About HueSense
HueSense is a free test that measures your color discrimination by having you find the one square whose color is just slightly different in a grid. As you level up, the grid grows from 2×2 to 8×8 and the difference between the two colors keeps shrinking.
Color differences are not eyeballed. They are calculated with CIEDE2000 (ΔE00), the color difference formula of the International Commission on Illumination (CIE). Each ΔE is re-measured from the 8-bit sRGB values actually shown on screen, so your smallest ΔE00 detected is the true difference between the two colors you saw.
All calculations and records stay inside your browser and are never sent to a server.
This is not a medical diagnosis. It measures how well you tell similar colors apart on a screen and is not a substitute for a color vision deficiency screening.
Key features
- Find the odd color out — 30 levels. The grid runs from 2×2 up to 8×8, and the target ΔE00 starts at 20 and shrinks by 15% per level, down to 0.5.
- CIEDE2000 measurement — The sRGB → CIELAB conversion and ΔE00 calculation are implemented from scratch and verified to four decimal places against the 34 test pairs published by Sharma et al. (2005).
- Hue family analysis — Rounds are drawn evenly from six hue families (red, yellow, green, blue, purple and neutral/lightness), and your accuracy and minimum ΔE per family reveal where you are weakest.
- Challenge and Practice modes — Challenge has a time limit and lives. Practice is adaptive with no time limit: a miss drops you one level. In Practice you can also focus on a single hue family.
- Saved records — Your best score, your last 20 runs and your all-time accuracy per hue family are stored on this device only.
How to use
- Keep your screen at its usual brightness and turn off night mode or blue light filters if you can. Look at the screen straight on.
- Choose Challenge or Practice and press Start.
- Find the square with the different color and tap it. On a keyboard, move between squares with the arrow keys and press Enter to select.
- At the end of a challenge you will see your smallest ΔE00 detected, the level you reached and your accuracy by hue family. Use Copy result to share your score.
Scoring rules
- Challenge mode: 12 seconds per round. A correct answer moves you up a level. A wrong answer or a timeout costs you one life and gives you a new round at the same level. The run ends when you have used all 3 lives or cleared level 30.
- Practice mode: No time limit and no lives. A correct answer takes you up one level and a miss takes you down one, so you keep practicing right around your own limit. Press Finish to see your results.
- Smallest ΔE00 detected: The smallest actual ΔE00 among the rounds you got right. The lower the number, the finer the difference you were able to see.
- Voided rounds: If you switch tabs or leave the page mid-round, that round is not recorded and a new one at the same level is served instead. Input in the first 0.15 seconds after a new round appears is ignored as a carry-over tap from the previous round.
- Display limits: An 8-bit screen has only 256 steps per channel, so differences below ΔE00 0.5 are hard to produce accurately. That is why the target ΔE00 bottoms out at 0.5, and for colors with coarse steps such as grays the actual ΔE may differ slightly from the target (the actual value is what gets recorded).
Frequently Asked Questions
What is ΔE (Delta E)?
It is a color difference expressed as a single number that describes how different two colors look. 0 means identical, and the higher the number, the more different they appear. The 1976 formula (CIE76) was simply the straight-line distance in CIELAB space; CIE94 and CIEDE2000 later corrected it to match human perception more closely.
Why use CIEDE2000?
The same CIE76 value can feel very different to people depending on the color, especially for highly saturated colors and in the blue region. CIEDE2000 weights lightness, chroma and hue separately and adds a rotation term for blues, so the same number corresponds to a similar perceived difference across hue families. That makes it well suited to comparing accuracy between hue families.
Can this tell me if I am color blind?
No. This test is not a medical diagnosis and does not replace a color vision deficiency screening. Color vision deficiency has to be confirmed by an eye care professional using standard tests such as Ishihara plates or an anomaloscope. A low score in one hue family does not mean you have a color vision deficiency.
I get different results on different devices.
That is normal. Every screen differs in color gamut, gamma, brightness and calibration, and features like night mode or True Tone change the white point itself. Ambient lighting and viewing angle matter too. Compare scores only from the same device, the same settings and similar lighting.
Why don't you show an average score or a top X% ranking?
There are no validated age- or sex-based norms for screen tests like this, and device differences are so large that ranking different people's scores against each other says little. Rather than invent percentiles with no basis, HueSense shows your own progress over time and your accuracy by hue family. The ΔE size scale is offered only as a rough industry rule of thumb.
How is the smallest ΔE00 detected calculated?
When a round is generated, colors are chosen to hit the target ΔE00, then rounded to the 8-bit sRGB values that will be displayed, and the ΔE00 between those two colors is recalculated. Your result is the smallest of these actual values across the rounds you answered correctly.
Why is the area around the test gray?
Surrounding colors change how the colors in the middle look (simultaneous contrast). Just as standard color evaluation setups use a neutral gray surround, the area around the grid is fixed to neutral gray so you see every round under the same conditions, whether you use the light or dark theme.
Where are my results stored?
Only in this browser's localStorage. Nothing is sent to a server, and you can delete it at any time with the Clear history button on the results screen or through your browser settings.
Standards and references
- CIE 15:2004 Colorimetry, 3rd ed. — definition of CIELAB (1976)
- CIE 142-2001 Improvement to industrial colour-difference evaluation · ISO/CIE 11664-6:2014 — CIEDE2000
- Luo, Cui, Rigg (2001), The development of the CIE 2000 colour-difference formula: CIEDE2000, Color Research & Application 26(5)
- Sharma, Wu, Dalal (2005), The CIEDE2000 color-difference formula: Implementation notes, supplementary test data, and mathematical observations, Color Research & Application 30(1)
- IEC 61966-2-1:1999 — sRGB color space (D65)