OCR problems

Overview

OCR turns images and scan-heavy PDFs into searchable text. When extraction fails or looks garbled, the usual fix is a clearer source file — not forcing the same photo again.

Why this feature exists

Readable text powers search, Timeline suggestions, and Ask Sera’s understanding of your materials.

Symptoms

• Document shows little or no extracted text
• Names and dates look scrambled
• Processing completes but search cannot find known phrases
• Message that OCR produced insufficient readable content

Likely causes

• Blurry, dark, or angled phone photos
• Handwriting-heavy pages
• Low-resolution scans
• Complex multi-column layouts
• Encrypted or image-only PDFs of poor quality

Step-by-step

1. 1

   Re-scan with better lighting

   Place the page flat, fill the frame, and avoid glare. Prefer a multi-page PDF when possible.

2. 2

   Upload a text-based PDF if you have one

   Native digital PDFs usually extract more accurately than photographs of paper.

3. 3

   Rotate and crop before upload

   Upright, tightly cropped pages OCR more reliably than tilted full-desk photos.

4. 4

   Spot-check then ask Sera

   Open the document, confirm key facts, and use Ask Sera only after the text looks reasonable.

5. 5

   Add Timeline notes manually when needed

   If a critical handwritten note will not OCR, create the Timeline event yourself and link the file.

Best practices

• Verify court dates and money amounts against the original image.
• Keep originals until you confirm extraction quality.

Productivity tips

• Batch clear scans overnight so morning review focuses on Insights, not re-uploads.

Common mistakes

• Assuming OCR equals legal accuracy without checking the source.
• Uploading many near-identical blurry shots instead of one good scan.

When to contact support

Contact support if high-quality text PDFs still produce empty extraction after re-upload while other files process normally.

Frequently asked questions

Screenshots

Helpful links

• Add Documents
• Ask Sera
• Timeline

Related articles

• How accurate is OCR?

  OCR works best on clear, upright, high-contrast scans. Blurry photos and handwriting reduce accuracy.

• Upload failures

  Recover when Add Documents fails validation, stalls, or reports a failed batch.

• Search not finding files

  Improve search hits when filenames, OCR text, or case context do not match your query.

• Slow processing

  Understand why organization takes time and what you can do while Sera works.