Clean up a scanned PDF

What gets removed

• Grey and beige paper casts, replaced with white.
• Uneven illumination — the bright centre and dark corners typical of phone scans.
• Speckle and dust: isolated dark pixels sitting on light paper.
• Faint show-through from text printed on the other side of a thin page.
• Shadow gradients cast by a book spine or the edge of a document feeder.

What is deliberately kept

Cleaning is only useful if it leaves the document intact. Two safeguards do most of that work.

First, the despeckle filter is only applied to pixels on the paper side of the histogram, so it never rounds the corners of small text. Second, the tone curve is constrained so the white point can never fall onto the measured ink level — a page whose text is only slightly darker than its paper gets darkened rather than erased.

Need only one of these jobs?

This page does them together and darkens the ink as well. If you want just one, with the ink left alone:

• Whiten PDF background turns grey or cream paper white.
• Remove scan shadows flattens uneven lighting.
• Remove PDF noise clears isolated specks.

Cleaning also shrinks the file

A grey, noisy background is expensive: every speck is detail that JPEG compression has to encode. Flattening the background to a uniform white gives the compressor almost nothing to store there, which is why cleaned scans are frequently smaller than the originals despite being re-rendered at print resolution.