This article is in active development! All contents are subject to change!

How to Correct Color Vision Deficiencies

Text in this style is marked as an important remark by the author of this article.

Color vision deficiencies (abbr.: "CVD", coll.: "colorblindness" or "color blindness"), which can come in as many different types and strengths as there are individuals possessing them, are anomalous color visions that restrictively and qualitatively deviate from normal trichromacy. Normal trichromacy constitutes the average color vision of humans that have three normally functioning cone types for detecting different (yet partially overlapping) wavelength ranges of the electromagnetic spectrum. Color vision deficiencies introduce a reduction of the dimensionality of color for those affected by one. A mild color vision deficiency might reduce the dimensionality of color only slightly, but a strong one will transform the normal trichromatic 3-dimensional color space closer to a 2-dimensional one. In severe dichromatic color vision deficiencies, perception collapses into just two primary hues that stand in stark contrast to each other—alongside a separate black-and-white (luminance) channel. From here on, we’ll concentrate on these strong deficiencies (i.e., dichromacies such as protanopia and deuteranopia) as our canonical examples, since mild and moderate forms vary so widely from person to person that they’re far harder to model.

In this article I aim to explain color vision deficiencies from a perceptual and color space oriented perspective. We'll also look at the retina and its cone cells as well as their simplified and normalized response spectra to spectral stimuli.

According to Machado et al. (2009) about 8,32% of Caucasians, 4,78% of Asians and 3.14% of Africans have some kind of color vision deficiency (CVD). As a rough estimate, this is about 200+ million affected people worldwide. This makes color vision deficiencies a common enough occurence to be categorized as siginficant. 

Unprecedentedly, this article presents a novel method to simulate color vision deficiency of any severity and functionally correct the three major dichromacies (protanopia/deueranopia/tritanopia) as well as the many transitional color vision deficiencies (protanomaly/deuteranomaly/tritanomaly) by constructively disrupting the chromatic redundancy of binocular color vision to create novel color experiences. 

In the following sections and their subitems, we will first examine what color vision deficiencies are, later explore how to simulate (and daltonise) them so that trichromats can better appreciate the challenges they pose, and finally discuss methods for functionally correcting these deficiencies to restore performance equivalent to normal trichromacy.

According to these models, a normal trichromat with three normally functioning cone types—S (short), M (medium) and L (long)—enjoys a 3-dimensional color space and a 1-dimensional circle of distinct hues. Each trichromatic hue can vary in saturation and luminosity. Each concept—hue, saturation and luminosity; or R (red), G (green) and B (blue)—introduces an additional dimension of color.

On the other hand, a dichromat with only two functioning cone types—S plus M, L or M'/L' (i.e. a spectrally shifted cone type; labelled as "Q" by Jessica Lee et al. (2024) in Figure 1.2/1)—enjoys a 2-dimensional color space and two 0-dimensional hues—often represented by yellow and blue. "0-dimensional" means that the hue experience itself only exists in a single untainted hue state that can vary in saturation and luminosity. To get from yellow to blue in red-green dichromacy (protanopia/deuteranopia) you have to go through either the achromatic (i.e. the perceptual white) point, black, or the line of grays in between, and vice versa. Because hue in dichromacy is 0-dimensional, adding the dimensions of saturation and luminosity results in a 2-dimensional color space. Naturally, the color space of a mild to moderate color vision deficiency falls somewhere between the 'full' 3-dimensional volume of normal trichromatic vision and the flat 2-dimensional plane of dichromatic vision.

Hue in monochromacy is a missing quality due to the principle of univariance. The experience of different hues is generated through the interactions of at least two distinct cone types. With only one functional cone type—S, M or L—there's only a 1-dimensional concept of luminosity—a line of colors with black at one and the achromatic point at the other end.

Interestingly, instead of adding a new concept similar to hue, saturation or luminance, hue in tetrachromacy becomes 2-dimensional. Tetrachromacy is a color dimensionality with four functioning cone types: S, M, L and M'/L'. For each trichromatic color there exists a line of distinct colors in functional tetrachromacy. This means that each tetrachromatic hue can be pictured as occupying a distinct location on the surface of a 3D hypersphere/-cube. This extends the 3D color space into a 4th dimension of color, because every tetrachromatic hue of this 2-dimensional hue plane can vary in saturation and luminosity.

By comparing these four color vision dimensionalities we can observe the following pattern: The dimensionality of one's color vision is directly correlated to the amount of functional cone types that one possesses. Hue, saturation and luminance seem to be the final color concepts. After trichromacy, only the dimension of hue increases and the other two concepts adjust accordingly. Crucially, the dimensionality of color does not exist in absolutes, but rather in transitional states. A reductive anomalous trichromacy can be 2,5-dimensional relative to a 3D-dimensional and a 2D-dimensional color space. An anomalous trichromacy with three spectrally better separated cone types would yield a more 3-dimensional color space compared to normal trichromacy.

