This Cyber Monday Tuts+ courses will be reduced to just $3 (usually $15). Don't miss out.
We tend to see color as an attribute of every material thing, and light as a factor that can change it. Tomato is red, grass is green, and light can only add a tint or shade to it, right...? Wrong.
Color doesn't exist universally - it's the effect of our vision mechanism, fueled by light. No light, no color, and you can notice this easily when it's dark. It's not that darkness "covers" the colors - it's light what creates them! If it sounds revolutionary to you, keep on reading - there's no more important thing to understand for an artist. Also, make sure to read the first article of the series before trying this one - it's a great introduction to shading.
What is Color?
Let's take a little physics revision. Don't worry, I'll make it as simple as possible! Some objects are able to emit radiation, what that means is they throw a bunch of particles (or waves) in various directions. Light is a kind of radiation, and every light source emits photons.
Photons are waves combined of various wavelengths (here x, y, x).
We're going to call the way the photons fly between the light source and a particular direction a ray.
Those were a couple of facts. But what happens when a human factor comes in? There's a lot of radiation everywhere around us, but our eyes are specialized to react to only a particular range of wavelengths. For example, we don't see heat until its wavelength comes into that range (red-hot metal suddenly becomes a light source). This part of electromagnetic radiation we can see is called visible light, and is commonly known as just light.
We've discussed it shortly in the first article of this series, but let's add a bit of detail now. There are two kinds of photoreceptor cells in our eyes: cones and rods. When a ray hits them, they react and transfer some information to the brain.
Rods are very light-sensitive and are responsible for night vision, seeing movement and forms. Cones, on the other hand, are much more interesting for us. They are able to separate the wave into particular wavelengths, that the brain interprets (roughly) as red (long), green (medium) and blue (short). Depending on what wavelengths the ray consists of, we perceive a color mixed of these three.
But where do various wavelengths come from, if they are all brought by the same light source? Most of rays hit some object on their way, and then they're being reflected somewhere else (for example, to your eye). Usually the object they hit doesn't reflect them perfectly like a mirror. Some of the wavelengths are being absorbed by the object and they never reach your eye. As a result, we receive only a part of the original ray from that object. These remains of the ray are then interpreted by your brain as the color of the object. Different colors come from different absorbing and reflecting properties of materials.
You probably wonder what it all has to do with color in painting. After all, we only paint with colors, we don't create them physically! I'm sure everything will become clear in a second.
Hue, Saturation, Brightness
Is there something more confusing than this? Our intuition tells us what hue, saturation and brightness is, but when it comes to painting, it's hard to guess how to use it. Hue is, well, color, right? Saturation is a level of vividness... and brightness tells us if something is dark or bright. But it only makes sense as long as you talk about a finished painting, and it's much harder to guess where to put it all when you do it yourself. However, all we need is to understand where all these values actually come from!
The Definition of Hue
Hue is a "type" of color. Red, purple, olive, crimson are all hues. They're based on the mechanism we've just talked over - the reflected wavelengths, mixed in various proportions, create a final color interpreted by brain. Therefore, putting it simply, hue is based on "the color of the object". An interesting fact: silver, gold or brown aren't hues. Silver is shiny gray, gold is shiny yellow, and brown is dark or unsaturated orange.
No matter how many names we invent for the hues, all of them base on red, green and blue. The further on the color wheel you are from any of them, the more "original" color you'll get. For example, 50% red + 50% green gives yellow, but change this proportion just a little bit and you'll see a greenish or reddish tint.
There's no greater or lesser hue, being put on a wheel they're all equal. Hence we describe them by degrees instead of a percent value.
The Definition of Saturation
Hue doesn't mean color (at least not formally). All the circles below have the same hue, the same exact position on the color wheel (the same brightness too!). So why do we perceive them as different colors?
The common definition of saturation is how much white there is in the color. But wait, wasn't that about brightness? You want brighter color, you make it whiter... But that would make darker areas more saturated. It's so confusing, isn't it? That's why we need more explanation.
Saturation is the dominance of color. The three samples below have the same brightness and hue. The only thing that changes is the proportion between the components. We're not "adding white" - we're reducing the distance between the components, so none of them stands out.
As you can guess, when there's no difference between the components, we've got no saturation, which gives us white (we don't include brightness yet).
