Skip to content

Color blindness simulator

Preview how photos, charts, and UI screenshots appear with Protanopia, Deuteranopia, Tritanopia, and Achromatopsia locally.

Drop an image, chart, or paste a UI screenshot (Ctrl+V)

Runs Viénot/Brettel cone simulation matrices locally in a Web Worker.

Your files are open only in this tab. Nothing is uploaded.

How to use Color blindness simulator

  1. Load a screenshot or chart. Drop an image, press Ctrl+V to paste from your clipboard, click Choose image, or click Load sample UI chart on the Image & UI screenshot tab.
  2. Pick a deficiency mode and severity. Select Deuteranopia, Protanopia, Tritanopia, Achromatopsia, Deuteranomaly, Protanomaly, or Tritanomaly, then drag the Deficiency severity slider.
  3. Inspect differences side by side. Drag the Split comparison handle from 0% to 100% or switch to 4-panel grid to view all four primary deficiencies together.
  4. Download or check hex palettes. Click Download simulated PNG to save your active view, or switch to Palette checker to test up to eight status swatches at once.

Your files are processed on this device and never uploaded. They stay private, and there's nothing to delete afterwards.

How does linear-RGB matrix math simulate color blindness?

Applying CSS feColorMatrix filters directly to gamma-encoded pixels skews brightness because sRGB channels store nonlinear display values rather than physical light intensity. This Protanopia, Deuteranopia, and Tritanopia simulator and tester decodes pixels into linear light inside a Web Worker before multiplying cone-confusion matrices.

Gamma-decoded sRGB lookup tables and alpha preservation

Following the W3C CSS Color Module Level 4 Specification, a 256-entry Float32Array lookup table linearizes each 8-bit channel around the 0.04045 inflection point (c / 12.92 below, ((c + 0.055) / 1.055)^2.4 above). Linearized values then multiply against Viénot, Brettel, and Mollon (1999) 3×3 confusion matrices whose rows sum near 1.0 to preserve neutral grays and pure white.

For Achromatopsia, pixels project onto ITU-R BT.709 relative luminance (0.2126 * R + 0.7152 * G + 0.0722 * B) before encoding through a 4,096-step linear-to-sRGB lookup table. Alpha transparency remains byte-identical (out[i + 3] = rgba[i + 3]) so transparent UI overlays and icons remain crisp.

Blending anomalous trichromacy from 0% to 100% severity

Full dichromacy modes run at 100% (1.0) severity, whereas anomalous modes (Deuteranomaly, Protanomaly, Tritanomaly) default to 60% (0.6). Before looping over pixels, the engine blends the 3×3 confusion matrix with the identity matrix ((1 - alpha) * I + alpha * M) so moving Deficiency severity from 0% to 100% requires only one pass per frame.

Hue confusion frequently compresses luminance separation between adjacent foreground and background tokens too. Verify simulated text pairs with our WCAG contrast checker to confirm 4.5:1 readability across dark and light modes.

Eight color vision modes and population prevalence

Under the NIH National Eye Institute Color Vision Deficiency Guide, inherited red-green deficiencies sit on the X chromosome and affect roughly 8% of males and 0.5% of females. Blue-yellow and complete monochromacy variants involve non-sex chromosomes and appear equally across populations.

Mode in simulator Physiological cone state Default severity Approximate prevalence Visual confusion axis
Normal Trichromacy Intact L, M, and S cones 0% ~92% of population Full trichromatic discrimination
Deuteranomaly Shifted M cone (green-weak) 60% ~5% of males (most common) Muted greens appear brownish or red
Protanomaly Shifted L cone (red-weak) 60% ~1% of males Reds appear darker and closer to green
Deuteranopia Missing M cone (green) 100% ~1% of males Red, orange, and green collapse
Protanopia Missing L cone (red) 100% ~1% of males Deep reds darken near black; red-green merge
Tritanopia Missing S cone (blue) 100% ~0.01% of population Blue blends with green; yellow turns pink-gray
Tritanomaly Weak S cone (blue-weak) 60% ~0.01% of population Subtle blue-green and yellow-red shifts
Achromatopsia Rod / luminance vision only 100% ~0.003% of population Luminance contrast without hue

Exported whitepapers and compliance packs frequently embed screenshots of analytics graphics checked in this CVD simulator. When visual charts appear inside downloadable reports, check tagged PDF reports for document accessibility alongside image checks so screen readers receive complete figure descriptions.

