Hi,
the basic settings are pretty good ones. What I did to do improve the
performance and prevent the false positives, I started to recorded every
call, and analyzed every incorrect detection :) Fairly soon I came with
optimal set for my environment:
initial_silence = 2500
greeting = 1500
after_greeting_silence = 300
total_analysis_time = 5000
min_word_length = 120
between_words_silence = 50
maximum_number_of_words = 4 ; it's usuall to pickup saying "Jon
Anderssen, hello" in here
silence_threshold = 384
by the way, for outgoing SIP calls you might want to do this Background
trick as it helped me a lot regarding AMD on SIP.
exten => _X.,n,Background(blank_audio)
exten => _X.,n,AMD
On Wed, Aug 4, 2010 at 5:08 PM, Tino <tino at sparksupport.com> wrote:
> Hello ,
>
> I would like to tweak my Answeing Machine Detection (AMD) in Asterisk. My
> current values are
>
> AMD(2500|1500|300|5000|120|50|5|256) and we were able to identify approx
> 25-30 % of all answering machines.
>
> Anybody have any suggestion to improve the accuracy of AMD.
>
> Thanks
>
>
>
>
> --
> _____________________________________________________________________
> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
> New to Asterisk? Join us for a live introductory webinar every Thurs:
> http://www.asterisk.org/hello
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
> http://lists.digium.com/mailman/listinfo/asterisk-users
>
--
Mvh,
Aurimas Skirgaila
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.digium.com/pipermail/asterisk-users/attachments/20100804/5c9e88d1/attachment.htm