Retuning the overtones of the voice: Helium and SF6

Hi there.

I think this is one of the ways to successfully confuse an overtone singer. :-D

This is the first time I've heard something like this. To make the example audio files sound like a normal voice, I had to slow them down to about 3/4 of the original speed (see the section "What Helium does to speech"): www.phys.unsw.edu.au/phys_about/PHYSICS!/SPEECH_HELIUM/speech.html

There was the event in the 1960s when a man who spent 30 days in the ocean in the depth of 200 feet was in a "decompression chamber" where he was breathing Heliox instead of Oxygen. Trying to make the recording of his speech sound like a normal voice, I realized I had to shift the pitch down at least by a sixth (IOW, 3/5 of the original speed): www.npr.org/programs/lnfsound/stories/991015.stories.html

Some websites say there's something called "Helium Speech Descrambler" which should make the Helium voice sound like normal. But I wonder how this can be done. The Helium only changes the pitches of the resonances in your mouth, not of the voice itself, and therefore if I apply a time compressing pitch-shifter to the sound, although the voice can then sound like normal, its pitch falls significantly.

BTW: Some people used SF6 to get the opposite effect. But interestingly, when I wanted to make such a recording sound like a normal voice, I had to raise the pitch at least by a fifth or a sixth -- a fourth was not enough. More here: www.youtube.com/watch?v=6KLNYlYGdYs

Anyway, I'm still wondering how these people do it that they don't faint, considering that these gasses are not suitable for breathing.

Okay, that's just about it. So ... Back to scales again. :-D

Petr

Petr Pa��zek wrote:

> Some websites say there's something called "Helium Speech Descrambler" which > should make the Helium voice sound like normal. But I wonder how this can be > done. The Helium only changes the pitches of the resonances in your mouth, > not of the voice itself, and therefore if I apply a time compressing > pitch-shifter to the sound, although the voice can then sound like normal, > its pitch falls significantly.

There are pitch shifters that keep the formant envelope constant, so I'm guessing you use them in reverse so that the pitch is constant but the formant moves. Kyma comes with examples -- maybe it's LPC resynthesis. These pages at least show that it's possible:

http://forum.cakewalk.com/tm.asp?m=899239&mpage=1&key=

http://www.flstudio.com/help/html/plugins/Edison_stretch.htm

Graham