Consonant recognition in on-harmonic versus between-harmonic noise.

Item

Title
Consonant recognition in on-harmonic versus between-harmonic noise.
Identifier
AAI9807996
identifier
9807996
Creator
Schoepflin, Janet Reath.
Contributor
Adviser: Harry Levitt
Date
1997
Language
English
Publisher
City University of New York.
Subject
Health Sciences, Audiology
Abstract
The consonant recognition performance of ten normally hearing adults was measured for twenty synthesized vowel-consonant nonsense syllables in speech-spectrum shaped noise, comb-filtered on-harmonic noise, and comb-filtered between-harmonic noise at four signal-to-noise ratios under full-band, low-pass, and high-pass filtering conditions to test the hypothesis that only those portions of the noise spectrum that lie in the same critical bands as the speech spectrum are effective in masking the speech.;Since the width of the critical band increases logarithmically with frequency, it was anticipated that speech recognition scores for the low-pass filter condition would be higher for between-harmonic noise than for on-harmonic noise because the speech harmonics and between-harmonic noise occupy different critical bands, but co-exist in the same critical bands for the on-harmonic noise. For the high-pass filter condition, no difference in masking effect between the on-harmonic and between-harmonic noise was expected since both of the noises occupy the same critical bands as the speech. For the full-band condition, speech recognition scores were again expected to be higher for the between-harmonic noise than for the on-harmonic noise because the between-harmonic noise only occupies the same critical bands as the speech in the high-frequency region of the spectrum.;Results supported the hypothesis in that speech recognition scores for the between-harmonic noise were significantly higher than those for the on-harmonic noise at the poorer signal-to-noise ratios in the full-band and low-pass filter conditions, and that there were no significant differences among the noise types at any of the signal-to-noise ratios tested in the high-pass filter condition. The implications of the findings with regard to current and future noise reduction strategies are discussed.
Type
dissertation
Source
PQT Legacy CUNY.xlsx
degree
Ph.D.
Item sets
CUNY Legacy ETDs