Text layer: what it is and what it is for

In 30 seconds

The text layer is the real text a PDF stores in addition to the drawing of each page.
Thanks to it you can search, select and copy; without it the page is just an image.
A scan does not include one, but OCR can add it on top of the image.
Flattening or converting to grayscale turns pages into images and removes it.

1What is the text layer of a PDF?

The text layer is the actual text content a PDF keeps in addition to its appearance. When you select a word with the cursor, copy a paragraph or search for a phrase with the viewer's find shortcut, you are using that layer. If it does not exist, what you see is pixels: letters that are drawn but that the program does not recognize as text.

What it is for is easy to see day to day: finding a figure in a fifty-page contract, copying a number into a spreadsheet, letting a screen reader read aloud, or having a file search tool index the content. All of that depends on the text existing as text.

2How does the text layer work under the hood?

In a PDF created from a program, each letter is stored as a character code with a position and a font. So that copying produces the right letter instead of a strange symbol, the file includes a table that translates each drawn glyph into the character it represents. If that table is missing or badly made, the text looks fine but odd characters show up when you copy it.

In a scan, OCR analyzes the image, recognizes the letters and places them as invisible text right on top of each word. The eye keeps seeing the original photo, but the cursor can select and search. If the recognition gets a letter wrong, that wrong letter is the one that gets copied, even though it looks correct on screen.

3What types of PDF exist based on their text layer?

There are four common situations. A native PDF has a complete, reliable text layer. A pure scan has none. A scan with OCR, sometimes called a searchable PDF, mixes image and invisible text. And there is a deceptive case: design documents where the text was converted to outlines or curves, which look like perfect text but have no layer at all.

There are also mixed documents, with native pages and scanned pages in the same file, typical of case files where copies were attached. In them, search works on some pages and fails on others, which is confusing until you check page by page.

4When does the text layer matter, and when is it better to go without?

It matters almost always: for searching, quoting, reusing content, converting to Word, accessibility and indexing. A file without it forces you to reread or retype by hand, and any later conversion gives worse results.

There is one situation where going without it is desirable: when you want certain information to be impossible to extract. There is a classic mistake here: covering a piece of data with a black rectangle does not remove the text underneath, and it can still be copied. To truly hide content you have to delete it, not cover it, and the right tool is Redact PDF.

5Common mistakes and myths about the text layer

A common myth: if I can see the text, I can copy it. False: a scan shows perfectly legible text and does not let you copy a single letter. Another myth is that OCR is flawless; its accuracy depends on sharpness, resolution, contrast and typeface, and handwriting is usually recognized poorly.

It is also believed that if strange symbols come out when you copy, the document is damaged. In reality it is usually a problem with the fonts' translation table, not the whole file, and it can be worked around by running the document through OCR.

6Common problems and how to fix them

If search finds nothing, first check whether you can select a word; if you cannot, apply OCR. If the text comes out scrambled when you copy, it is usually because of columns, tables or floating boxes, and it helps to copy block by block or convert to another format. If OCR makes too many mistakes, redo the scan with more sharpness and between 200 and 300 DPI, and straighten the page first.

Another problem is applying OCR to a PDF that already had text: it can duplicate the layer and make searches return repeated matches. Check beforehand whether the file is already searchable.

7Text layer and working in Rayo PDF GO

In Rayo PDF GO, OCR adds the text layer to scans and Extract text gives you the content as plain text, all on your device, without sending the document to any server. It is useful for contracts, identity documents or case files that you do not want to upload to an outside service.

Keep in mind which operations remove it: Flatten PDF and Convert to grayscale turn each page into an image, and the result looks the same but can no longer be searched. If you need it afterward, apply OCR to the result. And as a rule, always keep a copy with text before flattening.

Related tools

Frequently asked questions

How do I check whether my PDF has a text layer?

Try selecting a word with the cursor or searching for a phrase with the viewer's search. If it highlights or turns up, it has one; if not, the page is an image.

Why do strange symbols come out when I copy text from a PDF?

It is usually because the file's fonts do not include the table that translates each glyph into its character. The document looks fine, but the copied text comes out wrong. Running the PDF through OCR usually fixes it.

Does OCR change how the document looks?

No. It adds the text invisibly over the image, so the page looks the same.

Does covering data with a black rectangle remove it?

No. The text is still underneath and can be copied. To truly remove it, use a permanent redaction tool.

Can I get the text layer back after flattening a PDF?

Yes, by applying OCR to the result, although the text will be recognized again and may contain errors. It is better to keep an earlier copy.

Keep learning

See the whole knowledge base ›