We have investigated the auditory representation of vowels with low-frequency formants by recording the activity of auditorynerve fibers in anesthetized cats in response to Japanese /i/-/e/ synthetic-vowel continua. Vowels having either low (150 Hz) or high (350 Hz) fundamental frequency F0 were varied in either first-formant frequency F1 or the level of a “crucial harmonic” near F1 to span the /i/-/e/ continuum. Two different neural representations of the stimulus spectrum in the F1 region were examined: a population rate-place profile and a population interspike interval distribution. Characteristics of both representations depend on F0. When individual harmonics are resolved by the ear, as for high F0s, first formant frequency does not have explicit correlates in either ANF rate-place patterns or interspike interval distributions. Rather, both representations show clear patterns corresponding to individual harmonics, as well as the amplitude ratios of “crucial harmonics” near F1 that determine vowel identity in psychophysical tests. When harmonics are not resolved, as for low F0s, both rate-place and population-interval profiles of individual harmonics fuse to form broader, single peaks near F1, providing an explicit neural representation of formant frequency.
[1]
J. Pickles,et al.
Psychophysical frequency resolution in the cat as determined by simultaneous masking and its relation to auditory-nerve resolution.
,
1979,
The Journal of the Acoustical Society of America.
[2]
Brian R Glasberg,et al.
Derivation of auditory filter shapes from notched-noise data
,
1990,
Hearing Research.
[3]
D. D. Greenwood.
A cochlear frequency-position function for several species--29 years later.
,
1990,
The Journal of the Acoustical Society of America.
[4]
B. Delgutte,et al.
Neural correlates of the pitch of complex tones. I. Pitch and pitch salience.
,
1996,
Journal of neurophysiology.