How to scan a book to PDF
Use a phone app for a few books, a flatbed for a few delicate ones, or a mail-in service if you don't mind the book being cut. Scan at 300 dpi or more, then clean the whole PDF in one pass.
Key takeaways
- Capture is the quick part. The work most people underestimate comes after: crooked pages, grey paper, dark edges and no searchable text.
- Aim for at least 300 dpi. Tesseract, the OCR engine behind many free tools, says it works best from 300 dpi up.
- Phone apps such as vFlat flatten curved pages and split two-page spreads. A flatbed is slower but kind to fragile books.
- Mail-in services like 1DollarScan cut the spine off and recycle the book unless you pay for an exception.
- Scan everything first, then fix the whole book at once rather than page by page.
Pick your method
There are four realistic ways to turn a printed book into a PDF at home or near it. They differ mostly in what they cost you: money, time, or the book itself.
| Method | Good for | Watch out for |
|---|---|---|
| Phone app | A few books, quickly, anywhere | Uneven light and curved pages near the spine |
| Flatbed scanner | Fragile or valuable books | Slow, and the spine shadow is heavy on thick books |
| Overhead book scanner | Many books, kept bound | The price of the scanner |
| Mail-in service | A shelf of books you don’t need back | The book is cut and usually recycled |
With a phone
A phone camera is now good enough for most books, if the light is right. Apps built for book scanning do more than a general scanner app. vFlat, for instance, says it flattens page curves, removes fingers holding the page and splits two-page spreads. It runs on iOS and Android, and its App Store page lists optional text-recognition packs from $2.49 for 200 pages.
- Find even, soft light. A window on an overcast day beats a desk lamp, which throws a hot spot and a shadow.
- Hold the phone parallel to the pages. Tilt makes the far lines smaller and the page a trapezoid.
- Press the book open gently. A sheet of clear acrylic or your fingertips at the edge, not across the text.
- Shoot both pages of each spread. Let the app split them, rather than shooting each page separately and getting the order wrong.
- Check every 50 pages for a skip. A missing page is far easier to reshoot now than once the book is back on the shelf.
- Export one PDF for the whole book.
With a flatbed
A flatbed gives the most even light and the sharpest image, at the cost of time. Set it to 300 dpi or more. The Tesseract OCR documentation says its engine works best from 300 dpi up.
- Greyscale is enough for text pages. Colour helps only on illustrated pages, and makes the file larger.
- Don’t force the spine flat. Accept a little shadow at the gutter. It crops away later; a cracked spine doesn’t come back.
- Keep the book in the same corner of the glass. Pages then sit in roughly the same place, which makes one crop fit the whole book.
- Close the lid if you can, or cover the book with a dark cloth. An open lid lets in room light and a grey cast.
With a scanning service
Mail-in services do the scanning for you, at a low price per page, by cutting the binding off and feeding loose pages through a sheet scanner. 1DollarScan’s pricing is a fair example: $1.20 per set of 100 pages for basic scanning at 300 dpi in colour, plus $1.00 per set for OCR, with one-time orders taking about 60 days unless you pay for Express. Its FAQ says the book is cut and recycled unless a return exception applies.
That makes services ideal for paperbacks you’d otherwise give away, and wrong for anything you want to keep.
After capture: what every scan needs
Whatever you scanned with, the PDF you end up with has the same faults: margins and dark edges, grey or yellow paper, slightly crooked pages, open-book spreads in some cases, and no text layer. The Tesseract guide points out that dark borders get misread as extra characters and that skewed lines hurt recognition, so clean-up comes before OCR, not after.
Free desktop tools can do it in stages. ScanTailor Advanced splits pages, straightens them and sets margins on Windows or Linux, and then you add text with a separate OCR tool.
Or do it in one upload:
- Upload the PDF to FixMyScan. Or a DjVu file. Up to 1,000 pages.
- Draw one crop. 50 pages are layered on top of each other, so the box fits the whole book, not just page 1.
- Let it fix the rest. Spreads are split into single pages, crooked pages are straightened, paper turns white and text black, and pictures stay greyscale.
- Download a searchable PDF. Every page is in order with its printed page number.
Best suited for: people who would rather spend their evening scanning the next book than tuning the last one.
A few habits that save the most time
- Name the file as you go: author, title, year. You won’t remember which “Scan_0042.pdf” was which.
- Scan the cover and the title page. They identify the edition, which matters if you ever cite it.
- Keep the raw scan until you’ve read through the cleaned copy once.
Questions
What is the easiest way to scan a book to PDF?
A phone scanning app. Hold the book open, photograph each spread, and export a PDF. Apps built for books, such as vFlat, flatten the page curve and split the two pages for you.
How do I scan a book to PDF on my iPhone?
Use a book scanning app from the App Store, or the document scanner built into some note apps. Shoot in even daylight, keep the phone parallel to the page, and export as PDF.
What resolution should I scan a book at?
300 dpi is a sound minimum for text. The Tesseract OCR project says its engine works best on images of at least 300 dpi. 1DollarScan scans at 300 dpi in colour as standard.
Where can I get a book scanned to PDF?
Mail-in services such as 1DollarScan charge $1.20 per 100 pages for basic scanning. They cut the binding off to feed pages through a sheet scanner, so the book isn't returned unless you pay for an exception.
Can I scan a book without cutting it?
Yes. Phone apps, flatbed scanners and overhead book scanners all work with the book bound. Expect a shadow and some curve near the spine, which cropping and straightening deal with afterwards.
Why does my scanned book look grey and crooked?
Paper photographs as grey or yellow, and any tilt in the book or camera shows as a slanted page. Both are fixed after capture: level the paper to white and straighten each page.
How do I make the scanned PDF searchable?
Run OCR on it, which adds hidden text under each page image. Do it after cleaning the pages: OCR reads crisp black text on white far better than grey photographs.