In order to generate a more intuitive understanding of the dimensionality of color vision, let's take the models of Figure 1.2/1 and make them interactive and dynamic. The following Color Space Simulator (0D-4D) application visualizes 𝑑-dimensional color vision from total blindness (0D) to tetrachromacy (4D).

The Color Space Simulator (0D-4D) application visualizes how the color space and its colors change according to the functionality and distinctiveness of one's cone types. It uses total blindness, blue-cone monochromacy, protanopia, normal trichromacy and a non-functional simulated tetrachromacy as its absolute states, with many transitional states in between. A reduction in the dimensionality of this color space is always accompanied by a comparatively worse color discriminability.

Please note that this is a simplified, although mostly accurate visualization. For example, because a 4-dimensional, tetrachromatic color space is difficult to visualize, this application duplicates the existing 3D color cube, extrudes the duplicated cube into a 3-dimensional direction and generates connective lines between the original and new color dots. This is a common method to simulate a tesseract (i.e. a 4D hypercube) on a 2-dimensional screen.

In conclusion, the dimensionality of one's color vision is defined by the amount and distinctiveness of the functional cone types that one possesses. To understand color vision deficiencies is to understand how the dimensionality of color changes according to cone type functionality.

Next, we'll focus on the functionality of cone types and the many ways it can become impaired relative to normal trichromacy. Any impairment will reduce the dimensionality of one's color space.

There are four major anomalies through which the color vision of normal trichromacy can become impaired:

1. Spectral Peak (Shift): Here, a cone type spectrally shifts closer to another. This assimilation leads to less distinctiveness of the affected cone types and their color qualities. At a maximum overlap the two cone types can form one unified singular cone type.

2. Population Functionality (Efficacy): Here, some of a cone type's cones become non-functional. This means they don't respond to spectral stimuli anymore or only to a lesser extent. This results in a relative change in distribution/saliency of the cone types and makes the color quality of the affected cone type less salient. Overall brightness is reduced.

3. Population Change (Distribution): Here, some of the cones of a cone type are substituted by the cones of another cone type. This also results in a relative change in their distribution and makes the color quality of the affected cone type less salient. Overall brightness is maintained.

4. Colored Lenses (Spectral Dimming Filter): Here, either the natural lens or an artificial one is colored, which reduces the saliency of the colors opposite to the color of the lens (e.g. a yellow colored lens dims bluish light). It works by dimming certain wavelength ranges. In this case, the cones of more than one cone type can be affected at the same time.

Notably, there's a fifth anomaly: tetrachromacy alters the cone mosaic by adding a fourth cone type—but because retinal space is finite, this addition inevitably reduces the density of one or more of the existing cone types. Furthermore, brain damage can also lead to a color vision deficiency, which could count as a sixth anomaly.

The following Cone Type Functionality Simulator application visually demonstrates the first three anomalies (1-3) through spectral sensitivities (normalized responsivity spectra to spectra stimuli) of human cone cells (S, M, and L types) and an examplary segment of a retina's cone mosaic of a single (trichromatic) human eye.

The red-green assimilation in the Spectral Peak (Shift) option of the Cone Type Functionality Simulator application is modelled after the findings of Machado et al. (2009). They found that shifting the peak of either the L- or M-cone sensitivity curve by 20 nanometers (nm) towards the other induces a strong protanomaly/deuteranomaly (red-green color vision deficiency), nearly indistinguishable from full protanopia/deuteranopia (dichromacy). Even with the L-cone peak shifted by just 20 nm towards the M-cone—leaving about 10 nm of separation—color vision still exhibits a strong "-omaly" due to the increased overlap between the two cone types. Note that in the Spectral Peak (Shift) option, I’ve chosen to reduce the S-cone peak rather than shift it, since S-cone shifts are exceedingly rare. Furthermore, this is a simplified model of the normalized responsivity spectra to spectra stimuli of the S-, M- and L-cone types.

Interestingly, no matter through which way a color vision deficiency of a certain type is created, all four anomalies (1-4) can result in a similar severity.

In conclusion, there are not just different types of color vision deficiencies [protanomaly/-opia, deuteranomaly/-opia, tritanomaly/-opia; rarer: (red-, green- and) blue-cone monochromacies], but these types can also come in different severities. Additionally, each type and severity can be obtained through four different anomalies: increased overlap due to shifting of spectral cone curve peaks; decreased efficacy due to decreased cone population functionality; altered distribution due to a change in relative cone population; spectral dimming through colored lenses. More than one anomaly can exist in a single individual.

