What Is N-up? The Logic of Page Imposition and the Scaling Math

The request "print four pages on one sheet" looks simple, but behind it sit several geometric decisions: how much to shrink the page, how to center it in its cell, what to do when aspect ratios don't match, where the margins come from. This article explains the math of the N-up operation and the print-industry concept behind it: imposition.

Imposition: a concept from the print shop

In the printing industry, imposition is the process that determines the arrangement of pages on a large press sheet. When a book is printed, pages aren't printed one at a time; 8, 16 or 32 pages are printed on large sheets, which are then folded, trimmed and bound. For the pages to come out in the right order once folded, they have to be placed on the sheet in a very specific arrangement.

N-up is the simplest member of that family: it arranges pages in a grid in straight reading order, without any regard for folding. The aim isn't binding but saving paper and compact review.

That distinction has a practical consequence: you can't make a booklet with N-up. For an 8-page booklet the front of the first sheet needs 8 and 1, and the back needs 2 and 7. An N-up tool puts 1 and 2. When you fold it, the order is completely scrambled. If you want a booklet, you need a dedicated booklet imposer.

How the scale factor is calculated

Say you're placing A4 portrait pages (210 × 297 mm) onto an A4 portrait sheet in a 2×2 grid.

First the cell size is found. If no margin is left on the sheet, each cell is 105 × 148.5 mm.

Then the scale factor is calculated. While fitting the page into the cell, the aspect ratio has to be preserved — otherwise the content gets squashed. So each axis is checked separately and the smaller value wins:

• Horizontal scale: 105 / 210 = 0.500
• Vertical scale: 148.5 / 297 = 0.500

In this example the two come out identical, because A4's aspect ratio (√2) has a special property: split an A4 in half and you get A5, with the ratio unchanged. That's why N-up fits perfectly on A-series paper with no wasted space.

Now think about a presentation PDF: the pages are 16:9 landscape. Try placing them 2×2 on an A4 portrait sheet:

• Cell: 105 × 148.5 mm
• Source page ratio: 16:9 ≈ 1.78
• Horizontal scale limit: 105 / (16 units)
• Vertical scale limit: 148.5 / (9 units)

The horizontal axis is far more constraining. The page fits the cell exactly horizontally but uses only half the cell vertically. The result: wide white bands above and below every mini slide, and the slide shrinks needlessly.

That explains why the choice of orientation matters so much. Put the same slides 2×2 on a landscape A4 sheet and the ratios match far better, so the slides come out noticeably larger.

| Source page | Best output orientation (2-up) | Best output orientation (4-up) | |---|---|---| | A4 portrait | Landscape | Portrait | | A4 landscape | Portrait | Landscape | | 16:9 slide | Portrait | Landscape | | 4:3 slide | Portrait | Landscape or portrait | | Square | Doesn't matter | Doesn't matter |

The legibility math

The scale factor translates directly into point size. In N-up the linear reduction is roughly 1/√N:

• 2-up: 1/√2 ≈ 0.71 → 11-point text gives the effect of 7.8 points
• 4-up: 1/2 = 0.50 → 11 points becomes 5.5 points
• 6-up: 1/√6 ≈ 0.41 → 11 points becomes 4.5 points
• 9-up: 1/3 ≈ 0.33 → 11 points becomes 3.7 points

The comfortable lower limit in print is usually around 8-9 points; below 6 points is tiring for most people, and below 5 points is practically unreadable.

It's more useful to read that table backwards: look at your source document's point size, calculate what it becomes after N-up, and decide accordingly. 24-point slides become 12 points at 4-up and read comfortably. A 9-point technical document drops to 6.4 points even at 2-up and strains the eyes.

The effect on file size: why doesn't it shrink?

A common expectation is that when the page count drops to a quarter, the file will shrink too. It doesn't, and the reason has to do with how PDF works.

The N-up operation makes four XObject form calls on the target sheet. Each call says "draw that source page's content at this scale, at this position." The source pages' content — text streams, images, embedded fonts — stays in the file as it was.

Scaling is a transformation matrix, not data reduction. When a 3000-pixel photo is drawn at half scale, it's still a 3000-pixel photo in the file; it's just printed into a smaller area. From a print resolution standpoint that's a good thing (detail is preserved) but it delivers no gain in file size.

There can be a very small gain because the number of page objects drops, but that's at the level of a few kilobytes and goes unnoticed.

If you want to shrink the file, the right tool is compression — there images genuinely get re-encoded at lower resolution and data is discarded.

Page numbers and footers

The numbers, headers and footers on the source pages are part of the page content and get carried along, scaled. The result is that on a 4-up sheet you see four separate page numbers, each in the corner of its own mini page.

Sometimes that's exactly what you want (you know which original page you're looking at), and sometimes it's confusing. If you want a single numbering on the new sheets, you'll need to use a page-numbering tool after N-up — and then each sheet also gets a "sheet number" at the bottom.

Behavior with mixed page sizes

A PDF can contain pages of different sizes: an A3 technical drawing among A4 text pages, say. The N-up operation scales each page into its own cell individually, so technically it works without trouble.

But the visual result is unbalanced: the A3 page has to be scaled down more to fit its cell, so its text looks noticeably smaller than the neighboring A4 page's text. That's a common situation in scanned mixed archives and reports with appendices.

For a cleaner result you'd first normalize the pages to a single size (by running them through an operation that scales everything to A4, for example). On most documents that isn't worth the trouble, but if presentation quality matters it makes a difference.

Where the margins come from

In N-up output you see white areas between the mini pages and along the edges of the sheet. These have two sources:

The source pages' own margins. A Word document's 2.5 cm margin becomes 1.25 cm in N-up but doesn't disappear. Those margins are actually useful, because they stop the mini pages from running into each other.

The separator the tool adds. Some N-up tools put extra space or a separator line between cells. That helps legibility but narrows the usable area, so it lowers the scale factor a bit further.

The printer's unprintable area. Physical printers can't print in a band about 5 mm from the edge. If the N-up output comes very close to that limit, the edges of the mini pages along the border may get cut. Check it in print preview.

In summary

N-up is a geometric operation that scales pages into a grid while preserving their aspect ratio. The scale factor depends on how well the source page's ratio matches the target cell's — which is why it fits perfectly on A-series paper, and why orientation choice makes a big difference with 16:9 slides. Legibility drops with 1/√N, so the source point size should determine N. File size doesn't shrink because scaling discards no data, it just applies a drawing transform. And most importantly: N-up is not booklet imposition; if you want output that will be folded, you need a different tool.