The Stiles-Crawford colour change

Photoreceptor waveguides and effective retinal image quality

Individual photoreceptor waveguiding suggests that the entire retina can be considered as a composite fiber-optic element relating a retinal image to a corresponding waveguided image. In such a scheme, a visual sensation is produced only when the latter interacts with the pigments of the outer photoreceptor segments. Here the possible consequences of photoreceptor waveguiding on vision are studied with important implications for the pupil-apodization method commonly used to incorporate directional effects of the retina. In the absence of aberrations, it is found that the two approaches give identical predictions for an effective retinal image only when the pupil apodization is chosen twice as narrow as suggested by the traditional Stiles-Crawford effect. In addition, phase variations in the retinal field due to ocular aberrations can delicately alter a waveguided image, and this may provide plausible justification for an improved visual sensation as compared with what should be expected on the grounds of a retinal image only.

Comparison of cone directionality determined by psychophysical and reflectometric techniques

We measured the directionality of the cones with both a psychophysical (Stiles-Crawford I) technique and an optical technique. The two sets of measurements were made in the same subjects, with stimuli as similar as possible used. The two types of measurements gave similar estimates of the location in the pupil toward which the cones were optimally aligned. However, the two measurements gave quite dissimilar estimates of the width of the directional sensitivity. On average, optical measurements were half as broad as psychophysical measurements in the fovea, but there were substantial individual differences. At 2-deg retinal eccentricity the difference between techniques was even more marked.

Color matching at high illuminances: photopigment optical density and pupil entry

Changes in the effective optical density of the cones are sufficient to explain changes in color matches with retinal illuminance and pupil entry. We performed three experiments. In the first experiment, six observers made color matches under both bleached and unbleached conditions to a series of six standard wavelengths. The effects of bleaching could be modeled by a decrease in optical density of the L and M cone photopigments. Slight spectral shifts in the peak wavelengths of the photopigments were required for different observers. In the second experiment we varied retinal illuminance of the color-matching field from 2.4 to 5.4 log Td for a series of long-wavelength primaries. The shape of the color match versus the intensity function was unchanged by the wavelength composition of the matching field. In the third experiment we measured the change in color match with retinal illuminance for different pupil entry positions. At low luminances there was a marked dependence of the color match on pupil entry positions. At high illuminance there was only a small dependence. The half-bleach illuminance values varied as expected from the Stiles-Crawford I effect. We conclude that for wavelengths > 540 nm, changes in color matches with bleaching and pupil entry can be explained by changes in the effective optical density of the cones.

The change in color matches with retinal angle of incidence of the colorimeter beams

Differences between W.D.W. chromaticities of monochromatic lights obtained with all colorimeter beams incident on the retina "off-axis" and those found for lights striking the retina normally have been studied throughout the visible spectrum on 4 normal trichromats. The results are inconsistent with: (i) the assumption in Weale's theories of the Stiles-Crawford hue shift that the sets of absorption spectra of the visual pigments catching normally and obliquely incident photons are identical, and (ii) "self-screening" explanations for the change in color with angle of incidence on the retina. The color matching functions of a protanomalous trichromat are inconsistent with the hypothesis that the absorption spectra of the visual pigments catching normally incident photons in his retina are those catching obliquely incident photons in the normal retina.

The Stiles-Crawford effect of the second kind (SCII): a review

The background which led Stiles to the discovery that the color of a monochromatic ray of light varied with its angle of incidence on the retina, and the developments in the subject since 1937 when this discovery was reported are summarized. Stiles's original measurements of the 'hue shift' did not successfully quantify SCII in every part of the spectrum, since it provided only two degrees of freedom while the color changes sometimes require three. Nevertheless, they were the paradigm for most subsequent work for the next quarter of a century. Full quantification of the effect was first obtained by trichromatic matching on Stiles's NPL trichromator almost 25 years after the initial discovery. The phenomena were then also fully described quantitatively with an elaboration of Stiles's original theory ('self-screening' theory), on the assumption that the ordinary laws of additivity are valid for color matches of three primaries striking the retina at one angle of obliquity to a test incident at a different angle. More recent experiments suggest that this initial assumption may not be generally true and that 'self-screening' theory may not generally provide a satisfactory description of the color changes throughout the visible spectrum, even for measurements of the effect obtained under conditions in which the additivity assumption seems to be valid. However, the available data strongly imply that absorptions of photons obliquely incident on foveal cones depend upon spectra clearly different from those upon which absorptions of normally incident photons depend.