The Definition of Brightness
For our needs we can treat brightness as synonymous with value from the previous article. It defines the maximum of a value our eyes can perceive. There's no more blue than 100% blue, just like there's nothing brighter than 100% white.
The bars can't be filled over the maximum:
And, obviously, black comes from the lack of information.
An interesting fact: when it's dark, our cone cells get a little information, what makes us a bit color-blind. At this time rod cells, sensible to any light, will take over. However, since they're the most sensible to green-blue light, they'll make any green-blue object look brighter. It's called the Purkinje effect.
Despite having a certain, absolute brightness, every color has another property, luminance. While brightness tells us how much of color there is in the color, some of hues appear brighter to us - even when they're all 100% bright. Luminance is about how bright color is relative to white.
we turn 100% bright primary colors to grayscale, their brightness
suddenly drops. They still make white, but blue turns out to be very,
very dark, and green the brightest of them all. It comes from individual
sensitiveness of every cone, and that's why we perceive yellow (bright
red + very bright green) as the brightest of colors, or why cyan (dark
blue + very bright green) is sometimes called light blue. Luminance is important when you start your picture in grayscale - for example, yellow needs a brighter base than other colors of the same absolute brightness.
It's still a bit confusing, though. In reality we don't build the colors carefully, it would take too long! Fortunately, hue, saturation and brightness can be combined into a very useful tool. Take a look at the scheme below - you can notice there's a clear relation between colors. Why not use it?
If you're a digital painter, these should look familiar to you. It's a way of combining hue, saturation and brightness into one, consistent model called HSB. How does it work?
Once you've known what hue, saturation and brightness are, it's easy to locate them on the model. Hue wheel (or a bar, it doesn't matter) is independent and superior to SB square/triangle. Every hue possesses a range of saturation and brightness, and these two values are bound to each other. Together they define "richness" or "colorfulness" of a particular hue.
SB model can be divided into areas of different properties. If you learn to optically choose a proper color, you won't need to know anything about certain values of saturation or brightness - it's very helpful for spontaneous, fast painting.
While the square is much more intuitive, I personally prefer the triangle. It lets me control "richness" at a whole, not severally saturation and brightness (I've got separate sliders for that!). If you're like me and feel Photoshop could use a nice color wheel being opened all the time, check out this amazing, free plugin by Len White.
CMY and RGB
But what about traditional painters? They don't have a handy color wheel with neat sliders. How can you change a hue, saturation or brightness of a pigment?
First, let's think what's the difference between digital and traditional painting. They both use colors, right? The problem is digital painting uses colorful light sources, creating most perfect colors possible and shooting them right into our eyes, while in traditional painting we're limited to light reflected from a pigment. It's like using a middleman between what's painted and what you actually see! We can debate what medium is more artistic, but there's no doubt that digital painting does better with our vision mechanism.
So, to paint traditionally we need pigments. They don't emit color themselves, and instead they absorb some of the light hitting them, reflecting the wavelengths compatible with their names. For example, red paint absorb green and blue, reflecting only red.
The problem is we're not able to create perfect pigments reflecting the light exactly as it would be emitted, e.g. a pigment stimulating the "blue" cone only. CMY system is a kind of compromise: cyan doesn't reflect red, magenta doesn't reflect green, and yellow doesn't reflect blue. So, if we want to stimulate "blue" cone, we need to mix cyan and magenta - this pigment will reflect as little red and green as possible. "K", black, is added to CMY since the components are not perfect and they don't create pure black when mixed in equal proportions.
RGB is additive - the more values you add, the brighter color you get. CMY is subtractive - the less values you add, the brighter the color.
Four Rules of Color Mixing
Rule 1 - Hue Mixing
By mixing two hues you get a hue from somewhere between them, according to proportion. It works for both additive and subtractive mixing.
Rule 2 - Complementary Hue Mixing
You probably heard of complementary colors. They are hues laying in opposite to each other on the wheel. The contrast between them (when they've got the same brightness) is as striking as between black and white. However, when they're mixed, they neutralize each other.
Mixing complementary hues gives neutrality (gray or grayish). Additive mixing of 100% bright complementary hues will return white, subtractive - black.
In subtractive method, adding a bit of complementary hue is the easiest way to precisely reduce saturation.
Rule 3 - Saturation Mixing
In both methods, proportions between components equalize when mixing, and in result saturation is reduced.
