Reading With The Rod Visual System

People usually read under daylight illumination using their "day vision." Little is known about whether people are able to read exclusively with their night (i.e., rod) vision; and if they can whether reading with the night vision system differs from ordinary reading.

The present research investigates reading with rod vision. A special method of displaying text was devised such that the text could be read with rod but not cone vision. The key features of the method included the use of a blue chromatic filter and the selection of an appropriate luminance range. Results show that this method provides a 3 log unit luminance range over which the observer can read exclusively with the rods.

The first study investigated whether people can read exclusively with their rod visual system. The results show that this is not only possible but that normal subjects are also capable of reading at luminance levels only three times absolute threshold. It is possible that with the present text size that subjects can read at even lower luminance levels near threshold though reading rates would be very low (e.g., approximately 4 to 5 wpm). The reading performance of a rod monochromat was tested over a 4.5 log unit range to investigate the effects of luminance on rod reading performance. Because the monochromat has only rod photoreceptors reading can be tested at high luminance levels where normally cones would mediate reading. The rod monochromat was capable of reading with the rods over the entire luminance range though reading performance was found to plateau at relatively low luminance levels.

The purpose of the second study was to investigate whether people read best by fixating a passage of text (as one ordinarily does) or by viewing the passage with their peripheral vision (where rod night vision sensitivity is maximal). Reading performance was tested with text placed at the point of fixation as well as retinal eccentricities of 5, 10, 15 and 20 degrees below the point of fixation. The results show that, for three degree letter size. subjects read best when text was placed 5 degrees below or at the point of fixation. This finding was somewhat unexpected because at low luminances subjects are more sensitive in the peripheral retina than near the fovea and because subjects have a t'oveal scotoma.

In the third study, it was investigated whether reading with rods is faster or slower than reading with cone vision. Text was presented at five degrees below the subject's fixation. The text luminance was equated for rod and cone vision by scaling it relative to the rod and cone absolute threshold. Cone vision was isolated by testing reading during the cone plateau phase of the dark adaptation curve. The results show that subjects read faster when reading the text with cone vision than with rod vision.

In conclusion, reading is not an exclusive property of the cone visual system. Subjects can read even with an image available through the night vision system. Comparisons of rod and cone reading suggest that cone vision allows an observer to read faster than the rod visual system.

Whether reading with the rods or cones, performance of both day and night vision is dependent on the specific luminance level. Subjects' reading performance were found to improve rapidly with luminance at low luminance levels and to plateau once cone thresholds were exceeded.

The results also show that reading is greatly dependent on where in the visual field the text is presented. Subjects can read using the peripheral retina; however, reading performance is much poorer in the peripheral visual field than in the central visual field. This is true even though the subjects had a central scotoma.