Abstract
Detection threshold for small, brief rod-mediated probe flashes is proportional to the square root of adapting field intensity over a considerable range. Classically, this finding was explained in terms of a photon noise-limited ideal observer (Rose, 1942; de Vries, 1943). But retinal recordings from toad (Donner et al., 1990) and primate (Schwartz & Rieke, unpublished) support an alternative account based on gain changes in the rod network. The gain account is also supported by psychophysical data indicating that the brightness of small, brief flashes follows a square-root law for super-threshold flash intensities (Brown & Rudd, 1997). We used dichoptic brightness matching to study the time course of monocular square-root law gain changes in rod vision. Following 40 min dark adaptation, a 7.4 deg patch of each of the observer's retinas was exposed for 6 min to a 492 nm, 0.1 Rh*/rod-sec adapting field: midpoint of the square-root law range. The adapting level was then increased or decreased in one eye by a factor of 4. 10 msec, 0.55 deg, 492 nm flashes were presented simultaneously to the two adapted regions every 2 sec. The observer indicated in which eye the flash appeared brightest. Flashes were chosen to appear equally bright when the eye with the changed field had undergone a fixed proportion of the square-root law gain change associated with the new adaptation level. We varied this proportion to map the time course of light adaptation (159 and 178 sec; two observers) and dark adaptation (188 and 215 sec).
Meeting abstract presented at OSA Fall Vision 2012