How does brightness induction work at night?

Abstract

At low light levels, our visual sensation is mediated by phototransduction of rods. These photoreceptors are anatomically and functionally distinct from cone photoreceptors that mediate vision in daylight. However, it is not well understood how rod perception differs from the well-known cone perception. Brightness induction is a complex perceptual phenomenon that can produce two opposing effects (assimilation and contrast) and explanations range from low to high level theories. The aim of this study was to understand how brightness induction is achieved when only rod photoreceptors are active. We performed three experiments on young observers: 1) The brightness induction strength was obtained for two stimulus configurations: White's illusion (assimilation) and simultaneous brightness contrast, under different light adaptation conditions favoring cones or rods; 2) We used silent substitution on a tetrachromatic display to selectively stimulate rods or cones at a fixed light adaptation level, and measured the brightness induction for assimilation and contrast for isolated rods, isolated cones, and combined rods and cones; and 3) We estimated the perceived transparency of samples with different opacities at different light adaptation levels. Our results showed that compared to cone vision, rod-mediated brightness induction was stronger for assimilation, but doesn't change for contrast. Furthermore, transparency was underestimated at low light levels. These results suggest that brightness assimilation is mediated by a high-level transparency percept; following the scission theory (Anderson, 1997), the stronger induction is linked to a transparency underestimation at night light levels. In contrast, rod-mediated brightness contrast can be explained by low-level mechanisms.