We all know that having two eyes is what allows humans and many animals to see in stereoscopic vision and judge distance, but there are many depth clues in any scene that help us determine depth while only using one eye. This is called monocular vision, and it comes in very handy, especially if you are an artist working in two dimensions, such as with traditional movies and video.
This tutorial will explore the many depth clues that we use every day in the real world as well as how we use our two eyes to determine depth. We will also cover several of the technical aspects of shooting 3-D video (or film) and its related terminology.
Monocular depth clues
There are approximately 11 different depth clues that humans can use with only one eye. These are roughly broken down into three categories: lighting/shading/color, size and motion. Because we live in the real world, we have learned a few things about how familiar objects look and how to judge them.
Occlusion is where one object obstructs the view of another object, so we can conclude that the first one must be in front of the second. By changing our viewing position, we find that the second object comes into view. Occlusion is used in many video effects to provide a sense of 3-D space as one character moves out of the way to reveal something that was hidden behind him.
The relative size of objects is another way to judge their distance. We know that objects that are closer to us seem larger than more distant ones; therefore, we know that the man in the picture is not two stories tall, he’s just closer than the building behind him. This is used in forced-perspective shots in movies where objects and people are made to look smaller or larger than they really are. One of the first good examples of this is in the 1959 Disney movie “Darby O’Gill and the Little People,” in which they placed the actors on platforms above the leprechauns, who were standing on the floor below. With the correct lighting, the scenes look very convincing. (See Figure 1.)
The known size of objects is another clue. Looking down a line of parked cars, you can tell that the last one is far away because it is much smaller than the one you are standing next to. You know that all the cars are about the same size, so the smaller ones must be farther away. (See Figure 2.)
Focus or depth of field is another way to gage depth. A camera lens can be made to have a very short depth of field where only objects at prescribed distances are in focus. This change in focus is proof that object are at different distances or depth from the camera.
Perspective or parallax is where parallel lines seem to converge in the distance, such as is the case with railroad tracks. This was one of the first discoveries in the art world that allowed paintings to look more realistic. It allowed artists to more accurately place people and objects within the 2-D space of their canvas and make it look three-dimensional. (See Figure 3.)
Texture gradients are where surfaces and patterns close to us are more finely detailed than those farther away. You can see the dirt and rocks that makes up a dirt road close to you, but the detail is lost as it stretches into the distance, and it appears to be just a brown surface.
Lighting and shading of a scene gives us clues as to the distance of objects or areas within it. Shadows show us that objects are behind others, and brighter areas are usually closer to us. (See Figure 4.)
Aerial perspective is the effect the atmosphere has on the appearance of very distant areas. The atmosphere can cause a haze that reduces the contrast and clarity; it can also reduce the saturation of colors. (See Figure 5.)
Motion parallax is the effect of being in a moving vehicle and noticing that closer objects move past your window faster than farther objects, such as hills or mountains. When the moon is apparently following you is another example of this, because it’s so far away that it moves infinitely slow no matter how fast you are moving.
Kinetic depth is an example from mostly computer graphics in which the depth of an object or image is only apparent when it is in motion. A wire frame globe may only look 3-D when it is placed in motion and spinning.
Lastly, eye accommodation is where our own eyes must change the shape of the lens to keep the object of our attention in focus. It is an automatic function, but we are aware of the change and it allows us to know the relative distance of the object.
Stereoscopic depth clues
When we use both eyes to observe a 3-D scene, our two eyes work together with our brain to perceive the world and the actual depth of it.
Our two eyes are approximately 6.5cm apart, but there is a variance in this distance. Children, of course, have eyes that are closer together, and some adults have eyes spaced wider apart, but the mean average is 6.5cm. This distance provides each eye with a unique view of the world, separate from the view from the other eye. These two separate images our eyes perceive is called “retinal disparity” or “intraocular disparity.” Our brains process both images and note the differences between the two to develop a sort of map of the world that is made up not only of the colors and light we see, but also the depth of the objects within it.
Our eyes must move independently but in coordination with one another to follow or track objects in 3-D space. Our mind is aware of the movement of our eyes, and that helps us understand the distance of the object we are tracking.
As video moves beyond mere monocular vision depth clues to providing actual separate images to each eye, 3-D video tricks the brain into thinking it is observing an actual 3-D environment. But care must be taken when creating a 3-D video, so it doesn’t exceed the abilities of human stereoscopic vision. Artificial 3-D environments can create situations impossible in the real world, confusing the brain to the point that 3-D can trigger illness in viewers. Understanding how the human vision system works and what clues it uses to determine depth is the first step in making sure you create a satisfying, realistic and safe 3-D experience.
The next “Transition to Digital” tutorial will cover 3-D equipment showcased at the 2010 NAB Show.