SABLELANE PDF GUIDES
Cannot copy text from a scanned PDF? Make it searchable with OCR
Learn when a PDF needs OCR, how to choose a supported language, what credits it uses and how to check recognized text before relying on it.
This guide is available in English and Simplified Chinese.
Visible text is not always selectable text
A scanned PDF may contain only pictures of pages. You can see the letters, but the file has no text layer to search or copy. Try selecting a sentence and searching for a clear word on the page. If the page behaves as one image, OCR may help.
Copying problems can also result from document permissions, font encoding or the reader. OCR is not a solution for every case. Use documents you are authorized to process. If the source can provide a text-based PDF, obtain that version first.
Check the scan before uploading
Check page completeness, orientation and small-text clarity. Blur, glare, folds and skew can cause recognition errors. Rescan poor pages when possible; OCR cannot reliably reconstruct missing strokes.
Sablelane currently accepts up to 50 pages per OCR task and 20 MB per file. You can extract a few representative pages for a trial. Flatten fillable form fields before processing.
Choose a supported language
Sign in, upload the PDF and choose OCR. Select English, Simplified Chinese, Traditional Chinese or an available Chinese–English combination that matches the document. Interface languages and OCR recognition languages are different capabilities.
Check the page count and credit estimate before submitting. OCR costs five credits per input page, so ten pages cost 50 credits. Pages with existing text may be skipped during recognition; do not assume this changes the quoted input-page cost. Download the resulting PDF after completion.
Check the recognized text
Search for a word from the original, copy a short passage into a text editor and check both character accuracy and reading order. Compare important dates, amounts, names and reference numbers against the scan. Pay attention to similar characters such as 0 and O or 1 and I.
The output is still a PDF, with a searchable text layer added to the visible pages. Searchable does not mean error-free. Handwriting, complex layouts and poor scans are not guaranteed to be recognized accurately. Review important documents manually.
OCR is not document reconstruction
OCR does not directly deliver a Word document or Excel workbook. It does not automatically restore table cells or create fillable form fields. If you need another output format, verify the recognized text and check that the next tool supports your document structure.
Test a few representative pages before submitting the whole file. This helps reveal scan-quality, language and layout problems early.