In section 1. What is Color Vision Deficiency (CVD)? I've explained what color vision deficiency and dichromacy is and how it occurs. This simplified explanation has already offered a glimpse into the perceptual experience of color vision deficiencies. In section 2. Simulating Color Vision Deficiencies we're going to focus on how to functionally simulate color vision deficiencies as a normal trichromat. This allows normal trichromats to better understand the unique perspectives of individuals living with color vision deficiencies. It will also pave the way for creating a functional correction for dichromacies (protanopia, deuteranopia, tritanopia) and color vision deficiencies (protanomaly, deuteranomaly, tritanomaly) towards normal trichromacy, which we will be discussing in the next section 3. A Functional Correction for Dichromacies Towards Normal Trichromacy.

A protanope should perceive virtually no difference between the two spectra of Figure 2.2/1 and Figure 2.2/2a, but individual variance is normal. The protanopic spectrum of Figure 2.2/2a is equivalent to either the L-cone types missing or being severely impaired, disproportionally many L-cones being substituted by M-cones, the L-cone type's peak spectrally shifted closer to the M-cone type's peak (~20 nm), or reddish light being filtered out via a strongly cyan colored lens.

Likewise, a deuteranope should perceive virtually no difference between the two spectra of Figure 2.2/1 and Figure 2.2/2b, but individual variance is normal. The deuteranopic spectrum of Figure 2.2/2b is equivalent to either the M-cone types missing or being severely impaired, disproportionally many M-cones being substituted by L-cones or the M-cone type's peak spectrally shifted closer to the L-cone type's peak (~20 nm). In this case, dimming green light through a magenta colored lens does not simulate deuteranopia/-omaly for normal trichromats.

How the color conversion works:

The Custom Color Vision method exchanges every single hue of the normal 1-dimensional trichromatic hue spectrum and its related colors with the custom color on the same 1-dimensional position of a customized color spectrum. In other words, every trichromatic hue's respective color plane will adjust according to the selected custom color on the same 1-dimensional location. We can design the same color vision deficiency types and severities with this Custom Color Vision method as with, for example, the matrix oriented method of Machado et al. (2009). But the Custom Color Vision method is simpler, more modifiable and more easily shareable. You only need a single image (x: ~1530 pixel, y: ~4 pixel) to access your customized color vision. Additionally, you can design, save, load, share and re-design the specific settings of your customized color vision via the accompanying free online 1D Color Texture Designer (Multi-Spectrum) application.

I've developed the identically named Custom Color Vision application for Windows 11 (and with limited functionality also for Windows 10) that adjusts the screens' color information according to any custom color spectrum in real time. You will soon be able to purchase this application on Ooqui Sensory Lab's itch.io page (and later also on Steam). 

By default, as the only main application running, the Custom Color Vision application runs on 40-60+ FPS on a single 1080p-screen with an Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz CPU, a 16,0 GB main memory, and an NVIDIA GeFOrce GTX 1070 Ti graphics card. Apart from a few exlusive fullscreen applications—which you will have to set to windowed fullscreen—it applies the color transformation to the entire selected screen. Naturally, the more performance intensive the applications and games are that you're running concurrently, the more the FPS of the Custom Color Vision application will be restricted. As an example, I can play Elden Ring (2022) with my setup with an average of ~45 FPS at 1080p with no major lags.

Here are snapshots of the Custom Color Vision application, Version 1.20:

In Figures 2.2/6 you can see my test results for EnChorma's "New & Improved Color Blind Test" with the Custom Color Vision application running. I used the protanopia spectrum of Figure 2.2/2a, the deuteranopia spectrum of Figure 2.2/2b and the tritanopia spectrum, that you can find on the starter spectra page, in order to test my therethrough altered color vision.

Please note that me using the EnChroma test for color vision deficiencies is not an advertisement or endorsement of their products (e.g. their "color correction glasses"). The EnChroma CVD test is the first result to come up in a google search for "colorblind test". Hence, I'm using this test as an example because people are most likely to test their color vision with it.

In conclusion, both the Custom Color Vision method and application present a simple and highly modifiable digital technology that allows its users to simulate all color vision deficiencies below (and including) normal trichromacy. It also allows normal trichromats to simulate improved trichromatic color vision in digital contexts.

The Custom Color Vision application enables individuals with normal trichromatic vision to accurately experience any color vision deficiency below normal trichromacy. It also allows people with color vision deficiencies to apply Daltonisation—a color-correction method that shifts and alters the perceiveable hues to make them less confusing for color vision deficient people—to everything that's visible on the selected screen in real time. Although this method helps with distinguishing some previously indistinguishable colors, it's always at the expense of other colors. Metamerism—i.e. when colors that look identical but are actually distinct—is never reduced with this approach, because no new color experiences have been added. The normally available colors were only shifted or subsituted by other normally available colors. Although shifts are generally applied into a color dimension that might not normally be distinguishable by those with a color vision deficiency. Daltonization is a good technique, but vastly inferior to the technique I'm presenting in the next section 3. A Functional Correction for Dichromacies Towards Normal Trichromacy.

