You cannot edit the text of a scan the way you edit typed text, because a scanned page has no text: it is a picture of a page. Optical character recognition, OCR for short, can read the words in that picture and add them as an invisible text layer, which makes the scan searchable and lets you copy from it. It does not change the picture. “Editing a scanned PDF” therefore means one of three things: a tool that recognizes the text and redraws it as new text in a font it chooses, retyping the page in a word processor, or covering part of the picture and writing over it. Each has a cost. When the document matters, the better fix is a fresh scan of a corrected original, or the source file from whoever made it.

Why can’t I edit the text in a scanned PDF?

Because a scanner, a phone camera or a fax produces an image. The PDF wraps that image in a page, and the page’s only content is “draw this picture here”. There are no characters, no fonts and no lines for an editor to change: the letters are pixels, like the letters in a photo of a street sign.

You can tell in a few seconds:

  • Zoom in. Scanned letters get blurry or jagged at high zoom; real text stays sharp.
  • Try to select a word. If nothing selects, the page is a picture without an OCR layer.
  • If a word selects but the highlight sits slightly off the letters, the page is a picture with an OCR layer on top.
  • List the fonts. pdffonts file.pdf lists no font at all for a document made only of pictures.

Why can’t I edit the text in a PDF? covers the other reasons a PDF refuses edits, from permissions to text converted to outlines.

What does OCR add to a scan?

An invisible copy of the words, placed over the picture where each word appears. The PDF standard has a text rendering mode for exactly this: “Neither fill nor stroke text (invisible).” Viewers do not draw that text, but search, selection, copying and screen readers use it.

The two layers of a scanned page with OCR Below, the picture of the page, which is what you see and print. Above it, the invisible text layer OCR adds, which search and copy use. Changing the hidden layer does not change the picture. OCR layer invisible text for search, copy Picture what you see and print
OCR lays an invisible text layer over the picture. The two are separate: changing one does not change the other.

On a Mac you do not need extra software for this. Apple’s Preview guide says: “Preview can recognize text in an image-based PDF (like a scan or a picture of a document) and embed the text in the document.” Open the PDF, choose File > Export, then select Embed Text. Apple notes that the option appears only when no text was recognized in the document before. Separately, Live Text lets you copy words straight from a picture: “In Preview, you can use Live Text to copy and use the text that appears in a photo.”

OCR makes mistakes. Check the recognized text, especially numbers, names and anything in a small or decorative font; a misread “8” for “3” in the hidden layer means a search for the right number finds nothing.

What does “editing” a scanned PDF really mean?

Since the picture cannot be edited as text, every method replaces something:

  1. Recognize, then redraw. The tool runs OCR and turns the recognized words into real text, drawn in a font it picks to match, in place of the pictured words. Adobe’s help describes this for Acrobat: “Acrobat automatically applies OCR to your document and converts it to an editable PDF copy.” It states that “The new text matches the original font in your scanned PDF.” and adds: “If a font is missing, learn how to add fonts to your computer.” The edited parts are no longer a scan of the original: they are new text made to look like it, and recognition errors become text errors. Adobe advises: “Double-check the text for accuracy before editing.”
  2. Retype the page. Recreate the document in Pages or Word from the scan, then export a new PDF. Slow, but you control every word and font.
  3. Cover and write over. Draw a white box over the wrong part and add a text box on top. The picture under the box is unchanged and the new text sits in another font. This is a visible patch on a copy, and it removes nothing.

Some tools do not edit scans at all. Canva’s help, for example: “Scanned PDFS are imported as a single flattened image, so their text and design elements can’t be edited.”

Which Mac tools work with scanned PDFs?

A factual list by job, with what each vendor documents:

Job Tools on a Mac
Make a scan searchable and copyable Preview, with text recognition on export and Live Text, and PDF editors with an OCR feature
Change the words of a scan Editors whose edit tool works on OCR text, such as Acrobat; check their documentation for scans
Recreate the page Pages, Word or any word processor, then export to PDF
Change a typed PDF that is not a scan PDF editors that change existing text, PDF Textfix among them

Preview cannot change the text of any PDF, scanned or not; Apple’s guide says: “Although you can’t edit the text of a PDF in Preview, you can use the markup tools to provide feedback or write down something you want to remember.” PDF Textfix and Adobe Acrobat compares the two on typed documents.

What does PDF Textfix do with a scanned page?

It does not edit it. PDF Textfix changes text that is already in the PDF, one line at a time, in the fonts inside the document, and a scan has neither text nor fonts. When a page is a picture, the Lines pane says “This page is a scanned image” and “PDF Textfix edits text that is already in the PDF.” When a scan carries an OCR layer, the hidden text is locked, and the line says why: “This is hidden text, such as the text layer of a scan, so it cannot be edited.” The app does not run OCR.

A document can mix both: typed pages with a scanned signature page, or a typed letter with a scanned attachment. PDF Textfix edits the typed lines and leaves the scanned pages as they are. Hidden text in a PDF explains how to find an OCR layer you did not know was there.

When should I rescan or ask for the source?

  • You have the original file. Edit the Word, Pages or InDesign document and export a new PDF. The source file gives the cleanest result.
  • Someone else made the document. Ask them for a corrected file or the source. For a document they issued, only they can correct it.
  • You have the paper original. Correct it, or print a corrected version, and scan it again. The new scan is honest about what it is.
  • The text is short. Retyping a one-page letter takes less time than repairing its OCR.

Questions people ask about editing scanned PDFs

How do I know if my PDF is scanned?

Zoom in and try to select a word. Blurry letters and nothing to select mean a picture without OCR; a selection that sits slightly off the letters means a picture with an OCR layer.

Why can I select text in my scan but not edit it?

You are selecting the invisible OCR layer, not the picture. Editors that change typed text either lock that layer or ask to recognize the page again.

Can OCR keep the original font of a scan?

A scan holds no font, only pictures of letters. A tool that redraws recognized text picks a font to match, and for a font it does not have, Adobe’s own help says to add it to your computer.

Does PDF Textfix do OCR?

No. It edits text that is already in the PDF. Use Preview’s text recognition to make a scan searchable.