Why Scanned PDFs Are So Large — And How to Shrink Them
A ten-page digital report weighs 300 KB. A ten-page scan of the same report can weigh 40 MB. Nothing about the content changed — only the way it is stored.
Understanding that difference is the key to fixing it, because the settings that make scans huge are also the settings you control at scan time.
A scan is a photograph of paper
When a scanner 'creates a PDF', it is really taking a picture of each page and placing that picture in a PDF container. The PDF has no text in it, no fonts, no vector lines — just images, with perhaps an invisible text layer if OCR was enabled.
Image size is a matter of arithmetic. An A4 page at 300 dpi is roughly 2,480 × 3,508 pixels — about 8.7 million pixels. Stored uncompressed in colour, that is around 26 MB per page. Left uncompressed, a 40-page scan would be over a gigabyte.
What your scanner is actually choosing for you
Compression inside the scanner software is the other half of the story. JPEG compression is what brings a 26 MB page down to 1–3 MB, and the quality setting chosen there decides how the scan looks forever.
| Setting | Effect on size | When it is worth it |
|---|---|---|
| Resolution 600 dpi | 4× the pixels of 300 dpi | Rarely — fine for archival, overkill for reading |
| Resolution 300 dpi | Baseline | The right default for text documents |
| Colour | Largest | Signatures, stamps, colour charts |
| Greyscale | Roughly a third of colour | Most text documents, forms, contracts |
| Bitonal (black & white) | Tiny — often 1/10th of colour | Clean printed text with no colour marks |
How to shrink scans you already have
- Compress the PDF — re-encodes the page images at a sensible quality level. Scans typically drop 60–90% with no loss of legibility.
- Crop the black borders. Scanners frequently add a dark frame and a strip of scanner bed; cropping removes real pixels rather than compressing them.
- Convert to greyscale when colour carries no information — for printed documents it usually does not.
- Split the file if it needs to travel by email, rather than pushing compression to a point where text starts breaking up.
Scanning properly the first time
If you control the scan settings, half the problem disappears before the file exists. For standard paperwork — invoices, letters, ID copies, signed forms — 300 dpi greyscale with moderate JPEG compression produces a fully legible page at roughly 300–600 KB. Colour and 600 dpi should be reserved for material where colour and fine detail genuinely matter.
Taking thirty seconds to choose greyscale at scan time saves the entire compress-and-retry cycle later.
A note on searchable scans
OCR adds a text layer on top of the image so the document can be searched and copied from. This layer is small — text is compact — and usually increases file size by only a few percent. Making a scan searchable is almost always worth it; it is not the reason your file is 30 MB.
Frequently asked questions
Is 300 dpi enough for a legal document?
Does cropping a scan reduce quality?
Why is a colour scan so much bigger than greyscale?
Will compressing a scan make the text unreadable?
Tools used in this guide
Compress PDF →
Significantly reduce PDF file size while retaining crystal clear visual quality.
Crop PDF →
Trim margins and crop visual bounds of PDF pages precisely.
Split PDF →
Extract selected page ranges or separate every single page into distinct files.
PDF Analyzer →
Deep document inspection: page dimensions, font usage, image count, and metadata.