Simple Daltonization:

However, these simple color substitutions can still be very useful. Not all applications and games have a built-in Daltonisation or a fully customizable color/hue substitution system. People with color vision deficiencies who play games without custom color substitution support face more difficulties than normal trichromats, for example.

Figures 2.2/1a-c show a normal trichromatic hue spectrum, a protanopic hue spectrum and an altered protanopic hue spectrum with a custom Daltonisation design applied.

Specific Daltonization:

Since Custom Color Vision allows you to arbitrarily adjust the colors of the digital trichromatic hue spectrum, we can also apply more specific Daltonisations that aren't possible with convential applications designed for Daltonisation.

The custom spectrum of Figure 2.3/4 is merely an example. The Custom Color Vision application is highly modifiable. You can replace any hue with any color that's more distinguishable to you. How incredibly helpful this can be is left to your imagination and creativity.

Extreme Daltonisation:

Going several steps further to showcase the great modifiability of Custom Color Vision even better, let's design a strong Daltonisation spectrum for red-green color vision deficiencies that would be difficult to accomplish with conventional applications/methods.

Hue Focus Daltonisation:

You can also focus on specific hues using colors that are distinct for you. You can select which trichromatic hue you want to focus on via a custom 1D/2D texture. In the example of Figure 2.3/10 reds, oranges, yellows, limes, greens, turquoises, cyans, cobalts, blues, purples, magentas and pinks can be individually highlighted with colors/hues that are distinct to you. Since in this case no other color has a distinct hue, you can more easily identify the selected hues.

Hue focus can be very useful for monochromats. In Figure 2.3/11 you can see that even with monochromacy you can distinguish red, yellow and green simply by individually focusing on them. Here, too, creativity and personalization are your friends, since you can design your own custom spectra in any way you like.

Custom Color Vision’s highly adaptable Daltonisation—enabled by fully customizable spectral profiles—makes it a powerful tool for those with color vision deficiencies. It empowers users to enhance accessibility in any digital application while tailoring spectral corrections to their individual needs. The above examples merely scratch the surface of what an engaged color vision deficient community can devise to support its peers.

In conclusion, the Custom Color Vision application allows its users to apply a customized Daltonisation color correction to everything that's visible on the selected screen in real time—including games, applications, browsers, and so on (but excluding some exclusive full screen applications).

In this section, I'm going to present how to functionally correct dichromacies and color vision deficiencies so that people affected by them can develop a color discriminability that's equivalent to that of normal trichromacy.

The key insight of section 2. Simulating Color Vision Deficiencies is that we can customize our color vision through digital devices and intelligently designed applications that alter digital colors. This highly modifiable color vision customization paves the way for functionally correcting color vision deficiencies.

In order to increase the dimensionality of a dichromat's color space, we need a method to introduce subjectively new and distinct color experiences that are unlike any other color experience of their color vision. We need to fill the missing parts of this person's color space with these new colors. A color space cannot be extented without the introduction of new and distinct color experiences, because any extension without them would only result in a stretched version of the original—both the original and stretched have the identical amount of distinguishable colors. The only minor exception is when a cone type's spectral sensitivity spectrally shifts outside of the usually visible spectrum.

The problem we now face is that creating new color experiences seems impossible. The colors we can see are the all the colors we can see—as mundane as this sounds. Omitting the still largly uncharted and risky territories of gene therapy for improving the dimensionality of color vision, color vision seems to be a static system that doesn't change. However, in the previous section we've learned that we can customize our color vision with the help of digital technology and intelligently constructed applications. Therefore, color vision isn't a static system anymore in this digital age, but rather highly modifiable—if you have access to the right technology.

The next part will present the only currently available method that binocular humans can employ that can introduce subjectively new and distinct color experiences for binocular color vision and increase the dimensionality of color space.

The following images are using stereo viewing techniques to simulate normally impossible, binocularly differing color combinations. You have to cross your eyes (cross-/parallel-view) to overlap the frist and second eye's perspectives in order to see these new color experiences in the following exemplary cases. This is a skill and it takes training to become good at it.

Normally, each eye provides nearly identical color input, making one eye’s color contribution redundant—although the second eye is important for depth and visual acuity. You can see this chromatic redundancy in Figures 3.2/1a-b. 

Evolutionarily, this is a great advantage. Losing or closing one eye means that you can still distinguish colors normally. If you have normal binocular color vision, then closing one eye won't decrease your color discriminability or color space. This effect is simulated in Figures 3.2/2a-b by making all the colors that the second eye sees black.

