Fig 1.
Pure-tone audiometry for participants.
Audiograms for normal-hearing (NH; n = 7; dotted lines), normal-hearing at equivalent sensation level (NHSL; n = 8; dashed line), and hearing-impaired (HI; n = 9; solid lines) listeners for both left (blue X’s) and right (red circles) ears in dB HL. The HI listeners were screened to have roughly flat and symmetric mild-to-moderate hearing losses. Error bars represent one standard error of the mean.
Table 1.
Individual age and audiometric data.
Fig 2.
Spectrograms of modulated maskers used in the glimpsing task.
From left to right, the synchronously-modulated masker (Sync) and the asynchronously-modulated masker (Async) with 2, 4, 8, or 16 numbers of frequency bands. Modulation rates were set at 10 Hz, and initial phase of modulation was random.
Fig 3.
Schematics of monaural and dichotic masking conditions.
Primary conditions are represented on the top row, and control conditions are shown below. Only left-ear (L) schematics are visually depicted for monaural conditions, but right (R) ear conditions were also tested. As the legend indicates, each condition is represented as a 2-by-2 box in which the left and right columns represent stimulation of the left and right ears, respectively, and the top and bottom rows represent the speech and noise stimuli, respectively. In each box, frequency from 0.1 to 10 kHz is represented vertically, and a time span of 200 ms is represented horizontally. Speech is represented via spectrogram, and noise is represented in black. Amplitude modulation is performed at a rate of 10 Hz, and frequency bands are equally spaced on a logarithmic scale. The order of the primary conditions in the top row is an indication of the expected ranking in thresholds, with the worst performance starting on the left, with the Unmod-L and Unmod-R conditions, and the best performance on the right, with the Sync-L and Sync-R conditions. The numbers of bands tested per condition are given below each condition schematic.
Fig 4.
Masking release relative to unmodulated noise.
Mean masking release is plotted for modulated noise conditions relative to the unmodulated condition for hearing-impaired (HI; left panel) normal controls with stimulus presentation at either the same intensity (NH; middle panel) or same sensation level (NHSL; right panel). The difference in mean thresholds relative to the Unmod-M condition at 2, 4, 8, and 16 bands are plotted for the monaural asynchronous condition (Async-M; circles), the dichotic asynchronous condition (Async-D; triangles), the better of the dichotic control conditions (Control-D; bowties), and the mean of the synchronous conditions (Sync-M; dashed line). Error bars indicate standard error of the mean (n = 9 for HI group; n = 7 for NH group; n = 8 for NHSL group). The shaded region at the bottom of each panel indicates the range over which masking release might be accounted for by reductions in masker level associated with amplitude modulation (3.2 dB).
Table 2.
Mean speech recognition thresholds (in dB SNR) for each primary test condition and the better of the control conditions.
Fig 5.
The effect of hearing loss on dichotic benefit.
Dichotic benefit (i.e., the difference between Async-D and Async-M conditions) for hearing-impaired (HI) and normal hearing controls (NH and NHSL). Error bars indicate one standard deviation. Whereas the dichotic advantage increased significantly for NH listeners as band number increased, HI and NHSL listeners received similar benefit at all band numbers (NHSL group was only tested at 4 and 8 bands). Significant differences between groups are indicated by an asterisk (p < 0.05) or two asterisks (p < 0.001).