| Literature DB >> 27036256 |
Abstract
Frequency compression (FC) is an effective alternative to conventional amplification for patients with severe-to-profound middle- and high-frequency hearing loss and with some residual low frequency hearing. The present work investigated the underlying factors accounting for the intelligibility of FC-processed speech. Eighteen normal-hearing participants listened to FC-processed Mandarin sentences, and testing conditions included two compression strategies (i.e., linear and nonlinear), four degrees of compression, and two segmental conditions (i.e., vowel-center and vowel-onset). Results showed that the nonlinear strategy outperformed the linear strategy in enhancing the intelligibility of FC-processed sentences. A small frequency range of source compression region involved a lesser degree of compression, and favored a higher intelligibility score. Vowel segments carried a large amount of intelligibility information contained in nonlinearly frequency-compressed sentences, but could not fully account for the intelligibility improvement from nonlinear frequency-compression. The intelligibility benefit of FC-processed Mandarin sentences may be partially attributed to the additional F1-F2 transitional spectral envelope cue provided by nonlinear FC with the F1-included spectral region preserved.Entities:
Mesh:
Year: 2016 PMID: 27036256 DOI: 10.1121/1.4944037
Source DB: PubMed Journal: J Acoust Soc Am ISSN: 0001-4966 Impact factor: 1.840