Rate versus time representation of high-frequency spectral notches in the peripheral auditory system: A computational modeling study

A computational model of the peripheral auditory system is used to explain the paradoxical observation that discriminating between broadband noise sounds with and without high-frequency spectral notches is more difficult at mid-intensities than at lower or higher intensities [A. Alves-Pinto, E.A. Lopez-Poveda, Detection of high-frequency spectral notches as a function of level, J. Acoust. Soc. Am. 118 (2005) 2485-2469.]. The simulations suggest that the discrimination task in question relies on comparing the timing of auditory nerve spikes, hence that high-frequency sounds are represented in the auditory nerve by a time code. They further suggest that the improvement in spectral discrimination at high intensities is associated with inherent inner hair cell nonlinearities.