Modulation Detection in Subjects With Relatively Flat Hearing Losses

Abstract
Modulation detection thresholds were measured as a function of modulation frequency in 5 normal-hearing subjects and in 8 subjects with relatively flat, slight-to-moderate hearing losses. The carrier was a broadband noise that was sinusoidally amplitude modulated (SAM) in one of two observation intervals The spectrum level of the carrier ranged from –10 to 50 dB SPL, and, for a given carrier level, modulation frequency varied from 2 to 1024 Hz. The temporal modulation transfer functions (TMTFs) were fitted very well with a simple equation describing a low-pass filter function The TMTFs from the normal-hearing subjects were relatively independent of carrier level, although the derived time constant tended to increase slightly with decreases n carrier level, from an average value of 2.5 msec at 30 dB SPL to 6.0 msec at –10 dB SPL. In addition, sensitivity to amplitude modulation (AM) decreased by about 4 dB as the pressure spectrum level of the carrier was decreased from 0 to –10 dB SPL. The TMTFs from 7 of the 8 hearing-impaired subjects were similar to those from the normal-hearing subjects when the carriers were presented at equal SPLs, except that the derived time constants were slightly larger in the subjects with hearing impairment When comparisons were made at comparable sensation levels (SLs), however, the TMTFs from the two groups of subjects were quantitatively similar, with the exception that at the lowest SL (20 dB), hearing-impaired subjects typically were more sensitive to AM than normal-hearing subjects, and the derived time constants from their TMTFs were somewhat smaller These results, taken together with previously published results, suggest that a broad listening bandwidth is important for normal performance on a temporal resolution task. That the time constant from one of the hearing-impaired subjects was significantly longer than normal, regardless of whether the comparisons were made at equal SPL or equal SL, indicates that other factors can also be important

This publication has 25 references indexed in Scilit: