Cone-excitation Space Research Articles

Does spatially homogeneous color stimulation produce a single response point within the physiological cone excitation space?

Does spatially homogeneous color stimulation produce a single response point within the physiological cone excitation space?

The color cone.

While the notion of a color cone can be found in writings of Maxwell, Helmholtz, Grassmann, and other scientists of the nineteenth century, it has not been clearly defined as yet. In this paper, the color cone is understood as the set of points in the cone excitation space produced by all possible lights. The spectral curve representing all the monochromatic lights is shown not to entirely belong to the color cone boundary, since its ends turn into the color cone interior. The monochromatic lights represented by the fragment of the spectral curve lying on the color cone boundary make up what is called the effective visible spectrum. The color cone is shown to be a convex hull of the conical surface through the fragment of the spectral curve representing the effective visible spectrum. The effective visible spectrum ends are shown to be determined by the photopigment spectral absorbance being independent of the prereceptor filters (e.g., the spectral transmittance of the lense and macular pigment).

Spatio‐temporal selectivity of loss of colour and luminance contrast sensitivity with multiple sclerosis and optic neuritis

Colour and luminance-contrast thresholds were measured in the presence of dynamic Random Luminance-contrast Masking (RLM) in individuals who had had past diagnoses of optic neuritis (ON) some of whom have progressed to a diagnosis of multiple sclerosis (MS). To explore the spatio-temporal selectivity of chromatic and luminance losses in MS/ON, thresholds were measured using three different sizes and modulation rates of the RLM displays: small checks modulating slowly, medium-sized checks with moderate modulation and large checks modulating rapidly. The colour of the chromatic stimuli used were specified in a cone-excitation space to measure relative impairments in red-green and blue-yellow chromatic channels. These observers showed chromatic thresholds along the L/(L+M) axis that were higher than those along the S-cone axis for all display sizes/modulation rates and both red-green and blue-yellow colour thresholds were higher than luminance-contrast thresholds. The principal change in thresholds with spatio-temporal changes in the display was a reduction in thresholds for L/(L+M) and S-cones with increasing check size and modulation rate. However, luminance contrast thresholds did not change with display size/rate. These results are consistent with MS/ON selectively affecting processing in colour pathways rather than in the magnocellular pathway, and that within the colour pathways neurones with opposed L- and M-cone inputs are more damaged than colour-opponent neurons with input from S-cones.

Nonlinearity in color space measured by apparent motion.

We used an apparent motion technique to examine the intensity coding along the three cardinal axes of color space: achromatic (L + M + S), L-M cone, and S cone axes. Two horizontal bars of different colors were alternated to produce a vertical displacement. The color of the background was a mixture varied between the colors of the two bars. When the background color was close to either of the test colors, only the bar that was more salient appeared to jump. Observers adjusted the color of the background until they saw either the two bars moved equally frequently or both bars moved at once. If the color difference in a linear cone excitation space controls this apparent motion, the setting should be midway between the two colors. All of the three cardinal axes showed some deviation from linear behavior. The nonlinearity was less extreme than a logarithmic function for both the achromatic and S cone axes and could be attributed to a small compressive nonlinearity, possibly at the level of cone responses. However, the L-M stimuli showed a more extreme departure from linearity, which suggested a nonlinearity at an opponent site. A test of perceived contrast judgments did not show this nonlinearity for L-M axis, suggesting that it is specific to the L-M contribution to apparent motion.

Response of the human visual system to variable illuminant conditions: An analysis of opponent-colour mechanisms in colour constancy

Illuminant changes in a scene generally cause displacements in the chromaticity and luminance of the objects of which the scene is composed; however, the visual system is capable of maintaining a constant object colour appearance independently of the illuminant. This is the phenomenon traditionally known as colour constancy. A classical asymmetric colour-matching experiment is reported to address two aspects of colour constancy: successive colour constancy and the role of opponent mechanisms in it. To this end, colour matches are made with chromatically complex backgrounds under different illuminants and with an equal-energy light source as standard. The results are analysed in the cone-excitation space, separately along the L-2M and S dimensions. Data are presented showing that colour-vision mechanisms respond differently to illuminant changes when colour constancy is considered at both receptoral and post-receptoral levels. The L- and M-cones tend to adapt so as to support colour constancy, whereas S-cones are strongly influenced by the illuminant changes. In addition, the data suggest good approaches to colour constancy linked particularly to the yellow-blue mechanism.

