Effect of silence duration in intervocalic velar plosive on voicing perception for normal and hearing-impaired subjects.

For normally hearing subjects shortening the silence duration of an intervocalic voiceless plosive induces a misperception of voicing. The time boundary for this effect is about 60 ms, which corresponds to a possible forward masking effect at the frequency of voicing. If recovery from masking is indeed involved, hearing-impaired subjects, who may have prolonged forward masking, can be expected to show abnormally long time boundary for voicing misperception. This study investigated the perception of voicing of an intervocalic plosive for a natural speech sample "aka" as a function of occlusive silence duration for normally hearing and hearing-impaired subjects. To investigate a correlation with forward masking, a second test was performed on the subjects. The same first a of the "aka" was selected and at its end was concatenated a voiced murmur taken from an "aga" elocution from the same speaker, and the minimum duration of the voiced murmur necessary for it to be perceived was measured. About half of the hearing-impaired subjects needed an abnormally long silence duration to avoid voicing misperception. The data indicate a significant correlation between the results of the two tests with a slope of regression line close to unity, and thus support the hypothesis of a voicing perception ruled by recovery from forward masking. Increase in silence duration of voiceless plosives might then be a beneficial acoustical processing for some hearing-impaired subjects.