Abstract

A novel spatial method is used to measure the distinctness of peripherally viewed chromatic and achromatic borders. The method determines the degree to which sharp (step-edged) borders must be blurred to be just noticeably different, using Gaussian blur and forced choice. Stimuli consist of 0.5-s duration, centrally fixated disks inserted in backgrounds of another color. This configuration provides maximal border length and large color fields. One achromatic and three isoluminous chromatic borders are tested at five retinal eccentricities from 1.25 to 20° and at both low photopic and mesopic levels of illumination. Blur thresholds for low photopic achromatic borders can be scaled to inverse cortical magnification functions. Chromatic border distinctness for yellow-blue (tritan metamers) is poor below about 2.5° eccentricity and is very poor for red-green (isoblue cone) above 10°. Chromatic border distinctnesses are otherwise comparable with each other but are degraded relative to achromatic at all eccentricities. Mesopic blur thresholds are larger than photopic, the lower eccentricities being more affected in all cases other than red-green, which becomes worse mesopically at high eccentricities as well. The possibility of contributions by rods is discussed.

© 1987 Optical Society of America