There is a page from Bong Joon-ho’s Parasite notebooks that stops most filmmakers cold. It is a modest grid of hand-drawn panels, and it matches the finished film so closely that the comparison stops feeling like preparation and starts feeling like a prophecy. Bong boards his own features himself, shot by shot, long before anyone lights a set.
The reflex is to call this meticulousness. It is something stranger and more useful. What Bong is doing on those pages is not illustrating a screenplay. He is editing a film that does not exist yet.
That is the argument here. A storyboard has never really been a drawing of a movie. It is a movie’s first edit, committed to paper before a single frame is exposed. Animation understood this in the 1930s and built an entire review culture around it. Live action largely did not, and paid the difference in the cutting room. That split explains why some films feel inevitable and others feel assembled, and it explains why the current wave of AI Filmmaking tools arrived in pre-production before it arrived anywhere else.
What Is Actually on a Storyboard Panel
Ask most people to describe a storyboard and they describe a comic strip. The comparison is misleading, because the two forms are built for opposite purposes. A comic panel is a finished destination. A storyboard panel is a set of instructions.
Look closely at a working board and you find notation, not just pictures. Shot size is specified or implied by the frame itself: wide master, medium two-shot, insert on a hand. Arrows do double duty, one kind tracking where the subject moves inside the frame, another tracking where the camera moves in relation to it. Lens choices get scribbled in the margins, because a 24mm and an 85mm produce entirely different relationships between a face and the room behind it. Some artists mark the frame with numbers indicating how a shot subdivides.
Then there is the detail that gives the game away. The number of panels devoted to a moment is itself a statement about duration. Six panels for a two-second action means the sequence will be cut fast. One panel with a long note underneath means the camera holds and the actors carry it.
Arrows, tick marks, panel counts, holds. That is not illustration vocabulary. That is editing vocabulary. The board is already thinking in cuts, which is why calling it the first cut is less of a metaphor than it sounds.
How Animation Taught Hollywood to Edit First
Storyboarding as a formal discipline came out of the Disney studio in the early 1930s, when the shorts became too complex to hold in anyone’s head. The studio’s story department began pinning sequential drawings to boards on the wall so an entire crew could evaluate a sequence at a glance, argue about it, unpin a panel, and replace it.
That last part is the whole innovation. Not the drawing, the unpinning.
Animation was forced into this discipline by economics. Animating a sequence is punishingly expensive, so discovering that a sequence does not work has to happen before animation begins, not after. The board became a place where failure was cheap.
Live action inherited the technique and mostly discarded the culture around it. Boards went up on production office walls as logistical documents: what do we need to build, how many days do we shoot, which lens do we rent. The editorial question, does this sequence actually work, got deferred to the assembly cut, months later, after the money was spent. That deferral is the most expensive habit in the industry.
The Directors Who Board Like They Are Cutting
The filmmakers with the strongest visual signatures tend to be the ones who treat boards as authorship rather than logistics.
Akira Kurosawa took this to its extreme. In his later years he produced full production paintings for films including Ran and Kagemusha, works so complete they have been exhibited as art in their own right. They were not shot lists. They were the film’s palette, composition, and emotional temperature decided in advance.
Ridley Scott’s sketches, known on his sets as Ridleygrams, function differently. Scott trained as a designer and thinks in images before words, and his boards are less about locking a plan than about discovering what the scene is. The Coen brothers board tightly and famously shoot close to it, which is precisely why their films cut with such economy. There is very little coverage to hide behind because very little was shot.
The most instructive case is Mad Max: Fury Road. George Miller developed the film primarily as storyboards rather than as a conventional screenplay, working with artists to build thousands of panels. The logic is unarguable once you see the film: the story is chase geography, spatial relationships between vehicles moving at speed. Written prose is a bad container for that. Panels are a good one. The board was not a plan for the script. The board was the script.
The Artists Nobody Credits
Any honest account of storyboarding has to name the people whose work disappears into other people’s authorship.
Harold Michelson spent decades boarding for major directors, contributing to films including The Birds and The Graduate, and specific compositions now attributed to those films’ directors originated on his pages. His career, and his wife Lillian’s parallel work as a film researcher, became the subject of the documentary Harold and Lillian: A Hollywood Love Story, which exists partly because the credit never arrived any other way.
This matters for the argument, not just as fairness. If a storyboard were merely a technical drawing, the identity of the person making it would not affect the finished film. It plainly does. Which means the board is a site of creative decision-making, and the person making those decisions is doing directing work under a different title.
The Case Against Storyboarding
A piece that only praised boards would be worthless, because plenty of great filmmakers refuse them on principle.
