Jerry Malzan
14 Vermont Ave., Toronto, ON M6G 1X7, Canada
Ohm's law for hearing asserts that psychological phenomenon of pitch can be identified with the physical phenomenon of frequency. More precisely, the assertion is that perceived pitches are precisely those appearing in the Fourier spectrum of a periodic vibration having all frequencies in the audible range. Experiment shows, however, that both parts of this if and only if statement are false. Pitches may arise without a corresponding frequency (the ``pitch'' problem) and pitches may be absent when there is a corresponding frequency (the ``masking'' problem). Here an attempt is made to deal with these problems simultaneously, using only the fact that the ear is a short-time analyzer, rather than the long-time analyzer that Ohm's law for hearing would require. The model used is quite simple, requiring only one parameter, and is in reasonable accord with experimental results. Residue pitch is computed in the time domain, while other pitches are computed in the frequency domain with the possibility, however, of masking. [Work supported by NSERC.]