Phase and gain of the visual transient system

Some properties of the transient channel of the human visual system were investigated by applying a drif -correcting perturbation method and using a 1° stimulus with a dark surround. By assuming small-signal linearity and peak detection, the frequency and the impulse response of the system could be determined. The gains derived from both kinds of data were equal within the limits of experimental precision. The phase obtained from the frequency response was found to change almost linearly with frequency and the minimum phase behaviour assumed earlier had to be rejected. In the drift-correcting perturbation method, the starting moment of the impulse response is lost. However, phase functions can be obtained by choosing a starting moment which does not exceed the limits imposed by causality. The phase function obtained from the impulse response agreed acceptably with that obtained from frequency data.