Detection of gaps in sinusoids by frog auditory nerve fibers: importance in AM coding

Physiological studies were carried out in the frog (Rana pipiens pipiens) eighth nerve to determine: (i) whether the modulation rate or the silent gap was the salient feature that set the upper limit of time-locking to pulsed amplitude-modulated (PAM) stimuli, (ii) the gap detection capacity of individual eighth nerve fibers. Time-locked responses of 79 eighth nerve fibers to PAM stimuli (at the fiber's characteristic frequency) showed that the synchronization coefficient was a low-pass function of the modulation rate. In response to PAM stimuli having different pulse durations, a fiber gave rise to non-overlapping modulation transfer functions. The upper cut-off frequency of time locking was higher when tonepulses in PAM stimuli had shorter duration. The fact that the cut-off frequency was different for the different PAM series suggested that the AM rate was neither the sole, nor the main, determinant for the decay in time-locking at high AM rates. Gap detection capacity was determined for 69 eighth nerve fibers by assessing fiber's spiking activities to paired tone-pulses during an OFF-window and an ON-window. It was found that the minimum detectable gap of eighth nerve fibers ranged from 0.5 to 10 ms with an average of 1.23–2.16 ms depending on the duration of paired tone pulses. For each fiber, the minimum detectable gap was longer when the duration of tone pulses comprising the twin-pulse stimuli was more than four times longer. When the synchronization coefficient was plotted against the silent gap between tones pulses in the PAM stimuli, the gap response functions of a fiber as derived from multiple PAM series were equivalent to gap response functions deriving from twin-pulse series suggesting that it was the silent gap which primarily determined the upper limit of time-locking to PAM stimuli.Abbreviations MTF modulation transfer function - PAM pulse amplitude modulated - SAM sinusoidally amplitude modulated - SC synchronization coefficient - TW time window