Abstract
The perception of brightness is usually assessed under photopic viewing conditions, where vision is mediated by cone photoreceptors. However, under mesopic viewing conditions, rods have a profound effect on brightness perception. Here, we investigated how isolated cone and rod signals contribute to two spatial brightness induction phenomena: assimilation and contrast. We used an ad-hoc tetrachromatic projector (VPixx Technologies Inc.) and the silent substitution method to display isolated rod (Cmax = 72%) and cone (Cmax = 52%) stimuli. The stimuli consisted of White’s Illusion (assimilation) and Simultaneous Brightness Contrast displays, with embedded matching and reference patches. The brighter patch had to be chosen by participants in a constant stimuli paradigm. The normalized difference between the patches’ excitations at the point of subjective equality defined the induction effect. The stimuli were presented monocularly at 6.5° from the fovea. In a second experiment with a trichromatic monitor (no photoreceptor isolation), we analyzed the effect of light adaptation from scotopic to photopic conditions. For the rod condition, the induction effect was significantly higher for assimilation than for contrast. However, the assimilation and contrast effects were similar for the cone condition. These results mimicked the increase of assimilation with decreasing light levels from photopic to scotopic viewing conditions. In a subsidiary experiment, we tested the low-light-level-poor-spatial-resolution hypothesis: smoothed stimuli would maintain the same induction effect across light levels. However, we found similar trends with and without smoothing, rejecting this alternative hypothesis. Rod vision enhanced brightness assimilation, but it did not affect brightness contrast when compared to cone vision. These results indicate that brightness contrast and assimilation have different origins and highlight the important role of low-level mechanisms in brightness perception.