Pitch perception in electric and acoustic hearing

Abstract

In normal hearing (NH), timing and place-of-excitation cues to pitch are usually inextricably linked. In contrast, by varying the temporal pattern of electrical pulses applied to one channel of a cochlear implant (CI), researchers can investigate the usefulness and limitations of purely temporal cues to pitch. A related technique, which often produces very similar results, is to present NH listeners with trains of acoustic pulses that have been passed through a fixed bandpass filter so as to remove place cues. Furthermore, one can then measure how these temporal patterns are represented in the auditory nerve (AN) by recording, respectively, the electrically and acoustically evoked compound action potential. These physiological measures can then be compared to listeners’ pitch judgements to the same stimuli. We have used this approach to show how AN refractory effects can influence pitch, to evaluate competing pitch models, and to show how a code based on first-order intervals can provide a level-independent representation of pitch. We have also examined the possible basis for the relatively poor pitch perception of CI users, particularly at pulse rates above about 300 pps. [Work supported by RNID.]