Abstract
PURPOSE: Theories of brightness induction propose that the brightness of a test patch within a complex surround can be explained by local contrast or by a weighted sum of contrast from each point within the surround. In this study, an alternative hypothesis is tested that includes brightness assimilation. METHODS: Using an asymmetric matching task, observers set the brightness of a comparison ring to match the brightness of a test ring. The comparison ring always was presented on a uniform background (24 cd/m2). The test field was composed of one test ring and eight concentric inducing rings, four on either side of the test ring. The luminance of the test ring was fixed at 16 cd/m2. The luminance of the rings contiguous with the test was varied from 18 to 24 cd/m2. The luminance of the non-contiguous rings was set at 16 or 24 cd/m2. RESULTS: Surprisingly, changing the luminance in the non-contiguous rings caused a larger shift in brightness of the test than a similar luminance change in the contiguous rings. Further, assimilation was observed when the luminance of the non-contiguous rings was fixed at 24 cd/m2: in this case, the brightness of the test increased as the luminance of the contiguous rings increased. CONCLUSIONS: Varying the light level in non-contiguous regions of a patterned background can alter the brightness of a test more than a similar change in a contiguous region. Further, increasing the luminance in a contiguous region can result in assimilation. These results are inconsistent with the idea that brightness induction can be explained by either local contrast or a weighted sum of contrast from light through out the background. They can be explained, instead, by simultaneous processes of contrast and assimilation.