Purpose: To measure the effect of spatial frequency on luminous impulse response functions (IRFs).

Method: The stimuli were sinusoidal gratings in spatial frequencies from 0.43 to 13.35 cpd filtered by a Gaussian envelope.

Two brief luminous flashes were displayed successively at variable stimulus onset asynchronies (10 to 180 ms) on a white background (Luminance = 10cd/m^2). Detection thresholds were measured using a four-alternative forced-choice method with four observers. Burr and Morrone's model equation (J. Opt. Soc. Am. A 10, 1993) was used to obtain the impulse response function. In their model, the thresholds were determined by probability in detecting to the combined two impulse responses over time and the impulse response was described by exponentially damped frequency-modulated sinusoidal function with four parameters.

Results and Discussion: We used zero crossing time from excitatory phase to inhibitory phase as temporal property of the IRF. In the three lower spatial frequencies (0.43, 0.89 and 1.34cpd), the zero crossing time tends to be slowest (45ms). For five highest spatial frequencies (from 4.90 cpd to 13.35 cpd), it tends to be faster (35ms). For spatial frequencies around 3cpd, the zero crossing time of IRFs is much faster and it is fastest (about 18ms) at 3.12cpd among all spatial frequencies tested. These spatial frequency differences were paralleled by differences in the ratio of the excitatory to inhibitory amplitudes. These results suggest that a human spatial frequency channels centered around 3 cpd is faster than channels subserving spatial vision at higher and lower spatial frequencies.

David C. Burr & M. Concetta Morrone (1993). Impulse-response functions for chromatic and achromatic stimuli : Journal of the Optical Society of America A 10, 1706-1713

Keiji Uchikawa & Tatsuya Yoshizawa (1993). Temporal responses to chromatic and achromatic change inferred from temporal double-pulse integration : Journal of the Optical Society of America A 10, 1697-1705