Layout artist

What they do

On an animated feature the layout artist is the camera department. They receive the story reel and the built sets and characters, and return the actual shot: a camera with a focal length, a height, a path, and a duration, with characters blocked to hit their marks inside it. Rough layout replaces boarded panels in the reel with 3D shots, and from that point the reel is made of real cameras. Final layout tightens them after animation is approved, adjusting framing and adding the handheld or crane character the shot needs. This is the opposite of previs in one specific way: layout is not thrown away. The camera the layout artist publishes is the camera that renders, so animation, effects, lighting and compositing all inherit it. Days are spent in Maya, Blender or a studio tool, with proxy or subdivided assets and a USD scene assembled from published sets and characters. The unglamorous parts are the constraints: keeping scale consistent so a character does not drift a head taller between shots, keeping screen direction legal across a cut, checking that the set exists where the camera points, and flagging to the art department that a wall needs building because the boarded angle looks through it. In VFX rather than animation, the same job also inherits a tracked plate camera and has to stage inside it.

What they decide

The calls that belong to this role and no other.

• The lens, camera height, and camera move for each shot, presented to the director for approval
• Where characters stand and walk inside the set, and which marks they hit on which frame
• How the boarded sequence is translated into shot lengths and cut points, in conversation with editorial
• Screen direction and eyeline continuity across a sequence of shots
• Whether an approved board is achievable in the built set, and what has to be added if not
• The staging depth: which characters are foreground, which are mid, and what the set does behind them

Principles of the craft

• Eyelines and screen direction are contracts: once a character looks left to address someone, that geometry binds every shot in the sequence. Draw the overhead plan for the sequence before you set a single camera, and place the line of action on it, because you cannot fix a crossed eyeline in one shot without breaking the two around it.
• The lens is a dramatic choice, not a framing convenience: a 25mm puts the audience inside the character's space and bends the room; a 100mm flattens them into their background and takes the room away. Pick the focal length for what it does to the relationship in the shot, then move the camera to get the framing, rather than zooming to fix a composition.
• Stage in depth: a virtual camera has no cost per foreground element, which is exactly why layouts go flat. Put something in the near ground, put the subject in the mid, and give the background a reason to be there. Depth is what makes a CG frame feel photographed.
• Layout inherits the boards and owns the geography: the story artist decided what the shot is about, and layout decides where the audience is standing. When those conflict, the board wins on intent and layout wins on space, and the conversation happens in dailies rather than in a published camera.
• Camera motion needs a motivation and an operator: a move that starts because the shot got boring reads as a mistake. Decide what the move follows or reveals, then give it a plausible physical character, whether that is a dolly with mass, a crane, or an operator with a shoulder.

Common mistakes

• Cameras that are physically impossible: floating through walls, or a move with no acceleration, which reads as animation rather than photography.
• Losing scale continuity, so characters and set pieces shift relative size between shots in the same sequence and nobody can say which shot is wrong.
• Framing to the character and forgetting the set, then discovering the camera looks at an unbuilt wall or an empty backdrop.
• Pushing layout past its job into animation, blocking performance in detail and delivering a shot the animator then has to argue out of.
• The competent artist's mistake: a beautiful, complicated camera move that cannot be cut, because it starts and ends in places the neighbouring shots cannot match.

How to learn it

Exercises and routes in, not a reading list.

• Work through the Khan Academy Pixar in a Box sets and virtual cameras units, then rebuild one exercise in Blender with a camera set to a 36mm sensor. Shoot one staged conversation at 24mm, 50mm and 85mm from the same framing, matching subject size by moving the camera. Compare what happens to the room behind the characters.
• Draw an overhead plan of a three-character scene, mark the line of action, then lay out five shots that never cross it. Then deliberately cross it in shot three, cut the sequence, and watch someone else get lost.
• Download the Disney Animation Moana Island Scene data set, put a camera in it, and stage a shot with real foreground, mid and background. Then render the same shot with the foreground element deleted to see what you lost.
• Take a two-minute sequence from a live-action film, count the setups, estimate each focal length, and rebuild the geography as a layout in grey boxes. Where your estimate feels wrong is where you learn what the lens was doing.
• Lay out a walk-and-talk down a corridor twice: once as a single steadicam-style following shot, once in six static setups. Cut both. Note which one forced you to invent geography you had not built.