Modeling interaural-delay sensitivity to frequency modulation at high frequencies.

Interaural-delay sensitivity to high-frequency (> or = 3 kHz) sinusoidal-frequency-modulated (SFM) tones is examined for rates from 25 to 800 Hz and depths of -12 to 18 dB. Comparison is made to thresholds obtained for sinusoidal-amplitude-modulated (SAM) tones for the same observers and modulation rates. Both SAM and SFM threshold-by-rate functions are U-shaped with optimum sensitivity to SFM tones occurring at higher rates (fm = 200-400 Hz) compared to those for SAM tones (fm = 100-200 Hz). Effects of modulation depth were examined for rates from 50 to 300 Hz. In all cases thresholds improved considerably with increasing modulation depth. It is also shown that a hybrid dichotic signal composed of an SFM tone presented to one ear and an SAM tone to the other, can perceptually fuse and be lateralized, with the contingency that both stimuli have equal modulation rates but not necessarily equal carrier frequencies. Using bandpass noise to restrict off-frequency listening, it was shown that for this stimulus, observers can use information from filters either below or above the carrier frequency. Consistent with FM-to-AM conversion from cochlear bandpass filtering, several important differences between the SAM- and SFM-tone data can be predicted from a nonstationary stochastic model of binaural interaction whose parameters are uniquely determined from the SAM-tone data.