Effects of character size and chromatic contrast on reading performance

Recent work on the psychophysics of reading has concentrated primarily on achromatic (or luminance) contrast factors (Legge et al, 1985, 1986, 1987). In general, reading is fastest with text of high contrast defined by luminance contours. In the real world, however, chromatic as well as luminance edges abound. While several studies have examined the interaction of chromatic and luminance contrasts on stimuli near threshold (e.g., Mullen, 1987; Knoblauch et al, 1984; Jameson, 1985; Switkes et al, 1988), only a few have investigated suprathreshold levels. Legge et al, (1986) demonstrated that for observers with normal vision, the luminance contrast and not the color of text on a dark background determines reading rate. This finding is consistent with threshold studies that show that the color of a grating does not affect luminance contrast sensitivity either (Nelson and Halberg, 1979), provided that one avoids frequencies near the diffraction limit (Pokorny et al, 1968; Van Nes and Bouman, 1967).