Deskew PDF
Scanned pages that lean a degree or two are tiring to read and look unprofessional. This tool measures the tilt of every page — the angle at which its text lines line up most sharply — and rotates the page back by exactly that amount. It is completely free, needs no sign-up and adds no watermark, and it runs in your browser, so your files are processed on your device and never uploaded.
PDF → straightened PDFHow to use Deskew PDF
- Add your scanned PDF — it is read on your device, never uploaded anywhere.
- Tap Straighten pages. Each page's tilt is measured (search range ±8°) and corrected.
- Download the straightened PDF and check the report — it names the angle corrected on each page.
Good to know
- Price: free. There is no paid plan, no credits and no trial.
- Sign-up: none. The tool opens and works straight away.
- Watermark: none is added to anything you download.
- Privacy: Your files are processed on your own device, in your browser, and are never uploaded.
- Part of: the Scan & copy tools in PdfLuxe — 151 free tools in one place.
How Deskew PDF works
Deskew PDF measures each page's tilt before correcting it. The page is rendered in your browser and analysed with the projection-profile method: the page's dark pixels are projected onto an axis at a series of trial angles within a ±8° search range, and the angle whose projection is sharpest — where text lines stack into the strongest peaks — is the page's true tilt. The page is then rotated back by exactly that amount and written into a new PDF, and the tool reports the angle it corrected on every page, so you can see what was found rather than trusting a black box.
The correction rotation re-draws the page as an image at 200 dpi. That is the right engineering trade for scans — a scanned page is already an image, so straightening it as an image loses nothing meaningful — but it does mean the output's text is not selectable or searchable. If your workflow needs searchability, deskew first and then run OCR PDF on the straightened file: OCR is noticeably more accurate on level pages.
When to use Deskew PDF
- Scanner feeder skew. Pages pulled through a feeder at a slight angle come back level, page by page, each by its own measured amount.
- Phone-scanned documents. Hand-held captures lean; deskewing makes them look flatbed-clean before they are sent.
- Preparing scans for OCR. Recognition engines assume near-horizontal lines. A deskewed scan OCRs faster and with fewer errors.
- Professional presentation. A report or application whose pages lean looks careless; level pages look deliberate.
- Archiving mixed batches. In a big scan batch every page leans differently — per-page measurement handles what one fixed rotation cannot.
Privacy — your files stay on your device
Your files are processed on your own device, in your browser, and are never uploaded. There is no account, no sign-up and no cloud copy of your document: the work happens in this tab, and the result exists only as the file you download. Measurement and correction happen in this tab; your scans are never uploaded.
Questions people ask
Is this really free?
Yes. Every tool in PdfLuxe is free with no premium tier: no sign-up, no credits that run out, and no watermark on what you download.
Are my files uploaded anywhere?
No. The tool runs entirely in your browser, on your own device. The file you add is never uploaded to a server.
How does it know the angle?
It uses the projection-profile method: for trial angles it projects the page's dark pixels onto the vertical axis, and the angle whose projection is sharpest is the page's tilt. That angle is then rotated back out.
What tilt range can it fix?
Up to about 8 degrees either way — far more than any scanner feed error. Pages already straight (under 0.2°) are left untouched.
Does the text stay selectable?
No. Pages are re-drawn as images at 200 dpi to rotate them, so text in the result is not selectable or searchable. Page size and layout stay the same.
Why does the report show a different angle for every page?
Because every scanned page really does lean by a different amount — the feeder, the paper and the placement vary sheet by sheet. Measuring per page, rather than applying one guess to the whole file, is what makes the result level throughout.
Should I deskew before or after OCR?
Before. OCR engines work line by line and assume roughly horizontal text; a tilted page produces avoidable recognition errors. Deskew first, then run OCR PDF on the straightened result for the cleanest text layer.
Can deskew fix a page that is upside down or sideways?
No — that is rotation, not skew. Use Rotate PDF or Auto-Rotate for quarter turns and upside-down pages, then deskew for the small residual tilt.
Why straighten a skewed scan?
A page scanned at a slight angle forces the reader's eye to work diagonally, breaks text selection in any later processing, and makes OCR noticeably worse — recognition engines assume roughly horizontal lines. Fixing the tilt inside the file, once, means every later reader, printer and OCR run gets a level page. This tool measures the angle from the page's own text rather than guessing, and reports what it corrected on each page so you can see it worked.
Honest limits
The measurement needs visible text or line work on the page — a nearly blank page has nothing to measure and is passed through unchanged. Correction is one rotation per page at 200 dpi, and pages become images, so very fine print is slightly softer than the original vector scan.
- Pages are re-drawn as images at 200 dpi, so the result's text is not selectable or searchable, and very fine print is slightly softer than a born-digital original. OCR the result if you need a text layer.
- The search range is ±8° and pages already within about 0.2° of level are left untouched. A page rotated by more than 8° needs Rotate PDF first, then deskew for the fine correction.
- A nearly blank page, or a page of pure photography with no text lines or edges, gives the measurement nothing to lock onto and is passed through unchanged.
Related tools
Questions about this tool? Email support@pdfluxe.com.