How it works

How a watermark is removed

Most watermarks are a fixed image laid over every frame and partly see-through. The video underneath changes from frame to frame; the watermark doesn’t. That difference is what the tool measures.

1. 1Find it. The tool samples the area you selected across the whole video, around 80 frames. Edges that appear in the same place in every frame belong to the watermark; everything else is the moving background. This gives a precise outline of the watermark’s own pixels, so the rest of your selection is left untouched.
2. 2Measure it. For each watermark pixel, the tool compares what you see with what the background around it looks like, across all those frames. From that it works out the watermark’s colour and exactly how see-through each pixel is.
3. 3Reverse it. With colour and transparency known, the watermark’s effect can be subtracted from every frame, recovering the real picture underneath rather than blurring over it.
4. 4Fill what’s left. Where a watermark is fully solid, nothing survives underneath, so those pixels are filled in from their immediate surroundings.

Finding the watermark automatically

When you open a video, the tool samples small copies of about 36 frames and looks for edges that never move while the rest of the picture does. It groups them into boxes and places up to four for you. It skips anything colourful, because watermarks are almost always white, grey or black overlays, while scenery that merely holds still is coloured. If it can’t tell, it says so and leaves the box for you to place.

Watermarks that come and go

Some watermarks appear only for part of a video, or jump from one corner to another. Give each box its own time range: move the playhead and set when that box starts and ends. The tool learns each watermark only from the frames where it’s visible, and removes it only there.

Trim, format and quality

• Trim keeps only the part of the video you choose, which also makes processing faster.
• Save as MP4 (plays almost everywhere) or WebM (open format, often smaller).
• Quality sets the bitrate relative to your original: Smaller file, Balanced, or Best, which adds headroom because re-encoding always costs a little detail.

Several videos at once

After a video is done, you can add more clips from the same source. The same boxes and settings are applied to the whole of each one, one after another, and each gets its own download.

How cropping works

Crop keeps the largest part of the frame that doesn’t include your selection: it trims a strip from whichever edge loses the least picture. The result is always perfectly clean, but the video gets slightly smaller. Dimensions are kept even, which video encoders require.

Where the work happens

Everything runs in a background worker inside your browser tab, using the browser’s built-in WebCodecs video engine. Frames stream through one at a time and are released immediately, so memory use stays flat. Audio is carried over too, whenever your browser can encode it (if it can’t, the editor tells you). The page may only connect to this website and to its analytics services, and no request carries your video.

Offline use

In browsers that support it, the site saves its own pages and code in your browser the first time you open the editor. After that the editor, including automatic detection, works without a connection. Only the site’s code is saved, never your videos.

Limits

Browser support

The editor needs WebCodecs, Web Workers and OffscreenCanvas. It has been tested in recent desktop versions of Chrome and Edge. Other browsers are adding these features; when you open the editor it checks what your browser supports and tells you plainly if something is missing, instead of failing halfway through.

Keyboard controls

• Tab to a box, then use the arrow keys to move it (hold Alt for fine steps).
• Shift + arrow keys resizes the selection.
• Delete or Backspace removes the focused box when there’s more than one.
• On the result, focus the before/after divider and use the arrow keys to move it.