The human eye has two main types of cells that detect light for vision.
Rod cells simply detect light in general. But Cone cells only detect light
of specific frequencies/wavelengths and come in one of three varieties.
The frequencies that Cones pick up are roughly at red, green, and blue.
A lot cameras work in a similar way, by using sensors or photosensitive materials that detect these different frequencies.
Intermediate frequencies get detected by the combination of different Cones.
For example, if both the red and green cones are getting triggered, it might be interpretted as yellow, a frequency between red and green.
Cones need more intense light to work than Rods. This is why you can still make out shapes in dim light, but not colors.
If you've read the Dimension chapter already, you can kind of think of all colors existing in a 3D space, with red, green, and blue as axes that span them. This is the RGB color model. It's useful, but not always convenient to think of colors in terms of it's red, green, and blue components. There are other color models for navigating color space. You may or may not have heard of some of them. The Munsell color system, HSV, or HSL. They're some what similar. The latter two represent colors with an accuracy that as painters, we might not even need. I will be describing a system I use that generalizes different systems.
Value, lightness, brightness, intensity, luma, luminance. I'll use these terms interchangeably, but know that they're all subtly different.
The basic idea is how light or dark a color is. Black has low value, and white has high value. Value is the most important part of color.
Maybe it's because the rod cells can work when cone cells can't, or maybe the order the brain interprets visual information is related.
For whatever reason, value has the more impact in interpretting an image than the other two parts of color.
Often, a painter will do a value study for a painting before the actual piece. There are both traditional and digital workflows that involve doing the values of the piece before adding the color in on top. This way the values will work better from the start. You don't need to work this way, but considering value is still important.
Saturation and Chroma refer to how colorful or vibrant a color is. A very pure red is a saturated, high chroma color. Dull grey is desaturated and lacks chroma. The minor difference between saturation and chroma is that if you move that pure red closer to black or white, it can keep its saturation but will lose chroma.
The hue of a color is the color in its pure, highest chroma form. A bright orange and a neutral brown can share the same hue, but look very different because the value and saturation are different. Hue is the most flexible aspect of color. If the values are correct, you can do almost anything to the hue and the painting will stay mostly coherent.
Let's talk about Color Temperature. It's useful to separate hues into warm colors and cool colors. The colors with long wavelengths like red, orange, and yellow are warm. Higher frequencies of light like green, blue, and violet are cold colors. Saturation also affects color temperature. Value does too, but not by much. A desaturated orange is colder than a saturated orange. Dark values tend to be colder, and high values warmer, but this is easily cancelled out by hue and saturation.
Color is extremely flexible. What we see and what we perceive can vary quite a bit when it comes to color.
The brain naturally interprets colors relative to the colors around it and adjusts sensitivity.
A color looks warmer when surrounded by something colder than it. Or if they're more interspersed with each other, the area gets more neutral.
The same thing happens with value and saturation.
Using this, we can limit what parts of color space we use to a smaller gamut in order to evoke a certain mood or environment, while still giving the impression of colors outside the gamut.
Traditionally, color is taught with primary colors of red, yellow, and blue.
Combinations of two primaries make the secondary colors orange, green, and violet.
A primary and corresponding secondary that doesn't contain said primary are complementary.
Tinting with white, shading with black, and mixing complementary colors then allows the painter to control value and saturation.
Modern printing uses CMYK, representing cyan, magenta, yellow, and black on white paper. The white and black part works the same way, but the primaries are slightly different. They produce the secondary colors of red, green, and blue. Hopefully these sound familiar.
Using CMYK, magenta and yellow create red. Yellow and cyan create green. But earlier I mentioned red and green combining to make yellow.
This is because CMYK is a subtractive color model, while RGB is additive.
The subtractive model starts with white and produces a certain color using combinations of primaries to absorb unwanted frequencies of light. The additive model starts with black, the absence of light, and produces a certain color using combinations of primaries to add desired frequencies. Pigments are subtractive. Light itself blends additively.
When light hits an object, there are three basic things that can happen. Some of the light might reflect off of the object, some might be absorbed, and some might pass through the object. Everything that light interacts with will experience these things. The color of the sky comes from light reflecting off the molecules in the air. Similarly, clear water turns dark as it gets deeper due to absorbtion.
When light hits an object or surface, some of it can bounce off. On a smooth, reflective material, the light reflects like off a mirror.
This is called specular reflection; the angle of reflection will be equal to the angle of incidence.
The places where the light of the light sources reflect can reach the eye/lens create specular highlights, or we can just call them spec.
Light doesn't always bounce cleanly though. It might get scattered, or diffused, by a rough surface or refractive properties of the material. The light bounces off in many directions. A spec can still appear, but it will appear fuzzy and dimmer since power of the light gets split and spread out.
Specular highlights are the brightest parts visually. Since they're the reflections of the light sources themselves, there's less light lost from absorbtion and scattering between the the source and the viewer.
As a surface tilts away from an observer, more of the light reflecting off of it is able to reach the eye/lens.
This makes objects appear more reflective at its outer edges and sometimes when further away. This is known as the Fresnel Effect.
This is a good way to get materials to look more believable. Perfectly matte surfaces are rare, even very dull materials can have weak, diffused speculars
and will experience the Fresnel Effect.
This is also useful for composition/readability as a way to control the edge on an object.
When light hits a material, some of it can get absorbed. Different wavelengths and amounts are absorbed more than others depending on the material. The color of an object that we can see is a result of whatever wavelengths weren't absorbed and reach the eye.
Light experiences something called falloff. It gets weaker over distance from being absorbed or scattered away by what it travels through. The decrease is logarithmic. If the light source is already far away, it isn't as noticeable.
If light doesn't bounce away, or get absorbed, it will just pass through. Like reflection, diffusion/scattering can occur.
Let's talk about subsurface-scattering, or SSS. This is when light transmits into a material, some absorbtion and scattering happens, then the light comes back out. The result is brightness and saturation bleeding past shadow terminators, or high chroma glows where the light is coming out from. It's usually most apparent in translucent materials and thin materials. A lot of materials actually do this if you get them thin enough. This phenomenon is really useful for depicting organic and translucent materials.
I recommend that you have a good grasp of form and perspective for this section. You can learn about it in the Dimension chapter.
Let's render a simple scene: a glossy red ball sitting on a dull, flat, white box, lit from above and front by a simple white light. There is also a weak ambient white light diffused through the air. Imagine rays of light from our light source coming in and hitting our objects. Some of it bounces to our eye, some is absorbed or bounces away, and some of it bounces to the other object and the interactions happen again. But there are parts of the objects that the rays couldn't reach in the first place. On the ball we can see the shadow terminator, an edge between where light could hit and where the ball's surface is in shadow. The surface of the ball is perpendicular to the direction of the light where the terminator appears. The ball makes a shape of shadow on the box behind the ball, the cast shadow. The shadows aren't totally black because the ambient light fills some of it in. There's a small bit under the ball that is very dark though. This is an ambient occlusion shadow where not even the ambient light can reach because of how the ball blocks it. The white of the box and some of the background is visible in the Fresnel reflections on the ball.