Dramatic Auditory Pitch Change Due to Dynamic Intensity Change

Michael K. McBeath, John Neuhoff

ABSTRACT

Auditory pitch is one of the principal sources of information used in tasks ranging from localizing and parsing acoustic agents to deciphering complex passages in music and speech. Historically, pitch is considered to be principally a function of acoustic frequency with only a small effect due to absolute intensity. In the present work, we demonstrate that this view of pitch is dramatically violated in ways not previously described in the scientific literature. We found that when tones are Doppler shifted so that frequency only drops, the pitch dramatically rises and falls with dynamic intensity. We report two experiments using stimuli in which intensity and frequency are dynamically swept. First, we quantify magnitude of experienced pitch change of Doppler-shifted tones by comparing them to constant intensity tones that drop discretely in frequency. Then, we verify the independent influence of swept intensity and frequency on experienced pitch height. We show that dynamic intensity change can independently produce pitch variation comparable to a frequency change approaching an octave. This effect opposes and is an order of magnitude larger than the well known effect of discrete intensity change. Since changes in intensity in musical contexts do not distort the chromatic relation between tones, the experienced change in pitch due to swept intensity appears to operate on the dimension of pitch height. We propose the integrality of dynamic pitch and loudness reflects a natural correlation between change in frequency and intensity that is neurally encoded to facilitate processing of acoustic patterns.

	Figure 3. Plot of intensity, frequency, and the typical experienced pitch over time in experiment 1. The dynamic portion to the left is the Doppler stimuli with swept intensity and frequency. The discrete portion to the right is the comparison interval with constant intensity and a discrete frequency drop. The duration of the Doppler tone is six seconds. The duration of each discrete tone is 0.25 seconds. The pattern exemplifies the large discrete pitch change typically experienced to match the Doppler shifted tone.
	Figure 4. Results of Experiment 1. Graph shows percentage of times that the pitch drop of the Doppler tone was experienced as larger than that of the discrete comparison interval. The average match for the Doppler tone was a discrete drop of eight semitones, four times larger than the actual drop.
	Figure 5. Results of Experiment 2. Graph shows experienced rise in pitch for nine conditions in which frequency and intensity were each independently swept up, held constant, or swept down over a period of four seconds. The results indicate an additive effect in which changes in frequency and intensity were each independently correlated with changes in pitch. At this rate of change, an intensity sweep of 15 dB produced a pitch change roughly comparable to that of a frequency sweep of a semitone.