You know the drill from school. Red, yellow, and blue are the primaries. Mix two and you get a secondary. Simple. Sorted. Move on to maths.

Except - have you ever actually tried to mix a specific colour from scratch? Not "make something greenish", but hit a precise shade of sage, or a warm terracotta, or that particular dusty mauve you're imagining? It's humbling. Genuinely. You end up with a jar of something that looks like pond water and no clear idea where you went wrong.

The Gap Between Knowing and Doing

Colour theory, as most of us absorbed it, is basically a wheel and three rules. But the actual practice of mixing paint involves a load of variables that the classroom version quietly ignores.

Proportion, for a start. Red and yellow make orange - yes, fine - but which orange? Two parts red to one part yellow gives you a burnt, brick-leaning orange. Flip that ratio and you're suddenly into mango territory. The same two colours, completely different results depending on how much of each you use.

Then there's the shade/tint/tone business, which I think a lot of people genuinely don't have a firm grip on (I certainly didn't until fairly recently). A tint is a colour mixed with white to make it lighter. A shade is the same colour mixed with black to make it darker. A tone is mixed with grey - which mutes and desaturates it without necessarily making it lighter or darker. Three completely different operations that casual usage tends to smoosh together under "make it lighter" or "make it darker."

Adding black to yellow doesn't give you dark yellow. It gives you a sickly olive. Painters know this. The rest of us find out the hard way.

Why Proportional Mixing Breaks People's Brains

Here's the bit that really gets people. Imagine you're trying to mix a specific warm purple - say, something close to dried lavender. You need red and blue, but not equal amounts, and you might need to nudge the tint upward with a touch of white.

The problem is that colour mixing isn't linear in the way we intuitively expect. A tiny addition of one colour can swing the result dramatically. Add a fraction too much blue and your lavender becomes something cooler and flatter. Add the white too early and you've changed the base you're working from. The sequence matters as much as the quantities.

This is partly why professional paint matching exists as a whole industry. Humans are not naturally good at eyeballing proportions in a medium where the feedback is non-reversible. You can't un-mix paint. (Typically British approach: muddle through, make a mess, then act like it was intentional.)

The Puzzle Hiding Inside Every Colour

What makes this interesting - beyond the practical frustration - is that colour mixing is actually a routing and proportion problem dressed up as an art problem. To hit a target colour, you need to:

  • Know what primaries you're working from
  • Understand how they'll interact at different ratios
  • Account for any modifiers like white, black, or grey
  • Get the sequence right
  • And then actually execute it without overshooting

That's not a creative task. That's a logic puzzle. Which is probably why it appeals to a certain kind of brain that also enjoys optimisation problems - factory layouts, supply chains, that sort of thing...

It's also why I ended up building Kalabux, a puzzle game that puts colour mixing at the centre of a steampunk paint-works. You pipe red, yellow, and blue out of tanks through a hex grid of pipework, and where streams meet, they mix. Customers order specific colours - a particular shade of orange, a muted purple, a warm grey - and you have to route and proportion your pipes to hit the target exactly. The full tint/shade/tone system is in there too, once you research it. It's in testing right now, coming to Android and iPhone soon, and you can find out more at kalabux.com if you're curious.

The reason it works as a puzzle, I think, is that colour mixing has genuine rules - it's not arbitrary - but those rules are just non-obvious enough that working them out feels satisfying rather than mechanical.

Brown: The Colour Nobody Talks About

One thing colour theory classes tend to gloss over is how you actually make brown. The answer - mix all three primaries - feels like it should produce something dramatic and terrible. And it does, a bit. But the ratio determines whether you get a warm chestnut, a flat muddy brown, or something approaching a dark olive.

Brown is essentially what happens when you combine complementary colours or push all three primaries together. It's not a failure state. It's a legitimate outcome with its own range of variation. (Though I appreciate this is cold comfort when you were going for teal.)

Understanding this changes how you think about colour mistakes. A muddy result isn't random - it's the predictable outcome of certain combinations. Once you know the rules well enough to predict the mud, you're actually getting somewhere.

It's More Learnable Than It Feels

The good news is that proportional colour mixing is a skill, not a talent. People who are good at it got good at it through repetition and feedback, not through some innate gift.

The fastest way to learn it, apparently, is to work with a constrained palette - just the three primaries plus white and black - and try to hit specific targets. You build an intuition for what each adjustment does, and you start to see colour as a system rather than a mystery.

Whether you do that with actual paint, or through something more forgiving where you can experiment without wasting a tube of cadmium red - well, that's up to you.