The directional sensitivities of the Stiles' colour mechanisms

Field sensitivities of the three IIj (j = 3, 4, 5) mechanisms of Stiles were measured for monochromatic backgrounds of different wave numbers (mu)-1 traversing the eye through different points (r) displaced along a horizontal chord through the centre of the entrance pupil. Each mechanism shows an insensitivity to the direction of retinal incidence of short-wave backgrounds not previously described. The spectral densities of the centre-most part of the lens and of the macular pigment were measured on this eye. With reasonable assumptions the former allowed for correction at the receptor level of the directional sensitivity; together with the latter it allowed correction for the spectral sensitivity as well. No correction for the attenuation of the high spatial frequencies of the background as it traversed the pupil at different r was needed. The anomalies of section 2 (above) disappear after correction for losses in the eye media. After these corrections, for every mu and r, the results are well described by the parabola 'tentatively' suggested by Stiles (1939) for each mechanism, allowing only a small amount of variance attributable to experimental imprecision alone. Each mechanism is most sensitive to backgrounds going through essentially the same point of the pupil, independent of background. This result is inconsistent with a qualitative explanation of the 'hue shift' suggested by Safir, Hyams & Philpot (1971). The field sensitivity spectra for backgrounds traversing the pupil at this most effective point and at the 3.5 mm margin, are the data needed to predict this observer's brightness and colour matches of monochromatic lights passing through the entrance pupil at these two points according to a unified theory of the two Stiles-Crawford effects. In the following paper these predictions are quantified and confronted with results of the matching experiments (Alpern, Kitahara & Tamaki, 1982).

The dependence of the colour and brightness of a monochromatic light upon its angle of incidence on the retina

The changes in brightness and colour of a monochromatic test light as its angle of incidence on the retina was changed from normal (pupil centre traverse) to oblique (3.5 mm temporal pupil traverse), was measured by matching it with three normally incident primaries. Results on two normal trichromats were generally in accord with published data on the Stiles-Crawford intensity and colour effects. One observer was also the subject of the preceding paper (Alpern & Kitahara, 1983) in which the field sensitivities of his foveal IIj(mu) (j = 3, 4, 5) mechanisms for normally, and obliquely, incident backgrounds were reported. For normal incidence, the colour matching functions are in rough accord with expectation if the action spectra of the three cone mechanisms, which provide the photoreceptor basis for his trichromacy, were the same IIj mechanisms for normal incidence. A unified theory is developed for both Stiles-Crawford intensity and colour effects, assuming that the same visual pigments in the same set of univariantly signalling cones absorbs both the normal incident primaries and the obliquely incident test. Given no free parameters for curve fitting, the Stiles-Crawford intensity effect data are in reasonable agreement with the theory if the photoreceptor basis of these matches were the normally and obliquely incident IIj(mu) mechanisms. The Stiles-Crawford colour effect data contradict the expectations of the unified theory applied with these same IIj(mu) mechanisms. Either II3(mu) is not a valid operational definition of the action spectrum of his short-wave sensitive photoreceptors or at least one assumption of the unified theory is false.

The saturation of monochromatic lights obliquely incident on the retina

Foveal dark-adaptation undertaken to test the hypothesis that the excitation of rods causes the desaturation of 'yellow' lights in a 1 degree field traversing the margin of the pupil, fails to exclude that possibility. The desaturation is largest for a 1 degree outside diameter annular test, is still measurable with a 0.5 degree circular disk, but disappears for a 0.29 degree disk. The supersaturation of obliquely incident 501.2 nm test light follows the opposite pattern; it disappears with an annulus and is largest for a 0.29 degree circular field. It is unlikely that rods replace short-wave sensitive cones in the trichromatic match of an obliquely incident test with normally incident primaries. If rods as well as all three cones species are involved, the matches might not be trichromatic in the strong sense. Grassmann's law of scalar multiplication was tested and shown not to hold for the match of an obliquely incident test with normally incident primaries, though it remains valid whenever, both primaries and test strike the retina at the same angle of incidence (independent of that angle). The result in section 3 (above) cannot be due to rod intrusion. It persists (and becomes more conspicuous) on backgrounds (4.0 log scotopic td) which saturate rods. Moreover obliquely incident 'yellow' lights remain desaturated in intervals in the dark after a full bleach, whilst the test field is below rod threshold. The amount of desaturation does not differ appreciably from that normally found. The assumption of the unified theory of Alpern, Kitahara & Tamaki (1983) that the outer segments of only a single set of three cone species (with acceptance angles wide enough to include the entire exit pupil) contain the visual pigments absorbing both the normally incident primaries and the obliquely incident test is disproved by these results. Failure of Grassmann's law is most conspicuous under the conditions for which the changes in saturation upon changing from normal to oblique incidence are greatest and least when the saturation changes are the smallest. Either all unified theories of the Stiles-Crawford effects are wrong or all the effects of oblique incidence operate at a stage in the visual process at which the effects of radiation of different wave-lengths are no longer compounded by the simple linear laws.