Previs, storyboards, and animatics get used interchangeably in pitch meetings, and the confusion is expensive. Each one answers a different question about your film. A storyboard tells you what each shot looks like. An animatic tells you how the sequence plays in time. Previs tells you how the camera and the actors move through space. Pick the wrong tool and you either burn weeks planning a dialogue scene that needed an afternoon, or you walk onto a stunt day with a stack of drawings that never tested whether the shot actually works. This guide breaks down what each format is, when each one is enough, and how to decide scene by scene.
Storyboards: the composition layer
A storyboard is a sequence of panels, drawn or generated, arranged in shooting order. Each panel represents one shot and carries the information that lives inside the frame: shot size, camera angle, blocking within the composition, screen direction, and notes on movement or dialogue. Boards are where you decide that the scene opens on a wide, that the reveal lands on a slow push-in, and that the reaction belongs in a clean single.
Boards are the fastest and most universally understood planning document in filmmaking. Your DP reads them for lensing and lighting, your AD builds the schedule around them, your production designer sees what has to be in frame. A director who shows up with boards has already made most of the visual decisions that would otherwise get made under pressure on the day.
What boards cannot represent is time. A panel held on the page for two seconds and a panel that represents a forty-second oner look identical. That gap is where the other two formats come in.
Animatics: the timing layer
An animatic is your storyboard set in motion: panels sequenced on a timeline, each held for its intended shot duration, with scratch dialogue, temp music, and key sound effects layered underneath. Nothing about the imagery changes. What changes is that you can now watch the scene at its real length instead of imagining it.
The distinction matters because a storyboard tells you what each shot looks like, while an animatic tells you how each shot feels at its intended duration and how the sequence flows as a whole. Scenes that read beautifully as boards routinely die in an animatic. The establishing shot that felt essential turns out to stall the scene for six seconds. The three-cut escalation you planned needs a fourth beat. These are editing discoveries, and an animatic lets you make them for the cost of an afternoon instead of a reshoot.
Previs: the spatial layer
Previs, short for pre-visualization, traditionally means building rough 3D animated versions of scenes before production, using simplified models and environments to plan camera placement, lens choice, blocking, timing, and visual effects integration in three-dimensional space. It grew out of VFX-heavy studio filmmaking, where sequences on films like The Lord of the Rings were too complex for static boards to plan.
The discipline has layers of its own. Pitchvis sells a concept before financing. Standard previs works out shots during pre-production. Techvis converts creative previs into exact camera positions, lens specs, and rigging data for the crew. Postvis drops temporary effects into the edit so the cut can be judged before final VFX. For most independent productions, only the standard layer is ever in play, and even that is reserved for a handful of sequences.
Side by side
| Storyboard | Animatic | Previs | |
|---|---|---|---|
| Purpose | Lock composition, framing, and coverage per shot | Test pacing, rhythm, and scene length before shooting | Plan camera movement, blocking, and VFX in 3D space |
| Stage | Early pre-production, straight from the script | After boards are locked, before the shoot | Pre-production for complex sequences, often alongside boards |
| Fidelity | Static frames | Timed frames with scratch audio | Animated motion, traditionally rough 3D |
| Time cost | Hours to days per scene | Hours on top of finished boards | Weeks with a 3D team; hours with AI generation |
| Who uses it | Director, DP, AD, every department head | Director, editor, producers | Director, DP, VFX supervisor, stunt coordinator |
| When to skip it | Almost never | Simple dialogue coverage with no timing risk | Any scene without complex movement, stunts, or VFX |
The indie-budget reality
Traditional previs is a service business. Studios hire houses like The Third Floor, whose teams of 3D artists work in Maya and Unreal Engine alongside the director for weeks or months per sequence. On a blockbuster, that cost is trivial next to what it prevents: discovering a camera position does not work while two hundred crew members stand around on a live set.
That model does not scale down. An indie feature cannot carry a previs team, and learning Blender well enough to previs your own scenes is a production in itself. So the practical question for most filmmakers is not which format is best, it is which decisions actually need spatial simulation. The honest answer is that boards plus a timed animatic cover roughly ninety percent of the scenes in a typical script. Dialogue, walk-and-talks, standard coverage, montage, most interiors: none of these need a 3D camera. The remaining ten percent, the stunt sequence, the VFX shot, the complicated oner, are where previs traditionally earned its cost, and where its absence used to show up as wasted setups and missing coverage.
A decision framework by scene type
Dialogue scenes. Boards are enough. Frame your coverage, mark your eyelines, and move on. Add an animatic pass only when timing is the substance of the scene, such as comedy that lives or dies on a pause, or an interrogation where you want to feel how long the silence holds.
Action scenes. Boards plus an animatic, always. Action is rhythm, and rhythm is invisible on paper. Cut your boards to temp sound and you will immediately find the beats that repeat and the geography that confuses. Escalate to previs, or moving AI-generated shots, when vehicles, wire work, or stunt rigging make spatial planning a safety and budget issue.
VFX scenes. This is the closest thing to mandatory previs. Your VFX vendor needs to see motion, scale, and spatial relationships to bid accurately, and a static frame hides all three. Even rough moving reference dramatically reduces the back-and-forth that inflates VFX costs.
Oners. The trick case. A single panel cannot hold a moving camera, so board the keyframes of the move, the compositions the camera passes through, then time them in an animatic. If the choreography between camera and actors is genuinely complex, this is the one non-VFX case where generating motion for the shot pays for itself before you rehearse it.
How AI generation collapses the cost difference
The wall between these three formats was never conceptual. It was cost. Boards were cheap, animatics were cheap, and previs required a 3D pipeline, so previs stayed a studio luxury. AI generation has quietly removed that wall.
An AI storyboard generator produces boards from a screenplay in hours rather than days, with consistent characters held across every setup. In mStudio, you can import a finished screenplay as a PDF and get a frame-by-frame board of the script, then drop those frames onto a timeline, set per-shot durations, and mix scratch voiceover, temp music, and sound effects into a working animatic in the same tool. For the shots that need the spatial answer, image-to-video generation turns a board frame into a moving shot, which gives you the decision-making layer of previs, camera movement you can watch and judge, without touching 3D software. An animatic maker that generates its own frames and motion means the jump from boards to animatic to moving previs is a matter of minutes per scene, not a new budget line.
Be clear-eyed about the boundary. Generated motion will not give you techvis: there are no measured camera positions, no lens data extraction, no 3D scene to hand a crane operator. What it gives you is frame-accurate reference for the creative decisions, whether the move works, whether the coverage cuts, whether the scene earns its length, which is the ninety percent of previs value most productions actually need. If you are weighing dedicated tools for that job, our roundup of the best previs software compares the current field.
What your film actually needs
Board everything. Animatic anything with timing risk. Reserve full previs thinking for the scenes where the camera or the effects move through space in ways a flat panel cannot describe, and use generated motion to get there without a 3D team. The formats are not rungs on a ladder of seriousness; they are answers to different questions, and the discipline is asking the right question per scene.
mStudio was built so one workspace can answer all three, from screenplay import to boards, timed animatics, and generated motion, with an export your whole team can watch. If your next project has scenes that need more than paper, start with our previs software and see how far the boards-to-motion pipeline takes you before you ever consider a 3D pipeline.