Red-green mixture thresholds in congenital and acquired color defects.

A color television display was used to measure thresholds for mixtures of red and green on a white background; red and green components could be either incremental, decremental or zero. Ellipses are fitted to a plot of green contrast as a function of red contrast, and it is argued that the length of the ellipse is a measure of red-green color discrimination and the width of the ellipse is a measure of luminance discrimination. It is shown that the technique reliably distinguishes normals from congenital color defectives and also protan from deutan subjects. For some cases of acquired color defects (e.g. optic neuritis), there is a roughly equal loss of color and luminance discrimination whereas, in other cases (e.g., hereditary optic atrophies), the loss of color discrimination is much greater than the loss of luminance discrimination.

Color mixture thresholds measured on a color television--a new method for analysis, classification and diagnosis of neuro-ophthalmic disease.

A color television display can be used to determine color and brightness discrimination thresholds using identical adaptation conditions and experimental technique. The color discrimination threshold is measured by using an equiluminous test spot--i.e. one which differs in color from the surrounding screen but has the same luminance. Because there is no brightness clue, the subject is forced to detect such a spot by using color discrimination. It is shown how color and brightness thresholds may be determined from threshold measurements of different color-mixtures even though it is not known beforehand which stimulus will be equiluminous for the subject. Results are shown for normal subjects, congenital color defectives and for two patients having optic nerve disease who show respectively non-selective and selective loss of color discrimination compared to brightness discrimination. Normal control data are presented, illustrating the effect of eccentricity, optical blur, viewing distance, pupil size and age. It is concluded that the technique is relatively insensitive to moderate variations in these factors and that it is more sensitive in detecting selective color loss than a spectral sensitivity technique which has been described previously.