Two observers made equidistance settings in a two-rod apparatus with illuminated white and colored targets photometrically matched at 8–9 illumination levels ranging from 0.003 to 30 trolands. The equidistance settings were analyzed in terms of the angular magnitude of both the variable error, ηAD, and the constant error ηΔR. The ηAD vs retinal-illuminance curves for the white and colored targets show that at low retinal illuminances, ηAD is initially large. With increasing target illumination, ηAD progressively decreases to approach a final low asymptotic value. Each experimental curve shows a discontinuity at about 0.1 td. The curves representing the different wavelengths essentially overlap throughout the entire luminance range, indicating that, at both scotopic and photopic levels, wavelength has no differential effect on the variability of equidistance settings. Although the corresponding data for the angular constant error, ηΔR, are considerably less regular, no differential wavelength effects are evident; the absolute magnitude of the constant errors generally decreases for all wavelengths as target illumination is increased. The data are discussed in terms of the duplicity theory of vision, emphasizing the effect of brightness matches performed at sub-photopic (i.e., at scotopic and mesopic) levels.
© 1970 Optical Society of AmericaFull Article | PDF Article
OSA Recommended Articles
Alfred Lit and Harlyn D. Hamm
J. Opt. Soc. Am. 56(4) 510-516 (1966)
J. Opt. Soc. Am. 49(8) 746-752 (1959)
J. Opt. Soc. Am. 54(1) 83-88 (1964)