By deliberately disrupting this chromatic redundancy of binocular color vision and feeding your eyes two distinct color sets for the same visual scene, you simulate the effect of having extra cone types. This is because each of your eyes acts as an individual organ. Both eyes' perspectives are only combined later by the brain. A disruption of this chromatic redundancy of binocular color vision will double the amount of functional cone types that one possesses, because each eye has its own set of cone types and color channels. Only the quality of both eyes' colors are similar, which isn't a problem because of the unique way non-retinal (i.e. binocularly differing or "impossible") color combinations are experienced. For a dichromat, this can maximally generate a form of simulated tetrachromacy (4-dimensional color space). For a trichromat, you can extend color vision towards hexachromacy (6-dimensional color space), though intermediate stages like non-retinal tetrachromacy and pentachromacy are also possible.

Disrupting the chromatic redundancy of binocular color vision introduces new and distinct color experiences known as "impossible color combinations". These are colors where one eye sees yellow and the second eye sees blue for the same visual spot, for example. Any color of one eye can be combined with any color of the second eye. The distinctiveness of these new colors comes from the fact that the brain treats and combines the non-retinally (i.e. binocularly) generated color combinations differently than retinal color combinations. For example, you're able to see a normally impossible yellow/blue color combination by disrupting chromatic redundancy. A yellow/blue color would default to a white color in additive retinal color mixing, but becomes a unique hue in non-retinal color mixing. You can see blue substituted for an impossible yellow/blue color combination in Figures 3.2/2a-b.

Here, "retinal" means that only one eye is involved, and "non-retinal" means that both eyes are involved and binocular fusion of unequal color experiences can occur.

Any color can you can see with one eye can be combined with any color you can see with the second eye. The resulting non-retinal color combinations are impossible in retinal color mixing. But they can be stably and consistently seen and identified by experienced and trained viewers and evoke new and distinct color experiences. Naturally, someone with only one eye cannot see impossible binocular color combinations.

It's normal to not be able to stably and consistently see these impossible color combinations as a beginner. It takes time, training and a lot of exposure to these extraordinary colors for optimal binocular fusion to occur. Before the mastery of this skill binocular rivalry might occur, a phenomenon describing the initial instability of such impossible (non-retinal) color combinations. Your brain needs to get used to your eyes receiving different color information for the same visual spots in order to eventually and functionally combine the two eyes' different color experiences into a single, stable and consistent impossible (non-retinal) color experience. For most individuals, this is a skill that needs to be developed over a longer period of time. 

There's still an insufficient amount of scientific studies and research showing how long it takes until stable binocular fusion is achieved. But it's likely different for each individual and depends on many different factors (e.g. eye dominance; visual acuity; head rotation; eye focus; eye-crossing ability; eye muscle fatigue; utilized technology; color quality; wearing glasses; knowledge about colors; training time; and so on). The theory is sound, but putting it into practice and achieving functionality is a skill.

For people with strong color vision deficiencies this means that for every normal dichromatic color there exists a plane of new and impossible color combinations, because—again—every single color of one dichromatic eye's planar color space can be inflected by the entire plane of colors of the second dichromatic eye by disrupting the chromatic redundancy of binocular color vision.

In conlcusion, this method of disrupting chromatic redundancy of binocular vision is the only currently available and the safest method to reliably and functionally introduce new color experiences—excluding gene therapy—in the form of 'monocularly impossible' binocular (non-retinal) color combinations.

In the beginning of this section's subitem 3.1 Preface we've pondered whether it's impossible to generate new and distinct color experiences in order to functionally expand one's color space. 

However, subitem 3.2 Disrupting Chromatic Redundancy of Binocular Vision to Create New and Distinct Color Experiences has demonstrated that we don’t merely gain a few additional color experiences through impossible binocular color combinations—instead, a person with normal binocular vision can experience a color space dimensionality that is always twice that of their monocular vision.

The impossible color combinations that result from the disruption of chromatic redundancy of binocular color vision are the new and distinct color experiences that we were hoping for—even though it's a trained skill to correctly see them. They can be used to functionally expand the dimensionality of color.

Next up, we need to think about how to impossibly combine the colors of each eye in order to functionality expand color vision. Here, we'll focus on dichromacies (and strong color vision deficiencies) in order to demonstrate the functionality of the following method for the most extreme cases, and because expanding the inherently trichromatic digital color space (e.g. towards tetrachromacy) is impossible without additional color information. Fortunately, since digital color information is already trichromatic, we can exchange the trichromatic colors that a dichromat can't distinguish/see with impossible dichromatic color combinations that they can distinguish/see.

For this, the Custom Color Vision method that I've first introduced in subitem 2.2 An Easy and Highly Modifiable Method to Simulate Color Vision Deficiencies (CVD) becomes incredibly useful. Previously in the Custom Color Vision application, we've applied a singular customized spectrum to the color vision of both eyes identically. However, we can design a distinct custom spectrum for each eye. If done correctly, we can design an impossible trichromatic color vision that uses distinct impossible dichromatic color combinations instead of normal trichromatic colors, with no color duplicates. Such an "impossible trichromacy" only differs in the quality of its trichromatic colors, but its color discriminability is equivalent to that of normal trichromacy in experienced and trained viewers. This method of individually designing each eye's color vision is the main functionality of my PC/VR application Color in Color (Ooqui Sensory Lab) that's still in development.

