What vector scaling preserves in an N-up PDF

Vector text stays smooth when a PDF page is scaled into an N-up cell. The print still shrinks: in a four-up A4 example, 11 pt becomes about 5.19 pt. A sharp edge and a readable line are different things.

When an imported page is an oddly oriented shipping label, this label orientation and fit guide explains what page rotation data changes.

What the PDF keeps when it is scaled

Many PDFs describe letters, lines and shapes as instructions that a viewer or printer draws at a chosen scale. These are vector elements. They are not stored as a fixed grid of colored pixels for every page. A page may also contain photographs or scanned regions, which are raster images. A single PDF often mixes both kinds of content.

When vector text is scaled, its contours can be drawn cleanly at different zoom levels. That is why a line in a vector diagram can remain sharp on screen after enlargement. A raster picture has a fixed pixel grid; enlarging it reveals those pixels once the viewer has no additional image data. Neither format guarantees that the original content itself is correct or easy to read.

PdfNUp imports pages from the submitted PDF and places them in cells on an A4 output sheet. It does not first turn every full page into a screenshot. The service uses FPDI page import, uses the CropBox and respects page rotation. It preserves vector content when that content is present in the source. Photographs and scanned areas remain image content, and their resolution is not increased by the layout operation.

Small type stays small

On an A4 page placed four-up, an 11 pt line prints at 5.19 pt in the engine’s 99 × 142.5 mm cell. The outline remains smooth under a high screen zoom, but a person reading the paper sees a character less than half its original height. The space available for its strokes has fallen too.

The production grid is tighter than that rough half-size picture: with a 5 mm outer margin and 2 mm gutter, each four-up cell is 99 × 142.5 mm. The calculator fits a 210 × 297 mm portrait A4 page at 47.14%, so 11 pt becomes 5.19 pt. For six-up it chooses landscape, where a portrait A4 page fits at one third scale; 11 pt then prints at 3.67 pt.

Would you read a 3.67 pt table label on paper?

Use the small sheet to locate the label, then check the source.

This is why “vector preserving” describes the way page marks are represented, not a promise about accessibility. A dense table remains dense. A tiny footnote remains tiny. A diagram with thin lines can still become difficult to distinguish after scaling even if no pixelation occurs. Judge legibility at the intended printed size, and use a larger layout or the original file for close reading.

Page fit is proportional. If a source page is unusually wide, tall, rotated or cropped, one dimension may determine how large it can appear in its cell. A group of pages with mixed proportions can leave visible space around some page thumbnails. PdfNUp does not stretch a page to fill the cell if doing so would distort it. The source CropBox and rotation are therefore part of how the output looks.

The four-up A4 scaling guide gives an A4 scale example and suggests a practical test for small labels. A broader walk-through of N-up use cases weighs compact overviews against reduced type.

A scan needs its own text layer

A scanned paper page is usually an image inside a PDF. If that image contains no text layer, importing the page does not create one. PdfNUp does not perform optical character recognition, make a scan searchable, correct a crooked page, improve a faint photocopy, remove marks, or restore a photograph’s detail. It places the page as it is available in the source.

It also does not compress the file as a separate optimization step. The service is focused on page placement. It does not promise a smaller byte size, convert document formats, edit wording, remove or add pages, or reflow a layout. A PDF that looks like selectable text in one viewer may still include parts represented as images; the output follows what the importer can bring through from that source.

For a page with selectable text, vector preservation does not mean that the words are automatically better for a screen reader. Reading order, tags, language metadata and document structure are distinct properties. This service does not add accessibility tagging or rewrite the source’s reading order. If an accessible document is required, inspect the source and use a workflow designed for that requirement.

Why the service refuses some protected files

Some PDFs restrict printing through encryption permissions. PdfNUp checks whether printing is allowed and refuses a file when its policy forbids it. It does not remove a password or bypass the restriction. Encrypted PDFs are refused with a clear error. A damaged or unsupported source also produces an error rather than being presented as a successful result.

Those checks protect the document’s stated constraints and avoid creating a misleading output from an unreadable source. If you have the right to produce a print copy, obtain a version whose permissions allow printing from the document owner. Do not treat a converter as a way around those controls.

The current consumer accepts one PDF up to 25 MB. The service rejects invalid content even if the filename ends with “.pdf”; a filename or browser MIME label is not proof that a file is a valid PDF. It returns a downloadable output on A4 sheets, with a preview available on the page. It lays out imported content; it does not edit the source document.

Check the source and a printed sample

For a useful test, choose one text-heavy page, one diagram and one scanned page if the file contains one. Generate a modest N-up layout. First compare the vector text and lines at the output’s normal page view; then zoom in to see whether crispness is retained. Finally, print a sample at the intended paper settings. These checks answer separate questions: whether the import retained sharp marks, whether the physical size is readable, and whether a scanned image has enough source detail.

Do not judge legibility from a zoomed preview. Compare the same region in the source and output, then print a sheet if people will read it on paper. Existing texture in a scan does not show that vector text was rasterized.

If a diagram’s fine lines merge or the footnotes disappear, lower the number of pages per sheet. If the purpose is to find a page among many, a compact index may still be useful alongside the full document. Keep the roles clear: overview for navigation, original-size page for details that matter.

What the N-up conversion actually gives you

The setting is a whole number from 1 to 64 pages per sheet, with a default of 2. PdfNUp accepts a single PDF up to 25 MB and creates an A4 PDF that can be downloaded and previewed in the page. It keeps source page order, uses each page’s CropBox and rotation, and preserves vector elements where the imported source contains them.

It does not create OCR, compress or edit the source, repair damaged content or override printing restrictions. If an input is encrypted, damaged or does not allow printing, the service refuses it with an explanatory message. Choose the layout with the physical reading task in mind; vector sharpness is useful, but it is not a substitute for adequate size.

For print-run estimates, the worked paper-count table calculates the last partial sheet for 30, 100 and 250 pages. To inspect a real result, prepare a small N-up sample before printing.