SwitchPDF
All articles
Optimize June 11, 2026 4 min read

Why Your Scanned PDF Won't Compress (And How to Fix It)

You ran a scanned PDF through compression and got almost no size reduction. Here's why — and a 3-step workflow that actually works.

Why Your Scanned PDF Won't Compress (And How to Fix It)

You ran a 100 MB scanned PDF through a compressor and got back a 96 MB file. Almost no reduction. What's going on?

The culprit: how scanners save PDFs

Most scanners save each page as a lossless TIFF or PNG image wrapped in a PDF. Lossless means already-uncompressed. When you run a "compress PDF" tool over the result, the tool's first move is usually to recompress the embedded images — but they were already saved without compression artifacts that another pass can squeeze out.

Even worse: many scanners save at 600 DPI by default. That's print-shop resolution. For a document you'll only view on screen, 600 DPI is 4× more pixel data than you need.

The 3-step fix

Step 1: OCR the document first

Run SwitchPDF OCR PDF on the scan. This adds a text layer behind the image, making the document searchable. It also normalizes the page format in a way that downstream compression handles better.

Step 2: Compress at Medium

Now run SwitchPDF Compress PDF at the Medium setting. Medium downsamples images to 150 DPI — which is sharp enough for screen viewing but a quarter of 600 DPI's pixel count. You'll typically see a 60–75% size reduction.

Step 3: If still too big, try High

If Medium isn't enough, run again at High (72 DPI). For text-heavy scans this is still readable on screen. For documents with detailed images or fine print, stop at Medium — High will blur them.

Realistic before/after

A typical 100 MB scanned 50-page report:

StepFile size
Original (scanner output, 600 DPI lossless)100 MB
After OCR98 MB (slight overhead from text layer)
After Medium compression22 MB
After High compression9 MB

When compression still fails

If your scanned document was already low-resolution (150 DPI or below) when it left the scanner, there's not much to compress. The only option then is to remove pages with Remove Pages or split into smaller files with Split PDF.

Bottom line

Scanner-produced PDFs need a one-two punch: OCR first (which normalizes the file structure), then compress. Skip step 1 and you'll get the disappointing "barely smaller" result.

Related articles