Pitch of noise plus noise delayed
Ian B. Thomas, Richard D. Bostwich, A. Ravindran · The Journal of the Acoustical Society of America · 1974
Rosenberg has demonstrated [J. Acoust. Soc. Am. 47, 118 (1970)] that if white noise is added to itself delayed by time T, the composite stimulus gives rise to a perceived pitch equal to the reciprocal of T. Clearly, this experiment raises problems for place theorists, since the spectrum of the composite stimulus is also “white.” Rosenberg's results raise no less of a problem for those who propose periodicity in the stimulus waveform as the determinant of pitch. This stimulus has neither periodicity nor concentrations of spectral energy of any kind. We have applied the theory of pitch perception presented by Ravindran and Thomas at this meeting to this stimulus as follows. The stimulus waveform was generated with the aid of a computer for delay times of 1 and 2 msec. This signal was made available through D/A conversion for monitoring and recording and for input to an analog device whose output represents the waveform of basilar membrane displacement for various membrane locations and for two possible middle-ear transfer functions. The resulting signal was then further analyzed using the steps called for in the Ravindran and Thomas model. The result for any given delay is a “distribution of perceived pitches” which has a distinct maximum at the appropriate pitch (1/T). Changes in membrane location and in middle-ear characteristics have minor effects on the distribution and do not alter the frequency of its maximum. These results lend support to the proposed model. [Work supported by NIH.]