Open a PDF, try to select a word. Does the cursor highlight a single letter, or does it sweep across the whole page like a frame? That is the difference between an OCR PDF and an image PDF.
Here is what each is and when you need it.
Image PDF
Each page is essentially a photo wrapped in PDF. No selectable text, no searchability. Common output of phone scans and old scanners. Visually identical to a real document but functionally a picture.
OCR PDF
The image has had optical character recognition applied. A text layer sits beneath the image. You can select, search and copy text. The visual stays the same.
When you need OCR
Searchable archives. Copy-and-paste from scans. Form filling on scanned templates. Any time someone hands you a scanned PDF and asks you to find a name in it. Acrobat Pro, ABBYY FineReader and similar tools run heavy-duty OCR.
Where Flint sits
Flint handles editing of already-OCRed PDFs and standard text-layer PDFs. For converting a scanned image PDF to searchable, you typically want a dedicated OCR step before bringing it into the editor.
FAQ
How do I tell if my PDF is OCR'd?
Try to select a word. If it selects the word, OCR'd. If it sweeps the page, image.
Is OCR accurate?
For clean scans, 95%+ accurate. For poor scans, varies. Always proofread important text.
Can Flint OCR a scan?
Flint handles light OCR cases. For bulk scanned archives, dedicated OCR tools are stronger.
Scans for archival: OCR them first. Then edit and sign in Flint.