Steven Spielberg, arguably the most fluent storyboarder of his generation, deliberately abandoned boards for Schindler’s List. He wanted to arrive without a plan and find the film in the room, and the result is visibly different from his boarded work: handheld, searching, unresolved in a way his blockbusters never are. John Cassavetes built a career on that refusal. The Dardenne brothers rehearse extensively and then follow performance rather than a predetermined frame.
There is also a failure mode specific to over-boarding. When a sequence is locked into panels before the script has settled, production ends up executing images that the story has since outgrown. This is a chronic complaint on large effects-driven productions, where sequences get designed and previsualized while the writing is still moving. And in commercials and music videos, boards frequently stop being creative documents at all and become approval contracts, a set of frames a client has signed off on and will now hold you to.
Genre Decides How Much You Board
The right amount of boarding is not a fixed quantity. It is genre-dependent, and understanding why sharpens the whole concept.
Horror boards obsessively, because horror is built on withheld information. The scare lives in what the frame excludes, and exclusion has to be designed. Comedy often boards least, because comic timing is discovered in performance and in the cut, and a rigid frame can strangle it. Action and stunt work board out of necessity rather than aesthetics, since a sequence involving vehicles or falls has to be planned frame by frame for safety before anyone discusses beauty. Visual effects sequences board because downstream departments need to know what plate they are shooting. Dialogue-driven drama sometimes barely boards at all, because coverage is the plan.
The Animatic, Where the First Cut Stops Being a Metaphor
Everything above still describes static panels. The moment boards start moving, the phrase first cut becomes literal.
An animatic, or story reel, is a board cut together on a timeline with scratch dialogue, temp music, and real timing. Pixar’s process is built around screening these reels repeatedly and rebuilding them, and whole sequences get discarded at that stage. Wes Anderson records animatics for his live-action work, voicing characters himself, so he can hear the rhythm before he shoots it. Suddenly the questions you can ask change entirely. Not does this composition work, but does this scene work, does the joke land, is the second act forty seconds too long.
That is the real prize of pre-visualization, and until recently it was structurally out of reach for most filmmakers.
What Pre-Visualization Used to Cost
A boarded animatic historically required a storyboard artist, a schedule, and a budget line. That is straightforward on a studio production and impossible on a first feature, a short, or a proof-of-concept reel made to attract financing. The filmmakers who most needed to find their mistakes cheaply were the ones who could least afford the tool for finding them. They shot on instinct, discovered problems in the edit, and had no coverage or money left to fix them.
That access gap, more than any question of craft, is what is actually shifting.
What AI Filmmaking Changes About Pre-Production
The interesting thing about AI Filmmaking is where it turned out to be most useful. Not in replacing photography, where it remains conspicuously weak, but in the pre-production stage that always got skipped for cost reasons.
A director can now describe a scene in a line of text and get back a frame-by-frame visual breakdown, then expand a single shot into a connected sequence and watch it move. Tools such as ImagineArt’s AI Storyboard Generator are structured around exactly the vocabulary this article has been describing, with camera bodies, focal lengths, apertures, genre, and camera movement as explicit controls rather than adjectives buried in a prompt. Character and location references persist across shots, which is what makes a sequence readable as a sequence.
The practical consequence is unglamorous and significant. A filmmaker with no budget can now watch a rough version of their own film before scheduling day one, and can therefore discover in an afternoon what used to be discovered in an edit suite.
Where an AI Storyboard Generator Still Falls Short
Generated boards are good at coverage and continuity and bad at intention. A hand-drawn panel by an experienced artist contains a judgment about where your eye should go and what the shot means. Generated frames tend to be competent and slightly generic, and they cannot yet tell you that a scene is emotionally false, which is the most valuable thing a story reel reveals. Performance nuance, comic timing, and the specific weirdness that makes a filmmaker worth watching remain outside the machine.
The honest way to use a platform like ImagineArt, then, is the way animation studios have always used story reels: as a place to be wrong cheaply, early, and often, before the schedule makes being wrong expensive.
It is also worth being clear about labor. This does not replace storyboard artists on productions that can afford them, and the industry is right to litigate that question rather than wave it away. What it does is give boards to productions that previously had none.
The Panel Is Still a Decision
Strip away the drawing tools and the argument survives intact. A storyboard is a sequence of decisions about attention: where the audience looks, what is withheld, when the cut arrives, how long a moment is allowed to breathe.
Bong makes those decisions with a pencil. Miller made them with a wall of panels instead of a script. Spielberg made a decision by refusing to make them in advance at all. The tool is not the interesting variable. The judgment is.
What has changed is that the first cut is no longer a privilege of budget. Which means the only remaining excuse for arriving on set without having seen your film is that you did not want to look.
