Build a Film Lookbook Around a Visual Idea, Not a Pile of References
Build a Film Lookbook Around a Visual Idea, Not a Pile of References
You have a folder of images that would all work in a beautiful film. That is the problem. Each one would work in a slightly different beautiful film, and nothing in the folder has the authority to say so.
A lookbook stops being a pile the moment it holds a proposition: one sentence about how the film puts the viewer in relation to the story. With that sentence, references can be kept, cut and ordered by whether they demonstrate a particular choice. Without it, the only two criteria available are attractiveness and the reputation of the film an image came from, and both will cheerfully accept sixty images.
One practitioner account — We Make Movies' 2021 piece on pitch decks and lookbooks — treats images as a way of communicating a director's visual intention. That is a fair starting frame. It is also one practitioner's 2021 account rather than a standard, and it leaves you holding the hard part: specifying what your intention actually is.
Coherence is a relationship, not a palette
Most lookbooks that feel like six films feel that way for the same reason. They were assembled by resemblance. Each new image was chosen because it shared something with the last one — a color, a grain, a mood, a quality of light — and nothing was chosen because it repeated a decision.
You can match palette across forty pages and still be making six different films, because palette is not a point of view. Two images can share the same warm brown and hold opposite positions on where the viewer gets to stand. One lets you into the room with the person in it. The other keeps you at the door, watching. If your film is committed to the second, the first will fight it on every page, and no amount of color grading will reconcile them.
So the coherence you are after lives higher up than how the images look. It lives in a recurring choice about the viewer's position: intimate or withheld, steady or unsettled, expansive or confined. That choice can be demonstrated in a hundred visual ways. What matters is that it keeps being the same choice.
From adjective to relationship
"Lonely" is not a proposition. It describes a temperature, not an arrangement. If the lookbook says lonely, every mournful image you own is admissible and none of them is answerable to anything.
Convert the adjective into a relationship by asking three questions. Who is lonely — the character, or the viewer? What is the viewer allowed to see? What is the viewer denied?
Run it on an invented film. A caretaker lives inside rooms prepared for other people; she keeps them ready for owners who are elsewhere. The adjective wants to be "lonely." Answer the three questions and you get something usable: the viewer sees the rooms clearly and is never admitted into the caretaker's attention. That is a proposition. It has an inside and an outside, and it can be argued with.
The same conversion works elsewhere. "Claustrophobic" becomes a choice between confining the viewer in the room with the character and letting the viewer see the room clearly while the character cannot find the exit — two different films, one script. "Warm" becomes: the rooms are warm and the film is not, or the rooms are cold and the film is warm toward the person in them. "Intimate" is worth distrusting until you say intimate about what. The viewer is admitted to everything except the one thing that matters is a proposition. The viewer is close to everyone is a mood.
Notice what has stayed separate through all of this. Subject matter is the house, the keys, the made beds, the caretaker. Viewer position is the doorway. Change the position and the same subject matter becomes a film about property, a film about surveillance, or a film about a woman being slowly erased by other people's furniture. This is why the proposition, not the subject, is what makes the lookbook yours.
One caution: a proposition guides decisions, it does not prescribe a format. "Never let the viewer inside her attention" is not the same instruction as "shoot everything wide from doorways." The second is one solution; the first can be met in close shots, in sound, in what a scene refuses to cut to. Keep the proposition at the level of the relationship so it can keep working after you have used up your first idea.
What a proposition has to be able to do
A proposition has two jobs, and most drafts only do one of them. It has to keep images. It also has to kill them.
The keeping test is a sentence you finish for every candidate: This image is here so that the viewer ______. If the blank fills with a mood noun — feels the loneliness, senses the distance — the image is decoration. If it fills with a relation — sees the room fully but not her face — the image is doing work.
The killing test is harder and more revealing: what would this proposition make you throw away? A proposition that admits everything is not a proposition, it is a preference. If you cannot name an image you love and are prepared to lose, you have not decided anything yet.
There is a quick way to catch yourself. If you could swap one image for any of three others without changing your argument, none of them are doing a job. The argument is being carried by the captions.
Keep the vocabulary small. A lookbook is not a survey of good images; it is an argument conducted with a limited set of recurring choices. If a reader cannot summarize your proposition after three spreads, you have written a catalogue.
Give every reference a job, and say what you are not borrowing
For each image you keep, name two things: the feature you are borrowing, and the features you are not.
Take a doorway image — a sitting room seen through an open door at dusk, the room entirely legible, chairs and lamp and books all readable, the caretaker somewhere inside it, her face not. The borrowed feature might be the distance the frame measures, or the fact that the room resolves before the person does. The list of what you are not borrowing is longer: not the period, not the lamp, not the wallpaper, not the aspect ratio, not the film stock, not the way that particular director handles a doorway.
That second list is where your authorship lives. Without it, "we like this" is doing all the work, and the reputation of the source film is doing the rest. References borrowed for prestige always pull a project toward the film they came from, because the prestige is the other film.
Comparison is the tool that forces this into the open. Put two images side by side that share a palette and hold opposite viewer positions — the warm sitting room at dusk, and a warm close portrait of a person facing the camera in a lit kitchen. The color agrees. The access does not. If your proposition says the viewer stays outside, one of these two is decoration, and you can now say which and why. A single image can be persuasive on its own. A pair of images makes you show the difference you claim to see.
