Watermark Image

A watermark does three jobs, and it is worth being clear about which one you need. It deters casual reuse, because most people who would happily save an unmarked image will not bother with a marked one. It attributes the work, so that when an image does travel, your name travels with it. And it marks proofs, letting you show a client the picture while making the unlicensed version unattractive to use.

What it does not do is prevent theft. Anyone with editing software and a little patience can remove a watermark, and AI-based inpainting has lowered that bar considerably. Treating a watermark as copy protection leads to marks so heavy they ruin the photograph while still not stopping anyone determined.

The realistic goal is friction and attribution. Once you frame it that way, the design choices become much easier.

A mark in a corner is discreet and looks professional, which is why it is the default for finished work. Its weakness is obvious: a corner is trivially cropped away, and the image survives the operation looking fine.

Tiling the mark across the frame, or running it diagonally through the centre, solves that at a cost. Now removal means retouching over the subject rather than cropping around it, which is real work. But the photograph is harder to look at, so this belongs on proofs, previews and portfolio pieces rather than on delivered work.

A middle path many photographers use: place a single mark so that it overlaps the subject rather than sitting on empty background. It is far less intrusive than tiling, and much harder to remove cleanly than a corner mark.

Opacity around 30–50% is the range where a watermark reads without dominating. The trap is testing it on one image and applying it to a hundred: a white mark at 35% is clearly visible on a dark photograph and effectively invisible on a bright sky. If your batch is varied, check the brightest and darkest images before committing.

Size matters less than people expect. A mark occupying roughly 10–20% of the frame's width is legible on a phone without shouting. Very small marks are easy to crop and easy to miss; very large ones stop being a watermark and start being a poster.

For logos, use a transparent PNG. A logo saved as JPG carries a white rectangle with it, which looks like a mistake on any image that is not white behind the mark.

Watermarking is destructive — the mark becomes part of the pixels and cannot be lifted out later. Always work from a copy and keep the clean original, because the moment you need to license the image properly, print it, or supply it to a client, the watermarked version is useless.

The sensible order is edit, then resize for the destination, then watermark last. Watermarking before resizing means the mark gets scaled with everything else and often ends up illegible on the smaller output.

Everything here runs in your browser, so the unmarked original never leaves your device — which matters, since you are usually watermarking precisely because you do not want the image circulating freely.