Perception of Consonant Voicing in Noise

Publisher Website
Google Scholar
Abstract
Measurements are reported of the perception in noise of the occurrence of voicing in the English consonants /p, b, t, d, f, v, s, z/. The listeners' task was to report whether the consonant spoken was of the class /b, d, v, z/ (voiced) or of the class /p, t, f, s/ (unvoiced). The factors investigated were (1) the position of the consonant in the test utterance: initial, intervocalic or final; (2) the place of articulation: alveolar /t, d, s, z/, or labial, /p, f, b, v/; (3) the degree of occlusion: stop, /p, b, t, d/, or fricative, /f, v, s, z/, and (4) the spectrum of the masking noise: white noise or low-frequency noise. The absence of voicing was perceived better in alveolar consonants than in labials in low-frequency noise. Otherwise there were no large effects of position, place of articulation, or degree of occlusion, on voicing perception. The results are interpreted in terms of low-frequency cues to voicing which are independent of place of articulation and high-frequency cues which vary with place of articulation.

This publication has 12 references indexed in Scilit: