Date of Award

Spring 1-1-2019

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Aerospace Engineering Sciences

First Advisor

Allison P. Anderson

Second Advisor

Torin K. Clark

Third Advisor

Kathryn H. Arehart

Abstract

This research thesis explores improving auditory perception through the use of stochastic resonance (SR), a phenomenon in which the throughput of non-linear signals is enhanced using additive noise. While SR has been successfully explored in a variety of perceptual channels (visual, tactile, vestibular), past psychoacoustic experiments have yielded conflicting results. This study aims to understand how SR can be observed in the auditory system accounting for individual differences.

Two studies were carried out to investigate SR within the auditory system. Both studies observed how white noise magnitude influences perception of pure tone stimuli across the frequency spectrum. The Threshold Optimization Study aimed to correlate SR enhancement with a subject’s audiometric threshold, predicting that noise levels equal to the subject’s threshold at a specific frequency tone would yield the highest SR benefit. Ten subjects completed pure tone audiometry with and without noise. Observing auditory thresholds with subthreshold, at-threshold, and suprathreshold additive noise yielded insignificant results. The noise levels tested did not improve or worsen audiometric performance across the board, which led to changes in the experimental methodology, specifically the noise levels that were presented and signal administration. Using those changes a second Protocol Development Study was conducted to replicate the results found in past psychoacoustic studies.

The additional study also aimed to observe SR, but expanded the noise spectrum to also find the masking that was not discovered in the first study. Four lab members completed pure tone audiometry with and without the presence of noise over a broad range of noise levels. A qualitative analysis suggests masking existed for frequencies with low thresholds given the noise levels that were tested. For some subjects, SR benefits may have been observed, but for others they did not appear to be present. With these small subject numbers, this study did not yield conclusive results. A discussion of the results, as well as, further improvements into the experimental methods is given. Applying these lessons learned, more accurate perceptual threshold testing can be conducted within the lab, allowing greater reporting confidence for future studies.

The additional study also aimed to observe SR, but expanded the noise spectrum to also find the masking that was not discovered in the first study. Four lab members completed pure tone audiometry with and without the presence of noise over a broad range of noise levels. A qualitative analysis suggests masking existed for frequencies with low thresholds given the noise levels that were tested. For some subjects, SR benefits may have been observed, but for others they did not appear to be present. With these small subject numbers, this study did not yield conclusive results. A discussion of the results, as well as, further improvements into the experimental methods is given. Applying these lessons learned, more accurate perceptual threshold testing can be conducted within the lab, allowing greater reporting confidence for future studies.

Share

COinS