C. H. Graham

Effect of wavelength on foveal grating acuity.

Acuity–luminance relations for a grating test object were determined with red, yellow, green, and blue narrow-band chromatic illuminants for five subjects.At intermediate and high luminance levels, acuities for blue were lower than for the other chromatic illuminants in four of the five subjects. One of these subjects also showed lower asymptotic acuity for red at high luminances. The fifth subject, who did not show lower blue acuity, exhibited lower asymptotic acuities than the other subjects. The low blue acuities observed in four of the five subjects are attributed to neural rather than dioptric factors.

On Some Aspects of Real and Apparent Visual Movement

Some of the parameters of real and apparent movement are considered in situations where these two types of movement are encountered. Some basic properties of real movement are represented by displacement and rate thresholds as well as by limiting upper thresholds. The depth cue for monocular movement parallax can be analyzed as representing a situation in which differential angular velocity thresholds may be determined. These latter thresholds turn out to be quite small, of the order of 30 to 40 sec of arc/sec. They are influenced by a number of variables, including rate of standard stimulus, luminance, and visual axis. Differential angular velocities help to determine certain aspects of perceived movements in the third dimension, and attention is paid to the case of Ames’ trapezoid window which provides an illusion of movement. An account of apparent movement is given with special reference to Korte’s laws. The question then is asked: What mechanisms might be expected to occur in real and apparent movement? Tentative suggestions on this problem are advanced, with special attention to real movement.

Luminance Thresholds for the Resolution of Visual Detail during Dark Adaptation

Luminance thresholds for the visual resolution of various widths of alternating light and dark lines were determined at various times during dark adaptation. The finest gratings, representing high degrees of visual acuity, show only a single cone curve that drops from a high luminance threshold during the first moments of dark adaptation to a final steady level that is reached after about 7 to 10 minutes in the dark. Coarse gratings produce a duplex curve that shows an initial cone portion and a delayed rod portion. Visual acuity is a parameter that sets the position of a given curve on the log threshold axis. The higher the degree of resolution required, the higher the dark adaptation threshold. At a constant grating luminance, visual acuity rises rapidly to a maximum during dark adaptation; the higher the luminance, the earlier and more rapid the rise and the higher the maximum. Visual acuity increases at all dark adaptation times with increase in luminance.

Maintaining an absolute hue in the presence of different background colors

Abstract Subjects were instructed to make wavelength settings for various hues by an absolute method in the presence of surround-field colors or darkness. Surround- and test-field colors were equated in luminance. The subjects compensatory shift in setting for the maintenance of a test-color is taken to be the difference between his wavelength setting for the test-color in the presence of the surround-field color and the wavelength setting for the test-color with a dark surround. In general, the compensatory shift in wavelength setting is always in the direction of the background wavelength. A discussion regarding the nature of the shift setting is given.

Saturation and the Foveal Achromatic Interval

Abney’s curves for chromatic threshold are similar to curves for colorimetric purity as presented by Priest and Brickwedde. Determinations of chromatic threshold Ec and achromatic threshold Ea in two of our own subjects show that the curve for the chromatic threshold is similar to that of Abney and also similar to the curve for colorimetric purity. The curve for the achromatic threshold is below the curve for the chromatic threshold and reproduces the luminosity curve for the fovea. Six other subjects showed chromatic threshold curves similar to those of Abney. Defined as logEc/Ea the curve for achromatic interval has a maximum near 570 nm and, in general, falls to greatly reduced values at 420 in the blue and 700 in the red. This curve also gives chromatic thresholds in photopic photometric units based on achromatic thresholds (i.e., foveal luminosity). In the blue and red regions of the spectrum, the achromatic threshold has a value that approaches the chromatic. This result, not obtained by Purdy in the blue, is discussed. It is pointed out that colorimetric-purity thresholds and chromatic thresholds seem both to be reciprocal measures of saturation. Chromatic thresholds may have an advantage in not involving a physical mixture of color with white.

