Equating visibility of brief decrements: unconfounding duration and luminance

A common procedure in visual psychophysics involves equating the visual effectiveness of brief luminous displays. It may be equally important to equate the effectiveness of brief interruptions, as when two displays are presented sequentially, separated by a variable interstimulus interval (ISI). For example, in a procedure devised by Phillips and Singer [Expl. Brain Res. 19, 493-506 (1974)], the first display consisted of a random pattern of dots and the second display consisted of the same pattern, but with one added dot. Detectability of the added dot was presumed to be determined by interactions of transient neural events produced at the beginning and end of the ISI. Lengthening the ISI was believed to weaken progressively the magnitude of the neural interactions, resulting in poorer performance. But lengthening the ISI also increased its visual effectiveness (darkness). Using ISIs equated in visual effectiveness for durations from 10 to 320 msec, we found that the visual effectiveness of the interval, not its duration, was the prime determinant of performance. This finding requires a reinterpretation of the neural mechanisms being studied in the Phillips and Singer task.