The generation of psychoacoustic combination tones in relation to two-tone suppression effects in a computational model

The amplitude of the psychoacoustic distortion product 2 f1−f2 elicited by primaries f1 and f2 depends on the method of measurement. The cancellation-tone procedure gives consistently higher estimates of the distortion product than nonsimultaneous procedures. It was suggested that the difference could be attributed to suppression effects by the lower primary f1 [Smoorenburg, J. Acoust. Soc. Am. 52, 615–632 (1972)]. Simulations carried out with a computational model of the auditory periphery including a single, but distributed, compressive nonlinearity confirm the hypothesis; the amount of suppression produced at 2 f1−f2 by f1 equals the difference between simultaneous and nonsimultaneous methods of measuring the amplitude of the distortion product. This is in agreement with psychoacoustical data [Shannon and Houtgast, J. Acoust. Soc. Am. 68, 825–829 (1980)]. The model also reveals the importance of suppression effects by f1 to bring the phase of the traveling wave associated with the cancellation tone 2 f1−f2 in exact opposition with that of the distortion product 2 f1−f2 generated by the primaries. There is also evidence of suppression effects by the higher primary f2 when L2≫L1. In the model, cancellation occurs over a substantial portion of the basilar membrane near the characteristic place for 2 f1−f2.