HowtocompressaPDFtofitanemaillimit

Why scans are so much larger than documents

A PDF exported from a word processor stores text as instructions: this font, this size, these characters, at these positions. A page of it is a few kilobytes.

A scan stores a photograph of a page. At 300 DPI an A4 sheet is roughly 8.7 million pixels, and a scanner that saves it without much compression produces several megabytes per page. Nothing is wrong with the file — it is simply eleven photographs in a wrapper.

What compression trades away

Rebuilding each page as an optimised image is what makes a scanned document drop by 60–90%. The trade is the text layer: after compression the page looks the same but is no longer searchable, because what used to be text is now part of the picture.

That is the right trade for a scan, which never had a text layer to begin with. It is the wrong trade for a report you may need to search later — so keep the original, and treat the compressed copy as something you send rather than something you archive.

Choosing a level

• Medium for anything being read on a screen. This is the setting that clears most upload limits without anyone noticing.
• High when the document will be printed, or when small type has to stay crisp.
• Low only when you are up against a hard cap and legibility is all that matters.

Greyscale is the underrated option. A colour scan of black text carries three channels to describe grey paper; dropping them can take a chunk off the size before quality is touched at all, and it is the right answer for anything heading to an office printer anyway.

When compression makes the file bigger

It happens, and it is not a fault. A PDF that is only text is already about as small as it can be — a handful of kilobytes of instructions. Turning those pages into images replaces a few kilobytes of text with a few hundred kilobytes of picture.

If a text-only PDF is too large for a limit, the size is almost always coming from something else: an embedded font set, a full-page background, or a logo saved at print resolution on every page. Look there rather than compressing harder.

The order that gets the best result

Merge first, compress second. Compressing three files separately and then joining them gives a larger document than joining them and compressing once, because the second approach optimises the whole thing against one budget.

Crop before compressing if a scanner has left a black border — you are otherwise spending bytes describing the edge of the scanner lid.