Abstract
Purpose: To measure the contrast sensitivity function for adaptation to luminance and chromatic (red-green) flicker in the human cone system. Methods: Following pre-adaptation to a range of flicker frequencies (luminance: 5–50 Hz; chromatic: 4–40 Hz) and contrast levels (1–100%), contrast thresholds for a 10 or 30 Hz test were obtained from three observers. Test frequency was selected to ensure that both adaptation and detection were mediated by a single mechanism. For each adapting frequency, a nonlinear template was horizontally scaled for best fit to the function relating test threshold to adapting contrast. The contrast sensitivity function (CSF) for flicker adaptation was derived by plotting the best-fit scaling factor as a function of adapting frequency; this was compared to the conventional CSF. Results: (1) The CSF was notably shallower for adaptation than for perception: at CFF (40 Hz luminance, 20 Hz chromatic), attenuation from peak sensitivity was about 1.8 log units for perception, but only 0.9 and 0.5 log unit for adaptation to luminance and chromatic flicker, respectively. The adaptation CSF was shallower for luminance than for chromatic flicker, but in both cases, continued beyond the respective CFF, revealing adaptation to invisible flicker. A supra-threshold nonlinearity was ruled out as an explanation, because the functions relating test threshold to adapting contrast were of the same shape, but reached progressively lower peak values with increasing adapting frequency. Conclusions: In both the color and luminance pathways of the cone system, substantial neural filtering of high temporal frequencies occurs at sites both proximal and central to the site of flicker adaptation. The disproportionately greater filtering within the color pathway begins prior to the site of flicker adaptation.
Supported by NIH grant EY-01711