Build a Film-Deck Grid From the Copy and Images It Must Carry

By pitch.dogFoundation16 min read20 September 2026

Film

You have the type settled. Maybe a display face for the title, a text face that behaves at small sizes, a body size that reads and a heading size that outranks it. Then you open a new page for the synopsis and the page has no opinion about where anything goes. You drag a text box until it looks approximately right. On the next page you drag a different box, because there's a portrait there. By page six nothing lines up, and when someone asks why the third page has a wider margin than the fourth, you don't have an answer.

The usual response is to go looking for a better template. The better response is to derive the spatial rules from the material you already know you have to carry. That material is finite and mostly boring to look at, which is exactly why it works as evidence.

Collect the pages most likely to break a tidy system

Before you draw anything, gather the four kinds of content that will fight each other.

The longest continuous passage. Not a paragraph of placeholder text — the real synopsis, or the real director's statement, at its real length. Lorem ipsum has a flat distribution of word lengths and no proper nouns, so it never wraps the way your copy wraps. A name like "Anastasiya Volkov-Brenner" will break a narrow measure in places a Latin filler never does.

A portrait with its description. A character still, a headshot, a frame with a face in it, plus the paragraph that explains who this person is and why they matter.

A wide reference image. The frame that establishes the film's world — a landscape, a corridor, a crowd. It will be much wider than it is tall, which is the opposite problem from the portrait.

Short practical facts. Runtime, format, language, production company, contact, festival history. These live in small type and change line by line.

Keep the captions and the qualifiers attached to each of these while you work. "Concept art, not a final frame" is a line of copy with a job. If you drop it while you're testing, you'll build a grid that has no room for it and discover the problem on the page where it actually matters.

And establish the two things that determine everything downstream: the intended reading size and the presentation format. A deck that gets projected in a room and a deck that gets read as a PDF on a laptop are different documents wearing the same clothes. Decide which one you're building, or which one comes first.

What you should not do is start by choosing a column count. An empty cover looks balanced at almost any number of columns. The cover is the page with the least content and the most freedom, so deriving the grid from it produces a system calibrated for the one page that needs it least.

What follows is a paper example. The page described below has not been built in any application, nobody has looked at it at full size, and no comparison has been exported. The numbers are stipulated so the reasoning can be followed and repeated on your own content.

Set the usable area and the reading measure

Pick a page and write it down. Twelve inches by 6.75 inches is a 16:9 frame — 12 divided by 6.75 is 1.778 — and it's a workable size for a deck that will be shown and also printed.

Margins of three quarters of an inch all round leave a usable area of 10.5 by 5.25 inches. Now put the long passage into that area at its intended size.

Set the body at 11 point on 15 point leading. Fifteen points is 0.2083 inches, so the 5.25-inch usable height holds 25 lines of body text and a small remainder. Column height, in other words, is 25 lines, and it stays 25 lines however wide you make the measure, because a wider column doesn't make the page taller. That number is the budget for every text page in the deck, and it's worth knowing before you start moving anything.

How wide is a line? Estimate from average character width. For a typical text face at 11 point, an average character is about half an em, so roughly 5.5 points, or 0.0764 inches. A single column at 1.625 inches holds about 21 characters. Two columns, 45. Three columns, 68. Four columns, 91.

Those are estimates, and they shift with the face and with the specific words. The point isn't the exact figure; it's that you can predict the range before you typeset anything.

Measure and capacity are separate questions, and running them together is how a layout ends up promising a page more than the page can hold. The measure says how many characters fit on a line. The capacity says how many lines stack up in the height. A three-column measure at 11 on 15 is 68 characters wide and 25 lines tall. It does not hold 75 lines, because those lines run down the page one after another, not side by side.

Now test the synopsis. Six hundred words is about 3,900 characters once you count the spaces. At 68 characters per line, that's 58 lines. The page has 25. Fifty-eight lines of 11-point type is roughly 12 inches of text looking for somewhere to stand in 5.25 inches of space. The page at this measure holds 1,700 characters — about 260 words.

It doesn't fit at some other measure either, which is worth seeing once, because it tells you whether you have a format problem or a preference problem. The full usable width is 10.5 inches, about 137 characters, and 25 lines of 137 characters is 3,425 characters, about 525 words. That is the most 11-point type this page will ever hold, and it comes with a line no reader would want to follow. A 600-word synopsis is a two-page problem, and you want to know that before you promise it one page.

What the character counts do settle is the measure, and the trade there is real. At 45 characters, the same 3,900 characters run to 87 lines — three and a half pages. At 68 characters, 58 lines, about two and a third pages. At 91 characters, 43 lines, not quite two. A narrow measure costs pages; a wide one costs the reader, who has further to travel along each line and further back to find the start of the next. Twenty-one characters is a label. Forty-five is workable, cramped, and expensive in pages. Ninety-one is past the point where an eye returns comfortably. Sixty-eight sits between them. So three columns is the measure for continuous prose — not because it makes the synopsis fit, which it doesn't, but because it's the width at which the copy is worth setting at all.

