Listener perception of voice pitch can contradict measured pitch

Evidence

The claim that only the formants move, not the pitch, is reinforced by the analysis of the author’s voiced “a” as in cake, after breathing normal air and again after taking a breath of SF6. The pitch was varied while the vocal tract shape was kept approximately the same. An analysis of pitch and formant was done in Praat, a truly marvelous sound and voice analysis tool available free on the Internet for several platforms. Praat uses autocorrelation to find the pitch. The two a’s are startlingly different to the ear, and almost everyone describes the SF6 version as “lower,” but in fact the vocalization cycled through the same range of pitches—the gas did not change the pitch of my voice. The pitch change during the vocalization reveals that the formants do not depend on the pitch. As figure 17.12 shows, the formants are strongly dependent on the sound speed in the vocal tract, and the impedance mismatch at the SF6-air interface. — Eric J. Heller, Why You Hear What You Hear- An…, 1716

Categorization

• XREF: Reinforces the pitch/formant distinction established in 1713 and links to timbre discussion in chapter 24.
• SHARE: Send to anyone who insists helium actually raises the pitch of your voice — it demonstrably does not.