4aPP4. Spectral integration and the detection of tones in modulated and unmodulated noise.

Session: Thursday Morning, December 4


Author: Sid. P. Bacon
Location: Psychoacoust. Lab., Dept. of Speech and Hearing Sci., Arizona State Univ., Tempe, AZ 85287-1908, spb@asu.edu
Author: Nicolas Grimault
Location: Psychoacoust. Lab., Dept. of Speech and Hearing Sci., Arizona State Univ., Tempe, AZ 85287-1908, spb@asu.edu
Author: Jungmee Lee
Location: Psychoacoust. Lab., Dept. of Speech and Hearing Sci., Arizona State Univ., Tempe, AZ 85287-1908, spb@asu.edu

Abstract:

The intent of the present study was to determine whether spectral integration is similar in modulated and unmodulated noise. The signal consisted of one sinusoid of a triplet, or all three sinusoids together. Four sets of triplets were used. The masker consisted of one or three 100-Hz-wide bands of noise centered at the sinusoidal component(s) of the signal. They were unmodulated or sinusoidally amplitude modulated (in phase) at a rate of 8 Hz and a depth of 1.0. Thresholds were measured adaptively for each sinusoid in a triplet, and then for all three together, with the relative level of each determined by the threshold differences when measured individually. In unmodulated noise, the improvement in threshold (from one to three sinusoids) was about 2 dB, consistent with a [radical 3] improvement. In modulated noise, however, the improvement was often greater than that (5 dB). This greater improvement could not be explained by shallower psychometric functions in the modulated condition, as the psychometric functions for the detection of the individual sinusoids were similar for both backgrounds. Instead, it is likely due to the comodulated nature of the three masker bands. [Work supported by NIDCD.]


ASA 134th Meeting - San Diego CA, December 1997