OOJZ-6989 YS 53.00 + 0 00 Ctiprlghr ( 1965 Psrgsmon Press Lrd zyxwvutsrqpo RESEA RCH NOTE JUST NOTICEABLE INHOMOGENEITY CRITERION FOR DETERMINING WAVELENGTH DISCRIMINATION FUNCTIONS PETER K. KAISER and ,MMIYOSHI AYA.LIA Department of Psychology. York Uni\errity, North York. Ontario. Canada ?vl?J IP3 Abstract-The visual criterion of just noticeable inhomogeneity is described for determining wavelength discrimination functions. It involved determining the wavelength differences betueen reference and test fields required to produce a just noticeable inhomogeneity which cannot be eliminated by a brightness adjustment. The fields formed a checkerboard pattern the element size of which was variable. Tritanopic AL functions were obtained by using the small field insensitivity of the fovea. Just noticeable border. data obtained from bipartite field studies, bvere replicated with this checkerboard field. Wavelength discrimination Small field tritunopia border INTRODL’CTION In a recent paper, Kaiser and Boynton (1985) reported wavelength discrimination experiments per- formed under conditions of transient tritanopia. The purpose was to determine tritanopic wavelength discrimination functions from normal trichromatic observers for comparison with functions obtained using a just noticeable border criterion. Although the endeavor was successful, the painstaking difficulty imposed by the transient tritanopic procedure did not allow for a large number of reference wavelengths. This report describes the criterion of just noticeable inhomogeneity (JNI) which permits a more complete function to be easily and quickly obtained. Many experiments have been reported showing that stimuli subtending very small visual angles (e.g. 15’ and less) yield tritanopic-like visual data. Willmer and Wright (194.5) produced a set of dichromatic mixture coefficients using small fields and Thomson and Wright (1947) showed that the small field tri- tanopic effect could be obtained up to 40’ away from the fovea centralis. Williams et al. (1981) recently reported that the central 25’ of the fovea lacks functioning blue sensitive cones. They also reported that macular pigment or Troxler fading could not account for their results. Bedford and Wyszecki (1958) using rather high retinal illuminances (500 and 2000 td) and not fixating the stimulus found little tritanopic effect in a wavelength discrimination ex- periment. Williams et al. (1981) presented evidence that scattered light can mediate residual blue sensitive cone color discrimination. Therefore, it is reasonable to assume that Bedford and Wyszecki’s use of high Just noticeable inhomogeneity Just noticeable retinal illuminances resulted in sufficient amounts of scattered light to yield reasonably normal wavelength discrimination functions. McCree (1960) measured wavelength discrimination functions with several field sizes and luminances and, like Bedford and Wyszecki, without steady fixation. He reported the tritanopic effect to be greater than that found by Bedford and Wyszecki but less than that found by Willmer and Wright. He attributed his results to his lower lumi- nances than that used by Bedford and Wyszecki. Weale (1960) presented wavelength discrimination functions as a function of field size from I’ to 14’ visual angle. The 14’ field data appear similar to Wright’s (I 952) tritanopes. However, these functions are unpublished data of Forshaw (1960) given to Weale by private communication. No details of how the data were collected are reported by Weale and the figure only shows smooth line drawings without data points. Taking advantage of this so-called small field fovea1 tritanopia, and using the just noticeable in- homogeneity (JNI) criterion, with free viewing, we easily and quickly obtained complete wavelength discrimination functions similar to those obtained by the transient tritanopic method (Kaiser and Boynton, 1985) and from tritanopic observers (Wright. 1952). APPARATUS The essential part of the apparatus is a checker- board mirror beam splitter (Fig. I). There are a number of ways such a beam splitter can be imple- mented. We used a rectangular piece of glass upon which a checkerboard, first surface mirror was evap-