Watermark a PDF

Drawn into the page, not stuck on top

There are two ways to put a mark on a PDF and only one of them is a watermark.

The easy way is an annotation: a note object that sits above the page. It looks identical on screen and it can be selected and deleted in any reader in about two clicks. Several free tools do this, because it is far simpler to implement.

The right way is to draw the text into the page content itself, so it is part of the page in the same way the rest of the printing is. There is nothing to click on. That is what happens here.

How this tool works

The mark is drawn into each page as real text at the opacity you choose, so it stays sharp at any zoom and prints properly. Nothing is rasterised.

Placing a rotated mark takes more care than it looks. A PDF rotates text about the start of the baseline rather than the middle of the string, so a mark placed by its unrotated size and then turned swings away from where it was put: at 45 degrees a long word travels most of its own length. The size of the rotated box is worked out first and the mark is placed by that, which is why a diagonal mark lands in the middle of a wide landscape sheet rather than halfway off it.

The type is also shrunk if the mark would be wider than the paper, and sized per page rather than per document, because a pack can hold an A4 sheet and an A1 drawing and one size is not right for both.

What a watermark on a PDF is for

It is not security. It is worth being blunt about that, because a lot of tools imply otherwise. Anyone determined enough can redraw the page without your mark, and no free browser tool is going to stop them.

What a watermark does is make the status of a document impossible to miss. That is a real problem it solves properly:

• A draft that gets treated as final. The most common version. A price or a drawing goes out for comment, somebody forwards it, and three weeks later it is being worked to as though it were issued.
• NOT FOR CONSTRUCTION. A drawing that has not been signed off must be unmistakably marked, and a filename nobody reads is not marking.
• A copy, so the original is identifiable. When only one version is the record and the rest are circulating.
• A sample or a proof, usable enough to judge and not usable enough to keep.

Getting the strength right

This is the setting people get wrong, in both directions.

Too faint and it does not do its job. Somebody skim-reading on a phone will not register a five per cent grey mark. It has to be unmissable.

Too strong and you have made the document hard to work from. If you are asking for comments on a specification, burying the text under a solid red diagonal means you will get fewer and worse comments back.

Around 15 per cent is the sensible default for anything somebody has to read. Save the heavy settings for a sample or a proof, where making it awkward to use is the intention rather than an accident.

One mark, or tiled across the page

A single mark across the middle at 45 degrees is the classic, it reads well, and it is what most people picture. It is also the easiest to defeat: crop it, or paste a white box over it.

Tiling repeats the text across the whole page, including over the content. It is less elegant and considerably more stubborn, because removing it means removing it from everywhere. Use it when the document genuinely should not be reused rather than simply labelled.

Only some of the pages

You can mark a range rather than the whole document. That is more useful than it sounds: a pack where the drawings are draft and the covering letter is not, or an appendix that came from somewhere else and should be marked as belonging to them.

We do not add one of our own

Most free tools that put a watermark on your document also put a watermark on your document, theirs, in the corner, until you pay.

There is nothing here but what you typed. No logo, no URL, no small print in the footer, no reduced quality, no page limit and no account. The file is opened, stamped and saved by your own browser, and it never leaves your machine.