Abstract
In this study, differences between dynamic-range reduction by peak clipping and single-channel compression for phoneme perception through conventional hearing aids have been investigated. The results from 16 hearing-impaired listeners show that compression limiting yields significantly better phoneme identification scores than does peak clipping if a strong reduction of the dynamic range has to be effected. In addition, some qualitative differences in confusion patterns could be established: the perception of low-frequency information appears to be relatively important in compression limiting; the perception of plosiveness is less important.