No abstract

The contribution to speech intelligibility of the resonance effects of the ear canal and pinna has been of recent research interest. The present study investigated these effects by comparing monaural and binaural speech intelligibility between over-the-ear and in-the-ear microphone locations, utilizing a master hearing aid which controlled for other electroacoustic effects, The experimental design also allowed examination of the effects of azimuth on both monaural and binaural aided speech intelligibility. Items of the Tri-Word Test of Intelligibility [Sergeant et al., J. Acoust. Soc. Am. 65, 218–223 (1979)] were presented randomly from five loudspeakers equally separated on the front semicircle against an uncorrelated multiple-talker background. The results generally favored the in-the-ear microphone location. Binaural squelch was demonstrated.

The effects of frequency response smoothness and stimulus-type on the loudness discomfort level were examined. Two major questions were asked: (1) Does a smooth frequency response elicit a judgment of loudness discomfort at a higher sound pressure level than an irregular response? (2) Does a speech-spectrum shaped noise elicit a judgment of loudness discomfort at a higher sound pressure level than a multitalker bubble? The irregular response consisted of five peaks, one-third octave wide and 10 dB in amplitude. Both stimulus-types were processed through a master hearing aid with the electroacoustic parameters set to either the smooth frequency response or the irregular frequency response, and then presented to 20 normal hearing subjects listening in the sound field at zero degrees azimuth. The stimulus intensity was manipulated according to a simple updown adaptive procedure [H. Levitt, J. Acoust. Soc. Am. 49, 467–477 (1971)]. The smooth frequency response elicited loudness discomfort level judgments at significantly higher sound pressure levels. There was no significant difference in the measured loudness discomfort levels as a function of stimulus-type. The dependence of results on the calibration procedure for both frequency response and stimulus-type conditions will be discussed.