How accurate is automatic gutter detection?

The tool averages the brightness of every column (every row for top/bottom splits) and, within ±15% of the centre, looks for a clearly darker binding shadow or a continuous blank gap. Book scans usually have an obvious shadow, so detection is reliable; spreads with a picture across the middle, or pages printed very close together, may fall back to the centre. Every cut is shown on the thumbnails, so just drag the slider if one is off.

Does splitting reduce quality? Can I still search the text?

No quality is lost. Pages aren’t re-rendered: each original page is duplicated and the two copies show only the left and right halves (via CropBox and MediaBox). Scanned images keep their resolution, and digital text and vector graphics stay as they are — selectable, copyable and searchable.

How should I set it up for manga or vertical books?

Manga and vertical CJK books are read right to left, so the right half of a spread comes first. Tick Right-to-left reading and each spread will output its right half before its left, giving the correct page order. Covers are usually single pages, so tick Don’t split the first page and Don’t split the last page.

What if the file mixes single pages and spreads?

Tick Only split pages shaped like a spread: left/right splitting then only splits landscape pages and leaves portrait single pages alone (and the reverse for top/bottom). You can also untick individual pages in the per-page list. All processing happens in your browser and nothing is uploaded.

📔

📔Split Two-Page PDF Scans into Single Pages - Auto Gutter

Scanning a book with a flatbed or phone usually captures two facing pages at once, so every image in the PDF is a landscape spread — the text is tiny on a phone and awkward to print. This tool splits each spread into two single pages. It renders a thumbnail of every page and looks for the gutter within ±15% of the centre: on scans that’s the binding shadow (a dark band), on digital spreads it’s the blank gap between pages (a bright band); if neither is found it uses the exact middle. The cut is drawn on each thumbnail, and you can drag a slider to adjust it page by page or untick a page to leave it whole. Top/bottom splitting, right-to-left order (right half first, for manga and vertical CJK books), keeping the first and last pages (covers) whole, and splitting only spread-shaped pages are all supported. Nothing is rasterized: each half is a copy of the original page with a different visible area (CropBox / MediaBox), existing page rotation and crop boxes are respected, so quality is unchanged, text stays searchable and the file barely grows. Everything runs in your browser; files are never uploaded.

Free · No sign-up · Runs in your browserLast updated: 2026-10-26

Loading tool...

How to use Split PDF Spreads

1Upload a PDF of two-page scans — the gutter on each page is detected automatically
2Choose left/right or top/bottom, and options such as right-to-left order or keeping the covers whole
3Check the red cut line on each thumbnail, drag the slider if needed, then click Split & download

FAQ

How accurate is automatic gutter detection?: The tool averages the brightness of every column (every row for top/bottom splits) and, within ±15% of the centre, looks for a clearly darker binding shadow or a continuous blank gap. Book scans usually have an obvious shadow, so detection is reliable; spreads with a picture across the middle, or pages printed very close together, may fall back to the centre. Every cut is shown on the thumbnails, so just drag the slider if one is off.
Does splitting reduce quality? Can I still search the text?: No quality is lost. Pages aren’t re-rendered: each original page is duplicated and the two copies show only the left and right halves (via CropBox and MediaBox). Scanned images keep their resolution, and digital text and vector graphics stay as they are — selectable, copyable and searchable.
How should I set it up for manga or vertical books?: Manga and vertical CJK books are read right to left, so the right half of a spread comes first. Tick Right-to-left reading and each spread will output its right half before its left, giving the correct page order. Covers are usually single pages, so tick Don’t split the first page and Don’t split the last page.
What if the file mixes single pages and spreads?: Tick Only split pages shaped like a spread: left/right splitting then only splits landscape pages and leaves portrait single pages alone (and the reverse for top/bottom). You can also untick individual pages in the per-page list. All processing happens in your browser and nothing is uploaded.