Colorblind-Friendly Palette Generator

Its colors stay distinguishable for people with color vision deficiency, mostly the red-green types that affect about 8% of men of Northern European ancestry. That usually means varying lightness as well as hue, and avoiding red-green pairs of similar lightness. Even with a good palette, add labels, line styles or marker shapes so that color is not the only cue. (British readers search for a colour blind friendly palette; it is the same thing.)

For up to eight categories, Okabe–Ito or one of Paul Tol's qualitative schemes. For ordered values such as concentration or expression, a sequential scale such as viridis or cividis. For values above and below a meaningful midpoint, such as fold change, a diverging scale such as RdBu or Tol sunset.

Orange #E69F00, sky blue #56B4E9, bluish green #009E73, yellow #F0E442, blue #0072B2, vermillion #D55E00, reddish purple #CC79A7 and black #000000, proposed by Masataka Okabe and Kei Ito in 2008 and popularized in Nature Methods by Bang Wong.

Each color is converted with the Machado, Oliveira and Fernandes (2009) model for protanopia, deuteranopia and tritanopia at full severity, the strongest form. People with the milder anomalous forms see smaller shifts. The check then measures the CIEDE2000 color difference between every pair; pairs below 10 under red-green simulation are flagged. To simulate a whole image, use an image simulator; this page checks palettes.

Partly. Accessibility guidelines (WCAG 2.2) also ask for contrast: 3:1 for lines, bars and other graphics against the background, and 4.5:1 for text. The table marks colors below 3:1 against the chosen background, such as yellow on white, so you can give them an outline or use them only for fills.

Rainbow maps change lightness unevenly, which creates false boundaries, and their red and green ends are confused under red-green deficiency (Crameri et al., 2020). viridis and cividis increase lightness steadily, so they read correctly in color, in grayscale and with color vision deficiency.

Avoid red and green of similar lightness as the only difference between groups. If you need both, make one clearly darker, or use a red–blue or orange–blue pair instead. The check flags red-green pairs that collapse under deuteranopia or protanopia.

Use Code ▾ to copy an R vector or a ggplot2 scale_color_manual() call, a Python list or a matplotlib color cycle, or RGB values to type into Excel or PowerPoint’s custom color dialog.

Yes. No account is needed, and colors are processed only in this browser. Custom colors are included in the share link, so only share it if you are happy for others to see them.