First find out whether the text is real or a picture: try to select a word. If it selects, the text is sharp and your reader is rendering it badly — a zoom or a different reader fixes it. If it does not select, the page is an image, and no tool can add detail that was never captured.
1Test whether the text is realTry to select a word. Text that highlights word-by-word is vector text and is not blurry in the file — the problem is on screen.
2Zoom to 200% and look againReal text stays crisp at any zoom because it is drawn from outlines. An image gets visibly worse. This distinguishes the two cases in seconds.
3If the text is real, change reader or print itOpen it in a browser, or print it. A file that prints sharply was never blurry — you were looking at a rendering artefact.
4If it is an image, go back to the sourceRe-scan at 300 DPI in grayscale, or ask for the original digital file. This is the only thing that genuinely adds detail back.
5Avoid repeating itWhen you next compress, keep flattening off and check the before-and-after view, which scores what changed on each page.
Why does this happen?
It was scanned at too low a resolution. Below about 200 DPI, printed text loses the edges that make it legible. The detail was never recorded, so it cannot be restored.
It was over-compressed. Pushing a scan to a very small target forces heavy image compression, which softens edges and leaves halos around text.
The pages were flattened. Flattening to hit a size limit turns crisp vector text into a picture of text, at whatever resolution was chosen.
It is a photograph, not a scan. A page photographed by hand is rarely flat, evenly lit or in focus across the whole sheet, and one corner is usually softer than the rest.
The reader is rendering it poorly. Some viewers draw a cached low-resolution version first, or disable smoothing. The file is fine; the display is not.
Questions
Can a blurry scan be sharpened?
Not meaningfully. Sharpening raises local contrast at edges, which makes an image look crisper on screen while inventing detail that was never captured — and on text it usually makes the result harder to read, not easier.
Why did my PDF get blurry after I compressed it?
The target was small enough that pages had to be flattened into images, or the images were recompressed hard. Start from the original and use a larger target — at 1 MB or 2 MB this almost never happens.
What resolution should I scan at to avoid this?
300 DPI for anything that will be printed or read closely, 150 for reading on screen. Grayscale rather than color for text keeps the size down without costing legibility.
The text looks blurry but prints fine. Why?
Because printing uses the vector text, while your screen was showing a low-resolution preview. The file is correct and nothing needs fixing.