratic, and may occur when the digits are being  
sent while * is playing audio on the channel, even though it's not  
every time.

Does anyone have similar experience? Any additional info you may have  
would be helpful.

I'd like to get more info before filing a bug report.

Regards,
Mark G.