To better understand why color discriminability is more important than color quality (i.e. the qualia, the personal and subjective experience of a color), we can remember the notorious "Is my red your green and your green my red?" question that everyone has already heard or thought of at least once before. The quality of any color perceived by one individual is inherently incomparable and indescribable to that of another individual. You can only describe a color to someone else by using another medium (e.g. language, pointing at specific objects with that color, etc.). Therefore, we can only test for behavioral functionality of color vision in order to see whether people have a similar or dissimmilar color discriminability and compare the results against each other—and ultimately to a norm (i.e. normal trichromacy). Whether your red is a normal trichromatic red, an impossible yellow/blue color combination or a green, as long as you can distinguish the same colors as a normal trichromat the quality of your colors doesn't matter.

The color vision of Figure 3.3/1b is in no way inferior to that of Figure 3.3/1a. People with a red-green color vision deficiency might not immediately spot the difference, because some of their confusion colors are interchanged.

Now we need to find an intelligent method to construct two distinct custom spectra, one for each eye, that together generate an impossible trichromacy with only impossible dichromatic color combinations. Here, "impossible trichromacy" refers to a trichromacy generated through the implementation of distinct impossible dichromatic color combinations into a thereby artificially expanded color space. For this, it's important to recognize the following peculiarity:

An individual with a protanopia (red-blindness, anomalous L-cone type) in one eye plus either a deuteranopia (green-blindness, anomalous M-cone type), a tritanopia (blue-blindness, anomalous S-cone type) or a red-monochromacy (red and green blindness, anomalous M- and L- cone types) in the second eye can still develop normal trichromacy.

Because the red-green contrast is most optimally conserved in protanopia plus tritanopia or red-monochromacy, in protanopia plus deuteranopia it might not be salient enough—at least for a dichromacy correction that uses impossible color combinations of only the colors that said dichromat can see.

As a normal trichromat you can test this impossible trichromatic vision that's generated through two different dichromacies, one in each eye, by wearing a strongly yellow colored lens over one eye (which simulates a strong tritanomaly) and a strongly cyan colored lens over the second eye (which simulates a strong protanomaly). Even though each of your eyes will have a strong color vision deficiency on their own, together the two eyes can distinguish colors trichromatically because the brain can calculate the non-retinal color differences in its post-processing. Color quality will not be preserved, but color discriminability is maintained.

Here in Figures 3.3/2a-b, red is a red/black, magenta a red/blue, blue a black/blue, cyan a green/white, green a green/yellow, yellow a white/yellow and white stays white. These impossible color combinations are qualitatively different to normal trichromatic colors, but they still allow for normal trichromatic color discriminability, because each of them introduces a qualitatively distinct hue category.

In order to correct dichromacies and strong color vision deficiencies with this method we need to be a little bit more creative, because in binocular dichromacy both eyes see the same subjective color qualities. In Figures 3.3/2a-b we've had the luxury of both eyes having different color qualities, however, in binocular dichromacies this luxury is missing. We need to develop a method where each hue on the normal trichromatic spectrum can be replaced by a unique impossible dichromatic color combination in order to create an impossible trichromacy for dichromats.

For red-green dichromacies (protanopia & deuteranopia) such a fully functional impossible trichromacy with only impossible dichromatic color combinations looks like this:

Figures 3.3/4a-b are easier to overlap using stereo viewing techniques.

In both Figures 3.3/3a-b and Figures 3.3/4a-b color vision is functionally trichromatic. This design with two custom spectra, one for each eye, is the world's first that uses distinct impossible red-green dichromatic color combinations to evoke distinct trichromatic hue experiences. This custom color vision is functionally trichromatic for both protanopes and deuteranopes as well as for normal trichromats.

For tritanopia you can either use the same custom spectra of Figures 3.3/3a-b and change the yellow to red, and blue to green, for example, as depicted in Figures 3.3/5a-b.

Or you can re-design the custom spectra to be more faithful to the tritanopic color experience as in Figures 3.3/6a-b.

In the following Stereo Dot Viewer application you can draft a simplified version of a custom hue spectrum combination for your own color vision. Use the 1D Color Texture Designer for complete and exportable custom spectra.

In Figures 3.3/7 you can see the functional red-green dichromacy correction of Figures 3.3/3a-b applied to normal color images. Use stereo viewing techniques to overlay the lower left and lower right displays.

In Figures 3.3/8 you can see the functional red-green dichromacy correction of Figures 3.3/3a-b applied to Ishihara plates that test for protanomaly, deuteranomaly and tritanomaly. Use stereo viewing techniques to overlay the lower left and lower right displays.

