Summary:ASTERISK-06680: Monitor option b causes 1 second audio delay
Reporter:vortex_0_o (vortex_0_o)Labels:
Date Opened:2006-04-02 20:08:50Date Closed:2006-05-15 11:55:19
Versions:Frequency of
Description:Monitor with option b causes received audio to be delayed by approx 1 second.
Transmit audio is fine.

audio received from the dialed number is delayed by 1 second:
exten => _0[1-9].,1,Monitor(gsm,${CALLFILE},bm)
exten => _0[1-9].,2,Dial(SIP/44${EXTEN:1:10}@something)

received audio fine:
exten => _0[1-9].,1,Monitor(gsm,${CALLFILE},m)
exten => _0[1-9].,2,Dial(SIP/44${EXTEN:1:10}@something)
Comments:By: Andrey S Pankov (casper) 2006-04-10 13:36:01

Can you expalain the issue in more details?
How do you measure such a delay?

By: vortex_0_o (vortex_0_o) 2006-04-11 05:40:48

Very technical process to measure the delay ;-)
I have one phone handset on each side of my head
(speaking into one phone and hearing the delay on the other end)

No errors or anything seem to be showing up on the console / logs

By: Andrey S Pankov (casper) 2006-04-11 05:54:24

What codecs are you using? gsm?

By: Andrey S Pankov (casper) 2006-04-11 06:17:11

What happens when you enable constant delay (remove #if 0/#endif) in channel.c?

#define MONITOR_DELAY   150 * 8         /* 150 ms of MONITORING DELAY */

By: vortex_0_o (vortex_0_o) 2006-04-12 14:29:50

The recording itself seems to be ok it is the actual 'live audio' that is delayed

original test
alaw->alaw, recording in gsm
audio delay

tested alaw->alaw , recording in alaw
same result - audio delay

upgraded to current svn 1.2 branch
same result - audio delay

tested without m option (just option b)
same result - audio delay

I'll give the change to channel.c a go - does MONITOR_CONSTANT_DELAY effect actual audio or only the recording file?

By: Andrey S Pankov (casper) 2006-04-12 14:32:34

Please use trunk for your tests...

By: vortex_0_o (vortex_0_o) 2006-04-12 15:08:50

always using current branch-1.2 - should this be ok if we are talking bug fixes?

By: Serge Vecher (serge-v) 2006-05-06 15:09:55

vortex_0_o: yes, the fix will be applied to 1.2, once found. However, we have to find the fix first. Casper is suggesting you try running Asterisk trunk to see if the issue is present there. Because if it is not, then the fix maybe found relatively easily by examining code differences between trunk and 1.2 branch. Thanks.

By: Serge Vecher (serge-v) 2006-05-15 11:42:42

Ok, here is the deal; since nobody else can reproduce this, please test the latest 1.2 branch (rev > 27000). If the issue is still there, please test the latest trunk (rev > 27000).

Reopen the bug once you have the test results. Thanks.