Abstract
Extended viewing of blurred or sharpened imagery results in normalization of image appearance, sharpening imagery in response to blur and vice versa (Webster, Georgeson, & Webster, 2002), thus establishing that blur adaptation has a beneficial effect on perception. However, the question of how adaptation affects sensitivity to these qualities has not been addressed in a comparable paradigm. We measured observers' sensitivity to blur and sharp (negative or positive deviations in the slope of a stimulus' amplitude spectrum) during adaptation to blurred, normal, or sharpened video segments. Using a simultaneous 2AFC design, video adaptation was compared with blank adaptation by measuring matching functions for test videos presented in video- or blank-adapted locations. Results show that sensitivity to blur and sharp is significantly higher after adaptation to normal video, as compared with blank, blur, or sharp adaptation; or stated conversely, that blur and sharp sensitivity is significantly impaired by adaptation to sharpened, blurred or blank imagery. These results suggest that the normalization effects observed in earlier experiments come at a cost: the visual system is able to restore the normal appearance of distorted imagery, but in the process it loses some ability to sense further distortions. Furthermore, it is strongly suggested that adaptation to a blank display, often used as a null condition in vision experiments, results in suboptimal visual ability.