Re: [Phys-l] frequency and wavelength of sound in air

[John Mallinckrodt <ajm@csupomona.edu>]

Antti Savinainen wrote:

> Curtis Osterhoudt pointed out a good source explaining what helium  
> does to speech:
>
> <http://phys.unsw.edu.au/phys_about/PHYSICS!/SPEECH_HELIUM/ 
> speech.html>


This nice web page includes sketches of what the vocal tract  
resonances do to the sound produced by the vocal folds.  These days  
it's pretty easy to produce real sonograms and spectra of your own  
voice with inexpensive shareware.  Just now I fired up Amadeus Pro  
(for the Mac) and produced a sonogram of the two recordings on that  
page, "Pitch in Air" and "Pitch in Helium," that you can view at

   http://www.csupomona.edu/~ajm/special/heliumspeech.jpg

The sonogram shows very clearly the way the first 20 or so harmonics  
of the fundamental vocal fold vibration frequency (~120 Hz) are  
shaped by the resonances in the vocal tract ("formants") and how the  
formants change and generally move upward in frequency under the  
influence of helium.

I also produced a sonogram of myself singing each of the sounds "Ah,"  
"Oh," "Ee," and "Ooh" starting at around A-110 and sliding up a fifth  
to E-165.

   http://www.csupomona.edu/~ajm/special/formants.jpg

The sonogram and the accompanying individual spectra of the sounds at  
the two frequencies make clear that the formants are responsible for  
distinguishing one sound from another.  Moreover, one can see that,  
while the formants change with sung pitch, they do not move WITH the  
sung pitch so that a recorded voice can not be simply pitch-shifted  
by more than a very small amount without sounding unnatural.

John Mallinckrodt
Cal Poly Pomona