Today I decided to do a little experiment to see what a computer can do in terms of a female sounding voice. So I recorded myself as a normally speak. Not really feminine enough to convince, unfortunately. When I was in my early 20's I could keep up with female background vocals. But that was 20 years ago and these days I just don't have the same range as I used to, which hasn't helped my approximation of a female voice.
Anyway, I ran the recording through Adobe Audition (=computer sound processing program). After several "Donald Duck" variations I came up with something more or less realistic. Or at least, as best as I can get it without Donald Duck poking his head around the corner. The result is probably at best androgynous.
In the version I like best I raised the fundamental pitch by only 3 semitones (a minor third in musical terms). Any more than that and ol' Donald makes his appearance. I raised the "formants" by an octave. The difference with the formants to me is subtle, but I think important.
Now if only I could do that in real life.
