Abstract
Previous studies have shown that the number of photons detected by retinal cones can be estimated based on two contrast thresholds (with and without visual noise) measured under specific conditions. Indeed, contrast threshold can be limited by the variability in the number of photons detected (i.e., photon noise) under some conditions, in which case contrast threshold is inversely proportional to the square root of the luminance intensity. The present study aimed to adapt this psychophysical paradigm to estimate the number of photons detected by rods. Contrast thresholds to a motion direction discrimination task were measured in presence and absence of visual noise over a wide range of luminance intensities using either a red or blue monochromatic display. With a red display, contrast threshold in absence of noise was inversely proportional to the square root of the luminance intensity over a wide range of luminance intensities. At high luminance intensities, contrast thresholds were similar with the red and blue display at equal luminance intensities. At low luminance intensities, contrast thresholds were lower (i.e., better) with the blue display and inversely proportional to the square root of the luminance intensity. Given that rods are little sensitive to long wavelengths (i.e., red), contrast thresholds were undoubtedly limited by the photon noise of cones when using a red display and a blue display at high luminance intensities. The fact that contrast thresholds at low luminance intensities were better with the blue displays relative to the red display and since contrast threshold was inversely proportional to the square root of the luminance intensity under these conditions suggests performance was limited by the photon noise of rods. We conclude that the number of photons detected by rods can be evaluated by measuring motion contrast thresholds at low luminance intensities using a blue display.