Abstract
Using a two-stimulus afterimage procedure, VanHorn and Francis (VSS 2007) reported on a new type of color afterimage. When observers viewed two orthogonal bar gratings in sequence they reported seeing an afterimage that was the same orientation as the first image. Further, observers reported that the color of the bars in the afterimage had switched from the color of the bars in the first bar grating of the inducing sequence. When the first bar grating included blue and black bars, observers often reported that the afterimage contained black and blue bars with the colors in switched positions relative to the inducer. On other trials observers reported seeing a negative color afterimage rather than a switch afterimage. We now report that the probability of seeing a switch or negative afterimage depends on the intensity of the viewing surface. We varied the luminance of the viewing surface and asked observers to report on the afterimage color of one of the afterimage bars. Switch afterimages were reported more often when the luminance of the viewing surface was low, while negative afterimages were reported more often when the luminance was high. The role of viewing surface luminance is similar to its role for positive afterimages, which are more common when the viewing surface is dark (Robertson & Fry, 1937; Howard, 1960).