Evaluation of spectral enhancement in hearing aids, combined with phonemic compression

In this study, the separate and combined effects on speech perception of compensation of the reduced dynamic range by compression and compensation of the reduced frequency resolution by spectral enhancement is investigated. The study has been designed to compare the effects of signal processing on monosyllabic consonant-vowel-consonant words for hearing-impaired listeners in conditions of quiet, fluctuating noise, and continuous noise. Speech perception of spectrally enhanced speech was compared with unprocessed speech. In addition, a comparison was made between combinations of spectrally enhanced speech and two types of phonemic compression. In the past, the definition "syllabic compressor" is often used to indicate fast compressors. However, the time constants of the fast compressors used in this study are so short that syllabic has become an inappropriate term. Moreover, intelligibility tests were performed in which scores were acquired of monosyllabic words, and their constituent "phonemic" parts. Therefore, the definitions "phoneme" and phonemic will be used throughout this paper. In one condition, spectral enhancement produced significant improvements for vowel perception. But, this was counteracted by deterioration of the consonant scores for all but one subject. In general, the best overall scores for consonant-vowel-consonant words were obtained in the unprocessed condition. After the spectral enhancement, a single-channel phonemic compressor added no improvement. There are indications that a multichannel phonemic compressor and spectral enhancement have opposite effects, because the scores for this combination are, in general, the lowest.