Spectral distribution of /s/ and the frequency response of hearing aids

Abstract

The purpose was to determine a target for the upper frequency limit of a hearing aid that will provide access to the important spectral cues for all the sounds of English. The sibilant /s/ was studied because of its high-frequency content. Repeated tokens of /s/ were recorded from five men and five women before and between the vowels /u/, /a/, and /i/. Using fast Fourier transform analysis, the prominent spectral peak with the lowest frequency was identified and its center frequency determined for each token. This frequency averaged around 4.9 kHz for the /u/ context, 5.6 kHz for the /a/ context, and 6.0 kHz for the /i/ context. There were dramatic differences among talkers, with subject means ranging from 3.2 to 8.4 kHz. The women generated consistently higher frequency /s/ sounds than the men, but there were also large differences within gender groups. These data suggest that the upper frequency limit of a high-fidelity hearing aid should be in the region of 10 kHz. If this cannot be accomplished with direct amplification, an alternative might be the selective use of frequency transposition.