Two-Color Threshold as Test of Color Vision*

Abstract

The two-color threshold is obtained by finding the just-visible luminance of a test flash of one color, seen against a background of another color. For normal observers, threshold luminance for such heterochromatic conditions is lower than when the test and background stimuli are of the same color. The factor by which color differences alone reduce the increment threshold is defined as the heterochromatic threshold reduction factor (HTRF). A color test capable of measuring the HTRF is devised and applied to 67 subjects, 21 of whom are red-green defective according to five independent criteria. When filtered red and green stimuli are used, the HTRF is shown to be zero for most color-defective subjects, within the limits imposed by measurement error. For normal subjects, the average red-green HTRF varies from about 2 to 4, depending upon conditions. An examination of green-on-red thresholds plotted against red-on-green thresholds reveals a separation of normal subjects from the color-defectives, and also makes discriminations within the color-defective group. The test as a whole misclassifies no normals as defective, properly classifies all color defectives as such, and very clearly classifies all color defectives as to gross type (protan vs deutan). Further work must be done to evaluate the test’s ability to discriminate dichromats from anomalous subjects, although the results in hand are very promising. Within the color-defective group, the test seems to provide a quantitative assessment of the degree of color defect; it also appears capable of detecting weak normals.