How to compress a PDF without wrecking the quality
Compression is a trade, never a free win. Understanding what gets thrown away lets you decide how far to push it.
Last updated September 2026
Practical how-to · 7 min read
Why the file got big in the first place
- Scans at 600 DPI. A scanner default of 600 DPI produces roughly four times the data of 300 DPI with no visible benefit on screen.
- Photos placed at full camera resolution. A 12-megapixel image shown two inches wide is carrying about 40 times the pixels it can display.
- Embedded font sets. Full CJK or multi-weight families can add megabytes on their own.
- Layered exports from design software, which can keep hidden objects and unused colour profiles.
What compression actually does
There are two families. Lossless compression repacks the data more efficiently — nothing is discarded, and the savings are typically modest, in the 5–20% range. Lossy compression permanently reduces image resolution and increases JPEG compression. That is where the real savings live, and it is irreversible.
Be aware of a third approach some tools use, including ours: rasterising. The page is rendered to an image and rebuilt as a PDF. This shrinks complicated files dramatically and reliably, but it destroys the text layer, so the result is no longer searchable or selectable. Keep your original whenever you use it.
Sensible resolution targets
- Screen and email only: 100–150 DPI is plenty and looks fine on any display.
- Office printing: 200–300 DPI. Below 200 DPI, small text starts to look soft on paper.
- Professional print: 300 DPI minimum, and do not compress images aggressively at all.
- Archival or legal record: keep the uncompressed original as your master copy, always.
The limits you are usually fighting
Most email providers cap attachments around 20–25 MB, and many corporate gateways are stricter at 10 MB. Government and university upload forms often sit between 2 MB and 5 MB. Aim comfortably under the limit rather than at it, because encoding an attachment for email inflates it by roughly a third.
A workflow that avoids regret
- Keep a copy of the original before you compress anything.
- Compress once, from the original. Repeated compression stacks artefacts.
- Open the result and read the smallest text on the busiest page before you send it.
- If quality is unacceptable, go back to the original rather than compressing a compressed file.
- If the file is huge only because it has 200 pages you do not need, split or delete pages instead — that costs nothing in quality.