Page 1 of 1

AMD Detection

PostPosted: Wed Jul 16, 2008 9:14 am
by xCape
Hello,

I am currently TESTING an AMD campaign that leaves one message for a Human and another one for a machine.

It works perfectly on my cell phone (sprint) but fails to leave a message on my boss's cell phone (Cingular). Although it does indicate there is playback when I watch the call on the vicidial server.

Is this purely a timing issue or could it be related to some cellular carriers? Also, is there a way for me to listen to the auto-dialed call to see if it's left too early?

I am using VicidialNOW 1.1rc2 .

This is the error I'm getting for Cingular:

- Exiting with 2000ms silence >= 2000ms required
-- Executing Playback("Zap/1-1", "betty") in new stack
-- Playing 'betty' (language 'en')
== Manager 'sendcron' logged off from 172.20.100.5
-- Channel 0/1, span 1 got hangup request, cause 16
== Spawn extension (default, 8320, 2) exited non-zero on 'Zap/1-1'
-- Executing DeadAGI("Zap/1-1", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----16---------------") in new stack
-- AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... ---------- completed, returning 0
-- Hungup 'Zap/1-1'


Thank you.

PostPosted: Wed Jul 16, 2008 10:43 am
by mflorell
You will need to use the waitforsilence application in your edialplan, or just play the message twice if the "prease leave a message" is too long on your boss' cellphone.

PostPosted: Wed Jul 16, 2008 1:45 pm
by xCape
mflorell wrote:You will need to use the waitforsilence application in your edialplan, or just play the message twice if the "prease leave a message" is too long on your boss' cellphone.


In the following line:
exten => 8320,1,WaitForSilence(2000,2)

Should I adjust the 2000 value?

Thanks.

PostPosted: Wed Jul 16, 2008 1:46 pm
by mflorell
You could try adjusting both.

The 2000 value is milliseconds and the 2 value is iterations of the time before it will think everything is silent and exit, allowing the next dialplna entry to execute.

PostPosted: Wed Jul 16, 2008 6:11 pm
by JagFan
xCape,

I have been working for months on a project that includes AMD. My conclusion is that anything better than a 70% positive detection rate is a success. You will have many false-positives, there are too many variables involved. It is truly more art than science.