PDF / OCR / STUDY NOTES
A huge PDF, a usable notebook: the chapter-by-chapter route to Obsidian
Split a long PDF into focused sections, extract or recognize the text, and organize editable Markdown notes.
1. Diagnose the bottleneck
A large file and a scanned file are different problems. Try selecting text in the original: selectable text can usually be extracted without OCR. Image-only scans require recognition. A damaged text layer may produce missing or garbled output even though you can select text. Keep the original and compare a short test chapter before processing more.
2. Split by topic and keep the order
Use PDF Split & Merge to extract page ranges such as 1-15 or 16-30 into separate PDFs. The organizer accepts up to 20 files, 25 MB per file and 100 MB total, with 500 output pages maximum. It cannot accept a single oversized file or compress images. If a source exceeds 25 MB, reduce it in a trusted editor first. Keep numbered filenames to preserve the chapter order.
3. Extract text first; use OCR where needed
Open PDF to Markdown. Text PDFs can contain up to 200 pages and must be within 25 MB. For scanned pages, select the document language manually and use Fast recognition in the free workflow. Automatic and high-accuracy modes are planned paid options, currently unavailable. If a text layer is broken, enable OCR every PDF page; each processed page then uses the daily OCR allowance.
4. Review before building your vault
Check headings, line breaks and reading order in Source and Preview. Multi-column articles may need manual rearrangement. Fix names and numeric data against the PDF. Tables, images and reliable mathematical LaTeX extraction are not supported. Keep the source PDF alongside your notes so figures and equations remain available for reference.
5. Add focused notes to Obsidian
Download each .md file into a dedicated folder inside your vault. Use a filename such as 01-chapter-title.md, avoid overwriting existing notes, and add a short index note linking the chapters. You can edit the imported Markdown and add your own questions. The site does not sync with Obsidian or automatically generate a chapter index.
6. Keep the workflow economical
Free tools share 10 successful operations and 30 OCR pages per browser each day. Splitting and converting are separate operations. Failed or cancelled conversions do not count. Downloads do not use another operation. For sharing a polished note, use Markdown to HTML / Print PDF and review the print preview. Work chapter by chapter rather than expecting a lossless whole-book conversion.