Analysis of a Stroboscopic Illusion of Motion

Abstract

A visual illusion reported by D. M. MacKay occurs when a steady light is viewed against a surrounding intermittently-illuminated circle. The light appears to jump out of the circle when (1) saccadic eye movements are made to or from the stationary light and circle, and (2) when a subject visually tracks the moving light and circle. We measured the frequency of the illusion under condition (1), and under modified versions of condition (2) in which the circle was replaced by a vertical line to one side of the light or by two vertical lines at the same distance from the light on opposite sides. The effects of varying eye movement distances and rates of intermittent illumination on the frequency of the illusion are investigated. The normal deviate corresponding to the probability of occurrence of the illusion on a single illumination flash is found to be linearly related to ln(x/r), where x is the distance the light moves between flashes, and r is the radius of the background circle or the perpendicular distance of the line to the light. The illusion occurs because the changing position of the circle or line is not predicted during the interval between flashes, and the position of the light is perceived relative to the previous location of the circle or line.