Rule 4 - Brightness Mixing
Additive mixing returns brighter color, and subtractive - darker than the lighter one of the components.
The tradition to divide the color wheel into warm and cold halves is very strong. We know that warm colors are active and friendly, while cold colors are passive and formal. Whole books could be written about psychology of color, but the problem is this is not an objective division. What's the warmest color? Red, yellow? Is purple warm or cold? And where exactly should this border line be?
Look at the picture below. These are all reds, theoretically warm all the way. So why some of them appear colder than others? It's about contrast. A color can't be warm or cold, only warmer or colder. The color wheel is so easy to divide visually, because all these colors are put together and easy to compare. Cut red out of it and it's no more warm or cold. It's just red.
So, how to create a warmer or colder color? Every hue on the wheel has a neighbor. These neighbors are always colder or warmer than our sample (check their neighbors too, if you're not sure). To create a colder version of the sample, slide into direction of cold neighbors (and vice versa).
The Basic Rules of Shading
About time, huh? Give me a moment and you'll see this lengthy introduction was necessary to understand the whole process. If you memorize the rules only, you limit yourself to particular situations, but once you've understood where they come from, the sky is the limit!
The Local Color
The common base color, said not to be lighted by any light source, is called the local color. We already know an unlighted object can't have any color, so the better definition is a color not affected strongly by the light nor shadow. So a cherry's local color is red, even if it's illuminated with strong orange light on one side and reflected blue on the other. The local color should be the one you're starting your picture with.
What should be the saturation and brightness of the local color? The brightness is defined by imaginary scattered light that you start your scene with. To define the general brightness of the scene (the intensity of the scattered light) put your object on a white sheet. They're both illuminated by the same light, and the object can't be brighter than the white sheet under the same conditions.
The explanation is simple - the white sheet reflects 100% light. If the
object was brighter than it, it would mean the object reflects more than
100% light (so it's fluorescing or emitting light itself). It's all about contrast, so the darker is your base lighting, the more striking light source you'll be able to add later.
What about saturation? While brightness is about intensity of light, saturation comes from proportion between its components. This proportion stays the same when the intensity of light is changing (with a little exception we'll talk about in a second). It's like adding more water with every teaspoon of sugar - the drink is not going to become any sweeter!
The Direct Light Source
Here's a quick reminder about light areas from the first article:
Let's start with a simple scene not illuminated by any well defined light. The ground is green, the ball is red, and the sky... doesn't matter at the moment. If the background is very far away, it doesn't affect our object. We chose the brightness and saturation, and for now, without no directional light, it looks flat, 2D. That's why we call it flat colors, and it's the easiest part of painting.
When the light source is presented, it floods all the scene. Its intensity - brightness - is the highest where the light has a direct contact with objects (full light, half light) and the lowest where it cannot reach (core shadow, cast shadow). The brighter the light, the darker the shadow. Our local color becomes the terminator.
To keep the ball from floating, we need to add crevice shadow - the area where no light can reach. This is the darkest area of the picture.
The problem is the scene still looks... fake. It's colorful, merry, as if it came from a children book. But something's wrong... If you've read the first article carefully, you may notice we used only diffuse reflection. Every single ray hitting the ball was partially absorbed, reflecting only red. Therefore, in the area of maximum brightness we've got 100% red and there's no way of changing it! This is very natural state for matte materials, and decreasing saturation to get a "brighter" red is a mistake.
If it's natural, why does it look fake? It's because fully matte materials are very rare in nature. Almost everything reflects at least a bit of specular reflection, and it doesn't need to be a high gloss - usually it's very soft and subtle. Change your position when looking at some object close to you - if its "colors" move along to your movement (even subtly!), they're the effect of specular reflection. The ones independent to your position come from diffuse reflection.
Specular reflection, as we've learned before, is a reflection of the light source. The stronger it is, the clearer the image of the light source appears on the object. The biggest role here plays the proportion between specular and diffuse properties of the material. High glossy objects usually have a thin layer of transparent, strongly specular material on them, so both kinds of reflection don't mix (third ball).
To put it straight, when decreasing the saturation of a bright area ("adding white" to it), you're not brightening it - you're adding gloss.
