A Hypothesis About Whistle Voice

Ingo Titze

The hypothesis here is that whistle voice makes use of a source-vocal tract interaction based on acoustic inertance below the third formant. Acoustic inertance of the vocal tract helps to set the vocal folds into vibration.2 For a female vocal tract (see eleven vowel shapes in Figure 1a), the third formant occurs above 3000 Hz and vocal tract inertance is found for most vowel shapes if frequencies are in the 1500-3000 Hz range. This range is on the upskirt of the third formant F^sub 3^ (see Figure 1b, where formants F^sub 1^, F^sub 2^, and F^sub 3^ are labeled for the vowel /o/). Formants (the resonances of the vocal tract) are where the inertance curves make an upward turn followed by a sharp drop. When the inertance is high (above the zero line), source frequencies are reinforced. In particular, the fundamental frequency F^sub 0^ gets a boost. Note that for the /o/ vowel shape, the inertance curve is above zero for pitches slightly below and above C^sub 7^. But the higher harmonics (2F^sub 0^, 3F^sub 0^, 4F^sub 0^ ... ) are not systematically reinforced. They face variable inertances from the vocal tract. Note that the harmonics 2F^sub 0^, 3F^sub 0^, and 4F^sub 0^ lie in the midst of a cluster of formants above F^sub 3^. Similar situations occur for most of the other vowel shapes. Thus, the frequency range below and above C^sub 7^ (1500-3000 Hz) is particularly advantageous and consistent for the fundamental F^sub 0^, but quite irregular for higher harmonics.

A Hypothesis About Whistle Voice

View Online

Abstract

Details

Metrics