A watermark serves one of two purposes: identifying who created an image, or discouraging unauthorised use of a preview before a full resolution version is sold or delivered. Both uses depend on the watermark being baked into the pixels themselves, which is exactly what happens here, since the text is drawn onto the same canvas as the photo before the file is re encoded.
The photo stays on your device throughout. It is decoded by the browser, the watermark text is drawn on top of it using the canvas text API, and the combined result is re encoded locally, so an unreleased photo never has to be uploaded just to have a watermark added to it.
When to watermark a photo
Photographers commonly reach for a photo watermark tool to protect a preview gallery shown to a client before the final files are delivered, which lets the client review and select images without being able to lift a clean, usable copy from the preview itself. A visible credit line or a business name across a set of photos also serves as free advertising every time that image is shared or reposted elsewhere.
A subtle watermark in a corner works well when the goal is attribution rather than active deterrence, while a larger, more central watermark at moderate opacity is the standard choice when the goal is specifically to prevent a preview image being used without permission.
- Watermark client preview photos before delivering the final, unwatermarked files.
- Add a small corner credit to photos that will be shared publicly for attribution.
- Use a larger, more central watermark specifically to discourage unauthorised use of a preview.
- Watermark a batch of stock or portfolio images consistently with the same business name.
- Skip watermarking for final delivered files the client has already paid for and licensed to use freely.
Choosing position, size, opacity and colour
Position options place the text in one of the standard corners, along an edge, or centred across the whole image, which covers the placements people ask for most often without needing to drag text around manually. A corner placement is the conventional choice for a subtle credit, while a centred placement makes a watermark much harder to crop out of the frame.
Size is set relative to the image dimensions rather than as a fixed pixel value, so the same setting produces proportionally sized text whether the source photo is a small thumbnail or a full resolution original. Opacity controls how strongly the text shows through, and a value between 40 and 60 percent is usually the sweet spot: visible enough to identify or deter, but not so strong that it distracts from the photo itself.
Colour should generally contrast with the photo it sits on. White text with a moderate opacity works well on darker photos, while black or a dark grey works better on lighter ones, and this tool lets you pick any colour so a specific brand colour can be used consistently across a whole set of images.
Why a baked in watermark is harder to remove than an overlay
A watermark applied in a design tool as a separate layer can sometimes be deleted just by discarding that layer, if the file is ever shared in its editable format rather than flattened. This photo watermark tool never produces a layered file. The text is drawn directly onto the same canvas pixels as the photo and the whole thing is re encoded as a single flat image, so there is no separate layer to strip out, only pixels that would need to be painted over or cropped away to remove.
No watermark is completely unremovable with enough effort, since anything visible in an image can eventually be edited out by someone determined enough. The realistic goal of a watermark is deterrence and attribution, not an unbreakable technical guarantee, and a baked in text watermark meets that goal well for the vast majority of casual reuse.
How this photo watermark tool works under the hood
The source photo is drawn onto a canvas first, then the canvas text API draws your chosen text on top at the calculated position, using the font size, colour and opacity you selected. Opacity is applied through the canvas global alpha property, which affects only the text drawn while it is active, leaving the underlying photo at full strength beneath it.
Because text rendering and image drawing both happen on the same canvas using primitives built into the browser, there is no server step involved, and you can confirm nothing is uploaded by checking your browser's network panel while the watermark is applied.
How to add a watermark to a photo
- 1
Add your photo
Drag an image onto the drop area, or click it to choose a file. It stays entirely on your device.
- 2
Type your watermark text
Enter the credit line, business name or copyright notice you want stamped onto the photo.
- 3
Set position, size, opacity and colour
Choose where the text sits, how large it appears, how strongly it shows through and what colour it uses.
- 4
Apply and download
Press apply, review the preview, then download the watermarked photo.
Related tools worth bookmarking
Sources and further reading
- MDN: CanvasRenderingContext2D.fillTextThe canvas text API this watermark tool uses to draw the chosen text directly onto the photo's pixels.developer.mozilla.org
- WHATWG: Canvas text stylesThe living standard section defining how canvas text rendering and opacity combine when drawing a watermark.html.spec.whatwg.org
Frequently asked questions
Common questions about the photo watermark tool.
With enough effort, most watermarks can eventually be edited out, since the text is visible on the image and nothing visible is truly permanent. A watermark is meant to deter casual reuse and provide attribution rather than offer an unbreakable technical guarantee, and a text watermark baked directly into the pixels, as this tool produces, meets that goal for the overwhelming majority of cases.
No. The watermark text is drawn directly onto the photo using the canvas element in your browser, and the whole process happens locally. Your photo is never transmitted to a server, which you can confirm by checking your browser's network panel while the watermark is applied.
Somewhere between 40 and 60 percent suits most situations, since it keeps the text clearly visible for identification or deterrence without overwhelming the photo underneath it. Go lower for a subtle corner credit and higher for a watermark specifically meant to discourage unauthorised use of a preview.
This tool applies a watermark to one photo at a time so you can confirm placement and legibility on each image before downloading it. The same position, size, opacity and colour settings you configure carry over automatically as you process additional photos in the same session.
The watermark itself does not degrade the photo underneath it, since it is drawn as an additional layer of pixels on top before the final image is encoded. If you choose a JPG output, the usual JPEG compression trade offs apply to the whole image including the watermark text, the same as they would for any other JPG export.
Yes. There is no account, no daily limit and no cost to add a watermark to as many photos as you need in one session, since the whole process runs locally in your browser rather than on a paid server that would need to justify a subscription fee.
A centred or repeated placement across the image is much harder to crop out than a single corner mark, since removing it would also remove the centre of the photo itself. A single corner watermark is easier to crop away, which makes it better suited to attribution than to active deterrence.