Color Hue Test - Find the Square With a Different Shade
Color Hue Test
One square has a slightly different hue. Click it as fast as you can. Each find makes the grid bigger and the difference smaller.
Example result: level 9 · ΔE 2.0 — nine squares found, the last one 2.0 units different.
Ready when you are. The clock starts with the first grid.
The grid grows from 2×2 to 8×8 (8×8 from level 7). The difference shrinks by 20% per level: ΔE 12 on level 1, 0.42 on level 16 (OKLab ×100). Only the hue changes; lightness and chroma stay the same. 60 seconds in total; each wrong click costs 3 seconds.
Before you start
- Turn off night mode, blue light filters and automatic color adjustment (such as Night Light or True Tone): they shift every color on the screen.
- Look at the screen straight on, at normal brightness, in steady room light. The test colors are fixed: they stay the same in light and dark theme, but each screen shows them a little differently.
- This is a perception game, not a medical test, and it cannot detect color blindness. If you suspect a color vision problem, see an eye doctor. Screening plates: Color Blindness Test
How the color hue test works
This test measures how small a difference in hue you can spot on your screen. All squares share the same lightness and chroma; one is turned slightly around the color wheel. The difference is computed in the OKLab color space and shown as ΔE, so your result is a number you can compare from one try to the next.
How to take the test
- 1 Turn off night mode and blue light filters, sit in front of the screen and press Start. You have 60 seconds in total.
- 2 Click the square whose hue is different. The grid grows from 2×2 to 8×8 and the difference drops by 20% each level. A wrong click costs 3 seconds.
- 3 When time runs out, the result shows how many squares you found and the smallest difference you spotted, as ΔE and as a hue angle.
Why OKLab and ΔE?
Screen colors are stored as RGB numbers, but equal steps in RGB do not look equal to the eye. OKLab, published by Björn Ottosson in 2020 and now part of CSS, is designed so that distances match perceived differences more closely. ΔE here is the straight-line distance between two colors in OKLab, multiplied by 100. Because only the hue changes, ΔE = 2 × chroma × sin(hue angle / 2); with the chroma of 0.11 used here, ΔE 12 is a hue turn of about 66° and ΔE 2 about 10°.
What the screen adds to your score
Most screens show 256 levels per color channel, so the smallest differences cannot be drawn exactly. The test rounds both colors to whole RGB values and reports the difference that is really on screen, not the planned one. Beyond that, an uncalibrated or cheap panel, a wide viewing angle, low brightness, glare or a warm night filter all make hues harder to tell apart. Compare results on the same device and in the same light.
Why some rounds feel easier
Each round uses a random base hue. Human hue discrimination is not equal around the color wheel: classic experiments on wavelength discrimination found the finest steps near blue-green (around 490 nm) and yellow-orange (around 590 nm). OKLab evens this out only partly, so the same ΔE can be easy in one color family and hard in another. Playing several times gives a fairer picture than a single try.
Not a medical test
Clinics measure hue discrimination with standardized tests such as the Farnsworth-Munsell 100 Hue test, using physical color caps under controlled light. A screen game cannot replace that, and it does not diagnose color vision deficiency. Treat your score as a fun measure of you plus your screen. If colors look wrong to you in daily life, an eye doctor can test your color vision properly.