posted 1 month ago
So, long ago this thread opened my eyes to cmy(k) as a primary triad. Amazing! And I still totally love geeking out about it.
When I started learning painting, I discovered that CMY is extremely impractical and takes way more effort to get what I want.
To start with, paint is made with pigments. Each pigment brings personality to the paint, more than hue (colourness), saturation (muchness of colour), and value (lightness or darkness of a colour). The pigments influence transparency, light reflection indexy thingy, consistency, tinting strength, dry time, etc.
This may not have such an effect with modern media like acrylic and markers, but with the ones I work with (oils, watercolour, gouach, etc), pigments participate.
What's more, when a paint has more than one pigment (the gritty bits of rocks and stuff that make paint colourful), these pigments interact and, it means more to think about when mixing two paints together.
The key here, paint is more than colour. It exists in the physical world with physical properties that influences how the colour works.
The cyan I am currently working with has only three pigments, blue, white, and brown. ( PB15, PW4, PBR7). Most have five or more. All have white as there isn't really a pigment light (value) enough to be a cyan on its own.
And white in paint makes things tricky. White has an overpowering opaqueness and risks making any mixes chalky.
A pure magenta is also tricky to get. So any painting with a cmy colour palette already had limitations that require quite a bit of skill to overcome. Even then, the colour gamut one can get from cmy paint will not reach the maximum colour range promised by something like digital art.
We can also see these limits with professional art printing. Even the best cmyk printers in the world cannot get the desired range of colours and are specifically bad at matching reds.
But cmy is still fun to play with.
A much much bigger objection (other than the one where it doesn't behave as promised above), is that I dont like bright, modern colours in my work. To use cmy, i have to first mix a bright colour, then mix the complimentary to dull it down. Only it's not that easy.
I would much rather start with colours that are already limited and mix from there. This limitation is called Gamut mapping and is a realy cool thing to deep dive into.
Modern pigments can make a pretty good primary red, yellow, and blue, and as we can see from the pro-cmy camp, ryb colours have their limits in what colours they can mix (in theory they can't mix cyan, but that's not entirely true as we saw above it takes blue plus white and a touch of brown. They can't mix true cyan, but I have yet to see a paint that is true cyan, anyway). And I love the limits.
RYB limited palette is usually made with single pigment paints, and has a lot of choice on which pigments for each colour. The behaviour is easy to predict or adjust and gets rid of the problems of cmy paints. I don't have to mix so much to get the colour I want as I'm already mostly there. I can add white or not depending on what my aim is, so I don't have to battle milky shadows. I just love it and ryb is a much better match for my brain.
The colours we use dramatically influence the feeling of a piece. So I'll often choose red, yellows, and blues that are even more limited than the pure ryb primary colours. This puts physical limits on my colour gamut map that I cannot break. It keeps me within the mood I seek and I love it! If I start with a limited palette that can mix too wide a range of clours, like cmy, I will usually use them all. Because I can.
The limits of RYB are an advantage when painting.
So, colour theory. Yep. Both cmy and ryb are wonderful theories and even work well in a digital art. But when it starts interacting with the physical world, things get more complicated.
But I think, if one is starting out and wants the most colours for the fewest paints, a mixture of ryb and cmk and yellow ochre gives a great starting place.