Individual Colorimetric Observer Model

Development of an Innovative Pupillometer Able to Selectively Stimulate the Eye's Fundus Photoreceptor Cells

Recent advancements in clinical research have identified the need to combine pupillometry with a selective stimulation of the eye's photoreceptor cell types to broaden retinal and neuroretinal health assessment opportunities. Our thorough analysis of the literature revealed the technological gaps that currently restrict and hinder the effective utilization of a method acknowledged to hold great potential. The available devices do not adequately stimulate the photoreceptor types with enough contrast and do not guarantee seamless device function integration, which would enable advanced data analysis. RetinaWISE is an advanced silencing pupillometry device that addresses these deficiencies. It combines a Maxwellian optical arrangement with advanced retinal stimulation, allowing for calibrated standard measurements to generate advanced and consistent results across multiple sites. The device holds a Class 1 CE marking under EU regulation 2017/745, thus facilitating clinical research progress.

Deriving the cone fundamentals: a subspace intersection method

Two ideas, proposed by Thomas Young and James Clerk Maxwell, form the foundations of colour science: (i) three types of retinal receptors encode light under daytime conditions, and (ii) colour matching experiments establish the critical spectral properties of this encoding. Experimental quantification of these ideas is used in international colour standards. However, for many years, the field did not reach consensus on the spectral properties of the biological substrate of colour matching: the spectral sensitivity of the cone fundamentals. By combining auxiliary data (thresholds, inert pigment analyses), complex calculations, and colour matching from genetically analysed dichromats, the human cone fundamentals have now been standardized. Here, we describe a new computational method to estimate the cone fundamentals using only colour matching from the three types of dichromatic observers. We show that it is not necessary to include data from trichromatic observers in the analysis or to know the primary lights used in the matching experiments. Remarkably, it is even possible to estimate the fundamentals by combining data from experiments using different, unknown primaries. We then suggest how the new method may be applied to colour management in modern image systems.

Correction: Individual Colorimetric Observer Model

[This corrects the article DOI: 10.1371/journal.pone.0145671.].

Designing for sensory adaptation: what you see depends on what you've been looking at - Recommendations, guidelines and standards should reflect this

Sensory systems continuously recalibrate their responses according to the current stimulus environment. As a result, perception is strongly affected by the current and recent context. These adaptative changes affect both sensitivity (e.g., habituating to noise, seeing better in the dark) and appearance (e.g. how things look, what catches attention) and adjust to many perceptual properties (e.g. from light level to the characteristics of someone's face). They therefore have a profound effect on most perceptual experiences, and on how and how well the senses work in different settings. Characterizing the properties of adaptation, how it manifests, and when it influences perception in modern environments can provide insights into the diversity of human experience. Adaptation could also be leveraged both to optimize perceptual abilities (e.g. in visual inspection tasks like radiology) and to mitigate unwanted consequences (e.g. exposure to potentially unhealthy stimulus environments).

Formulae for generating standard and individual human cone spectral sensitivities

Normal color perception is complicated. But at its initial stage it is relatively simple, since at photopic levels it depends on the activations of just three photoreceptor types: the long- (L-), middle- (M-) and short- (S-) wavelength-sensitive cones. Knowledge of how each type responds to different wavelengths-the three cone spectral sensitivities-can be used to model human color vision and in practical applications to specify color and predict color matches. The CIE has sanctioned the cone spectral sensitivity estimates of Stockman and Sharpe (Stockman and Sharpe, 2000, Vision Res) and their associated measures of luminous efficiency as "physiologically-relevant" standards for color vision (CIE, 2006; 2015). These LMS cone spectral sensitivities are specified at 5- and 1-nm steps for mean "standard" observers with normal cone photopigments and average ocular transparencies, both of which can vary in the population. Here, we provide formulae for the three cone spectral sensitivities as well as for macular and lens pigment density spectra, all as continuous functions of wavelength from 360 to 850 nm. These functions reproduce the tabulated discrete CIE LMS cone spectral sensitivities for 2-deg and 10-deg with little error in both linear and logarithmic units. Furthermore, these formulae allow the easy computation of non-standard cone spectral sensitivities (and other color matching functions) with individual differences in macular, lens and photopigment optical densities, and with spectrally shifted hybrid or polymorphic L- and M-cone photopigments appropriate for either normal or red-green color vision deficient observers.

