Watermark PDF

How to use Watermark PDF

Key features of Watermark PDF

About Watermark PDF

Watermark PDF draws your text or image onto every page with pdf-lib. Text is set in bold Helvetica at 44 pt, rotated 45 degrees and centred using its measured width, so it sits in the middle of the page whatever the wording. An image is embedded once and drawn on each page at half the page width, capped at half the page height.

Mark a draft before it goes round for comment, label a sample copy, or put a logo behind a price list. The mark becomes part of the page content, so it prints and shows in every viewer and cannot be switched off as a layer. There is no setting for size, position or a single page, and the font cannot draw characters outside the Latin (WinAnsi) set. Use the Remove Watermark tool if you need to take a text watermark out again.

When to use Watermark PDF

A tender goes to three bidders and each copy should carry "COPY 1", "COPY 2" or "COPY 3" so a leak can be traced back. Run the tool three times on the same original with different wording. Each run marks every page, and the original stays untouched on your computer.

A photographer sends a PDF of proofs to a client. Choose the studio logo as the image, set the opacity to 30%, and it is centred on every page at half the page width, faint enough to judge the pictures but hard to crop out.

A price list that is only valid for a month: "VALID UNTIL 30 JUNE" in blue at 50% runs diagonally across every page, and a customer who prints it cannot miss it. Once the month is over, Remove Watermark takes the wording out again.

You need the mark small, in a corner, upright, or on the first page only. This tool has none of those settings. Add Text puts small upright text in one of nine positions and Add Image does the same with a logo, and both can limit the mark to the first or last page.

Tips and troubleshooting

• Leaving the text box empty gives you CONFIDENTIAL, the default wording; there is no way to apply a blank text watermark.
• The size is fixed at 44 pt and the text is centred, so a long sentence runs past both edges of the page. A few words is the practical limit; check with Preview Document, which shows the result in your browser's built-in PDF viewer.
• "Failed to apply watermark: WinAnsi cannot encode" means the wording contains a character bold Helvetica does not have, such as Cyrillic, Chinese, an emoji or a Central European letter like ł or č. Western European accented letters (é, ü, ñ) and the euro sign are fine.
• "Failed to apply watermark" with a message that the input document is encrypted means the PDF has a password. Remove it with Unlock PDF and watermark the result.
• The image field accepts PNG and JPEG only. While an image is chosen the text and colour settings are ignored, and the opacity applies to the image instead. To go back to text, reload the page and start again.
• The mark is drawn after the existing content, so it sits on top of pictures as well as text, and on a scanned page it is drawn over the scan. Each page is centred using its own size, so mixed page sizes are handled page by page.
• The download is named watermarked_ followed by the original filename. A text watermark is real text that a reader can select and search for, and Remove Watermark can take it out by matching the wording; an image watermark cannot be removed that way.

How Watermark PDF works in your browser

pdf-lib parses the file in memory. For text it registers the standard Helvetica-Bold font, measures the wording at 44 pt, and for each page works out where the text origin must sit so that after a 45-degree turn the text is centred: rotation in a PDF happens about the start of the text, not its middle, so the origin is offset from the page centre by the rotated half-width and half-height. The text is then drawn in the chosen colour with a transparency state at the opacity you picked. For an image, the picture is decoded once, stored as a single image object, and drawn on every page at half the page width (capped at half the page height) with the same opacity. No server is involved and neither file leaves your browser.

The document is saved to a new byte array, wrapped in a Blob and shown on the result card, whose message says whether a text or an image watermark was applied and to how many pages. The existing content, bookmarks, links, form fields and metadata are written back unchanged; each page gains one drawing instruction and one transparency state. The font is not embedded, so a text watermark adds only a few bytes per page, while an image adds its own size once for the whole document.

Frequently asked questions