When OCR is worth it
A scanned contract, an old letter or a receipt photographed with a phone is just a picture inside a PDF: you cannot search it, copy a sentence from it or let a screen reader read it aloud. After OCR the file behaves like a normal digital document while still looking like the original paper.
Tips for the best result
- Scan at 300 DPI if your scanner lets you choose; 200 DPI is the minimum for small print.
- Keep pages straight and evenly lit; a crooked scan can be levelled first with Deskew PDF. Shadows and folds are the most common cause of missed words.
- Choose the document's language. If English and Turkish are mixed, "English + Turkish" reads both.
When the result is not right
- Letters are swapped or missing. Look at the scan at full size. If you cannot read it comfortably yourself, OCR will struggle too; scan again with more light and the page lying flat.
- Accented letters come out plain. "é" turning into "e", or Turkish "ş" into "s", usually means the wrong language was chosen. Run the file again with the matching language.
- Handwriting is ignored. OCR is built for printed text. Neat block letters sometimes work, but signatures, margin notes and cursive writing are not recognised reliably.
- Copied tables lose their columns. The recognised words sit exactly where they appear on the page, so searching works. When you copy them out, though, columns can run together line by line; check tables after converting to Word.
- A long scan takes minutes. Recognition runs on your own processor, page by page. Keep the tab open, or pull out the pages you need first and run OCR only on those.
Checking the result in a minute
You do not need to read the whole file to know whether recognition worked. Open the new PDF and press Ctrl+F (Cmd+F on a Mac), then search for a word you know is on a page — a name, a company, a heading. If the viewer highlights it in the right place, that page has a working text layer. If nothing is found, look at that page: it is usually faint, skewed or photographed at an angle.
Pay extra attention to numbers. A misread letter in a word is easy to spot from context, but a misread digit is not, and look-alike characters are the most common mix-ups: zero and the letter O, one and a lowercase L, five and S, eight and B. When you copy an account number, an amount or a date out of a recognised scan, compare it digit by digit with the image before you use it.
What to do next
If what you have is a photo or screenshot rather than a PDF, Image to text gives you the words directly as text.
Once the PDF is searchable you can convert it to Word to edit the text, extract the text as plain words, or compress it — scans are often large — before sending it by email.
Scanning paper on a Mac? How to scan a document to PDF on a Mac covers your iPhone camera and Image Capture, then brings the scan here for recognition.