Two very different kinds of PDF
A native (digital) PDF is created by software, such as exporting from Word, saving a web page or printing to PDF. Its text is real text: characters with fonts and positions.
A scanned (image) PDF is created from paper, by a scanner or a phone camera. Each page is a photograph. What looks like text is really just a picture of text.
They can look identical on screen, but they behave completely differently.
The five-second test
Open the PDF and try to select a single word with your cursor:
- The word highlights: it's native, or a scan with an OCR text layer.
- A box draws across the page instead: it's a scan without a text layer.
For a second check, press Ctrl+F (or Cmd+F) and search for a word you can see. If it isn't found, there's no searchable text.
Other clues
| Clue | Native | Scanned |
|---|---|---|
| Zoom in to 400% | Text stays perfectly sharp | Text gets fuzzy or blocky |
| File size per page | Usually 20–200 KB | Often 0.5–5 MB |
| Page edges | Clean white | Slight shadows, tilt or speckles |
| Copy and paste | Gives the words | Gives nothing, or garbled text |
Why it matters
| Task | Native PDF | Scanned PDF |
|---|---|---|
| PDF to Word, Text, Excel | Works well | Needs OCR first |
| Compress PDF | Small savings | Big savings, often 80–90% |
| Search and copy | Yes | Only with OCR |
| Screen readers | Can read it if tagged | Can't read it |
| Applicant tracking systems for CVs | Can parse it | Usually can't |
| Redact PDF using Find | Finds matches | Draw boxes by hand |
Scans with a hidden text layer
Many modern scanners and phone apps run OCR (optical character recognition) automatically. They keep the page image but add an invisible layer of recognised text behind it. These files pass the selection test and can be searched, but:
- The OCR text may contain errors, especially with numbers (0 and O, 1 and l, 5 and S).
- Converting to Word uses the OCR text, so errors carry through.
- When redacting, the hidden text must be removed as well as the visible image. PDFWisely's Redact tool rebuilds the page, which removes both.
Making the right kind of PDF in the first place
- Sending a document you typed? Export it as PDF from Word or Google Docs. Never print and scan it.
- Sending a CV? Always native. Scanned CVs often can't be read by recruitment software.
- Sending paper documents? Scan at 200–300 DPI if you'll need OCR, or 150 DPI grayscale if you just need a small, readable copy. See How to get accurate OCR.
- Photos of documents? Convert them with JPG to PDF, cropped to the page edges.
Frequently asked questions
Can a PDF be both native and scanned?
Yes. A bundle might combine a typed letter (native) with a scanned signature page. Check pages individually if tools behave differently on different pages.
Why can I select text in my scan?
Many scanners and apps run OCR automatically and add an invisible text layer. The page is still an image, but the text layer makes it searchable.
Is a scanned PDF less legally valid?
No. A scan of a signed document is generally accepted as a copy. The difference is only about what you can do with the file digitally.
How do I turn a scan into an editable document?
Run OCR to recognise the text, then convert to Word. Accuracy depends heavily on scan quality.