Compress a scanned PDF without losing readability
Why is a scan so much bigger than a typed document?
When you export a document from a word processor, each character is stored as text and costs almost nothing. A scanner or phone camera does something completely different: it records the whole sheet as a grid of pixels, including the paper texture, the shadow at the fold and every speck of dust.
Colour, high resolution and grey backgrounds multiply the size further, so a handful of scanned pages can outweigh a hundred typed ones. The good news is that this kind of file responds better to compression than any other.
Does your scan have hidden text?
Open the file and try to select a single word. If you cannot, or the selection covers the whole page as one block, the PDF is images only.
If you can pick out individual words, your scanner or scanning app added a layer of recognised text (OCR) behind the images. That layer is what lets you search the document or copy a paragraph out of it. Keep it in mind: one of the options described below removes it.
Compression level or target size?
Compress PDF offers two modes. With Target size you say how big the file may be and the tool tries quality settings until it fits. With Compression level you choose low, medium or high depending on whether sharpness or savings matter more.
| Setting | Good for | Trade-off |
|---|---|---|
| Low compression | Small print, detailed stamps, handwriting | Smaller savings |
| Medium compression | A first attempt on almost any scan | Changes are rarely visible |
| High compression | Getting the smallest possible file | Images may look softer |
| Target size | Meeting a stated maximum | Not every file can reach every target |
From heavy scan to lighter file
All of this runs in your browser, so a scanned lease, a medical letter or a bank statement never leaves your device.
- Open Compress PDF and add the scan.
- Choose Target size if you have a limit to meet; otherwise choose Compression level and start at medium.
- Press Compress and note how much smaller the file became.
- Open the copy and zoom in on the hardest parts to read: footnotes, stamps, signatures and handwritten notes.
- If anything has become hard to read, step back: a lower level or a larger target.
Trim pages before you compress
Double-sided scanning often produces blank backs, and each blank page weighs as much as a full one. Duplicates and upside-down pages are common too. Removing them in Organize PDF lightens the file before compression even starts.
While you are there, rotate any sideways pages. Rotation does not change the size, but whoever opens the document will thank you.
Still too big? Fix it at the source
Some scans are made at such a high resolution that compression alone cannot bring them down far enough. If you still have the paper, check the scanner settings. A "document" or "text" profile produces far lighter files than "photo", and greyscale or black and white makes a big difference when colour adds nothing. A medium resolution is normally plenty for reading; the maximum mostly adds weight.
If you scan with a phone, look at the quality setting in your scanning app and crop to the edges of the sheet. The table top around the page is just extra pixels.
The last resort is the option "Shrink by converting pages to images". On a scan without an OCR layer you lose almost nothing. If your file does have OCR, though, that layer is thrown away and the document is no longer searchable, so use it only when you will never need to search or copy the text.
Frequently asked questions
- Does compressing a scan remove the OCR text layer?
- Normal compression re-encodes the images and leaves the recognised text in place. Only the option that converts pages to images removes it, and the tool warns you about that before you use it.
- What happens if the target size is not reached?
- You are told how far the file could be reduced instead of getting a false success. Remove unneeded pages, raise the target or rescan with lighter settings, then try again.
- My scan contains an ID number. Can I hide it before sending?
- Yes, but not automatically, because a scan without an OCR text layer has no text to search. Open Redact PDF and draw an area over the number; that page becomes a new image with the area blacked out. Redact first, then compress.