Why a PDF looks different on another computer
Why does my PDF look different on someone else’s computer?
Save your favourite tools
Create a free iBuildPDF account to keep your favourite tools and find them quickly anytime.
Free account. The PDF tools themselves never need one.
Short answer
Almost always because its fonts were not embedded. A PDF can either carry the typefaces it uses inside the file, or merely name them and hope the reading machine has them. When it only names them and the machine does not, the reader substitutes something else — and substituted type is a different width, so the text moves, lines rewrap and the layout you carefully arranged comes apart.
What embedding actually means
A PDF describes text as a sequence of glyph codes plus a font reference. The font reference can take two forms. In the first, the file contains the font program itself — the actual outlines — and any reader can draw the text exactly as intended, on any machine, forever. In the second, the file contains only a description: a name, a weight, a flag saying whether it is serif, and enough metrics for a reader to find something approximate.
What happens when a font is missing
- The document as written, with its font carried inside the file.
- The same document where that font was not embedded. A substitute has different letter widths, so the lines break elsewhere.
- Embedded, or not. This one difference is behind almost every "it looks wrong on their computer".
The second form exists because early PDFs were small for a reason, and because fourteen fonts (Helvetica, Times, Courier and their variants, plus Symbol and ZapfDingbats) were once guaranteed to be present in every reader. That guarantee no longer holds in the way people assume: modern readers substitute for these too, and a substitute is not the original.
Everything else follows from which form your file uses. This is also why the same PDF is stable across a decade of readers and then suddenly is not — nothing changed in the file; a machine that happened to have the font was replaced by one that does not.
Subsetting, and the surprise it causes
Embedding a complete font family into every document would be wasteful, so most producers subset: they embed only the glyphs the document actually uses. A file that uses forty characters of a font carries those forty outlines, not the several thousand in the full family.
Subsetting is good practice and it has one consequence worth knowing. A subset font cannot render characters that were not in the original document. If you or a tool later edits the text and types a character the subset does not contain — an accented letter, a currency symbol, a dash of a different length — that character has no outline to draw. Depending on the editor you get a blank, a box, or a silent substitution back to a non-embedded font for that one character.
Subset fonts are usually visible in a file by their names, which carry a six-letter prefix and a plus sign: ABCDEF+Garamond. If you look at a font list and see prefixes like that, the file is subsetted.
How to check before you send it
The reliable test is not "does it look right on my machine" — your machine is the one place it is guaranteed to look right, because that is where the fonts are installed.
- Look at the document properties. Most desktop readers list the fonts under a Fonts tab, and say for each one whether it is embedded and whether it is a subset. Anything listed without "Embedded" is a risk.
- Open it somewhere without your fonts. A phone, a different operating system, or a browser’s built-in viewer. A browser viewer is a particularly good test because it ships with a deliberately small font set.
- Check the file information. The metadata editor shows what the producing application recorded, which often explains how the file was made and therefore how it was likely to have handled fonts.
Fixing a file whose fonts are missing
Honest answer first: the reliable fix is to re-export the document from whatever created it, with font embedding switched on. Every serious word processor, layout application and print driver has that option, and it is the only approach that keeps the text as text.
If the source document is gone and you only have the PDF, the options are narrower:
- Flatten it. Flatten PDF resolves interactive layers into the page content. This helps with form fields and annotations rendering inconsistently. It does not embed a font that is not there.
- Turn the pages into images. PDF to JPG renders each page as it appears on your machine, with your fonts, and the result looks identical everywhere. The cost is total: the text stops being text. It cannot be searched, selected, copied or read aloud, the file is usually much larger, and the change cannot be undone. This is a last resort for a document that must look exactly right and will never need to be read by a machine — a signed-off proof, a poster — and the wrong choice for almost everything else.
- Repair it. If the layout problem is corruption rather than fonts, repair PDF rebuilds the file structure. It cannot add missing outlines either, but it rules the other cause out.
For documents that must stay readable for years, the standard that exists for exactly this is PDF/A, which makes embedding mandatory. If a file matters in a decade, that is the format to ask your producing application for.
Tools this article covers
Sources
Primary documentation for the claims above.