The image you like most is often the one to cut
Now take the portrait. She is close, at ease, facing the lens, lit warmly. It is probably the best photograph in the folder. It also argues for a film that hands the viewer access early and often, which is the opposite of the film you have proposed.
Name what it contradicts, and name it in writing, in the working notes if not on the page: this grants the intimacy the film is withholding. That is not a criticism of the photograph. It is a diagnosis of the gap between the photograph and your film.
Rejection is not a verdict on the source image. If the project were about a woman who runs other people's houses and is completely present in them, that portrait would be your opening image and you would be right to lead with it. When you cut it, you are not saying it is bad. You are saying it belongs to a different film, and that you know which one.
There is a version of that image that belongs here, and it is worth understanding why. Move it to the moment when the story has changed — when she invites someone into her own room — and something very like it becomes legitimate, because the access has been earned by an event rather than asserted by a photograph. The same image can be a rejection or a departure. What changes is whether the film has produced the conditions for it.
Which brings up the tone page. Many lookbooks have a spread of beautiful images that are not doing anything except setting a mood. Sometimes that page has a real job. Often it is where the pile grows back. Ask what the page would forbid, and if the answer is nothing, you know what you are holding.
One departure, and what stays fixed
A proposition that never varies is a wallpaper pattern. The reader needs to see, once, that it can develop — that the visual relationship is a rule the film can earn its way out of.
In the invented example, the changed condition is the invitation. She opens her own door to someone. The departure is a close view of her face: nearer than the film has allowed itself to be, at the one moment the threshold rule is deliberately broken.
What makes it a departure rather than another favorite is the anchor. The doorway is still there. The composition still measures access; the camera has simply moved because she opened the door. Without that continuity the close view is just an image you like more than the others. With it, the reader can see a rule, and can see the exact moment the film chooses to break it — which is information about authorship, not decoration.
One departure is enough for this task. The lookbook does not need to plot the whole arc; establishing that the proposition can move is different from sequencing the movement scene by scene. How a lookbook shows progression across a finished film is its own problem, and it is a later one. Your job here is deciding what belongs at all.
Assemble an argument, not a shooting plan
Four or five spreads will carry it. The proposition, stated plainly. The pair of images that share a palette and disagree about access. The rejected image with its reason. The departure with its anchor. In the invented case, that is three candidate images described in words — the doorway sitting room at dusk, the kitchen portrait, and the close view. One states the rule, one is refused for contradicting it, and one is the exception the rule allows. The arithmetic is worth doing, because a set that only ever adds is not a set.
Mark reference material as reference. If a page shows a still from someone else's film, say so; do not let it drift into looking like footage from yours, and if you have test material of your own, mark that too. Whether you may use a given image is a real question with real answers, but it is a separate one, with its own basis to establish — do not settle it in a caption.
Be honest about what a still can and cannot show. A photograph demonstrates framing, what is legible in the frame, and where the viewer appears to be standing. It cannot show how long the moment is held, whether the camera moves, whether a door closes behind her, whether anyone speaks, or what the room sounds like. That matters here, because with a threshold rule the duration of the withheld moment is probably part of the proposition. Write a line beside the image that carries it, or plan the piece of moving or sounding material you would need. The still alone will quietly imply a rhythm the film has not decided on yet.
Three sentences to finish with
If your proposition is working, you should be able to write three sentences about the book without looking anything up.
This image is in because it puts the viewer here.
This image is out because it puts the viewer there, which is a different film.
This one is unlike the rest because the story has changed at this point, and it is still measured by the same rule.
Write those three, keep the proposition short enough to say out loud in a meeting, and let every page answer to it. Leave the equipment out. A camera package may follow from the intention later, but a lookbook that turns into a gear list has stopped communicating a film and started communicating a purchase.
Frequently asked questions
What turns a folder of strong reference images into a lookbook?
A proposition: one sentence about how the film puts the viewer in relation to the story. Without it, selection falls back on attractiveness and source-film reputation, and both will accept too many images.
Why is a consistent palette not enough to make references cohere?
Palette is not a point of view. Two images can share warm brown and hold opposite viewer positions—one admits the viewer into the room, one keeps the viewer at the door. Coherence is a recurring choice about the viewer's position, not a color match.
How do you convert an adjective such as lonely into a usable proposition?
Ask who is lonely—the character or the viewer—what the viewer is allowed to see, and what the viewer is denied. In the caretaker example, the viewer sees the rooms clearly but is never admitted into the caretaker's attention.
What tests should a proposition pass?
It must keep images and kill them. For each keeper, finish: This image is here so that the viewer ___. A mood noun means decoration; a relation means work. If you cannot name a loved image you are prepared to lose, you have not decided anything yet.
What should you name for each reference, and what should you do with a favorite image that contradicts the film?
Name the borrowed feature and the features you are not borrowing; the not-borrowing list is where authorship lives. If the favorite grants an intimacy the film withholds, cut it and write why it belongs to a different film. A later departure can be legitimate if the story earns access and the doorway anchor remains.