Two facing pages at that measure hold fifty lines, about 520 words. At 600 words, the synopsis is either trimmed by about 75 words or given a third page. Both are reasonable answers, and both are better than discovering the question on the day the deck is due.

One caveat on size. If the 12-inch page is displayed at 100% on a 1920-pixel-wide screen, an inch is 160 pixels, so 11-point type renders at about 24 pixels. That's comfortable. If the same deck gets shown smaller — say at 60% — those 24 pixels become about 15. Choose the size for the smallest place the deck will be read, which is usually not the projector.

Divide the 10.5-inch usable width into six columns of 1.625 inches with five gaps of 0.15 inches. Six times 1.625 is 9.75, plus 0.75 of gaps, is 10.5 exactly. Nothing is left over, which means the outer column edges sit precisely on the margins and every internal edge falls on a repeatable line.

Six columns, and a one-column measure of about 21 characters. That is too narrow for prose and perfectly right for a label. "Runtime 94 min," "Language: Turkish," "Contact: press@ —" all fit. The grid now has a unit for prose and a unit for facts, and they're the same unit.

The test of the modules is whether different shapes can share edges without being forced into the same box.

Take the wide reference image at 1.85:1 — a flat frame. At the full six columns, 10.5 inches wide, it would be 5.68 inches tall. The usable height is 5.25. It doesn't fit. At five columns the width becomes 8.725 inches and the height 4.72 inches, which does fit, leaving 0.53 inches. Put a two-line caption at 8 point on 10 point — 0.278 inches — plus the standard 0.15 gap below it, and the total is 5.15 inches. It fits with a tenth of an inch to spare.

Now take a 2.39:1 scope frame. At the full six columns it's 4.39 inches tall, and with the same caption treatment it comes to 4.82 inches. It fits at six columns.

So the scope frame runs the full width of the grid and the flat frame stops a column early. They share the left edge and the top. They do not share the right edge, and they shouldn't: cropping the flat frame to a wider ratio to make the right edges match would change what the frame shows. The grid preserves the relationship at the left and lets the right land wherever the image's ratio puts it.

The portrait is a different problem again, and it needs a different span. A 3:4 portrait at two columns is 3.4 inches wide and 4.53 inches tall — inside the 5.25-inch height. Beside it, across the remaining width, you can have a three-column description at 5.175 inches and a one-column side strip at 1.625 inches. Three plus one is four, and two columns of image plus four columns of content is six. Everything shares the baseline and the top edge.

That side strip is worth keeping. It holds the role, the performer, a pull quote, or a single line of context — short items that would clutter a prose column. It's a one-column measure, so it wants a smaller size and a leading that matches, and it should never be asked to hold a paragraph.

Upsetting the portrait to four columns gives you a 6.8-inch image and a 91-character description, which is worse on both counts. Six columns of image leaves no room at all. Two is right, and the arithmetic says so rather than the mood board.

Use baseline alignment only where it helps

A baseline grid is a rhythm for aligning text baselines. It is not the same thing as the six-column spatial structure — it organizes lines within the spaces the columns define.

Adobe's help page for InDesign describes baseline grids as a way of aligning text baselines across columns or pages, and documents two sets of controls: a document grid and a frame-based custom grid. It also documents the relationship between the grid and body leading. That's a narrow, factual observation about how the tool works. It does not tell you what margins to use, how many columns to have, or anything about whether your deck is accessible, and it shouldn't be read as doing any of that. The controls have also shifted between versions, so check the documentation for the version you're actually running rather than the one you remember.

With body leading at 15 points, set the document baseline grid to 15 points and let the prose columns sit on it. Now the lines of the synopsis and the lines of the character description, on different pages, start at the same heights. You can see it in the deck: flip between two text pages and the lines land in the same places. That's a real gain, obtained for about thirty seconds of setup.

Where the baseline grid stops helping is captions, fact rows, and headings.

A caption set at 8 on 10 doesn't share a rhythm with 15-point body leading. If you force it onto the 15-point grid, the caption occupies 10 points of text and 5 points of unexplained gap, which reads as a mistake in the spacing rather than a decision. The tidier move is a frame-based grid at 10 points for the caption block and the fact column, and let those sit where they need to sit.

Here's a detail worth knowing before you decide the caption is out of step forever. Ten and fifteen meet at thirty. A 10-point frame grid realigns with the 15-point document grid every 30 points — every two body lines. So if you start the caption block on a document grid line, it drifts for one line and comes back into alignment on the third. The drift is bounded and predictable, which is a much easier thing to live with than a caption set in the wrong size to satisfy a grid.

Headings should generally not be pinned to the baseline rhythm either. A heading's job is to group, and grouping is spatial: it belongs close to the text beneath it and further from whatever precedes it. Forcing a 30-point heading onto a 15-point grid rarely damages it badly, but forcing a 22-point heading onto a 15-point grid makes it either cramped or floating, and neither of those helps the reader understand what belongs together.

