| Popis: |
We have previously shown that the temporal patterns of auditory‐nerve fiber discharges provide a robust representation of the spectra of periodic steady‐state vowels [Young and Sachs, J. Acoust. Soc. Am. 66, 1381–1403, (1979)]. The purpose of the study described here was to determine whether the spectra of a whispered (aperiodic) vowel could be represented by such a temporal code. A terminal analog synthesizer was excited by white noise. The first three formant frequencies of the synthesizer were set to 530, 1840, and 2480 Hz which correspond to the vowel /e/. The resulting whispered vowel was presented in 400 ms bursts once per second. Responses to the vowel were recorded in as many as 350 auditory‐nerve fibers in individual cats. Interval histograms of fiber responses were constructed from the data and discrete Fourier transforms with a frequency resolution of 64 HZ were computed for each interval histogram. For each fiber the magnitude of the Fourier transform component nearest its CF was used as a mea... |
