Release from upward spread of masking in regions of high-frequency hearing loss

The upward spread of masking was compared for 500-Hz quasifrequency-modulated (QFM) and sinusoidally amplitude-modulated (SAM) maskers. The modulation rate was 20 Hz. These maskers had identical magnitude spectra but different envelopes, which were relatively flat for the QFM masker and strongly fluctuating for the SAM masker. At signal frequencies more than an octave above the masker, masked thresholds for the SAM masker were lower than for the QFM masker, revealing "masking release" (QFM-SAM masked threshold differences) exceeding 30 dB in normal-hearing ears. In ears with high-frequency sensorineural hearing loss, but normal hearing in the region of the masker, masking release was markedly reduced or completely absent in regions of hearing loss. The data were evaluated with a model of masking based on the linearized response growth (LRG) of basilar membrane transfer functions associated with cochlear damage in animals. The LRG model predicted more gradual slopes of the growth of masking and reduced amount of masking in regions of hearing loss. The reduced masking release seen in regions of hearing loss could be largely accounted for by a more rapid growth of response to the probe tone in regions of hearing loss.