Dichromacy · Blue–yellow
Tritanopia Simulated in sRGB
It is the absence of working short-wavelength cones, the ones most sensitive to blue light. It is inherited on chromosome 7, not the X, so it affects men and women about equally, and it can also be acquired through eye disease. It is rare: estimates put it at around 1 in 10,000 people or fewer.
- Category
- Dichromacy
- Type
- Blue–yellow
- Cones
- S cones missing
- Simulation
- Machado et al. (2009)
Live lab
Everyday interfaces with tritanopia
Four everyday interfaces, or an image of your own, simulated as you watch. Drag the divider, change the severity, and turn on the fix to see the same interface designed so color is never the only clue.
The live lab needs JavaScript. The sections below show the same simulation, color by color.
A line chart of weekly sign-ups in four channels, told apart by color.
The same scene through all nine types, each at its usual severity.
- Closest pair
- Typical vision
- Tritanopia
- Verdict
ΔE2000 between the two key colors drawn closest together. Under 5 is too close for thin lines and small marks; 10 or more is safe at any size. A rule of thumb, not a standard.
With the fix, words, icons, numbers and patterns carry the meaning too, so the scene still reads even where two colors stay close.
A photo has no key colors to compare, so the lab simulates every pixel instead. Save the result to share it.
Everyday colors
How tritanopia changes common colors
Blue and green merge, and yellow and violet or pink become hard to tell apart; yellows can take on a pinkish cast.
Each chip is the color on the left and the simulation on the right. Open one for its full page.
Machado et al. (2009) built their tritan model as a shift of the S cones and say that even at full severity it is not tritanopia, so treat this simulation as an approximation.
Confusion lines
Colors on one line look alike
This map holds every color's hue and saturation, with the spectrum round its curved edge. Each straight line runs through colors that tritanopia cannot tell apart by hue; only lightness separates them. All the lines meet at one point, at the violet end of the spectrum. The line through white meets the spectrum at about 566 nm, the neutral point: light of that wavelength looks gray. Measured in tritanopes, it is about 569 nm.
Seen as one hue
The numbers are the closest pair in each group, typical vision first.
The lines meet at points worked out from the cone measurements of Smith and Pokorny; strictly those belong to a corrected version of this diagram, so the lines here are close, not exact. Smith & Pokorny 1975
- Confusion lines
- The line through white
- White (D65)
- Neutral point 566 nm
- sRGB gamut
Confusions
Colors tritanopia confuses
CSS named colors that are far apart to typical vision but nearly identical once simulated. Never tell these apart by color alone.
The cones
No working S cones
Each curve is how strongly one cone pigment absorbs light across the spectrum, peaking at about 419, 531 and 558 nm. The dashed curve is the pigment that is missing. Below the curves, the visible spectrum as typical vision and as this type see it.
- S
- —
- M
- 0.91
- L
- 0.99
Typical vision
Tritanopia
The curves are the visual pigment template of Govardovskii et al. (2000) at the peaks Dartnall, Bowmaker and Mollon measured in human eyes. They show the pigments themselves; at the cornea the eye's lens and macular pigment shift them. Govardovskii et al. 2000 · Dartnall et al. 1983
Genes
How it is inherited
The genes behind it, where they sit, and the odds of passing it on. Pick the parents to see a Punnett square: each square is one chance in four.
- Gene
- OPN1SW
- Where
- 7q32.1
- Inheritance
- Autosomal dominant
- What changes
- Changes in the S-opsin gene that disrupt the S cones. The gene has no look-alike neighbor, so hybrids like the red–green ones cannot form.
The calculator needs JavaScript.
If the parents are
A is the changed copy. One is usually enough, but not everyone who inherits it shows it, so these are the odds of inheriting it, not of having it. Tritan-like loss is also often acquired, with age or eye disease.
Diagnosis
How it is tested
A screening test says that something is different; a diagnostic one says what. Only an eye care professional can diagnose a color vision deficiency. Plates on a screen, including this site's, show the idea but cannot diagnose.
| Test | What it is | For this type |
|---|---|---|
| Ishihara plates | Plates of colored dots hiding numbers: the standard screen for red–green deficiency. It cannot grade severity or tell dichromats from anomalous trichromats. Ishihara instructions | Misses it |
| HRR plates | Plates with shapes instead of numbers. They screen for red–green and blue–yellow loss and grade it as mild, medium or strong. Bailey et al. 2004 | Detects and grades it |
| Farnsworth D-15 | Fifteen colored caps to put in order. Almost every dichromat fails; many people with a mild anomaly pass. The pattern of errors shows the axis. Birch 2008 | Detects it |
| Farnsworth–Munsell 100 Hue | Eighty-five caps in four trays, scored for errors. Good for measuring discrimination and following an acquired loss; weak at telling inherited types apart. Birch 1989 | Partly |
| Nagel anomaloscope | Match a yellow light with a mix of red and green. The reference test for red–green deficiency: it tells protan from deutan and dichromacy from anomalous trichromacy, and grades severity. Birch 2008 | Misses it |
| Moreland anomaloscope | A blue–green match for tritan defects. Its power is limited, because the eye's lens and macular pigment shift the match. Sharpe et al. 1999 | Partly |
| CAD test | A computer test that measures red–green and yellow–blue thresholds against moving luminance noise, and reports severity in standard units. Rodriguez-Carmona et al. 2012 | Detects and grades it |
| Cambridge Colour Test | A computer test: find the gap in a C drawn in colored dots among dots of random lightness. It measures thresholds along the protan, deutan and tritan axes. Regan et al. 1994 | Detects and grades it |
Status colors
Red, amber and green
A traffic-light status set leans entirely on hue. Give each state a word or an icon too, as WCAG 1.4.1 asks.
WCAG 1.4.1 Use of Color- Error
- Warning
- Success
Top: typical vision. Bottom: tritanopia. The closest two status colors end up ΔE2000 20.0 apart.
Your palette
Check your colors
Paste a palette, or start from one below. The bars show the closest pair under every type; the table shows every pair under the one you pick.
The checker needs JavaScript.
Hex codes, separated by commas or spaces, up to 12. Skipped:
Closest pair under each type
Too close Borderline Distinct
Design for it
Designing so it does not matter
- Never let color carry meaning alone: add a label, an icon, a pattern or a position.
- Separate colors by lightness as well as hue, so they stay apart in every simulation.
- Keep text contrast at 4.5:1 or more, and 3:1 for icons and the edges of controls.
- Check designs with a simulator before you ship them.
Simulate a whole page in the Color Blindness Simulator, or check a pair in the Contrast Checker.
Other types