Voice stress analysis

Special Interest Groups Audio Processing
1

Here I am again. Audacity is a great program but i guess I would not be happy even if you hung me with a brand new rope.

Voice stress analysis performs Fourier Analysis of vocal frequencies and displays the shift of vocal frequencies which accompany even the mildest of physiologic stress. I am an engineer but I don’t claim to understand all of the implications of this statement but I have experimented with it.

And the source code is available here:
http://members.tripod.com/JusticeMachines/Prevaricator/freeware.html

I do not have the programming skills to incorporate the souce code into a plug in but I would bet someone on the audacity team does and I would be willing to do the testing and evaluating. Most of the time people think that i may be crazy when I propose an off the wall programming suggestion but after a little time of working with me they are absolutely sure that i am crazy. But the project always eventually works.

2

Not wanting to detract from your suggestion, but with all these ideas that you are coming up with, I think you should start looking at developing your own plug-ins.
I’ve been playing around with Nyquist for a little while now, and even though I have virtually zero previous experience in programming, I am finding that I can make things that work.

There are some links that will help you get started with Nyquist here: http://audacityteam.org/wiki/index.php?title=Nyquist_Basics:_The_Audacity_Nyquist_Prompt

(but you will probably be better starting with something a little more simple than this idea)

3

Thanks that is a good suggestion but I really was hoping that I could be lazy. Before I even get started, is there a way to interface “C” code into the Nyquist programming language.

4

Yes, if you want to. Most of the basic functions in Nyquist are written in C, though Nyquist itself is a kind of LISP.

5

Is the indicator of stress simply an increase of energy in the 6 - 16 Hz range (which the Audacity Spectrum view will show, or is the calculation of a “norm” for the audio critical?

And how do you tell in Prevaricator at what time point the lie is supposed to occur without a time axis?

Gale