How to watermark a video without losing quality
Any watermark that's part of the picture means the video has to be encoded again, and every encode loses a little. With the right settings that loss is too small to notice, and the sections below explain what to choose and what to avoid.
Updated · 5 min read

Why a watermark means re-encoding
Video files don't store every frame as a full picture. They store some frames in full and the rest as changes from the frames around them, compressed with a codec such as H.264 or H.265. To put a mark into the picture, every frame has to be decoded, drawn on and compressed again.
Each compression throws away some detail, which is why a copy of a copy looks worse than the original. A tool that claims to add a watermark without re-encoding is adding a separate overlay that players can hide and anyone can strip out. A burned-in watermark always means one more encode, so the aim is to make that encode a good one.
This is also why the watermark's size barely matters for quality. A small logo and a large one cost the same encode; what matters is the settings and the file you start from.
Start from the best file you have
Quality you've already lost can't come back. Use the original file from the camera or phone, or the high-quality export from your editor, never a copy downloaded from a social platform or sent through a messaging app, since both are compressed heavily.
If you edit in a video editor, export your master as high-bitrate H.264 or H.265, since editing formats like ProRes won't open here. If you're already exporting from an editor, adding the logo there saves one encode altogether; a separate watermarking step makes most sense when you're starting from finished files.
Check what you have before you start. A file straight from a phone or camera is usually far larger per minute than the same clip saved from a social app, and file size is a rough but useful sign of how much detail is left.
Keep the original size
Scaling a video down loses detail for good, so leave the size at Original unless you need a smaller file. Platforms make their own smaller versions from whatever you upload, and they have more to work with when you give them the full size.
Downscaling does make sense in some cases: a 4K clip you'll only post to a phone-first app, or a batch that has to fit in an email or a shared drive. Then 1080p is a sensible floor, and 720p is for previews and quick shares.
Never scale a video up to make it look better. Enlarging adds pixels but no detail, and it makes the file bigger for nothing.
Pick Standard for the most quality
In Watermark Ready, Standard encodes H.264 at the original size with high quality settings. H.264 plays everywhere, so it's the safe choice for a master copy or anything you'll upload.
Smaller file uses H.265 (HEVC) where your browser can encode it, otherwise H.264, and caps the size at 1080p. H.265 needs fewer bits than H.264 for a similar picture, so files come out noticeably smaller, but this setting gives up some detail compared with Standard, and some older devices can't play H.265.
If your browser can't encode H.265, Smaller file still uses H.264 at up to 1080p. You get a lighter file from the size cap and lighter settings, just without the extra savings of H.265.
For a copy you'll keep or hand to a client, use Standard at Original size. Save Smaller file for sharing and storage.
Avoid stacking generations
The biggest quality losses come from repeated encodes, not from the watermark itself. Export once from your best file, upload that one export, and don't watermark a copy that has already been watermarked or downloaded from somewhere.
Keep the clean original. If you want a different mark later, or a version without one, go back to it rather than working from the watermarked copy. The mark is part of every frame of the copy and can't be taken out.
The same goes for editing after watermarking. Trimming or color grading a watermarked copy in another app means another encode, so do those first and watermark last.
What changes no matter what
A few things change in any burned-in export. HDR footage becomes standard range, so bright highlights look less intense than on an HDR screen. Sound is copied as it is when MP4 can hold it and re-encoded when it can't. The frame rate and length stay exactly as they were.
Your device sets the limits, too. Video needs a recent Chrome, Edge, Safari or Firefox, and long or 4K files go better on a computer than on a phone, which has less memory. The video is processed on your device and never uploaded.
If a long clip fails on a phone, try it on a computer before lowering any settings. The extra memory is often all it needs, and you keep the full quality.
Check the result
Compare the copy with the original at full size on the scenes that show problems first. Smooth gradients like skies and walls show banding, fast motion and fine textures such as water, leaves or confetti show blockiness, and dark scenes show blotchy shadows.
If you see any of these with Standard at Original size, the source was probably already heavily compressed. Go back to a better source if you have one. If not, the copy is as good as that file allows.
Listen to the copy as well as watching it. Sound in a format MP4 can't hold is re-encoded, so check a quiet passage and a loud one on clips from unusual cameras or recorders.
Pause on a fast-moving scene rather than judging at normal speed. Problems that are hard to spot in motion show up clearly in a paused frame.