In traditional film production, multiple cameras are often used to shoot each scene. (See the camera positions A, B, and C in the figure on the right). Actors play the same scene a number of times until the director is satisfied with their performance and also until each scene is recorded using all the camera positions that may be needed.

In the editing phase, selected shots are assembled into a scene, scenes into the final piece. The editing is done in a such way that the story is told most effectively. A large amount of what’s recorded will not be included in the final edited work, not even the deleted scenes on DVDs!

In 3D computer animation production, you *could* render each shot using multiple cameras and decide how to edit each scene later, but rendering would take a considerable amount of time, i.e. you would waste a lot of time and resources if you render frames without a good plan. 

Before you start animating or even start modeling, while you are in the pre-production stage,

• think about each scene and shot, and
• think about transitions between shots, and
• make a storyboard.

Once you start animating, use hardware rendering, such as Maya's Playblast. Render, play and tweak your animation sequence over and over until you are satisfied with your characters’ actions and everything else in the animation and also find the best camera position for each shot, based on your storyboard.

Before starting rendering full-size frames, do a pencil test by using a smaller frame size (e.g., 1/2 or 1/4 size of NTSC or HD) and rendering every other frame.

At this point, make a 3D animatic by compiling pencil tested shots with the transitions and sounds. The main goal of making an animatic is getting timing right. Animatics also help you make the final determinations on which camera position to use for each shot.

When you are ready to render full size frames, render each shot using only the selected camera. A good planning minimizes the production time and maximizes the amount of fully rendered frames that will be in the final animation.

A shot is a continuous view filmed by one camera without interruption. 
A scene is a place or setting where the action takes place.  A scene may consist of one shot or series of shots depicting a continuous event.

Two major categories of shots: Commonly used shot types in film, video, and animation can be categorized into two major groups: static shots and dynamic shots.

In a static shot, the camera does not move or change its aim within the shot, although the camera may move from the shot to the next shot.

In a dynamic shot, the camera moves or changes its aim within the shot.

Static shots

There are many types of static shots and can be categorized in multiple ways. We will see how static shots are identified by A) scale and then B) angle. And see C) the point of view shots, D) two shot sand over the shoulder shots.

Static shots: A) scale. One set of shots are identified by their scale or shot sizes. The shot size determines how large the area that’s visible within the frame. Among the following common shot sizes, the distance between the camera and subject varies:

Extreme close-up shot Close-up shot Medium shot Medium wide shot (American shot) Wide shot (full shot) Extreme wide shot (long shot)
1. Extreme close-up shot shows only a part of a character's face. It fills the screen with the details of a subject.
2. Close-up shot shows a character's face and shoulders. It is close enough to show subtle facial expressions clearly.
3. Medium shot shows a character's upper-body, arms, and head.
4. Medium wide shot (American shot) shows a character usually cut off across the legs above or below the knees. It is wide enough to show the physical setting in which the action is taking place, yet it is close enough to show facial expression.
5. Wide shot (full shot) shows an entire character from head to toe.
6. Extreme wide shot (long shot) shows a broad view of the surroundings around the character and coveys scale, distance, and geographic location.

The images are from "The Art of Technique: An Aesthetic Approach to Film and Video Production"
by John Douglass and Glenn Harnden

Indecisive cut and shock cut 
If you cut from one shot to another shot of a slightly different size, the size of the subject does not change enough and you create an indecisive cut. The audience thinks it's a mistake and gets distracted by it.

If you cut from one shot to another shot of a significantly different size (e.g., from a wide shot to a close-up shot), you produce a shock cut that shocks the audience. A shock cut can be avoided by having a shot of a shot size that is between the sizes of the two shots (e.g., from a wide shot to a medium shot, and then to a close-up shot).

Watch indecisive cut and shock cut.

Static shots: B) Angle. Another set of shots are identified by their camera angles. Changing the camera angle changes the appearance and function of your shot.

Horizontal camera angles. Moving the camera around the subject horizontally while aiming at the subject creates different camera angles below:

1. Frontal. The frontal angle tends to flatten the three dimensionality of facial features and environments.
2. Three-quarter front. The three-quarter front angle is more often used than the frontal angle or profile because it shows more depth and volumes.
3. Profile.
4. Three-quarter rear.
5. Rear.

Vertical camera angles. Moving the camera around the subject vertically while aiming at the subject creates different camera angles below:

1. High angle. The camera is placed above eye level, looking downward. A high angle shot (downshot) can make a character look smaller, younger, weak, confused, or more childlike.
2. Eye level. Most commonly used.
3. Low angle. The camera is placed below eye level, looking upward. A low angle shot (upshot) can make a character look bigger, stronger, or nobler. It also gives the impression of height.

