Image Watermark
Put text or your own logo over a picture, once in a corner or tiled across the whole frame, and save it at the picture's own resolution. The mark is sized in fractions of the image rather than in fixed pixels: its width is a percentage of the picture's width, the gap to the edge is a percentage of the shorter side, and the tile step is a percentage of the mark. That is what lets one setting suit a folder of photographs at different sizes, so a queue of pictures can be marked together and each one comes out looking the same. Everything happens in this browser tab: the files are decoded here, the text is measured in a font your device already has, the canvas is drawn here and the finished file is handed straight back to you. There is no upload, no account and no size limit in bytes; the only ceiling is 50 megapixels a picture, which is where a browser canvas starts to struggle.
your picture, untouched
the marked picture appears here
· or drop as many as you like on the stage
No picture chosen yet. The watermarked preview appears here once you add one.
The mark is 12% of each picture's own width, so one setting suits a folder of photographs at different sizes. Tab to the placement grid and the arrow keys walk the mark around the nine cells. Text is drawn in a font your device already has, up to 200 characters, and each picture may hold up to 50 megapixels.
Common questions
- Are my pictures uploaded anywhere?
- No. There is no server in this. Every file you open, the photographs and the logo alike, is read by the browser's own image decoder, drawn onto a canvas inside the page and handed back to you as a download. No image byte is sent to an endpoint, to analytics or to error reporting. The only thing written to your browser's storage is the settings row: the mark type, the words, the font, the bold switch, the colour, the placement, the layout, the opacity, the size, the inset, the tile spacing, the angle and the format. No picture is ever stored, and if storage is blocked the tool still opens and still works.
- Can I watermark a batch of photos at once?
- Yes. Choose or drop as many pictures as you like and they queue up under the stage. One set of settings covers the queue: Save all writes every one of them, a beat apart so the browser does not drop the downloads, and each file keeps its own name with -watermarked and the format on the end. Each file is checked as it opens, and one the browser cannot decode is named in a message and left out of the queue instead of stopping the rest. Click a row to bring that picture onto the stage and see the mark on it before you save.
- Does one setting really work on photos of different sizes?
- Yes, and that is the thing the tool is built around. Nothing in the placement is a fixed pixel measurement. The mark's width is a percentage of the picture's width, the inset is a percentage of the shorter side, and the tile step is the mark plus a percentage of the mark. Double the picture and every number doubles with it, so a 1600 by 900 photo and a 3200 by 1800 one from the same shoot come out with the mark the same relative size in the same relative place.
- Can I use my own logo, and does its transparency survive?
- Yes. Choose a logo file and it is drawn instead of the text, scaled to the width you set and keeping its own shape. A PNG with an alpha channel keeps that alpha: the holes let the photograph show through, and the whole mark is then drawn at the opacity you chose. Saving as PNG keeps transparency in the picture as well. JPEG has no alpha at all, so when you save as JPEG a white matte is painted under the picture first, which is why transparent areas come out white rather than black.
- Does a watermark stop someone from using my photo?
- No, and the tool does not pretend otherwise. A mark baked into the pixels is a visible credit: it says whose picture this is wherever the file travels, and it makes casual reuse awkward. It is not a lock. Anyone holding the file can crop the corner off, paint the mark out or cover it, and a tiled mark across the frame is harder to remove than a small one in a corner but is still only harder. Choose the setting that fits what you are posting rather than one that promises something a picture cannot do.
- What do the opacity and the scale numbers actually mean?
- Opacity is the alpha the mark is drawn at, from 0 to 100 per cent, and it does what the arithmetic says: at 0 per cent the canvas draws the mark with the alpha at zero, so the saved file is the original picture's appearance, unchanged. Scale is the mark's width as a percentage of the picture's width, so 12 per cent on a 1600 pixel wide photo is a mark 192 pixels across. For text, the font size follows from that width and the measured shape of the words, which is why a long line comes out in smaller type than a short one at the same scale.
- Which fonts can the text use?
- Three stacks your device already has: a system sans, Georgia or another serif, and a monospace face, each with a bold switch and a colour you pick. Nothing is downloaded for the mark, which is why the text appears the instant you type it. The words are measured with the same font the canvas draws them in, so what the preview shows is what the file holds. Up to 200 characters are drawn; a longer mark is refused in words that name the ceiling and your length rather than being cut down without saying so.
- Is there a limit on how big the pictures can be?
- Each picture may hold up to 50 megapixels. That ceiling is on pixels rather than on bytes, because the canvas has to hold four bytes for every pixel, and a picture past that point is more likely to fail in the browser than to be marked. The size of the file on disk is not checked at all: nothing is uploaded, so the bytes are nobody's business. A picture over the ceiling is refused in a sentence that names both the ceiling and how many megapixels the picture holds. A tiling is capped too, at 2,000 marks, which only a very small mark with no spacing can reach.
- Can I place the mark without a mouse?
- Yes. The placement control is nine buttons in a three by three grid. Tab to it and the arrow keys walk the mark around the nine cells, stopping at the edge rather than wrapping around. Opacity, scale, angle, inset and spacing are all sliders, which the arrow keys move a step at a time, and every control carries a name a screen reader announces. The current mark is also written out in words under the stage, so the result is readable without looking at the canvas.
The watermark is drawn onto a canvas in your browser and baked into the image you save. Neither the photo nor the logo is uploaded. Placement, size and tile spacing are exact arithmetic on that picture's own pixels, and a baked-in mark is a visible credit rather than a lock: anyone holding the file can crop it off or paint it out.