However, the balls above still look fake! (so many ways to paint fake colors, huh?). This time they look like taken from a 3D modeling exercise. This is because we used neutral white light that doesn't occur in nature either. Sunlight, before it can reach our eyes, needs to break through the layers of atmosphere. The previous article explained what happens here, so let's just add color to this mechanism.
Short and medium wavelengths are being scattered the most easily. The longer their way through atmosphere, the more of them stray and never reach your eyes (at least, not from initial direction). Therefore, a "white" ray becomes mostly red and green, and even in the highest point it has a bit of blue deficit - sunlight is warm.
So why would reflection of a warm light source be neutrally white? To avoid that fake 3D model effect, decrease the saturation and increase the temperature at the same time when adding warm gloss (no matter strong or subtle). As we noticed before, there are cold and warm reds, so it doesn't mean that a red surface becomes orange or yellow instantly!
It's important not to use gloss as a universal way to make the picture more attractive. When you feel you're getting closer to white, it means your object is shiny or wet. Think about it when painting skin!
The Indirect Light Sources
But what happens to all this blueness that gets scattered? It makes the sky blue, of course, but if we can see this bright blueness, it means it reaches our eyes - and not only our eyes. All the objects around get "touched" by this indirect light, and then it can be reflected to us too. It's not as bright as the direct sunlight, but it still makes the surface a bit brighter. Also, if it's not fully matte, the surface loses a bit of saturation and becomes colder (since our indirect light source is cold). Keep in mind that the direct light is always stronger than indirect one, so these two will never mix - indirect reflection can't cross the terminator line.
The most intense reflections are created by glossy surfaces, but matte ones, like our "ground", affect the objects too.
As we noticed in the previous article, contrast decreases with distance. But what about hue, saturation and brightness of the receding object? Well, it's a little bit more problematic. When the object recedes into background, the information from it is mixed with the light reflected from the sky, right? It means that:
- Hue gradually changes temperature in the direction of the sky's hue;
- Brightness gradually grows until it reaches the value of the sky;
- Saturation is mixed with the noise, therefore it decreases. However, if the light source is actually in the background (the foreground is dark), the saturation may increase gradually with coming close to it.
The clearer the atmosphere, the less this effect occurs. Respectively, when there's a lot of dust, smoke or humidity around, even close object change their properties drastically. The common trick of artists (and movie creators too!) is to render aerial perspective even in smaller scale, for example drawing one leg of a monster bluer, brighter and less saturated. For our brains it means it's further, and therefore a depth is achieved. However, keep in mind it also thickens the atmosphere - it will not work in clear air.
Color and Value
Proper coloring creates correct values, so to say, involuntarily. Beginners often start their pictures with values only to define them properly, but the truth is with the rules we've just learned you shouldn't have any problems with color painting. How can it be?
- The initial brightness of the local colors sets an uniform brightness for all the scene;
- Diffuse lights and shadows are as saturated as the local color - unsaturated shadows would look brighter as value!
- The more gloss, the more value brightness;
- Indirect lights are never brighter than direct one, so they can't be confused with main light source;
- The local color becomes a terminator, with shadows on one side and lights on the other, what creates a proper contrast.
How to check if more lights or shadows should be added? It's a matter of contrast and you need to choose yourself which is the best for your picture's atmosphere. Generally, it's good to put your main object on three backgrounds: white, black and 50% gray. If it looks OK on every one of them, you're fine. Converting your picture to grayscale for a test is a good idea too.
Points to Remember
- Highly saturated, bright colors are rare in nature - reserve them for flowers, birds and magic things;
- Put lights on lights, never lights on shadows! If you want to put a light on a dark area, brighten it gradually;
- If the shading looks too colorful, take a break, get some distance. There's a chance your eyes are just too focused on them after hours of work and the colors are actually OK. Rotating the picture or looking at it indirectly, in the mirror can help too;
- Save pure white for highlights and 100% black for crevice shadows - overusing them drastically decreases their power.
No More Guessing!
Once you've realized that color is just a signal, a kind of information, it's so much easier to imitate the real world with your paintings. You don't need to memorize hundreds of rules - once you've understood the fundamentals, you can calculate reality with a great accuracy! Of course, don't treat them as a universal recipe for success - art is art, and sometimes you get the best effects when actually breaking the rules.
Stay tuned for the last article of the series, where I'll present you more tricks, such us multiple and colorful light sources, transparency, subsurface scattering, light emission and refraction, and show you what's the fuss about textures.