ASA 127th Meeting M.I.T. 1994 June 6-10

4pSP30. Perception of back vowel height: Effects of varying F1-F0 bark distance.

Richard P. Fahey

Dept. of Psychol., Univ. of Texas, Austin, TX 78712

Using front vowels, Hoemeke and Diehl [J. Acoust. Soc. Am. 93, 2422(A) (1993)] confirmed that F1-F0 (in bark) is a perceptual correlate of vowel height [Traunmuller, J. Acoust. Soc. Am. 69, 1465--1475 (1981)], and that the [+high]/[-high] boundary falls at 3--3.5 Bark F1-F0 [Syrdal and Gopal, J. Acoust. Soc. Am. 79, 1086--1100 (1986)]. The present study replicates Hoemeke and Diehl's experiments, using back vowels. Listeners identified three sets of synthetic vowels varying orthogonally in F1 and F0, ranging between /u/--/(small capital you)/, /(small capital you)/--/(open oh)/, and /(open oh)/--/(open aye)/. In a variable higher-formant condition, F1-F0 better predicted labeling than did F1 for both the /|pA/--/(small capital you)/ and /(small capital you)/--/(open oh)/ series. Scatterplots of F1-F0 vs labeling scores were most steplike for the /u/--/(open oh)/ ([+high]/[-high]) series, with the boundary at 3--3.5 Bark. In a fixed higher-formant condition, F1-F0 better predicted labeling than did F1 for the /u/--/(small capital you)/ series. Although less clearly so, the most steplike scatterplot was again that for the /u/--/(open oh)/ series, with a boundary at 3--3.5 Bark F1-F0. [Work supported by NIDCD.]