Deuteranopia Accessible Palette Generator

Accessible palettes for deuteranopia: what actually works

Deuteranopia is the most common form of color blindness — caused by missing or non-functional M (medium-wavelength, green) cones. It affects about 5% of males and 0.4% of females (X-linked inheritance). For deuteranopes, certain color pairs collapse into one perceived color: red ↔ green, brown ↔ olive, light blue ↔ pink, and dark blue ↔ purple. A palette is "deuteranopia-accessible" when no two semantically distinct colors collapse into the same perceived color under deuteranope simulation.

Battle-tested palettes researchers actually use

Several palettes have been engineered and validated for color-blindness accessibility. Pick from these instead of designing from scratch:

• Wong palette — Bang Wong's 2011 Nature Methods paper proposed 8 colors safe across the three main color-blindness types: black, orange, sky blue, bluish green, yellow, blue, vermillion, reddish purple. Still the gold standard for scientific figures.
• Okabe–Ito palette — 8-color set used widely in Nature Methods; nearly identical to Wong and often credited together.
• Viridis — perceptually uniform colormap, matplotlib's default since 2017. Safe for all forms of color blindness, including monochromacy. Available as a sequential gradient for continuous data.
• Cividis — viridis variant optimized specifically for deuteranopia and protanopia. Use it when you know your audience leans toward red-green deficiency.
• ColorBrewer 2 — Cynthia Brewer's set (used in cartography for decades), with sequential, diverging, and qualitative palettes all tagged for color-blind safety. colorbrewer2.org has a filter for "colorblind safe".
• IBM Design accessible palette — IBM's corporate color system ships with documented accessible variants.

Design rules that go beyond palette choice

A safe palette is necessary but not sufficient. Three habits multiply accessibility:

• Vary luminance, not just hue — two colors with the same lightness can become indistinguishable under deuteranopia. Spread your palette across the lightness axis as much as across hue.
• Add redundant encoding — shape, pattern, position, or direct labels. A scatterplot with circles, triangles, and squares stays readable even fully desaturated.
• Avoid pure red + green together — the classic confusing pair. Hollywood-style teal + orange grading is incidentally color-blind safe; that is part of why it became dominant in cinema.

Tools to test under simulated deuteranopia

Don't ship without testing under simulation:

• Chrome DevTools — Rendering panel → "Emulate vision deficiencies" → Deuteranopia / Protanopia / Tritanopia / Achromatopsia. Built in, zero install.
• Stark — Figma plugin with real-time CVD simulation and contrast checking.
• Sim Daltonism — free macOS app with live screen simulation.
• Color Oracle — free cross-platform (Mac, Windows, Linux) full-screen filter.
• Coblis — web-based image simulator from color-blindness.com.

Standards and legal context

WCAG 2.1 — Success Criterion 1.4.1 "Use of Color" requires that color is never the only means of conveying information. Under Brazil's Lei Brasileira de Inclusão (LBI, Law 13.146/2015), accessible information design is a legal requirement for public-facing services. Newsrooms have caught up: Folha de São Paulo and O Globo's data desks have published color-blind-safe charts since 2020, and Tableau and Power BI both ship with built-in color-blind palettes by default.

FAQ

Can I use vibrant colors at all? Yes — viridis and cividis are vibrant and gradient-friendly and remain safe under simulation. Vibrancy is not the enemy; uncontrolled hue clashes are.

Do I need to test with a real person with deuteranopia? Simulators (Stark, Color Oracle, Chrome DevTools) catch 90% of issues. For high-stakes public-facing work, recruit color-blind reviewers — confusions that look subtle in simulation are jarring in lived experience.

Is color enough on its own? No — WCAG 1.4.1 explicitly forbids relying on color alone. Always add shape, pattern, position, or label as a redundant channel.

What about protanopia and tritanopia? Protanopia (missing L cones, ~1% of males) has very similar confusions to deuteranopia, so the same palettes generally work. Tritanopia (missing S cones, <0.01%) is rare and confuses blue/yellow rather than red/green — viridis and Wong stay safe.