Comparing UI screenshots in split view and 4-panel grid

Checking UI layouts requires testing realistic components rather than isolated swatches. The Image & UI screenshot tab lets you test dashboards, maps, and charts locally via drag-and-drop, Choose image, or direct clipboard pasting.

Pasting screenshots and running the built-in sample chart

Pressing Ctrl+V (or Cmd+V on macOS) pastes any clipboard image immediately without saving temporary files to disk. If you do not have a screenshot ready, click Load sample UI chart to generate a local 720 × 420 px bar chart featuring five status hues and ONLINE, DEGRADED, and CRITICAL badges.

  • Split comparison: Drag Split comparison handle from 0% to 100% (default 50%) to sweep between Original on the left and the active simulation on the right.
  • 4-panel grid: Switch to 4-panel grid to dispatch four concurrent Web Worker simulations for Deuteranopia, Protanopia, Tritanopia, and Achromatopsia at 100% severity.

Inspecting 1,200 px worker renders and exporting PNGs

Uploaded images scale proportionally so the longest edge caps at 1,200 px before extracting a 4-byte-per-pixel RGBA buffer defined by the MDN ImageData Web Canvas Reference. Transferring ownership of each copied ArrayBuffer into simulateCvd keeps main-thread scrolling responsive while multiplying millions of subpixels.

In Split comparison, click Download simulated PNG to export a lossless <filename>-<mode>.png artifact for design reviews or pull requests. If your color-checked graphics head toward commercial offset presses rather than screens, prepare print-ready CMYK artwork before commercial trimming with vector bleed and crop marks.

Testing brand and status palettes in Palette checker

Switch to Palette checker whenever you want to audit design-system status tokens before writing UI components or taking screenshots. Serving as an accessible palette simulator, this tab lets teams check UI color groups simultaneously across the four primary deficiencies on one screen.

Spotting red-green collisions across up to eight swatches

The table loads five default status swatches (#dc2626, #16a34a, #2563eb, #eab308, #9333ea) and lets you pick or enter any HEX value. Click Add swatch to expand up to 8 rows or Remove down to 2 rows, immediately spotting when green #16a34a and red #dc2626 collapse into nearly identical olive-brown tiles under Deuteranopia and Protanopia.

Each swatch row renders four live simulated color tiles paired with shifted hex codes for rapid side-by-side comparison. Teams can catch colliding positive, warning, and danger states early without exporting mockups or guessing raw RGB offsets.

Designing CVD-safe charts with non-color cues

Under W3C Understanding WCAG 2.2 Use of Color (Success Criterion 1.4.1), color cannot be the sole visual channel conveying state, required fields, or chart series. Pair hue shifts with direct data labels, distinct icons, or dashed line patterns, and adjust lightness separation using our OKLCH and HEX design token converter before locking design tokens.

Tip: Shifting success green slightly toward blue-teal (#0d9488) and warming error red toward vermilion (#ea580c) creates distinct luminance and blue-channel separation that survives Protanopia, Deuteranopia, and even Achromatopsia simultaneously.

Questions

How accurate is this simulation compared to simple CSS filters?

Every pixel is decoded from sRGB to linear light using a 256-entry lookup table, transformed with Viénot, Brettel, and Mollon cone-deficiency matrices (or BT.709 luminance for Achromatopsia), and re-encoded inside a local Web Worker.

Can I test status badges or chart palettes without uploading a screenshot?

Yes. Switch to Palette checker at any time to evaluate up to 8 custom hex swatches across Protanopia, Deuteranopia, Tritanopia, and Achromatopsia simultaneously, or click Load sample UI chart for instant testing.

Does my screenshot leave my computer?

No. Image decoding, linear-RGB matrix multiplication, 4-panel grid rendering, and lossless PNG export run strictly inside a dedicated browser Web Worker with zero network uploads or file telemetry.

What is the difference between deuteranopia and deuteranomaly?

Deuteranopia indicates complete absence of green-sensitive M cones, simulated at 100% severity. Deuteranomaly means M cones are present with a shifted spectral response, defaulting to 60% severity and affecting roughly 5% of males.

Why are uploaded images scaled to 1,200 px during simulation?

Capping the longest dimension at 1,200 px keeps 4-byte-per-pixel RGBA buffers compact so the 4-panel grid can render four parallel Web Worker simulations rapidly without exhausting browser tab memory.

How do I fix chart bars or status badges that look identical under red-green deficiency?

Pair color differences with text labels, icons, or hatch patterns, and shift green hues toward cooler teal while warming red hues toward orange-vermilion so luminance contrast separates adjacent bars.

Updated