was covered by groups of cochlear implant recipients, hearing-aid users, and normally hearing listeners. Experiment 1 measured the shape of the internal spectrum using a forward-masking paradigm. Experiment 2 tested the ability of subjects to discriminate these sounds in all combinations of pairs using a 4IFC design. Experiment 3 consisted of a closed-set identification task with the same sounds.

Following widespread success with speech communication, the recipients of cochlear implants often express a desire for the improved perception of music. Inspired by the ultimate aim of delivering better musical appreciation to implant users, this project investigated the perception of musical timbre. Previous research with implants and music perception had concentrated on pitch and rhythm, making timbre a logical choice for further study. The shape of the steady-state frequency spectrum was investigated, thereby focussing the project on a single dimension of timbre.
Three psychophysical experiments were used to study the internal spectrum. This is the internally perceived frequency spectrum, which is thus a powerful measure for comparing the effects of different degrees of hearing loss. To represent a range of typical musical sounds, five instruments and five sung vowels were chosen for study. A range of hearing impairments

"Inspired by the ultimate aim of delivering better musical appreciation to implant users, this project investigated the perception of musical timbre� "

The similarity of an internal spectrum to its corresponding physical spectrum was quantified with the spectral correlation measure. A hypothesis that this relationship would be significant for the hearing-impaired subjects as well as the normally hearing subjects, though weaker for the former, was sup

(Continued on page 5)