On this page
Identify the page structure first
Try selecting text and inspect whether the editor finds bounded text runs. Native text usually provides coordinates and font information; a scan provides pixels and may have no searchable layer.
Mixed PDFs can contain native pages, scanned pages and OCR overlays in the same file, so make the decision per page.
Treat OCR as assisted recognition
OCR estimates characters and reading order from pixels. Recognition quality depends on resolution, skew, language, contrast and layout complexity.
An OCR-assisted edit should preserve the surrounding scan, avoid duplicating visible text unnecessarily, and remain labeled as OCR-assisted rather than claiming native-font fidelity.
- Review names, numbers and punctuation.
- Check columns and reading order.
- Keep the original scan available.
- Verify the final searchable layer.
Choose the workflow that matches the source
Do not silently convert a native document into page screenshots to avoid a difficult edit.
| Source | Recommended workflow | Primary check |
|---|---|---|
| Selectable native text | Direct text replacement | Font, baseline and old-text removal |
| Image-only scan | OCR-assisted editing | Recognition accuracy and visual repair |
| Outlined lettering | Add text or rebuild carefully | Visual alignment and truthful labeling |
Put the workflow into practice
Edit supported PDF objects, preserve the original until export, then verify the edited result before downloading or continuing to another workflow.
Try Edit PDFFrequently asked questions
Clear answers before you edit
Why can I see text that the editor cannot select?
The page may contain a scan or vector outlines instead of native PDF text objects.
Is OCR text identical to the original font?
No. OCR recognizes characters; matching the original visual typeface is a separate approximation.
Does OCR change the whole scan?
A careful OCR workflow keeps the source image and adds or edits a distinct recognized layer.