Abstract

Cochlear implants (CI) enable patients to perceive sound by the generation and transmission of electrical impulses. Voice encoded (vocoded) audio is commonly used as a simulator of CI speech. Past studies of vocoded English concluded that coded speech intelligibility is independent of whether a sinusoidal or noise-based carrier is used. Based on this work, recent CI research for tonal languages like Mandarin, have opted for a sinusoidal carrier without considering the impact of the carrier on intelligibility. The importance of spectral cues in Mandarin speech necessitates further analysis of the relationship between the carrier and intelligibility. This study explored intelligibility differences between English and Mandarin vocoded speech. This approach assessed speech recognition of randomly presented phrases to normal hearing English and Mandarin listeners. Available frequency channels and carrier type were varied to compare their effects on Mandarin word and tone identification. Results indicated that audio processed with a sinusoidal carrier led to significantly lower Mandarin intelligibility scores. In comparison, no effect was observed between English and Mandarin intelligibility when a noise-based carrier was used. The data suggest that the nature of the carrier type affects tonal language intelligibility and warrants further research as an experimental consideration in vocoded speech studies.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call