Merge PDF
Combine PDFs and reorder pages by dragging. Nothing is uploaded.
Stamp text or an image across pages with opacity control.
The goal is a mark that is unmistakable at a glance and still lets the document be read. Four settings decide that.
| Setting | Recommended | Why |
|---|---|---|
| Opacity | 15–25% | Below 10% is missed; above 35% obscures text |
| Rotation | 45° diagonal | Crosses lines of text so it cannot be cropped out |
| Size | 60–80% of page width | Covers the content area without spilling into margins |
| Color | Mid grey, or a desaturated brand color | Saturated red at 20% reads as pink and prints badly |
| Position | Center | Diagonal centered is the convention readers recognize |
A useful test: print the result in black and white. Many watermarks that look right on a backlit screen either vanish or turn into an unreadable smear on paper, and paper is where DRAFT and CONFIDENTIAL usually matter most.
No, and it is important to be clear about this. A watermark is a deterrent and a provenance marker, not a protection mechanism.
Anything drawn onto a PDF page can be removed by someone with the right tools. The overlay is a separate object in the content stream, and editing software can delete it. Even a flattened watermark can be removed by rendering the pages, inpainting, and rebuilding.
What a watermark does well is establish intent and attribution. A document stamped DRAFT cannot be honestly presented as final. A document stamped with a recipient’s name discourages casual forwarding and identifies the source if it leaks — which is why per-recipient watermarking is standard practice for pre-release financial and legal material.
For actual protection, you need encryption with a user password, or a rights-management system. Neither is what a watermark is for.
Text is the right choice for status marks — DRAFT, CONFIDENTIAL, SPECIMEN, INTERNAL USE ONLY, a date, a recipient name. It renders as vector text, stays sharp at any zoom, adds a few hundred bytes total, and remains legible when printed.
An image is the right choice for a logo or a signature. Use a PNG with a transparent background; a JPEG watermark brings an opaque rectangle with it, which will cover the text underneath regardless of the opacity setting.
For image watermarks, size matters for file weight: the image is embedded once and referenced on every page, so a 2 MB logo adds 2 MB to the document, not 2 MB per page.
Yes. The page range field accepts the same syntax as the split tool — 1-5, 12, 20- — and there are odd and even shortcuts.
The common case is skipping the cover page. A report with a designed cover usually looks better with the watermark starting at page 2, and applying it to 2- does exactly that.
By someone with PDF editing software, yes. A watermark is a deterrent and a provenance marker, not encryption. For real protection, use a password.
You choose. In front at low opacity is the convention, because it reads as an overlay. Behind is subtler but can be hidden entirely by opaque page elements.
Yes. Upload a PNG with transparency — a JPEG carries an opaque background that will block the content under it.
Yes. The watermark is drawn on top of the page content, whether that content is text or a scanned image.
Not in one pass. Apply one watermark per range and repeat for each variation, or use the page-number token to insert a per-page value.
A text watermark adds a few hundred bytes total. An image watermark adds roughly the size of the image once, since it is embedded and referenced rather than duplicated per page.
No. The overlay is applied by pdf-lib in this page.