Purpose. We extracted spatial channels underlying scotopic contrast sensitivity functions (CSFs) of normal adult observers, and compared them to channels underlying photopic contrast sensitivity functions. Methods. Scotopic CSFs were measured on 50 observers between the ages of 20 and 88 years (Schefrin et al. 1999). These CSFs were measured psychophysically using 2AFC at 7 spatial frequencies ranging from 0.2 to 3.0 c/deg, with mean retinal illuminance equated for observers at -0.85 log scotopic Trolands. Results. A covariance analysis of individual differences was applied to the data. We computed statistical sources of individual variability, used them to define "covariance channels," and determined the number and frequency tuning of these channels. We found evidence for 3 discrete channels operating below 3 c/deg. Two covariance channels operated primarily above 1 c/deg, and were predicted and well fit by a computational model of photopic vision (Wilson &Gelb, 1984; mechanisms "A" and "B"). A third channel, unlike channels found for photopic vision, was optimally tuned to very low spatial frequencies (approximately 0.3 c/deg). Conclusions. (1) The results are consistent with previous investigations indicating that channels tuned below 1 c/deg exist for scotopic but not photopic vision. However, the difference between photopic and scotopic mechanisms appears to be a single low spatial frequency channel. (2) The results have implications for understanding channel development. Channel tuning data from infants are generally consistent with the idea that channels shift in scale from lower to higher spatial frequencies with age, due to cone migration into the fovea and changes in eye size. However,the presence of a low spatial frequency channel in infants could merely reflect a shift from rod- to cone-dominated vision.
- Motion &deficits posters
ASJC Scopus subject areas