Similarity of dynamic range adjustment in auditory nerve and cochlear nuclei

1. Rate versus level functions were recorded for responses to best-frequency (BF) tones of 116 cochlear nucleus units and 53 auditory-nerve fibers in the presence of interrupted tone backgrounds and continuous noise backgrounds of various intensities. The backgrounds shifted the dynamic ranges of rate-level functions to higher test intensities, so in the presence of backgrounds, rate saturation occurred at higher intensities than in quiet. 2. The shift in saturation intensity evoked by each background was measured by comparing the rate-level function recorded with the background to one recorded without. The relation between change in saturation intensity and background intensity could be approximated by the formula ΔI(sat) = 0 I < θ; ΔI(sat) = A(I - θ) I ≥ θ. ΔI(sat) is the shift in saturation intensity, I is the background intensity, θ is the threshold for evoking shift, and A is the ratio of shift to background intensity re θ. In the appendix, it is shown that A is a measure of a unit's ability to avoid saturation by the background stimulus. The optimal value of A is unity, at which point a unit's operating range is infinite. 3. The value of A depended on BF for interrupted tone backgrounds, but not for continuous noise backgrounds. For BF < 10 kHz, the mean value of A for tone backgrounds was 0.33 in the auditory nerve, 0.37 in the ventral cochlear nuclei (VCN), and 0.47 in the dorsal cochlear nucleus (DCN). The difference between auditory nerve and VCN was not statistically significant. For BF > 10 kHz, the mean A was 0.16 in auditory nerve and 0.30 in VCN. 4. The mean value of A for noise backgrounds was 0.79 in auditory nerve, 0.86 in VCN, 0.86 in DCN units of response types II and III, and in 1.04 in DCN type IV units. Only the differences between DCN type IV and the non-DCN unit groups were statistically significant. 5. The qualitative changes produced in rate-level functions by tone and noise backgrounds were similar in auditory nerve and cochlear nuclei except for DCN type IV units. The shifts in rate functions produced by interrupted tone backgrounds did not prevent saturation of the rate response at background intensities above the dynamic range of the unit as recorded in quiet. However, the rate response to test tones was preserved in the presence of all noise background levels used (up to a 30-dB spectrum level). The shift in rate function produced by the noise was almost sufficient to allow the unit to encode test intensity relative to noise background intensity. 6. On the basis of the quantitative and qualitative similarities between the effects of backgrounds on auditory-nerve responses and cochlear nuclei responses, we conclude that dynamic range adjustment occurs primarily in the cochlea. 7. On the basis of the preservation of rate response to test tones embedded in noise backgrounds, we conclude that rate is an adequate code for conveying fine intensity discrimination in the presence of noise maskers.