Abstract:
A computational model of the precedence effect is proposed by assuming that the precedence effect is caused by the inhibition of sound localization which depends on the estimated sound-to-echo ratio. It is found that a generalized pattern of impulse response with delay and decay features can be used for echo estimation. As a real neural implementation, the pattern of an after-effect in the human auditory system, based on the previous psychoacoustic just-noticeable-difference tests of interaural delay and intensity, is adopted. This pattern has delay and decay features similar to those in the generalized pattern of impulse response, and the difference in decay scales is shown to cause little difference in echo estimation. The results of psychoacoustic experiments, e.g., equal-level and unequal-level paired click tests [P. M. Zurek, J. Acoust. Soc. Am. 67, 952--964 (1980); K. Saberi and D. R. Perrott, ibid. 87, 1732--1737 (1990)], Haas's tests [H. Haas, J. Audio Eng. Soc. 20, 146--159 (1972)], Franssen's tests [J. Blauert, Spatial Hearing (MIT, Cambridge, MA, 1983)], and others, can be interpreted consistently. This model can also give criteria for an available onset and an explanation of why a transient onset is required for the precedence effect.