Watch out ladies and gentlemen, this post is about physics! Some painters might disagree that you should understand something about optics in order to paint, and if you are painting directly from life, or from a perfect photographic memory maybe I agree. But for painting from your imagination, it is essential to understand how light bounces around and through things if you want your images to appear realistic.
Looking at the room around you, you can sense intuitively where the various light sources are coming from. Our brain is fantastic at picking up the subtle cues of shadows, shading, and highlights to tell us not only the light source, but also the type of material. Is it hard and smooth? Furry? Rough? All of these textures can be visually identified with no conscious thought.
As a painter, you have to remember that the viewers of your art are all fluent in the way light acts without even knowing it. We’ve seen it our whole lives. They might not be able to tell you why, but if your painting has clearly impossible or incorrect lighting, they will immediately sense that something is wrong. The majority of the times that I see a painting which appears “amateurish,” closer inspection reveals it is a failure of shading that has caused my brain to say “this isn’t realistic”. I’m just scratching the surface here too.
Consider a room with a chair next to a window (as in this photo “empty chair” by Alf Caruana)
Simply looking at this photo, you have a very reasonable sense of how far away from the window both you and the chair are. How? Because we intuitively understand the falloff of the intensity of light from a source. Further more we understand the intensity of daylight, and the way a room is softly filled with ambient light. If you painted this scene and the chair was geometrically located in the same place, but illuminated too brightly, as if right under the window, your brain would have conflicting messages about the chair’s location, and the painting would feel wrong.
I’ll rest my case now – that understanding light is important to a painter – and move on to the fun stuff.
First of all, a very basic and misunderstood concept. The northern hemisphere is tilted away from the sun during the winter, which causes colder temperatures. We all know this, but why is it true? Many people (me included at one point) erroneously think it’s because we’re further away. This is false. The fact is that the distance moved is insignificant compared to the total distance from the sun. The real reason is that the angle of the sun at its peak is reduced with respect to the ground.
Consider one square meter of sunlight traveling towards the earth. If you take a one square meter board and hold it normal (perpendicular in all directions) to the incoming beam, the board will capture 100% of the incoming light. However, if you angle the board at 45% for instance, it’s area to the sun is reduced and only a fraction of the light hits it. This means that the light per unit area is strongly dependent on the angle of incidence.
Another way to think of it is a flashlight shone at a wall. Shine it straight on and the total intensity of the light is concentrated on a very small area. Shine it at a very steep angle and the *same amount* of light is spread much thinner, meaning each square inch is far dimmer.
This is the first of two fundamental principles you must understand. It governs how much light a surface receives from a given source. If the surface is directly facing the light, it gets 100% of the intensity. At a 45% angle, roughly 50%. At a 90% angle and anything beyond it is in shadow. We will call this the amount of incident light.
Now we must consider what happens to the light after impact. Assuming we have an opaque surface (so no light passes through), it is either absorbed or reflected. Some light is always absorbed, so remember a reflected light source is never as intense as the original, and often significantly less so (in the case of a dark matte surface).
When rays hit a surface they bounce like billiard balls, off at an angle of reflection equal to their angle of incidence. So a parallel light source hitting a sphere reflects something like this:
The shades in the middle represent the amount of incident light at that point on the sphere. The middle of the sphere most directly faces the source, and so there is the most incident light. Near the edges the incoming light has been spread thin like the flashlight over the wall, so even though the rays still reflect, there is much less light to start with.
This is a simplification. In reality, the light rays scatter as they reflect (another reason they become less intense). This part is really important to getting your shading right! Unless you are looking directly along the line of reflection at the point you are seeing, you are only seeing diffusely scattered light! You can check the line of reflection by imagining that the object you are looking at is mirror coated. The line of reflection goes through the point on the object where you would see a reflection of the light source. It may not even be visible depending on the orientation. Due to scattering, each of the lines in the diagram above actually looks more like this:
The further away from directly on, the fewer percentage of photons at that point have been scattered, so you can imagine each ray as having the greatest intensity when viewed directly on, and diminishing intensity as viewed from a greater and greater angle off center.
rate of the diminishing intensity depends on the surface quality. For glossy surfaces, the majority of the incoming light reflects directly along the line of reflection. For matte surfaces, light is scattered widely, and the difference in intensity between viewing directly at the line of reflection or not is minimal.
One critical non-intuitive thing to point out here is that, due to reflected light, the shading of an object absolutely depends on where you view it from! You can see this in highlights, which will move across a surface as you walk around it, even though the object and light remain stationary. Of course, incident light levels remain constant, so the amount of light which could possibly be reflected from any given point is limited depending on it’s orientation to the source. But the percent of that potential incident light which actually makes it’s way into your eyeball is 100% dependent on the position of your eyeball.
Going back to the original reflection diagram, we can redraw it, but also include the angle of scattered light (smaller arrows) which actually makes it to our eyeball. Now you can see what I mean about how most of the light we see is indirect scattered light.
You can also see how the brightest point to the observer will NOT be the point which receives the most light. It has taken a while to get here, but this is the biggest mistake I see in shading. Circular objects lit from the side have their brightest and darkest points inside the edge, not at the edge.
A cylinder viewed in this orientation will have a light color at it’s leading edge, but much less intense than the total incident light because it is being viewed at a steep scattering angle. This is especially true of a glossy object which can appear very dark even on the illuminated edge. It will then get lighter going down the curve. Even though the amount of incident light is diminishing, the viewing angle is getting closer and closer to a direct reflection (the highlight) which is the brightest point.
If you can draw or imagine rays of direct or scattered light when you paint (called ray-tracing) you will have a much easier time locating the shading gradients. You’ll also understand more concretely interesting phenomenon such as rim lighting, and reflected lighting in the shadows (not covered at all in this post, but absolutely essential to realistic painting). Of course this is an incredible simplification of illumination – but a reasonable attempt to cover the subject would easily fill a whole book. Speaking of which! All of these concepts and many more can be found in much more detail here: http://www.amazon.com/Light-Artist-Ted-Seth-Jacobs/dp/0823027686 (I have no association with this book, other than I really really like it).