The images are from "The Art of Technique: An Aesthetic Approach to Film and Video Production"
by John Douglass and Glenn Harnden

Another type of an Indecisive cut:
When cutting from a shot to another shot of a different angle (e.g, from a frontal shot to a three quarter front) while framing the same subject, the difference between the two camera angles must be greater than 35 degrees. If the change in the camera angle is less than 35 degrees, the appearance of the subject does not change enough and you produce an indecisive cut.  The audience thinks it as a mistake and gets distracted.

Watch indecisive cut and shock cut.

Static shots: C) Point of view shots. In a point of view (POV) shot, the camera is placed at the eye position of a character. (Birn, 8.2.4 POV Shots, pages 180-1)

Static shots: D) Two shot & over-the-shoulder shot

• Two shot shows two characters.
• Over-the-shoulder shot is a close-up of a character as seen over-the-shoulder of another person in the foreground.

Dynamic shots. The camera position is often animated in computer animation for no good reason or no reason at all simply because the virtual camera can be moved easily. If you must have dynamic shots, make the camera move realistic and effective. First, study popular types of possible camera moves with a real camera. Try the following:

• Tilt. The camera rotates to aim upward or downward without changing the location. Tilt is sometimes called "pitch".
• Boom. The camera's actual position changes. It travels up and down.
• Pan. The camera rotates from side to side, so that it aims more to the left or right. The camera does not change the location.
• Dolly. The camera's actual position changes, such as to move alongside a moving subject.
• Roll. The camera rotates in such a way that the horizon tilts. Commonly seen in fly throughs.
• Truck. Dolly in moves the camera closer to the subject. Dolly out backs the camera away from the subject. Dolly in and dolly out together are called "truck".
• Zoom. The camera's lens is adjusted to increase or decrease the camera's field of view, magnifying a portion of the scene without moving the camera. Zoom is often confused with truck (dolly in/dolly out). Zoom is not a camera move.

The image is from "The Art of 3-D Computer Animation and Imaging" by Isaac Kerlow.

Dolly vs. Zoom
The difference between dolly and zoom is that when you dolly, you are moving the camera in space, while zoom refers to changing the camera's focal length. When you move the camera, the perspective changes. Objects far from the camera change in relative size at a slower rate than objects which are close to the camera. That is what you see through your human eyes as you walk around, your perspective changes. On the other hand, when you zoom (i.e., when you change the focal length of your camera), your camera does not move and perspective does not change. A technique in which the camera dollies in and zooms out at the same time, or zoom in and dollies out simultaneously is called "Zolly."

180 degree rule
If you plan to use multiple camera positions for an animation scene, i.e., in the editing phase, you intend to edit shots rendered from multiple cameras into a seamlessly sequenced scene, probably the most important rule is to place all the cameras on one side of a line of action. This rule is called 180 degree rule.

A common type of a line of action is an imaginary straight line between two characters that are interacting. (See the figure below). Another type is the path which your character is traveling along. A line of action of the latter type can be a straight line or a curved line. (See the last figure in this section). No matter which type of a line of action you have in your scene, remember to place all the cameras on one side of the line of action.

Look at the following camera placements:

Line of action type 1: An imaginary straight line between two characters that are interacting

The images are from "The Five C's of Cinematography" by Joseph V. Mascelli

Line of action type 2: A path which your character is traveling along.

When your character travels either on a straight line or along a curved path in a scene, placing all the cameras on one side of the path (i.e., one side of the line of action) will create continuity in moving directions among all the shots in the scene.

Composition Rules: The followings are useful guidelines you can use when composing a shot.

1. Rule of thirds
Rule of thirds divides the frame into thirds both horizontally and vertically. The points where the vertical and horizontal lines cross are aesthetically pleasing spots to place subjects or to have perspective lines converge. It is usually best to avoid placing horizon lines exactly in the middle of a frame, but to place the horizon either above or below center, approximately one-third or two-thirds up the height of the frame.

The images are from "Digital Lighting and Rendering" by Jeremy Birn and "The Art of Technique"

2. Avoid tangent
Tangents in composition tend to destroy the depth.

3. Avoid frontal angle
The frontal angle tends to flatten the three dimensionality of facial features and environments. Angling the shot produces more depth and volumes.

Safe Areas
Text, e.g., the title of your animation, should be kept in the central 80% of the screen, within a guideline called the title safe area (the inner green line in the image below). To make sure that your audience will not miss any important action in your animation, the vital parts of your scene should take place within the central 90% of your frame, a guideline called the action safe area (the outer green line in the image below). In Maya, to view the safe areas, in a viewer go to View -> Camera Settings -> Safe Action and Safe Text.