Abstract

Eight listeners with severe to profound hearing loss read a six-sentence passage under spectrally altered and/or delayed auditory feedback. Spectral manipulation was implemented by filtering the speech signal into either one or four frequency bands, extracting respective amplitude envelope(s), and amplitude-modulating the corresponding noise band(s). Thus, the resulting auditory feedback did not preserve intonation information, although the four-band noise signal remained intelligible. The two noise conditions and the unaltered speech were each tested under the simultaneous and three delayed (50 ms, 100 ms, 200 ms) feedback conditions. Auditory feedback was presented via insert earphones at the listener’s most comfortable level. Recorded speech was analyzed for the form and domain of the fundamental frequency (f0) declination, the magnitude of the sentence initial f0 peak (P1), and the fall–rise pattern of f0 at the phrasal boundaries. A significant interaction between the two feedback manipulations was found. Intonation characteristics were affected by speech delay only under the spectrally unaltered feedback: The magnitude of P1 and the slope of the f0 topline both increased with the delay. The spectral smearing diminished the fall–rise pattern within a sentence. Individual differences in the magnitude of these effects were significant.

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