FIG. 9.

Sensorineural hearing loss is predicted to degrade both fine-structure and envelope coding of speech in neural responses to psychoacoustic stimuli that contain only speech fine structure. Spike trains were obtained from model AN fibers with CF = 550 Hz. Responses are compared from three versions of the AN model: normal hearing (filled squares), 30-dB hearing loss due to selective outer-hair-cell (OHC) damage (open triangles), and 30-dB selective inner-hair-cell (IHC) hearing loss (open circles). Stimuli were the same as used in Figure 8. Dashed lines at ρ = 0.1 represent the estimated noise floor for uncorrelated conditions (see Fig. 3). The overall speech levels (35 dB SPL for normal hearing, 60 dB SPL for impaired conditions) were chosen based on the AN-fiber best-modulation level for the original speech for each model condition.