While you might not be able to tell these impossible colors and hues apart yet and many impossible color combinations might still be unstable and inconsistent as a beginner, if you train everyday for this impossible trichromacy you will naturally become better at it.

According to Mancuso & Neitz et al. (2009), even still unvailable and risky gene therapy towards trichromacy in trained monkeys only resulted in a functional form of trichromacy after about 20 weeks post-injection. According to Adams et al. (1998), even infants don't perfectly see colors at first and only develop their full color vision after some time. Impossible color vision is different from these two examples, but they demonstrate that color vision augmentations need some time to become effective and functional. By the time a hypothetical gene therapy for humans would become effective, you are likey to already have had enough training and adaption time to stably and consistently see most impossible color combinations—and this without any risks involved.

However, use this impossible color vision with caution because prolonged eye-crossing and impossible color viewing can lead to feelings of discomfort for some people—especially for beginners. A generally good advice is to start with simple impossible colors (e.g. equal brightness combinations of differing hues, e.g. yellow/white) and incrementally introduce more and more difficult impossible colors (e.g. unequal brightness combinations of differing hues, e.g. black/white) for longer and longer periods and in different contexts so that your eyes and brain can get used to them. Don't except to see every impossible color perfectly and stably at your first or second and for some not even at your tenth try. It takes time to develop this skill of stably and consistently seeing and correctly identifying impossible color combinations.

While this approach provides a functional correction for dichromacies and color vision deficiencies in digital contexts, it’s important to note that it is not a true biological cure (only gene therapy has this potential). For example, you still won’t see “red” in the exact quality a normal trichromat does; instead, you’ll see a specially remapped impossible color combination that serves the same purpose, because it's a new and distinct color experience unlike any other. As subjective color perception is ultimately personal and unverifiable between individuals, the key point is that you can now reliably identify and differentiate colors that were previously indistinguishable as someone with a color vision deficiency.

In conclusion, by presenting a different, custom-designed spectrum to each eye, we can encode otherwise “impossible” dichromatic color combinations within a virtual 3-dimensional (trichromatic) color space, allowing people with dichromacy to replace indistinguishable hues with these novel and unique color experiences and thereby expand their inherently two-dimensional color perception into a full three-dimensional experience.

For people with only a single functioning eye impossible color combinations do not work. You need at least two eyes to see non-retinal color mixes. Additionally, some people find it difficult or impossible to use stereo viewing techniques, or get a feeling of discomfort when looking at impossible color combinations for a longer period of time.

For those people who cannot view impossible color combinations, temporal colors are the next best option to augment their color vision. Instead of using your second eye to introduce new color dimensions, temporal color vision introduces new colors and greater color dimensionality through the dimension of time. Temporal colors are colors that cyclically flicker between two or more different colors. In this case, colors aren't linked non-retinally, but across time.

Stability in temporal color vision is not an option, since flickering is inherently unstable. However, in temporal color vision color discriminability and dimensionality can exceed the limits of impossible binocular color vision due to the infinite flow of time. Impossible binocular color combinations only allow for a maximum of two colors to be visible at once at the same visual spot (e.g. red in one eye, green in the second eye), but temporal color vision allows for more than two colors to be visible at the same spot across time. When the flicker timing and the temporal colors are chosen adequately, color vision can be augmented to more than 2-times the dimensionality of a single eye.

Videos 3.4/1-2 show how temporal colors can be used to augment both dichromatic and monochromatic color vision towards normal trichromacy. While these temporal colors are more unstable than impossible color combinations (this refers to the flickering), their inherent instability is more predicatable, can be viewed monocularly and can augment even monochromacy towards trichromatic vision. You have to get used to this temporal color vision and color flicker, but once you're used to it your options for combining colors and augmenting your color vision expand manifold across time.

Temporal Color Vision Example: Temporal Trichromacy for Red-Green Dichromats  

Temporal Color Vision Example: Temporal Trichromacy for Monochromats 

In conclusion, rapid temporal color modulation (i.e. temporal colors) can improve color perception at least as much as impossible color combinations; and its ultimate reach is bounded only by the 3-dimensional volume of the digital trichromatic (RGB) space and by the inherent dimensionality of the viewer’s own color vision.

In this digital and technology-driven age color vision isn't a static sense anymore. With the right technology color vision becomes incredibly modifiable and can be augmented beyond normal color dimensionality and discriminability. This article has demonstrated that using the Custom Color Vision method and application any color vision can be augmented through custom Daltonisations, modifying the color vision of each eye independently to create normally impossible binocularly differing color combinations, as well as by utilizing the dimension of time to introduce temporal colors that augment color vision through color flicker.

