Perceptual information for voicing of postvocalic fricatives
Sigfrid D. Soli, Ann D. Lassalle, W. Van Summers
Abstract
Sigfrid D. Soli, Ann D. Lassalle, W. Van Summers
Abstract
As voicing in a word-final fricative varies, the duration and the internal temporal and spectral structure of a preceding vowel also vary. We examined the importance of vowel structure variations as a fricative voicing cue. Stimuli were prepared by computer editing of vowel nuclei from utterances [jus] and [juz] which lacked final first formant transitions and differed in both vowel duration and structure. Durations were systematically varied from 350–550 ms to produce pairs of vowel nuclei with equal durations and different structures. Duration of unvoiced friction was also varied from 100–250 ms making a total of 40 stimuli. Identification responses to these stimuli by 15 listeners revealed that friction duration was a weak voicing cue, while vowel duration functioned as a much more effective voicing cue, replicating previous findings [P. Denes, J. Acoust. Soc. Am. 27, 761–764 (1955)]. However, vowel structure was also a strong voicing cue, having a greater perceptual effect than 50 ms changes in duration. Acoustic analyses of the experimental stimuli suggested that proportional durations of the formant transitions and total vowel duration together specify voicing information.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
As voicing in a word-final fricative varies, the duration and the internal temporal and spectral structure of a preceding vowel also vary. We examined the importance of vowel structure variations as a fricative voicing cue. Stimuli were prepared by computer editing of vowel nuclei from utterances [jus] and [juz] which lacked final first formant transitions and differed in both vowel duration and structure. Durations were systematically varied from 350–550 ms to produce pairs of vowel nuclei with equal durations and different structures. Duration of unvoiced friction was also varied from 100–250 ms making a total of 40 stimuli. Identification responses to these stimuli by 15 listeners revealed that friction duration was a weak voicing cue, while vowel duration functioned as a much more effective voicing cue, replicating previous findings [P. Denes, J. Acoust. Soc. Am. 27, 761–764 (1955)]. However, vowel structure was also a strong voicing cue, having a greater perceptual effect than 50 ms changes in duration. Acoustic analyses of the experimental stimuli suggested that proportional durations of the formant transitions and total vowel duration together specify voicing information.
Key concepts: Voice, Vowel, Duration (music), Formant, Acoustics, Perception, Audiology, Voice-onset time