Abstract

Using the SYNTH serial software synthesizer [Mattingly, Pollock, Levas, Scully and Levitt, J. Acoust. Soc. Am. Suppl. 1, 69, S83 (1981)], the voiced fricatives [v] and [ð] were produced in initial position before the vowels [iy], [a], and [uw]. Frequencies of both a consonant pole and formant transition onsets (F2 and F3) were varied along continua in order to examine their effect on the perception of the voiced fricatives [v] and [ð]. The range of the formant transition onsets differed for each vowel but the consonant pole varied from 2400 to 4400 Hz in every case. The stimuli were presented in forced choice labelling tests. Preliminary testing indicates that the boundary shift from [v] to [ð] in each vocalic context depends significantly more on the change in formant transition than on the position of the consonantal pole. That is, with rising or relatively straight formant transitions, subjects reported hearing [v] whereas they heard [ð] with falling transitions regardless of the pole position. These results support the findings of previous studies with natural and synthetic speech that formant transitions are indeed important cues for distinguishing between labiodental and dental fricatives. [Work supported by NSF Grant PFR 8006144.]

Full Text
Paper version not known

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