Maintaining an absolute test hue in the presence of different background colors and luminance ratios

Abstract Measures are made of changes in test wavelength required to compensate for a contrast effect introduced by a background color. The test hue remains constant when the contrast effect is modified by a shift of the test wavelength, usually toward the direction of the background. Ratio of test-to-background luminance has an imprecisely specifiable influence on compensatory wavelength changes in the test area. Wavelength settings for similar test hues made in a dark surround at two levels of luminance, 1.1 and 12.0 mL, demonstrate a Bezold-Brucke shift due to intensity level (data of Akita, Graham and Hsia (1964)).

Visual discriminations of a subject with acquired unilateral tritanopia.

Abstract Binocular hue matches, luminosity functions, color naming and neutral point determinations are reported in a case of acquired unilateral tritanopia. Comparisons are made with the normal vision in the patients non-affected eye, and with predictions from data on congenital tritanopia. The affected eye shows a loss in luminosity in the region centering on the neutral point at 569 mμ. In all functions but luminosity the results are similar to those found in the congenital tritanope. The various implications of these findings are discussed.

Analysis of Photopic and Scotopic Function in an Incomplete Achromat

Spectral sensitivity and color vision tests were performed on a subject who was totally color blind at low photopic luminances, but exhibited hue discrimination at relatively high photopic levels. Chromaticity confusion loci showed that the residual color vision was abnormal; a tritan defect was superimposed on the generalized reduction of cone sensitivity. Although cone function was markedly depressed, a photopic spectral sensitivity curve was obtained for the light-adapted fovea. Dark-adapted foveal measurements, on the other hand, gave a scotopic function almost identical to that of the dark-adapted periphery. However, it was possible to demonstrate objectively that this subject shifts fixation to an eccentric position under scotopic conditions, i.e., the dark-adapted “foveal” results were, in fact, for a para-macular region. Irrespective of the degree of light adaptation, cone function was not detectable in the peripheral measurements.

Some visual functions of a unilaterally color-blind person. I. Critical fusion frequency in various spectral regions.

A new case of unilateral dichromatism is described. She is a young woman with normal color vision in one eye and dichromatic vision of a primarily deuteranopic type in the other.Critical fusion frequency functions for a centrally fixated 28-min field were determined in ten spectral regions, ranging from one having a spectral centroid at 452 mμ to one with a spectral centroid at 682 mμ, on both eyes of this unilaterally dichromatic subject. Determinations were also made with white light. Measurements extended over a range of approximately 3.5 log millilamberts. For all colors except red, the curve of critical fusion frequency vs log luminance for the color-blind eye is displaced downward on the critical fusion frequency axis with respect to the curve for the normal eye. For any given spectral region, the displacement is approximately constant over the luminance range tested and the two curves do not reach the same maximum fusion frequency. The magnitude of the shift varies with wavelength. It is greatest in the green, next in the blue-green, yellow-green, blue, and yellow; there is a slight loss in the orange and no detectable loss in the red. The data for white light data also show reduced critical fusion frequencies for the color-blind eye.These findings are taken to reflect a reduction in the dichromatic eye of the number of receptors (of a type especially sensitive to green) available for excitation by the spectral range from about 450 mμ to 625 mμ.Some measurements of critical fusion frequency with a 1° green and a 2° white field are also reported. They display the same general trends as the data for small fields, but the extent of the downward shift of the color-blind function is in each case less than that for the corresponding pair of curves with the 28-min area. The reduction in the amount of downward displacement with larger test fields comes about through a proportionately greater increase in critical fusion frequency with area for the color-blind than for the normal eye. The results are formulated in terms of a nonuniform distribution of color receptors across the fovea.

Occurrence of theoretically correct responses during rotation of the Ames window

It has been hypothesized (Graham, 1963) that reversals in the Ames window are the outcome of a resolution of ambiguous differential angular velocity cues by linear perspective cues. This theory was tested. A parallel projection of the window on an opal glass screen was used as the stimulus. Ss almost always reported two apparent reversals per rotation. The long vertical side of the figure was always apparently in front of the short vertical side. These results were interpreted to be in line with theoretical expectations.