Split two-page scans into single pages
If you scanned the book open, with two pages in every image, FixMyScan finds the fold and cuts each scan into a left page and a right page, in order.
Key takeaways
- An open-book scan puts two pages side by side, so on an e-reader each one shows at less than half the screen width.
- Splitting at the fold turns one wide image into two pages, left first, then right, so the book reads in order.
- FixMyScan finds the spreads and the fold for you, and shows the fold line on the crop screen so you can move it.
- Choose Split, or Split and crop to give left and right pages a box each.
- Books that read right to left come out in the wrong order. Choose Don’t split for those.
Why books get scanned two pages at a time
Laying a book open and photographing both pages at once is the quick way to scan it. It halves the number of shots, it’s easier on the spine than pressing each page flat, and it’s how a lot of phone apps, overhead scanners and library digitisation rigs work.
The result reads fine on a monitor. On an e-reader it doesn’t. A portrait screen showing a landscape image of two pages gives each page well under half the width, so the text is tiny and you end up zooming into the left page, panning to the right, then turning.
What a good split needs to get right
Cutting an image in half sounds simple. On a real book, four things go wrong:
- The fold drifts. The book shifts on the glass, so the gutter is rarely in the exact middle of every image.
- Not every image is a spread. The cover, a fold-out plate or a page scanned alone shouldn’t be cut.
- Some content crosses the fold. A map or a two-page table cut down the middle is worse than no split.
- The order matters. Left page, then right page, then the next spread.
How FixMyScan splits a book
First it decides whether the book was scanned in spreads at all, from a sample of pages across the book rather than page by page, so one odd page doesn’t flip the decision. A wide-page test on its own isn’t enough, because some spreads are barely wider than they are tall, so it also looks for the fold.
The crop screen then shows all the sample pages layered into one picture, with the fold line. Drag it if it’s off, and choose Split, or Split and crop to draw a box for left pages and one for right. Each spread is cut at its own fold near your line, left page first.
After the split, every single page goes through the rest of the clean-up: one crop for the book, straightening, white paper and black print, and searchable text.
Steps
- Upload the scan. A PDF or DjVu file, up to 1,000 pages.
- Check the fold. Drag the line to the fold if it’s off.
- Choose the split. Split, or Split and crop with a box for left pages and one for right. Don’t split for books that read right to left.
- Review flagged pages. Pages with pictures, odd crops or unsure reading are marked. Swap in the original page wherever you want it.
- Download a PDF with one book page per PDF page.
Splitting by hand, and why it’s slow
Without a tool built for books, people do it one of three ways:
| Method | What it does | The catch |
|---|---|---|
| Crop twice | Duplicate the PDF, crop one copy to the left half and one to the right | You then interleave hundreds of pages by hand |
| “Split page in half” tools | Cut every page at the exact centre | Off-centre scans lose letters at the gutter |
| ScanTailor Advanced | Detects and splits pages, then crops and straightens | Desktop on Windows and Linux, and no searchable text |
| FixMyScan | Finds the fold per spread; you can move the line and crop left and right pages apart | Right-to-left books need Don’t split |
If you’re still scanning, some phone apps split at capture time. vFlat, for example, lists splitting two-page spreads among its features, along with flattening the page curve.
Best suited for: books scanned open on a flatbed or overhead scanner, phone scans of both pages at once, and downloaded library scans that show the book open.
Page numbers after a split
A 250-image scan of spreads becomes a PDF of about 500 pages. Your reader’s page counter then counts PDF pages, not printed ones, so “page 120 of 500” on screen won’t match the number at the foot of the page. The printed numbers are still there, and every page keeps its place in the book, so references to the printed page still hold.
When not to split
Leave splitting off for:
- right-to-left books, which come out in the wrong order
- art books, atlases and comics where most spreads are meant to be seen together
- scans that are already one page per image, where splitting has nothing to do
Questions
How do I split a PDF page into two?
For a scanned book, upload it to FixMyScan and choose Split on the crop screen, and each spread is cut at the fold. General PDF tools that split a page in half cut at the exact middle, which only works if every scan was centred.
Do I have to mark which pages are spreads?
No. Spreads are found from a sample of the book's pages, and the crop screen shows where the fold is.
What if the fold isn't in the middle of the scan?
The cut follows each page's own fold, not the centre line. You can drag the line on the crop screen, and a page whose fold can't be found is cut at your line.
Will a picture across both pages be cut in half?
Yes, if you split the book: every two-page scan is cut. For a book whose spreads are meant to be seen whole, choose Don’t split.
Does it work for manga or Japanese, Arabic or Hebrew books?
Not yet. Right-to-left books come out in the wrong order. Choose Don’t split on the crop screen.
My page numbers don't match the PDF any more. Why?
Splitting doubles the number of PDF pages, so PDF page 40 is printed page 80 or so. The printed numbers on the pages are untouched, so you can still cite by them.
Can I get the split pages straightened and cleaned too?
Yes. After the split, each page is cropped, straightened, cleaned to white and black, and made searchable in the same pass.