PySilSub: An open-source Python toolbox for implementing the method of silent substitution in vision and nonvisual photoreception research

The normal human retina contains several classes of photosensitive cell-rods for low-light vision, three cone classes for daylight vision, and intrinsically photosensitive retinal ganglion cells (ipRGCs) expressing melanopsin for non-image-forming functions, including pupil control, melatonin suppression, and circadian photoentrainment. The spectral sensitivities of the photoreceptors overlap significantly, which means that most lights will stimulate all photoreceptors to varying degrees. The method of silent substitution is a powerful tool for stimulating individual photoreceptor classes selectively and has found much use in research and clinical settings. The main hardware requirement for silent substitution is a spectrally calibrated light stimulation system with at least as many primaries as there are photoreceptors under consideration. Device settings that will produce lights to selectively stimulate the photoreceptor(s) of interest can be found using a variety of analytic and algorithmic approaches. Here we present PySilSub (https://github.com/PySilentSubstitution/pysilsub), a novel Python package for silent substitution featuring flexible support for individual colorimetric observer models (including human and mouse observers), multiprimary stimulation devices, and solving silent substitution problems with linear algebra and constrained numerical optimization. The toolbox is registered with the Python Package Index and includes example data sets from various multiprimary systems. We hope that PySilSub will facilitate the application of silent substitution in research and clinical settings.

Using equiluminance settings to estimate the cardinal chromatic directions for individuals

Color information is processed by the retina and lateral geniculate along principal dimensions known as the cardinal directions of color space. Normal differences in spectral sensitivity can impact the stimulus directions that isolate these axes for individual observers and can arise from variation in lens and macular pigment density, photopigment opsins, photoreceptor optical density, and relative cone numbers. Some of these factors that influence the chromatic cardinal axes also impact luminance sensitivity. We modeled and empirically tested how well tilts on the individual's equiluminant plane are correlated with rotations in the directions of their cardinal chromatic axes. Our results show that, especially for the SvsLM axis, the chromatic axes can be partially predicted by luminance settings, providing a potential procedure for efficiently characterizing the cardinal chromatic axes for observers.

Improving the reliability of smartphone-based urine colorimetry using a colour card calibration method

Objective

Urine colorimetry using a digital image-based colorimetry is potentially an accessible hydration assessment method. This study evaluated the agreement between urine colorimetry values measured with different smartphone brands under various lighting conditions in patients with dengue fever.

Methods

The urine samples were photographed in a customized photo box, under five simulated lighting conditions, using five smartphones. These images were analyzed using Adobe Photoshop to obtain urine Red, Green and Blue (RGB) values with and without colour correction. A commercially available colour calibration card was used for colour correction. Using intraclass correlation coefficient (ICC), inter-phone and intra-phone agreements of urine RGB values were analyzed.

Results

Without colour correction, the various smartphones produced the highest agreement for Blue and Green values under the 'daylight' lighting condition. With colour correction, ICC values showed 'exceptional' inter-phone and intra-phone agreement for the Blue and Green values (ICC > 0.9). Red values showed 'poor' (ICC < 0.5) agreement with and without colour correction in all lighting conditions. Out of the five phones compared in this study, Phone 4 produced the lowest intra-phone agreement.

Conclusions

Colour calibration using photo colour cards improved the reliability of smartphone-based urine colorimetry, making this a promising point-of-care hydration assessment tool using the ubiquitous smartphone.

Factors affecting colour matching between displays

A colour matching experiment was conducted to study and observe metamerism between different displays. The goals were to investigate the parameters of the display primaries (spectral power distributions (SPDs)), display types (OLED and LCD), and the colour matching functions (CMFs). The results showed that the use of the CIE 2006 2° CMFs can give better agreement to the visual results, especially matching between OLED against LCD displays, mainly due to the SPDs of the primaries. The results also showed that a simple color correction model improved the matching performance between displays, regardless of the display type.

Do You See What I See? Diversity in Human Color Perception

In our tendency to discuss the objective properties of the external world, we may fail to notice that our subjective perceptions of those properties differ between individuals. Variability at all levels of the color vision system creates diversity in color perception, from discrimination to color matching, appearance, and subjective experience, such that each of us lives in a unique perceptual world. In this review, I discuss what is known about individual differences in color perception and its determinants, particularly considering genetically mediated variability in cone photopigments and the paradoxical effects of visual environments in both contributing to and counteracting individual differences. I make the case that, as well as being of interest in their own right and crucial for a complete account of color vision, individual differences can be used as a methodological tool in color science for the insights that they offer about the underlying mechanisms of perception. Expected final online publication date for the Annual Review of Vision Science, Volume 8 is September 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Using Smooth Metamers to Estimate Color Appearance Metrics for Diverse Color-Normal Observers

