Development of a method for determining binaural sensitivity to temporal fine structure

Christian Füllgrabe,Andrew J Harland,Aleksander P Sęk,Brian C J Moore

doi:10.1080/14992027.2017.1366078

Abstract

Objective: To develop and evaluate a test of the ability to process binaural temporal-fine-structure (TFS) information. The test was intended to provide a graded measure of TFS sensitivity for all listeners. Design: Sensitivity to TFS was assessed at a sensation level of 30 dB using the established TFS-LF test at centre frequencies of 250, 500 and 750 Hz, and using the new TFS-AF test, in which the interaural phase difference (IPD) was fixed and the frequency was adaptively varied. IPDs varied from 30 to 180°. Study Sample: Nine young (19–25 years) and 23 older (47–84 years) listeners with normal hearing over the tested frequency range. Results: For the young listeners, thresholds on the TFS-AF test did not improve significantly with repeated testing. The rank-ordering of performance across listeners was independent of the size of the IPD, and moderate-to-strong correlations were observed between scores for the TFS-LF and TFS-AF tests. Older listeners who were unable to complete the TFS-LF test were all able to complete the TFS-AF test. Conclusions: No practice effects and strong correlations with an established test of binaural TFS sensitivity make the TFS-AF test a good candidate for the assessment of supra-threshold binaural processing.

Highlights

Broadband signals like speech are decomposed in the peripheral auditory system into a number of bandpass-filtered signals corresponding to the outputs of different auditory filters
Since the adaptive procedures used steps that were based on fixed factors rather than fixed arithmetic steps, all means were calculated as geometric means, and statistical analyses were based on the log-transformed data
One exception was the mean threshold for the second block for YNH9, which was over 2000 Hz

Summary

Introduction

Broadband signals like speech are decomposed in the peripheral auditory system into a number of bandpass-filtered signals corresponding to the outputs of different auditory filters. TFS information is conveyed by neural synchrony (phase locking) to individual cycles of the TFS, and phase locking becomes very weak for frequencies above 4000–5000 Hz, the exact upper limit in humans is not known (Verschooten and Joris 2014). It has been known for many years that TE information is important for speech intelligibility (Dudley 1939), and more recent evidence suggests that medium-rate modulations (from 4 up to about 16 Hz) are especially important for the intelligibility of speech in quiet (Drullman, Festen, and Plomp 1994; Shannon et al 1995), while a somewhat wider range of modulation rates may be important for speech in background sounds (Fullgrabe, Stone, and Moore 2009; Stone, Fullgrabe, and Moore 2009, 2010)

Methods

Results

Conclusion