I support Herman and would love to jump on this evidence in his defense but voice stress analysis is voodoo BS. Independant study after independant study shows their reslts are no more reliable than random chance. They have gotten popular because they are much cheaper to use than polygraphs, and are used for the same reason - to coerce confessions. But for reliability - you might just a well toss a handful of JuJu beans into a circle filled with chicken feathers drenched in goat’s blood.
Are you saying that doesn't work?
;-)
“But for reliability - you might just a well toss a handful of JuJu beans into a circle filled with chicken feathers drenched in goat’s blood.”
Actually, I know more about this than anyone here I suspect.
In 1992 the company I owned wrote and marketed a voice stress analysis software package that was sold internationally. We had two versions, one that was sold as more of a game of fun package, and another version that was technically the same that we sold to the Israeli’s.
It was not layered, but it compared a statistical sample of the subjects normal speech, to the speech in question. We accomplished this by doing a Fourier analysis, looking for micro-tremors as well as pitch changes in the speech pattern. PC’s were slow back then so we were limited to our sampling ability. With modern computers it would “rock”.