Artists and art teachers are famous for giving enigmatic advice, artsy phrases that may or may not have any bearing on reality. One such gem is “Paint your colors,” which is when your professor insists you paint your red ball by using colors like green. It seems asinine, but painting your colors is actually an essential part of realistic art. Artists have been striving for realism or centuries, and many of the techniques they developed have their basis in math and neuroscience. Perspective, for example, took hundreds of years to develop properly. But now that we understand it, the geometric framework is obvious. Painting your colors is also based in science, though this science is a little more complicated. Instead of geometry, we involve the physics of light and neuroscience.
The physics of the situation say: light that hits your eye is almost never pure. Light is made up of photons, each of which is a certain color. Every time a photon interacts with matter, it could change its color. Quantum physics determines the exact rules of color-changing, but basically, by the time the light has actually gotten to your eye, it has passed through, bounced off, been absorbed and re-emitted so many times by so many objects that it the resulting spectrum is mixed.
The technical term for this is color complexity. Your eyes and brain are used to the patchwork of information that comes at them. Since color complexity and color noise are ubiquitous, the brain has layers upon layers of processing that takes the jumble of color information and assembles it into useful stuff. The process of filtering and simplifying starts at the retina, even before the information travels to the optic nerve!
The visual systems in the brain are built to allow you to function, to identify threats, food, friends, and so on. Much of the raw data is useless, in terms of survival; you would drown in excess information if your brain didn’t selectively throw most of it away before it gets to your conscious mind. The conscious mind gets the Cliff's Notes, not the whole book.
From right to left: simplest red ball; with shading; with shading detail; with hue variation. Click to enlarge for a better look. Copyright 2014 Lauren Ellenberg.
Okay: so, what difference does color complexity make-- why should artists bother with it? Take a look at the image above. The balls range from least to most complex, and not coincidentally from least to most realistic. Adding color complexity by adding detail, noise, and hue variation greatly increases the realism. Why? Because this mimics the appearance of a real ball. It has the natural color variation of a real ball with all kinds of light bouncing around.
Your brain is used to "chewing" on the visual data to extract the essential information. When the data doesn't have enough variation, the brain is not fooled into thinking the art is real. 3D computer art tends to run into this problem. The edges and colors are too neat and clean; it's too perfect. Texture artists actually have to draw on dirt and scratches because the brain knows to look for them! This problem is hard to solve for computer art, and has led to both the development of ray-tracing and brown video games. Cartoons are very selective about their color complexity, but are overall very simple. The brain hardly has to "chew" on them at all, which makes them very quick and easy to decipher.
As a painter, when you choose the paint to put on the canvas, you have to be able to see that the ball is actually a wild patchwork of colors that averages out to “red.” Artists learn to break down those filters to see the original colors, so that we can paint them better. So when your art teacher tells you to put purples and browns in your shadows, listen, and more importantly: look. There's a whole spectrum of colors living there, waiting to be seen.