You converted a JPG or image to PDF. Looks fine. Try to select text — nothing happens. Try to search — no results.
What's actually going wrong
Conversions from images (JPG, PNG, scans) produce image-only PDFs by default. The 'text' in the PDF is part of the embedded image — there's no separate text layer for selection or search.
To make image-based PDFs searchable, you need OCR — character recognition that reads the visible text and adds an invisible text layer underneath.
The quick fix
Run the converted PDF through convert PDF to Word. Flint's OCR reads the text and produces a Word document. Open in Word, verify recognition is good, export back to PDF.
The new PDF has a real text layer — selectable, searchable, copyable. The visible content is identical to the original.
If that didn't work
If recognition is poor (low-resolution source), the OCR text will have errors. Proofread in Word before re-exporting.
For PDFs converted from clean digital sources (Word documents that became PDFs), the original conversion should have produced a text layer. If it didn't, redo the conversion via Flint — most digital-to-PDF conversions preserve text natively.
Prevent it next time
Convert from sources with text layers (Word, Excel, PowerPoint) and the PDF inherits the text. Convert from images (JPG, PNG) and you need OCR to add a layer. Plan accordingly.
FAQ
How can I tell if my PDF has a text layer?
Try selecting text. If selection works and highlights words individually, you have a text layer. If selection doesn't work or selects pages as blocks, you don't.
Does OCR change visible content?
No — the visible content stays identical. OCR adds an invisible text layer behind the image. Selection and search work without changing what users see.
Why doesn't Flint always OCR by default?
OCR adds processing time and isn't always needed. For digital-source PDFs, native text layers already exist. Flint OCRs when converting from images (PDF-to-Word) or on demand.
Can I make any PDF searchable?
Yes — any PDF can have OCR applied to add a text layer. Quality depends on source legibility; for crisp sources, accuracy is high.
Make image PDFs searchable with Flint's OCR — add text, keep visuals.