It is a genuinely confusing experience: you compress a PDF, check it on your monitor, it looks perfectly sharp, and then it prints noticeably soft. Neither the compression nor your screen check was wrong — screens and printers simply reveal image quality differently, and a compression level that is invisible on one can be visible on the other.
Why screens are more forgiving of compression than print
A screen displays at a relatively low pixel density and viewing distance combination that tends to mask the subtle softening light-to-medium compression introduces. Print reproduces detail at a much higher density and is often viewed close up, which exposes exactly the kind of fine softening that a screen check would never reveal. The same compressed file is genuinely capable of looking fine in one context and soft in the other — this is not inconsistency, it is two different tests with two different sensitivities.
Steps for compressing a document you know will be printed
- Use a lighter compression level than you might for a purely on-screen document, since print has less tolerance for softening.
- If possible, do a physical test print of one representative page before committing to the whole document at a given setting.
- Check the test print at the actual distance it will be read from, not right up close beyond how it will normally be viewed.
Why checking only on screen is not a reliable print test
If a document's final use is print, checking it only on a monitor is testing the wrong thing — it tells you how the file will look on a screen, which may not be its actual destination at all. For any document where a physical printed copy matters, a real test print is the only way to genuinely confirm the compression level is appropriate, since no amount of screen zooming reliably simulates how print reproduces fine detail.
The specific content most likely to show the difference
- Small text, particularly footnotes or fine print, which screens render with anti-aliasing that print does not replicate the same way.
- Detailed diagrams or technical drawings with fine lines close together.
- Photographs with subtle gradients, where compression artefacts are more visible under print's higher effective resolution.
Choosing compression levels differently for print versus screen versions
If the same document needs both a screen-friendly small file and a print-ready version, it is often better to maintain two separate compressed versions rather than trying to find one setting that serves both equally well. Compress more aggressively for the version that will only ever be viewed on screen, and use a lighter setting, prioritising quality over size, for the version intended for print.
Why resolution matters alongside compression level
Compression level is only one factor; the underlying resolution of the scanned or embedded images matters just as much for print quality, covered in more detail in our guide on resizing images for print versus screen. A document scanned at too low a resolution to begin with will print poorly regardless of compression level, since compression cannot add detail that was never captured.
A practical rule of thumb
If you know a document will be printed and the print quality matters, default to a light compression setting and verify with an actual test print rather than a screen check. If the document will only ever be read on screen, a stronger compression level is usually completely safe, since the screen genuinely will not reveal the difference the way print would.
What to do when you genuinely do not know if it will be printed
Not every document's eventual use is known at the time you compress it — a report might be read on screen by most people but printed by one person who wants a physical copy for a meeting. When the destination is genuinely uncertain, defaulting to a lighter compression setting than you might otherwise use is a reasonable hedge, since it costs relatively little extra file size while protecting against the case where the document does end up printed.
A note on how printers themselves vary
Beyond the file's own compression and resolution, different printers reproduce the same file with genuinely different results — a well-maintained, higher-quality office printer can make a moderately compressed document look perfectly acceptable, while an older or lower-quality printer can make even a lightly compressed document look rougher than expected. If a print came out looking worse than you anticipated, it is worth considering whether the printer itself, not just the file's compression, played a role before assuming the compression settings need adjusting.
Frequently asked questions
Is there a compression level that is always safe for both screen and print?
Light compression is the safest general choice when print quality matters and you are unsure; it has the least visible effect in either context.
Does this apply to compressing a standalone image the same way it applies to a PDF?
Yes, the same screen-versus-print sensitivity difference applies to any compressed image, not just one embedded in a PDF.
Can I tell from the file size alone whether a PDF will print well?
Not reliably; file size reflects overall compression and content, but print quality depends on the specific resolution and compression level applied to each embedded image.
Is a physical test print always necessary?
For anything where print quality genuinely matters, yes, it is the only reliable check; for a low-stakes print, skipping it is a reasonable risk to take.
Does printer quality also affect this, separate from the PDF itself?
Yes, a lower-quality printer can make even a well-prepared file look softer than expected, which is a separate factor from the compression level of the file itself.
Should I always keep a separate print-ready master?
For any document you may need to print in future, keeping a lightly compressed or uncompressed master alongside a smaller screen-optimised copy is a sensible habit.
Is the tool free and private?
Yes. No sign-up, no watermark, and files are removed from the server automatically about an hour after processing.