Why colour judgement matters more than colour theory
Every designer learns colour theory. Complementary pairs, analogous schemes, the 60-30-10 rule. Very few ever train the underlying perceptual skill that theory sits on top of: the ability to look at two colours and say, accurately, how far apart they are.
That skill is what fails in the moments that cost money. The brand red that drifts across three touchpoints because nobody caught it. The hover state that reads as a rendering bug rather than a deliberate change. The print proof that looks fine until it sits next to last year's.
What the professional tests actually measure
The Farnsworth-Munsell 100 Hue Test
The reference standard for colour discrimination since 1943. You arrange 85 colour caps in order of hue across four trays, and the score reflects how many you place out of sequence. It is used in industries where colour accuracy is a job requirement — printing, textiles, paint, quality control.
What it measures is not whether you can see colour, but how finely you can distinguish neighbouring hues. Most people with normal colour vision still make errors, and the errors cluster: nearly everyone is worse in the blue-green region than in the reds.
Ishihara plates
Often confused with the above. Ishihara plates screen for colour vision deficiency — the red-green confusion that affects roughly 8% of men and 0.5% of women of northern European descent. It is a pass/fail medical screen, not a measure of skill, and no amount of practice changes the result.
Delta E — the number behind "close enough"
ΔE quantifies the distance between two colours in a perceptually uniform space. The rough scale worth memorising: ΔE below 1 is invisible to the human eye; 1–2 is visible only to a trained observer under controlled conditions; 2–10 is noticeable; above 10 the colours read as different.
Print houses commonly work to a ΔE tolerance of 2 or better. If you have never calibrated your own eye against that scale, you are approving work against a standard you cannot personally verify.
Why your eye is probably worse than you think
Three reasons, none of them about talent.
You almost always judge colour in context. Simultaneous contrast means the same hue looks different against different backgrounds — the effect behind every optical illusion your feed has ever shown you. Designers spend their day judging colours embedded in layouts, which is exactly the condition where judgement is least reliable.
Your screen lies, consistently. An uncalibrated monitor, a phone with adaptive brightness, a room with warm evening light. You adapt to your own environment within minutes and stop noticing its bias.
You rarely get feedback. Skills improve with immediate, accurate feedback on error. Most colour decisions never produce any — the work ships, nobody complains, and you learn nothing about how far off you were.
How to actually practise
The feedback problem is the tractable one, and it is why a scored matching game is more useful than it sounds. You can try it right now in this browser — the colour perception test runs five rounds and scores you with real CIEDE2000 ΔE, no install. Chroma Quest is the same idea on Android, with a camera mode and multiplayer, and it tells you your precision as a percentage. That last part is the whole point — it closes the loop that professional practice leaves open.
Three habits worth building around it:
- Fix your conditions. Same screen, same brightness, same time of day. Otherwise you are measuring your environment rather than your eye.
- Use Camera Mode on real materials. Photograph fabric, packaging, a painted wall. Reproducing a colour pulled from the physical world is harder and closer to the job than matching a generated swatch.
- Track over weeks. Colour discrimination improves slowly. A single session tells you about your sleep and your screen, not your skill.
Where the honest limit is
Practising colour matching will make you better at colour matching. Whether that transfers cleanly to approving a brand palette is not something anyone has demonstrated rigorously, and you should be sceptical of any product that promises it — including this one.
What it does reliably provide is calibration: a repeatable, scored check on a judgement you otherwise make blind. That is worth something on its own, and it costs two minutes.