Video Watermark
Burn a text watermark into a video at one of nine positions and an opacity you choose. Re-encoded, so not lossless.
What the watermark tool burns in
One line of text, drawn into the picture at a corner, an edge or the centre, with an opacity between 10 and 100 per cent and a font size between 12 and 96 pixels. The text sits on a translucent box so it stays readable over both a bright sky and a dark room.
Because the text becomes part of the frames, the picture is re-encoded with H.264 at CRF 23. The audio is copied untouched. That is the honest description: a burned-in watermark is not a separate layer that can be removed later, and the encode is lossy.
How to use it
Type the text, choose one of the nine positions, set opacity and font size, then press Run. The summary repeats the text and position that were applied.
How the text is handled
The string is written to a file inside the job's temporary directory and read by the draw engine from there. It is never spliced into an argument or a filter expression, so quotes, colons, percent signs and semicolons are ordinary characters rather than syntax. Up to 80 printable characters fit on the line.
Limits
20 MB in, 2 minutes of duration, 24 MB out, 180 seconds of processing. A missing font on the server is a refusal with that reason, not a failed encode.
Upload a video, type a watermark and choose one of nine positions, an opacity and a font size; the server burns the text into the picture with ffmpeg and returns a new MP4.
The picture is re-encoded, so this is not a lossless operation. The audio is copied through unchanged, and the video is H.264 at CRF 23 with yuv420p. A 60% white watermark is drawn over a translucent dark box so it stays readable on a light or a dark frame.
The text is written to a file and read from there by the filter, so a quote, a colon or a space in your watermark cannot become filter syntax. It may contain up to 80 printable characters on one line; the font size runs from 12 to 96 pixels and the opacity from 10 to 100 per cent.
Placement and appearance
The nine positions are the corners, the middle of each edge and the centre, each inset by 24 pixels from the frame edge. The position is fixed at encode time; there is no drag-to-place preview on this page.
Opacity applies to the white text, and the box behind it stays at a fixed 35 per cent black. A low opacity over a busy area can still be hard to read, which is why the box exists and why the size is adjustable.
- Nine positions, all inset 24 pixels
- Opacity 10 to 100 per cent, default 60
- Font size 12 to 96 pixels, default 28
What the encode changes
The video is decoded and encoded again, so a source that was already compressed may lose a little detail. The summary reports the output size, the position, the opacity and the text, which is enough to confirm what was burned in.
The input is limited to 20 MB and two minutes, one encode has 180 seconds, and the output is capped at 24 MB. The temporary directory, including the text file, is removed whether the run succeeds or fails.
How to use the watermark form
Choose the video in the Upload step, type the watermark text on one line of at most 80 characters, and pick one of the nine positions, an opacity from 10 to 100 per cent and a font size from 12 to 96 pixels. Press Run; the server burns the text into the picture and returns a new MP4.
Check the summary's position, opacity and text against what was intended before sharing the file, since the mark is fixed at encode time. Re-run with a higher opacity when the text sits over a busy background and reads poorly.
What to Expect
Best for
- Marking a draft clip with a channel name
- Adding a visible credit before sharing
- Burning a short label into a preview
Not ideal for
- Removing a watermark, which is a different and harder problem
- A clip over 20 MB or two minutes
- Keeping the picture bit-for-bit, since the re-encode is lossy
What this tool keeps
- The full duration and the audio track
- The aspect ratio and frame size
- The exact text you typed, up to 80 characters
What may need cleanup
- A watermark over a busy background, which may need a higher opacity
- A position that covers an existing caption
- The fixed 24-pixel inset, which cannot be moved per video
Common errors
- Typing more than 80 characters and finding the text refused
- Expecting the audio to be re-encoded with the picture
- Choosing an opacity below 30 over a bright sky
Example use cases
- A credited clip for social media
- A labelled preview sent to a client
- A draft marked as not final
Sample input
A 25-second clip watermarked AllTools at bottom-right, 60 per cent opacity and 28-pixel text.
Sample output
An H.264 MP4 with the text burned into the bottom-right corner, the audio unchanged and the output size reported.
Who this is for
- Content creators marking a draft
- Small teams adding a credit
- Anyone labelling a preview
Frequently asked questions
What does the video watermark tool do?
It burns the text you type into the picture at one of nine positions, with the opacity and font size you choose, and returns a new MP4.
Can it remove a watermark?
No. Removing one is a different and harder problem.
Does it keep the picture bit for bit?
No. The re-encode is lossy, so the picture changes even where no watermark is drawn.
Build custom tools for your store
Join the SDK program to create useful ecommerce utilities for your store.