Color-normal subjects sometimes disagree about metameric matches involving highly structured SPDs, because their cone fundamentals differ slightly, but non-negligibly. This has significant implications for the design of light sources and displays, so it should be estimated. We propose a broadly applicable estimation method based on a simple adaptive "front-end" interface that can be used with any selected standard color appearance model. The interface accepts, as input, any set of color matching functions for the individual subject (for example, these could be that person's cone response functions) and also the associated tristimulus values for the test stimulus and also for the reference stimulus (i.e. reference white). The interface converts this data into tristimulus values of the form used by the selected color appearance model (which could, for example, be X, Y, Z), while also carrying out the needed transform, which is based on an estimate of the subject's likely previous long-term adaptations to their unique cone fundamentals. The selected standard color appearance model then provides color appearance data that is an estimate of the color appearance of the test stimulus, for that individual subject. This information has the advantage of being interpretable within that model's well-known color space. The adaptive front end is based on the fact that, for any selected input SPD and the subject's unique color matching functions, there can be many different SPDs that are metameric for that individual. Since observer-to-observer color perception differences are minimized for spectrally smooth SPDs, smooth metamers predict color appearances reasonably accurately.

Personalizing image enhancement for critical visual tasks: improved legibility of papyri using color processing and visual illusions

This article develops theoretical, algorithmic, perceptual, and interaction aspects of script legibility enhancement in the visible light spectrum for the purpose of scholarly editing of papyri texts. Novel legibility enhancement algorithms based on color processing and visual illusions are compared to classic methods in a user experience experiment. (1) The proposed methods outperformed the comparison methods. (2) Users exhibited a broad behavioral spectrum, under the influence of factors such as personality and social conditioning, tasks and application domains, expertise level and image quality, and affordances of software, hardware, and interfaces. No single enhancement method satisfied all factor configurations. Therefore, it is suggested to offer users a broad choice of methods to facilitate personalization, contextualization, and complementarity. (3) A distinction is made between casual and critical vision on the basis of signal ambiguity and error consequences. The criteria of a paradigm for enhancing images for critical applications comprise: interpreting images skeptically; approaching enhancement as a system problem; considering all image structures as potential information; and making uncertainty and alternative interpretations explicit, both visually and numerically.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10032-021-00386-0.

Quantifying observer metamerism of LED spectra which chromatically mimic natural daylight

When LEDs are used to mimic daylight, a side-by-side comparison of the chromaticity difference between the LED spectrum and natural daylight will be perceived differently by individual observers. The magnitude of this effect depends on the LED light's spectral power distribution and can be assessed by using individual observer functions. To minimize the computational effort, an observer metamerism index can be utilized. Here, we compare three methods from the literature to define an observer metamerism index by carrying out a correlation analysis, in which reference spectra of the whole daylight range (1600 K to 88000 K) are used together with an empirical study. The recommended metric is based on a principal component analysis of 1000 individual observers' color matching functions to define a deviate observer. Using the proposed metamerism index significantly simplifies the calculation of the observer metamerism evaluation. Thus, this metric can be applied in spectral optimization pipelines, which are embedded in smart and adaptive multi-primary LED luminaires.

Maximum possible contrast level for silent substitution: a theoretical model applied to melanopsin stimulation

For any given set of light sources stimulating the photoreceptors of the retina, the theoretical levels of illumination producing the smallest and the largest expression of one photoreceptor with fixed stimulation for the others are analytically computed. The cases of four, five, and more light sources are studied. We show that, for contrast optimization, only as many light sources as photoreceptors do matter and that, in the case of four light sources, the maximum contrast achievable for melanopsin lies at the intersection of the lines joining the sources in the CIE xy chromaticity diagram. This result is used to obtain the optimal position of four Gaussian primaries of equal bandwidth. In addition, we derive a procedure to construct level maps for melanopsin contrast overlying the diagram. In the second part of the paper, the interpersonal variability of the perceived stimulation is shown to be globally reduced if the bandwidth of the light sources is increased and, under some assumptions, if a light source is added.

