[Asterisk-Dev] Liar liar - Voice stress analysis

[John Todd]

At 9:48 AM +0100 1/26/04, Florian Overkamp wrote:
>Hi,
>
>I found this on FreshMeat:
>
>http://freshmeat.net/releases/149150/
>(or http://liarliar.sourceforge.net/)
>
>Could it work on * telephony channels ? :-)
>
>Florian

That would be really cool, but I suspect that compression would 
significantly "fuzz" the ability of that trick.  Now, if you were 
just using PRI channels inbound or G711, maybe there is a chance, but 
someone who knows more about the audio spectrum of microtremors and 
the human voice would be able to answer more definitively.

Let me propose this:  I would love to see a channel driver 
(pre-channel driver?) that allows for modulation of  voice on either 
end of the channel, to allow for voice disguising via a handful of 
parameters that can be changed on a per-call basis.

JT