How Add watermark works
Image Watermark overlays a text string or a second image onto your photo directly in the browser, then exports the composite as a single flat file. You control the opacity, position (nine-point grid or free placement), font, size, colour and, for tiled watermarks, the repeat spacing. The original photo and the watermark asset are both read from your local disk and composited using the Canvas API; neither file is sent to any server at any point.
Text watermarks are rendered using the browser's text rendering engine, so any font available on your system can be used. Image watermarks (PNG logos with transparency, for example) are drawn with the alpha channel intact, blending naturally onto the photo. The flattened output contains only the merged pixels; there is no embedded watermark layer that could be stripped by removing metadata.