Color appearance model incorporating contrast adaptation - implications for individual differences in color vision

Color appearance models use standard color matching functions to derive colorimetric information from spectral radiometric measurements of a visual environment, and they process that information to predict color perceptual attributes such as hue, chroma and lightness. That processing is usually done by equations with fixed numerical coefficients that were predetermined to yield optimal agreement for a given standard observer. Here we address the well-known fact that, among color-normal observers, there are significant differences of color matching functions. These cause disagreements between individuals as to whether certain colors match, an important effect that is often called observer metamerism. Yet how these individual sensitivity differences translate into differences in perceptual metrics is not fully addressed by many appearance models. It might seem that appearance could be predicted by substituting an individual's color matching functions into an otherwise-unchanged color appearance model, but this is problematic because the model's coefficients were not optimized for the new observer. Here we explore a solution guided by the idea that processes of adaptation in the visual system tend to compensate color perception for differences in cone responses and consequent color matching functions. For this purpose, we developed a simple color appearance model that uses only a few numerical coefficients, yet accurately predicts the perceptual attributes of Munsell samples under a selected standard lighting condition. We then added a feedback loop to automatically adjust the model coefficients, in response to switching between cone fundamentals simulating different observers and color matching functions. This adjustment is intended to model long term contrast adaptation in the vision system by maintaining average overall color contrast levels. Incorporating this adaptation principle into color appearance models could allow better assessments of displays and illumination systems, to help improve color appearances for most observers.

Modeling individual variations in equiluminance settings

Recently, we reported measurements of heterochromatic flicker photometry (HFP) in 22 young observers, with stimuli that (nominally) modulated only L- and M-cones and were kept at (approximately) a constant multiple of detection threshold. These equiluminance settings were represented as the angle in the (L, M) cone contrast plane, with the greenish peak of the flicker in quadrant II and the reddish peak in quadrant IV; equiluminance settings were reported as the greenish angle. The mean equiluminance angle was 116.3° (an M:L cone contrast ratio of -2 at equiluminance), but individual differences in the settings were substantial, with the variation across individuals almost five times larger than the within-subject precision in the settings. In the present study we sought to determine the degree to which we could account for our observers' HFP settings by plausible variations in the macular pigment optical density (MPOD), the lens pigment optical density (LPOD), the cone photopigment optical densities (PPOD), and serine/alanine polymorphism in L-cone opsin (λmax shift). Most of the range of our measured equiluminance angles could be accounted for by these factors, although the largest two angles (smallest |ΔM/M: ΔL/L| ratio at equiluminance) could not. Individual differences in HFP have sometimes been taken to indicate variations in the ratio of L:M cone number; our results suggest that most of the individual differences in HFP might be equally well ascribed to physiological factors other than cone number. Simple linear models allow predictions of equiluminance angle, cone adapting level, and artifactual S-cone contrast from the values of the four factors considered here.

Statistical Properties of Color Matching Functions

In trichromats, color vision entails the projection of an infinite-dimensional space (the one containing all possible electromagnetic power spectra) onto the three-dimensional space that modulates the activity of the three types of cones. This drastic reduction in dimensionality gives rise to metamerism, that is, the perceptual chromatic equivalence between two different light spectra. The classes of equivalence of metamerism are revealed by color-matching experiments in which observers adjust the intensity of three monochromatic light beams of three preset wavelengths (the primaries) to produce a mixture that is perceptually equal to a given monochromatic target stimulus. Here we use the linear relation between the color matching functions and the absorption probabilities of each type of cone to find particularly useful triplets of primaries. As a second goal, we also derive an analytical description of the trial-to-trial variability and the correlations of color matching functions stemming from Poissonian noise in photon capture. We analyze how the statistical properties of the responses to color-matching experiments vary with the retinal composition and the wavelengths of peak absorption probability, and compare them with experimental data on subject-to-subject variability obtained previously.

Optimizing methods to isolate melanopsin-directed responses

The intrinsic melanopsin photoresponse may initiate visual signals that differ in spatiotemporal characteristics from the cone-opsin- and rhodopsin-mediated signals. Applying the CIE standard observer functions in silent-substitution methods can require individual differences in photoreceptor spectral sensitivities and pre-receptoral filtering to be corrected; failure to do so can lead to the intrusion of more sensitive cone processes with putative melanopsin-directed stimuli. Here we evaluate heterochromatic flicker photometry (HFP) and photoreceptor-directed temporal white noise as techniques to limit the effect of these individual differences. Individualized luminous efficiency functions (V(λ)) were compared to the CIE standard observer functions. We show that adapting chromaticities used in silent-substitution methods can deviate by up to 54% in luminance when estimated with the individual and standard observer functions. These deviations lead to inadvertent cone intrusions in the visual functions measured with melanopsin-directed stimuli. To eliminate the intrusions, individual HFP corrections are sufficient at low frequencies (∼1Hz) but temporal white noise is also required at higher frequencies to desensitize penumbral cones. We therefore recommend the selective application of individualized observer calibration and/or temporal white noise in silent-substitution paradigms when studying melanopsin-directed photoresponses.

