Why Is My Scanned PDF Blurry, and Can It Be Fixed?
The document opens and the text is soft, smeared, or so mushy that you are guessing at digits. You need to read it, or send it to someone who will judge you by it.
Before looking for a tool to sharpen it, there is one question that decides whether this is repairable at all — and most advice on this subject skips it entirely.
Was it captured blurry, or made blurry?
These look identical on screen and have opposite answers.
Captured blurry. The scanner or camera never recorded the detail. The information does not exist in the file, and nothing can retrieve it. No tool can sharpen a scan into detail that was never captured, whatever the marketing says — sharpening only increases contrast at edges that are already there, which on a soft scan produces a crunchy, haloed version of the same illegible text.
Made blurry. The original capture was fine and something degraded it afterwards — almost always compression, sometimes a resize, sometimes a screenshot of a screenshot. Here the detail existed at some point, so the fix is to go back and get it.
The test: find the earliest version of the file you still have — the scanner’s output, the original email attachment, the file before you ran anything on it. Open it and look at the same page.
- The old file is sharp → it was made blurry. Recover from that copy and do not repeat whatever softened it.
- The old file is equally soft → it was captured blurry. Skip to the rescan section; everything before it will waste your time.
If it was made blurry: over-compression
This is the common self-inflicted case, and the tools on this site can cause it, which is worth saying plainly.
Compressing a scan works by re-rendering each page as an image at lower resolution. That is genuinely how you shrink a scanned document — but push the quality setting too far and you are throwing away the detail that makes small text legible. The file gets smaller and the document gets worse, and the trade is not reversible.
The fix is to start again from the pre-compression original and use a gentler setting. Compress PDF offers three levels: Higher quality keeps the most detail, Balanced suits ordinary printed text, and Smallest file is where small print starts to suffer. If your compressed copy is unreadable, you chose one level too aggressive.
Check the result at full zoom on the smallest print in the document, not at page-fit on a heading. Page-fit hides exactly the damage you are looking for. And keep the original until you have checked — why is my PDF so large covers what compression does and when it is the wrong operation entirely.
A specific warning: never compress twice. Compressing an already-compressed scan re-renders an image that is already degraded, and the damage compounds.
If it was captured blurry: the four causes
Resolution too low
The most common by far. Scanning at 150 DPI or letting a phone app decide produces character shapes without enough pixels to be distinct — and once the page is captured that way, the letters genuinely are ambiguous rather than merely soft.
300 DPI is the working threshold for ordinary printed text. Below it, quality falls away quickly and text recognition accuracy falls with it. Above it you gain very little for documents and pay a great deal in file size: a letter-size page at 300 DPI greyscale carries about 8.4 million values, while the same page at 600 DPI in colour carries about 101 million — twelve times the data, describing nothing extra on a page of black text.
It is a photograph, not a scan
Phone photographs of documents fail differently from scanner output. The camera focuses on part of the page and lets the rest drift, the paper is not flat, the angle introduces perspective so one edge is smaller than the other, and a hand-held shot in room light picks up motion blur you cannot see on a phone screen.
Phone scanning apps that flatten perspective and correct lighting are markedly better than the plain camera. If you have access to a real scanner or a photocopier with a scan-to-email function, use it.
The wrong colour mode
Many scanners default to a black-and-white mode that decides each pixel is either black or white with nothing in between. On crisp laser print that is fine. On faint text, a photocopy, or anything with grey tone, it destroys the intermediate shades that made the characters readable, producing text that looks broken rather than soft.
Greyscale is the safer default for documents. It keeps the tonal information, works better for recognition, and costs only a third of what colour costs.
The paper was already poor
A third-generation photocopy, a faded thermal receipt, or a document printed with a dying toner cartridge is degraded before it reaches the scanner. Scanning at a higher resolution captures the degradation more accurately; it does not undo it.
What can still be done with a bad scan
If rescanning is impossible — the paper is gone, or someone else produced the file — a soft scan is not necessarily useless.
Make it searchable. Recognition often reads text that a human finds tiring, and even imperfect recognition means you can find pages by searching rather than squinting through all of them. OCR PDF adds an invisible text layer and leaves the picture untouched, so it cannot make the appearance any worse — see how OCR makes a scanned PDF searchable. Expect more errors than on a clean scan, and never trust unchecked recognition for account numbers or amounts.
Straighten it. A page that is also rotated or skewed is harder to read than one that is merely soft, and straightening costs nothing in quality — how to rotate PDF pages.
Crop the noise away. Scanner edge shadows and black borders compete for attention and hurt recognition accuracy. How to crop PDF margins.
Read it larger. Obvious and frequently forgotten: a soft scan at 200% zoom is often perfectly legible, and printing it at a larger paper size can be too.
Be sceptical of anything promising to “enhance” it
Tools advertising AI upscaling or enhancement for documents deserve a moment of caution, because what they do and what people assume they do are different things.
They generate plausible detail. On a photograph of a landscape that is often pleasant and harmless. On a document it is a genuine hazard, because the model is inventing character shapes based on what it expects to see — and a confidently rendered wrong digit in an account number or an amount is far more dangerous than a soft one you knew to squint at.
A blurry scan at least announces its own unreliability. An enhanced one does not, and nothing marks which strokes were captured and which were guessed.
If you use one, treat the output as a reading aid rather than a record, keep the original as the authoritative copy, and never let an enhanced version be the one you file or forward.
Sharp on screen, blurry on paper
Worth separating, because it is a different fault with a different fix.
A scan can display acceptably and print badly, because screens and printers work at different densities. A page captured at 150 DPI has enough pixels to look reasonable on a screen and not enough to print crisply — printing exposes a shortfall that the screen was hiding.
If a document reads well on screen and prints soft, the capture resolution is the cause and the only fix is a better capture. If it prints wrong in other ways — clipped edges, missing content, wrong size — that is a different problem entirely, covered in why won’t my PDF print correctly.
When to rescan, and how
If you still have the paper, rescanning is almost always faster than anything else in this guide and produces a better result than all of them combined.
- 300 DPI, greyscale. Higher only for fine detail you genuinely need to enlarge.
- Clean the glass. Flatbed scanners accumulate dust and smears that appear on every page.
- Flatten the paper. A curled or folded sheet is out of focus wherever it lifts off the glass.
- Use the feeder for volume, the glass for quality. Feeders introduce skew; the glass does not.
- Check the first page before scanning eighty. The single most valuable habit here.
If the paper is gone and the file is your only copy, accept it, run recognition so it is at least searchable, and file it. Chasing sharpness that was never captured is time spent on an outcome that is not available.
Preventing it next time
Everything above reduces to one decision made at the scanner. Set it to 300 DPI greyscale, and the overwhelming majority of blurry-scan problems never happen — along with most oversized-file problems, since the same setting is the one that keeps documents from becoming enormous.
For records intended to last, the archival format is PDF/A, standardised as ISO 19005 — and what PDF/A is and whether you need it explains what conforming to it costs, which requires a document to be self-contained so it renders identically in decades to come. The PDF Association publishes a readable introduction to what the profile involves. Its scan-oriented variant keeps the page image alongside the recognised text — the same two-layer arrangement described above, which is a good sign the arrangement is correct.
The whole capture-to-archive sequence, in the order that works, is set out in how to scan and file paper records.