Discrimination of rippled spectra in listeners with hearing loss at two experimental paradigms

Abstract

Frequency resolving power (FRP) of hearing was measured in listeners 30 to 82 years old as a maximal resolved ripple density (ripples/oct) of rippled-spectrum signals. The rippled-spectrum signal should be distinguished either from another rippled signal with different positions of ripples on the frequency scale or from a “flat”-spectrum signal with no spectral ripples. Signals had either 2-oct wide frequency band centered at 1, 2, or 4 kHz, or had a wide frequency band 0.125–8 kHz. Listeners older than 60 had substantially increased tonal thresholds (haring loss). If the listener should distinguish a rippled test signal from a rippled reference signal, FRP correlated with the tonal threshold: the higher threshold, the lower FRP. For all the signals, the trend of the dependence was from –0.04 to –0.07 ripples/oct·dB. It is suggested that in this case, the test and reference signals were discriminated by the spectral mechanism and depended on the interrelation between the ripple density and acuteness of the frequency-tuned filters. If the listener should distinguish the ripple test signal from a flat reference signal, the FRP dependence on tonal threshold was not statistically significant for 2-oct-wide signals on condition that the signal level was above threshold. Or the wide-band signal, the FRP dependence on tonal threshold was statistically significant with a trend of –1.23 ripples/oct·dB. It is suggested that in this case, the test and reference signal were discriminated by the temporal-processing mechanism and depended on the perceived frequency range.