There's still insufficient scientific data and studies on how long it takes until stable binocular fusion of impossible binocular color combinations is achieved. It appears to be a skill, an ability that needs to be trained over a longer period of time. Some people will naturally pick up this skill more easily and quickly than others. Hence, you should not use the Custom Color Vision application with the expectation to immediately and perfectly see impossible/temporal colors, but at first rather as a tool to learn how to see them. I (the author; i.e. Kilian-Roy Lachner) personally achieved full functionality in true-red non-retinal tetrachromacy, for example, only after weeks and months of exposing myself to these novel color experiences everyday and across many different contexts. Even to this day, the functionality of my true-red non-retinal tetrachromacy continues to increase with my understanding of the true-red tetrachromatic 4D color space and my growing experience with true-red tetrachromatic colors.

The Custom Color Vision application uniquely allows you to augment your color vision beyond your biological limits. It is the first application that allows you to functionally augment your color vision through the use of novel color experiences.

Read more about the Custom Color Vision application by clicking the following image or button:

• Adams, R. J. & Courage, M. L. et al. (1998). "Human newborn color vision: measurement with chromatic stimuli varying in excitation purity". In: J. Exp. Child Psychol., 1998 Jan, 68(1):22-34. Link: https://pubmed.ncbi.nlm.nih.gov/9473313/.

• Fong, J. et al. (2025). "Novel color via stimulation of individual photoreceptors at population scale". On: Science, ScienceAdvances. Link: https://www.science.org/doi/10.1126/sciadv.adu1052#con1.

• Jacobs, G. H. (2018). "Photopigments and the dimensionality of animal color vision". In: Neuroscience & Biobehavioral Reviews, Volume 86, March 2018, p. 108-130. Link: https://www.sciencedirect.com/science/article/pii/S0149763417307479.

• Jenny, B. & Kelso, N. V. (2007). "Color Design for the Color Vision Impaired". In: Cartographic Perspectives, (58), p. 61–67. Link: https://doi.org/10.14714/CP58.270.

• Judd, D. B. (1944). "STANDARD RESPONSE FUNCTIONS FOR PROTANOPIC AND DEUTERANOPIC VISION". In: Journal of Research of the National Bureau of Standards, Volume 33, December 1944, p. 407-437. Link: https://nvlpubs.nist.gov/nistpubs/jres/33/jresv33n6p407_A1b.pdf. This is an older source!

• Kröger, J. et al. (2013). "Analysis and Improvement of the OpenStreetMap Street Color Scheme for Users with Color Vision Deficiencies". Link: https://www.researchgate.net/publication/255978894_Analysis_and_Improvement_of_the_OpenStreetMap_Street_Color_Scheme_for_Users_with_Color_Vision_Deficiencies.

• MacEvoy, B. (2015). "Color Vision". "Light and the eye". Link: https://www.handprint.com/HP/WCL/color1.html.

• Machado, G. M. et al. (2009). "A Physiologically-based Model for Simulation of Color Vision Deficiency". Link: https://www.inf.ufrgs.br/~oliveira/pubs_files/CVD_Simulation/Machado_Oliveira_Fernandes_CVD_Vis2009_final.pdf.

• Machado, G. M. et al. (2009). "A Physiologically-based Model for Simulation of Color Vision Deficiency". Link: https://www.inf.ufrgs.br/~oliveira/pubs_files/CVD_Simulation/Machado_Oliveira_Fernandes_CVD_Vis2009_final.pdf.

• Mancuso K. & Neitz, J. et al. (2009). "Gene therapy for red-green colour blindness in adult primates". In: Nature, 2009 Oct 8, 461(7265):784-7. Link: https://pmc.ncbi.nlm.nih.gov/articles/PMC2782927/.

• Sabesan, R. et al. (2015). "Characterizing the Human Cone Photoreceptor Mosaic via Dynamic Photopigment Densitometry". PLoS ONE 10(12): e0144891. DOI: 10.1371/journal.pone.0144891. Link: https://www.researchgate.net/publication/286996700_Characterizing_the_Human_Cone_Photoreceptor_Mosaic_via_Dynamic_Photopigment_Densitometry.

• Tillem, M. & Gün A. (2023). "Color Blindness in the Digital Gaming Landscape: Addressing Critical Issues and Research Gaps". In: Proceedings of the 17th European Conference on Games Based Learning, Vol. 17 No. 1 (2023). Link: https://papers.academic-conferences.org/index.php/ecgbl/article/view/1749.

• Wikipedia (last update access: May 2025). "Color blindness". Link: https://en.wikipedia.org/wiki/Color_blindness#. Some elements of this source are refutable! E.g., the chart with the subtitle "These color charts show how different color blind people see compared to a person with normal color vision." is wrong.

As always, you can't trust the words of anyone on the internet. Always conduct your own research and experiments also. I try my best to provide correct information. If I fail to do so, please correct me using constructive criticism.