Stress the construction and keep purposeful exceptions

Build all four pages, then make them harder.

Lengthen the character description from 150 words to 300. At 68 characters per line, 150 words is about 975 characters, 14 lines, something under three inches of the 25-line column beside the portrait. At 300 words it is 29 lines, about six inches, and the space beside the portrait is 5.25 inches tall. The block absorbs about 260 words before it runs out of height.

That is still a useful result, because it replaces a reflex with a number. When a description grows, shrinking the type becomes a decision you make against a known capacity: below roughly 260 words there is room, and above it you are choosing between a second page and a smaller size.

Now substitute a less accommodating image. Replace the 3:4 portrait with a 9:16 still — the tall, phone-shaped frame that shows up whenever someone grabs a vertical reference.

At two columns, 3.4 inches wide, a 9:16 image is 6.04 inches tall. The usable height is 5.25. It doesn't fit. The page is 6.75 inches tall, so the image does fit on the page, with 0.355 inches of clearance above and below — but only if it extends past the top and bottom margins, which the text column beside it does not.

You have three options and they are not equivalent.

Crop it to the box. To fit a 3.4-by-5.25-inch space, the image ratio has to become about 0.648. From 0.5625, that means keeping about 87% of the frame's height — you lose roughly a seventh of the picture, top and bottom. On a wide establishing shot that's nothing. On a face composed with the eyes high and the hands at the bottom edge, it's the difference between the frame you chose and a different one.

Shrink it to one column. At 1.625 inches wide, the image is 2.89 inches tall and fits comfortably. It's also about 13% of the page width, which is a thumbnail rather than a portrait.

Or let it extend into the top and bottom margins, keeping its ratio and its full frame, with the text column beside it holding the margin as usual.

The third option is an exception, and it should be treated as one. An exception is a decision you make for one page because the meaning requires it. If you find yourself making it on four pages, you don't have exceptions; you have a second template you haven't admitted to. Two image formats with two different spatial treatments is a legitimate answer — many decks need exactly that. Repeated improvisation is not.

Record the rules you actually kept, in a form short enough to remember:

The left column line and the top edge are the same on every page.

Continuous prose runs at three columns, about 68 characters. One column is for labels, never for reading. Four or more is for nothing you expect anyone to read.

Images take whole columns and keep their aspect ratio. The right edge may land on different column lines from page to page. The left edge may not.

Gaps are 0.15 inches — between columns, and between an image and its caption.

Tops align across a page; bottoms vary unless a page has a reason to line up along the bottom instead.

Body text sits on a 15-point baseline grid. Captions and fact rows use a 10-point frame grid and rejoin the document grid every 30 points. Headings aren't pinned to either.

An image whose ratio cannot fit the height at a workable width may extend into the top and bottom margins, once, when cropping it would misrepresent the frame.

Six rules, and only one of them is an exception. That's the shape a usable system has: mostly things you don't think about, plus a small allowance for the fact that films come in more shapes than pages do.

None of this has been built or looked at. The arithmetic holds, and it holds whether or not you set the type in the application you had in mind. But numbers can't tell you that a caption reads as belonging to the wrong image, or that a synopsis at 11 point on a laptop screen makes people lean in. That check is the part no calculation replaces. Build the four pages with your own copy and images you're cleared to use, open them at the size they'll actually be read, flip between them, and change the grid where it fails. When it fails twice in the same place, that's the grid's problem. When it fails once, on the page where the frame needed the room, that's the system working.

Frequently asked questions

Why not derive a film-deck grid from the cover page?

The cover has the least content and the most freedom, so it looks balanced at almost any column count. Deriving the grid from it produces a system calibrated for the page that needs it least.

What four kinds of content should be gathered before drawing the grid?

The longest continuous passage at its real length, a portrait with its description, a wide reference image, and short practical facts such as runtime, format, language, company, contact, and festival history. Keep captions and qualifiers attached.

In the article's example, how do measure and capacity differ?

Measure is how many characters fit on a line; capacity is how many lines stack up in the height. A three-column measure at 11 on 15 is about 68 characters wide and 25 lines tall. It does not hold 75 lines, because lines run down the page one after another.

What does the example say about fitting a 600-word synopsis?

At 68 characters per line, about 3,900 characters become 58 lines, while the page has 25 lines and holds about 260 words. Even full width holds only about 525 words with an unwieldy line. A 600-word synopsis is a two-page problem; two facing pages hold about 520 words, so trim about 75 words or give it a third page.

When may an image extend into the top and bottom margins?

When its ratio cannot fit the height at a workable width and cropping would misrepresent the frame. Treat it as an exception for one page. The 9:16 example at two columns is 6.04 inches tall, taller than the 5.25-inch usable height, but it can fit on the 6.75-inch page with 0.355 inches clearance if it extends past those margins.