Plasticity in perception: insights from color vision deficiencies

Inherited color vision deficiencies typically result from a loss or alteration of the visual photopigments absorbing light and thus impact the very first step of seeing. There is growing interest in how subsequent steps in the visual pathway might be calibrated to compensate for the altered receptor signals, with the possibility that color coding and color percepts might be less severely impacted than the receptor differences predict. These compensatory adjustments provide important insights into general questions about sensory plasticity and the sensory and cognitive processes underlying how we experience color.

Observer metamerism in commercial displays

Two uniform patches presented on two displays under identical viewing conditions can appear as the same color to one observer but as mismatched colors to another observer. This phenomenon, called observer metamerism (OM), occurs due to individual differences in color matching functions. To avoid its potentially adverse impacts in display calibration and characterization, it is desirable to have a predictive model of OM. In this work, we report the computational results of how to use existing metrics to quantify the potential OM between commercial display pairs and a proposed OM metric that is verified through a psychophysical experiment.

Predicting color matches from luminance matches

Color vision and spectral sensitivity vary among individuals with normal color vision; thus, for many applications, it is important to measure and correct for an observer's sensitivity. Full correction would require measuring color and luminance matches and is rarely implemented. However, luminance matches (equiluminance settings) are routinely measured and simple to conduct. We modeled how well an observer's color matches could be approximated by measuring only luminance sensitivity, since both depend on a common set of factors. We show that lens and macular pigment density and $L/M$L/M cone ratios alter equiluminance settings in different ways and can therefore be estimated from the settings. In turn, the density variations can account for a large proportion of the normal variation in color matching. Thus, luminance matches may provide a simple method to at least partially predict an observer's color matches without requiring more complex tasks or equipment.

The Verriest Lecture: Adventures in blue and yellow

Conventional models of color vision assume that blue and yellow (along with red and green) are the fundamental building blocks of color appearance, yet how these hues are represented in the brain and whether and why they might be special are questions that remain shrouded in mystery. Many studies have explored the visual encoding of color categories, from the statistics of the environment to neural processing to perceptual experience. Blue and yellow are tied to salient features of the natural color world, and these features have likely shaped several important aspects of color vision. However, it remains less certain that these dimensions are encoded as primary or "unique" in the visual representation of color. There are also striking differences between blue and yellow percepts that may reflect high-level inferences about the world, specifically about the colors of light and surfaces. Moreover, while the stimuli labeled as blue or yellow or other basic categories show a remarkable degree of constancy within the observer, they all vary independently of one another across observers. This pattern of variation again suggests that blue and yellow and red and green are not a primary or unitary dimension of color appearance, and instead suggests a representation in which different hues reflect qualitatively different categories rather than quantitative differences within an underlying low-dimensional "color space."

Individual differences and their implications for color perception

Individual differences are a conspicuous feature of color vision and arise from many sources, in both the observer and the world. These differences have important practical implications for comparing and correcting perception and performance, and important theoretical implications for understanding the design principles underlying color coding. Color percepts within and between individuals often vary less than the variations in spectral sensitivity might predict. This stability is achieved by a variety of processes that compensate perception for the sensitivity limits of the eye and brain. Yet judgments of color between individuals can also vary widely, and in ways that are not readily explained by differences in sensitivity or the environment. These differences are uncorrelated across different color categories, and could reflect how these categories are learned or represented.

Human perception of light chromaticity: short-wavelength effects in spectra with low circadian stimulation, and broader implications for general LED sources

Although light sources are designed assuming the same color sensitivity for all viewers, inter-user variability can in fact cause significant discrepancies in individual perception. Here, perception variability related to short-wavelength effects is investigated. An experimental study is reported on LED sources with reduced blue content, which cause reduced circadian stimulation. Perceived chromaticity is strongly dependent on the viewer's age and spectral shape, in excellent agreement with a model based on modern colorimetry. Broader implications for LED sources in lighting and displays are discussed, and significant effects are found. These results confirm the inadequacy of conventional colorimetry and support the use of modern color science in the design and engineering of lighting products.