Photoreceptor Nonlinearities Can Account for the MacAdam Ellipses

Colour perception is known to be nonlinear, as highlighted by D L MacAdam (1942 Journal of the Optical Society of America32 247). He showed that the just noticeable difference (JND) between two colours is not constant and depends nonuniformly on the values of these two colours: under constant illumination, the constant-JND loci, drawn in the xy colour space around various centres, appear as ellipses of different orientations and eccentricities. Though this phenomenon encompasses the whole perception chain, from receptors to higher levels, we show that it can be accounted for mainly by the nonlinearity of receptor transduction. We start from the colorimetric space XYZ defined by the CIE (1931) and its linear transformation into the cone-excitation space LMS, using data from V C Smith and J Pokorny (1996 Color Research and Applications21 5). Then we consider a perceptual colour space obtained by the nonlinear correspondence between the L, M, and S cone inputs and the l, m, and s cone outputs, through a compression law of Michaelis - Menten type, eg: l = L/( L + L0). In this space, the JNDs are circles, drawn on the surface corresponding to the nonlinear transformation of the constant-luminance plane in LMS, around the corresponding centres. By looking at the projections of those circles onto the constant-luminance plane in LMS, and their representations in the colorimetric space xy, we naturally obtain the ellipses of MacAdam. The formulas, fitted to the data, provide the parameters of the LMS-to- lms transformations.

Perfect Von Kries Contrast Colours

Under certain viewing conditions, the colour of a light in a surround is determined by the vector from the surround to the colour in a 3-D colour space. This behaviour has been explored with a computer graphics tool that allows sets of such vectors to be drawn from two surrounds in any plane in any colour space, and allows the resulting lights-in-surrounds to be compared haploscopically. In a log cone-excitation space contrast colours with the same surround-to-colour vectors match well, over a surprisingly wide range of chromaticities and luminance contrasts. This implies that the contrast colours are determined by the three cone contrasts; the absolute chromaticities and luminances do not matter. Some limits and implications of this behaviour are discussed.

Chromatic-discrimination axes, CRT phosphor spectra, and individual variation in color vision.

There is a growing use of color monitor systems in visual research and a parallel growth in the use of cone-excitation space to define stimuli and to report data. Color specification in monitor systems is accomplished by combination of the phosphor chromaticities. The effect of interobserver variation on color specification is highly dependent on the spectroradiometric properties of the primaries. We review potential sources of biologic variability and its effect on the nominal axes in a cone-excitation diagram for a color monitor system. Variation in preretinal pigment (lens and macular pigment), in the effective optical density and the spectral sensitivity of the visual photopigments, and in the cone weighting used to derive the spectral luminosity function are considered. The consequences of such biological variability are rotation and translation of the axes for a given observer relative to the nominal axes that the observer used for color specification. The importance of such rotations can be viewed within the framework of a particular experimental paradigm.

Evaluation of color-discrimination ellipsoids in two-color spaces

We obtained the color-differential thresholds for 20 stimuli and 3 observers within 2 color-representation spaces. First, in the CIE-1931 color space we compare the results of our experiments, which followed chromatic-discrimination criteria, with those obtained by other authors that followed color-matching criteria. Second, the representation of our data in the cone-excitation space shows that all the discrimination ellipses are nearly oriented along the S-cone-variation axis. In our experiments the variations in the threshold of each color channel according to its level of excitation are similar to those found by other authors using different experimental conditions and methods, but certain discrepancies appear when the influence that one color channel may exert on the other is analyzed. The characteristics of luminance discrimination are shown to be independent of the chromaticity of the stimuli.

Interobserver variability of chromaticity discrimination and color representation spaces

The authors have evaluated the interobserver variability of color discrimination data in three different color representation spaces: CIELUV,CIELAB and cone-excitation space. They determined differential color thresholds for an achromatic stimulus and eight observers and represent them in the three color spaces. The results show a very little variability in the cone-excitation space, based upon Boynton's color vision model. The differences are a little higher in CIELUV space, although can be considered as moderated, whereas these are quite higher in CIELAB.

Analysis of color-matching ellipses in a cone-excitation space.

Color-discrimination ellipses derived from the variability of color-matching data of six observers are analyzed in a normalized constant-luminance cone-excitation space. The analysis shows that the ellipses do not vary significantly in shape with chromaticity, observer, or experimental conditions. The discrimination contours are predictable from the thresholds on the two cardinal axes of this space; these are used to normalize the data at each chromaticity for each observer. Thresholds on these two axes vary with chromaticities, individuals, and experimental conditions in accordance with simple and familiar laws.