You highlight a sentence. Right-click. Copy. Paste into your doc — and nothing arrives. Or only a single weird character. Or the paste pastes a tab and that's it.
The PDF is either restricted, image-only, or has badly encoded text.
What's actually going wrong
Copy restriction. The PDF's permissions block copying. Image-only. There's no real text to copy — only pictures of text. Bad text encoding. The text is real but its character map is broken, so copy produces garbage.
Quick test: can you select the text at all? If selection works but paste produces nothing or gibberish, it's encoding or restriction. If selection doesn't work at all, it's image-only.
The quick fix
For image-only or badly-encoded PDFs, run convert PDF to Word. Flint's OCR reads the displayed text directly and produces a clean Word document. Copy from there.
For permission-restricted PDFs you own, unlock the PDF first, then copy normally. For PDFs with valid restrictions that aren't yours to override, request the content from the sender.
If that didn't work
If OCR produces errors on specific words, those are likely the harder-to-recognise items. Open the Word output, fix manually. Even a partly-imperfect OCR beats retyping a long document from scratch.
For PDFs where the text encoding is completely broken (rare but happens), the only reliable route is OCR. Don't bother trying different viewers — the encoding is in the file, not the renderer.
Prevent it next time
When creating PDFs you'll work with, export with text-as-vector rather than text-as-image. Don't enable copy restrictions unless you specifically need them. And when receiving PDFs you'll need to extract from, ask for source documents alongside.
FAQ
Why does copy from PDF paste as a tab character or symbol?
The PDF's character map is broken — the file shows recognisable letters but copies meaningless codes. OCR via Flint rebuilds the text layer from the visual output and produces clean copyable text.
Can I bypass a copy restriction in a PDF?
OCR works regardless of restrictions because it reads the visual output. You're not bypassing encryption — you're re-deriving the text from what's visible. Legally varies by file and jurisdiction.
Why does some text copy fine but other text doesn't?
Mixed-source PDFs have vector text in some places and image text in others. Vector text copies; image text doesn't. OCR fixes the image portions.
Is OCR perfect?
Very good on crisp source, imperfect on faded or low-resolution scans. Proofread for important content. For most office documents, accuracy is over 99%.
If copy fails, OCR succeeds. Convert PDF to Word in Flint and the text comes out clean and copyable.