How to Report Light Exposure in Human Chronobiology and Sleep Research Experiments

Exposure to light has short- and long-term impacts on non-visual responses in humans. While many aspects related to non-visual light sensitivity have been characterised (such as the action spectrum for melatonin suppression), much remains to be elucidated. Here, we provide a set of minimum reporting guidelines for reporting the stimulus conditions involving light as an intervention in chronobiology, sleep research and environmental psychology experiments. Corresponding to the current state-of-the-art knowledge (June 2019), these are (i) measure and report the spectral power distribution of the acute stimulus from the observer's point of view; (ii) measure and report the spectral power distribution of the background light environment from the observer's point of view; (iii), make spectra available in tabulated form, (iv) report α-opic (ir)radiances and illuminance; (v) describe the timing properties of stimulus (duration and pattern); (vi) describe the spatial properties of stimulus (spatial arrangement and extent), and (vii) report measurement conditions and equipment. We supplement the minimum reporting guidelines with optional reporting suggestions and discuss limitations of the reporting scheme.

The Method of Silent Substitution for Examining Melanopsin Contributions to Pupil Control

The human pupillary light response is driven by all classes of photoreceptors in the human eye-the three classes of cones, the rods, and the intrinsically photosensitive retinal ganglion cells (ipRGCs) expressing the photopigment melanopsin. These photoreceptor classes have distinct but overlapping spectral tuning, and even a monochromatic light with a wavelength matched to the peak spectral sensitivity of a given photoreceptor will stimulate all photoreceptors. The method of silent substitution uses pairs of lights ("metamers") to selectively stimulate a given class of photoreceptors while keeping the activation of all others constant. In this primer, we describe the method of silent substitution and provide an overview of studies that have used it to examine inputs to the human pupillary light response.

Individual differences in visual science: What can be learned and what is good experimental practice?

We all pass out our lives in private perceptual worlds. The differences in our sensory and perceptual experiences often go unnoticed until there emerges a variation (such as 'The Dress') that is large enough to generate different descriptions in the coarse coinage of our shared language. In this essay, we illustrate how individual differences contribute to a richer understanding of visual perception, but we also indicate some potential pitfalls that face the investigator who ventures into the field.

The human visual cortex response to melanopsin-directed stimulation is accompanied by a distinct perceptual experience

The photopigment melanopsin supports reflexive visual functions in people, such as pupil constriction and circadian photoentrainment. What contribution melanopsin makes to conscious visual perception is less studied. We devised a stimulus that targeted melanopsin separately from the cones using pulsed (3-s) spectral modulations around a photopic background. Pupillometry confirmed that the melanopsin stimulus evokes a response different from that produced by cone stimulation. In each of four subjects, a functional MRI response in area V1 was found. This response scaled with melanopic contrast and was not easily explained by imprecision in the silencing of the cones. Twenty additional subjects then observed melanopsin pulses and provided a structured rating of the perceptual experience. Melanopsin stimulation was described as an unpleasant, blurry, minimal brightening that quickly faded. We conclude that isolated stimulation of melanopsin is likely associated with a response within the cortical visual pathway and with an evoked conscious percept.

Variations in normal color vision. VI. Factors underlying individual differences in hue scaling and their implications for models of color appearance

Observers with normal color vision vary widely in their judgments of color appearance, such as the specific spectral stimuli they perceive as pure or unique hues. We examined the basis of these individual differences by using factor analysis to examine the variations in hue-scaling functions from both new and previously published data. Observers reported the perceived proportion of red, green, blue or yellow in chromatic stimuli sampling angles at fixed intervals within the LM and S cone-opponent plane. These proportions were converted to hue angles in a perceptual-opponent space defined by red vs. green and blue vs. yellow axes. Factors were then extracted from the correlation matrix using PCA and Varimax rotation. These analyses revealed that inter-observer differences depend on seven or more narrowly-tuned factors. Moreover, although the task required observers to decompose the stimuli into four primary colors, there was no evidence for factors corresponding to these four primaries, or for opponent relationships between primaries. Perceptions of "redness" in orange, red, and purple, for instance, involved separate factors rather than one shared process for red. This pattern was compared to factor analyses of Monte Carlo simulations of the individual differences in scaling predicted by variations in standard opponent mechanisms, such as their spectral tuning or relative sensitivity. The observed factor pattern is inconsistent with these models and thus with conventional accounts of color appearance based on the Hering primaries. Instead, our analysis points to a perceptual representation of color in terms of multiple mechanisms or decision rules that each influence the perception of only a relatively narrow range of hues, potentially consistent with a population code for color suggested by cortical physiology.