Physiological recordings were made from single auditory fibers in the frog eighth nerve to determine quantitatively how the different behaviorally relevant temporal parameters (the signal rise-fall time, duration, and rate of amplitude modulation) of complex sounds are encoded in the auditory periphery. Individual temporal parameters were varied. Response functions (RFs) were constructed with respect to each of these parameters using each unit's best excitatory frequency (BF) as the carrier. In response to a change in signal rise-fall time, auditory nerve fibers showed little change in the mean spike count or firing rate, i.e., all fibers displayed ALL-PASS RF (rft)s. But the transient components, particularly the early phasic component, of responses varied with rise-fall times; these components were more pronounced in the responses to stimuli with shorter rise-fall times. In response to an increase in signal duration, auditory nerve fibers showed a corresponding increase in firing duration and thus in the mean spike count, giving rise to HIGH-PASS RF(dur)s. The shape of response curves differed among fibers; the difference appeared to be related to the fiber's temporal adaptation characteristic. When the firing rate was measured, all fibers displayed higher mean firing rates in response to shorter duration stimuli than they did to longer duration stimuli, thus giving rise to LOW-PASS response functions. To determine the response transfer functions to modulation rate, pulsed (PAM) and sinusoidally (SAM) amplitude-modulated signals were used. These signals differed substantially in terms of their envelopes and how they varied with AM rate. Data were analyzed by 1) plotting spike counts against the AM rate to derive modulation transfer functions (MTF(spk)s) and 2) plotting synchronization coefficients (SCs) against the AM rate to generate MTF(sc)s. In response to PAM stimuli, all fibers showed an increase in mean spike count with modulation frequency over the range examined, giving rise to HIGH-PASS MTF(spk)s. For SAM stimuli, the average energy and duty cycle are independent of AM rate. Most (79%) auditory fibers showed little selectivity for AM rate over a range of 5-400 Hz, giving rise to ALL-PASS MTF(spk)s. The remaining auditory fibers displayed LOW-PASS MTF(spk)s, i.e., there was a distinct decline in the mean spike count with increasing AM rate. In response to PAM stimuli, most fibers showed good response synchrony at low AM rates but the SC declined with an increase in the AM rate (i.e., LOW-PASS MTF(sc)s). The cut-off frequency was typically very high, averaging 90 pulses/s. In response to SAM stimuli, auditory fibers exhibited little change in the SCs when the AM rate was altered, giving rise to ALL-PASS MTF(sc)s. When the time-locked response of a fiber to a modulating waveform was compared with the phase-locked response to the carrier waveform, it was clear that for most fibers these responses were distinctly different. Most high-frequency sensitive fibers exhibited poor phase-locked response to low frequency tones because of their insensitivity to such stimuli, but displayed good time-locked responses to AM signals over a wide range of modulation frequencies. Conversely, many low-frequency and some high-frequency fibers exhibited precise phase-locked responses to low-frequency carriers but gave poor time-locked responses to the modulating waveform at equivalent frequencies. In general, our data are in agreement with eighth-nerve data obtained from other vertebrate species, showing that auditory nerve fibers function primarily as 'envelope detectors.' The data generated in this study, when compared with those derived from central auditory neurons, give us some insights into the algorithms by which the nervous system computes the different temporal auditory parameters.
