Plasmas: A New Class of Motion-Induced Brightness Illusion

When a white figure is set in smooth continuous motion against a dark background, a ‘shadowy’ region forms around and behind it, somewhat similar to the wake that a boat leaves in water. Conversely, when a dark figure moves against a white background, this ‘wake’ appears lighter than the background. When several such figures move, these ‘wakes’ can link up in surprising ways. For example, if four white circles of radius r are positioned on a dark background at the vertices of an imaginary square whose width is, say, 4r, and this imaginary square rotates about its centre, bands darker than the background form between the four circles, linking them through the centre. This ‘black cross’ is nonrigid, and is accompanied by dark ‘wakes’ on the outside of the circles. I call these illusory brightness regions ‘plasmas’.

The relation of this effect to other motion-induced brightness illusions is considered. In particular, this effect has different properties than either induced gratings (McCourt, 1982 Vision Research 22 119 – 134) or phantom gratings (Tynan and Sekuler, 1975 Science 188 951 – 952). Plasmas are not due to retinal afterimages or persistence, since they form in regions never traversed by the moving figure. Nor are they an artifact of the computer screen, since the illusion occurs equally well in displays constructed from paper. Plasmas have the opposite phase relative to the inducing figure, disappear when a figure stops moving, are strongest at high background/inducer contrast, and are most visible when the luminance of the inducers is in the photopic range. Experiments measuring the strength of plasmas as a function of angular velocity and visual angle are described. A model is proposed according to which plasmas are due to a phase lag in luminance-edge induced lateral inhibition.