4pPP24. Discrimination of harmonic- and log-spaced profiles and of static and dynamic profiles by good and poor profile listeners.

Session: Thursday Afternoon, May 16

Author: Ward R. Drennan
Author: Charles S. Watson
Location: Dept. of Speech and Hear. Sci., Indiana Univ., Bloomington, IN 47405


Most profile experiments have employed static profiles with logarithmic spacing. Many naturally occurring sounds have harmonic spacing and vary dynamically in time. Watson and Drennan [J. Acoust. Soc. Am. 97, 3272(A) (1995)] examined profile discrimination using both static and frequency-glide profiles with harmonic and logarithmic component spacing. Subjects detected an intensity increment in the middle component of 11-component, 400-ms profiles with a frequency range of 200--2200 Hz. Differences among the seven subjects were considerably larger than differences among the types of profiles, confirming earlier observations of a large range of abilities to discriminate profiles. Another experiment was therefore conducted to estimate the distribution of profile discrimination abilities for normal-hearing listeners. Forty-six subjects were screened using static-log profiles. The distribution of thresholds was roughly normal with a range of -2 to -26 dB (signal level relative to component level) and an s.d. of 4.8 dB. No dichotomy in profile discrimination ability was found. Subjects were selected from each tail of the distribution and tested using the static-log, static-harmonic, dynamic-log, and dynamic-harmonic profiles. The order of presentation of conditions significantly affected the results; however, whatever the order, static profiles yielded lower thresholds than the frequency-glide profiles. [Work supported by NIH/NIDCD and AFOSR.]

from ASA 131st Meeting, Indianapolis, May 1996