How to blur other players' names in gameplay footage

Last updated 2026-09-22 · 8 min read

What is the short version?

You want to post the clip without exposing the people you played with. The names are not in one place: some are fixed to the interface, some travel with players across the screen, and some appear for a second and scroll away.

1. Tell Vizard Agent to blur every username in the recording.
2. Say what must stay readable — the action, the score, your own tag.
3. Expect the moving ones to be tracked rather than boxed over.

What do you need before you start?

The recording and a decision about your own name. Vizard Agent can cover everything or leave your tag visible, and it is worth saying which, because "blur all the player names" usually means everyone except the person posting.

• The recording. At the resolution you will publish.
• Whose names go. Everyone, or everyone but you.
• What must stay readable. Score, timer, objective, chat.
• The game. Its interface decides where names live.
• Where it will be posted. It sets the final crop.

What do you type into Vizard Agent?

Ask for all of them rather than pointing at one. Naming a single example anchors the job to that one kind of overlay, and the kill feed at 1:40 will still be carrying four usernames when you upload it.

Prompt

Variants worth knowing:

• "The tags floating above players." The kind people forget.
• "Keep my own tag visible." Usually the actual intent.
• "Do not cover more than you have to." Stops a blurred band across the top.

What does Vizard Agent actually do?

Here is the Vizard Agent sequence on a round of team shooter footage. Most of the list is spent finding where the names actually appear, because the answer turns out to be four different interface systems that each behave differently on screen.

1. Builds a contact sheet of representative frames.
2. Extracts high-resolution crops of the HUD to map the name positions precisely.
3. Analyses the whole match to identify every type of username overlay and when it appears.
4. Builds a focused sheet around the moments names were detected — gameplay, kill feed, spectator views.
5. Enlarges each candidate region at original resolution before sizing any mask.
6. Confirms the frame geometry and the exact pixel locations of the fixed overlays.
7. Checks the centre-screen and death-card regions for overhead and spectator names.
8. Makes coordinate-labelled frames so the moving tags can be followed.
9. Maps the tag trajectories through corridors, doorways and stairwells without obscuring the action.
10. Builds a close montage of the mid-screen tag area across the whole timeline.
11. Handles the overlapping cases where two tags cross in a corridor.
12. Tracks the kill feed separately, since it scrolls rather than moves.

Step three is the step that makes the rest possible. Rather than blurring what it can see on one frame, Vizard Agent asks what kinds of name overlay this game has and when each appears, which turns an endless frame-by-frame hunt into four bounded problems.

Step nine is where the craft is. A floating tag follows a player behind cover, through a doorway and up a staircase, and a box big enough to cover it everywhere would cover the part of the screen you are posting the clip for.

Step eleven is the case that catches automated approaches. Two teammates cross, their tags overlap for a few frames, and a tracker that assumes one box per tag hands you a moment where a name is legible.

What does the result look like?

The clip you wanted to post, with other people's names unreadable in all four places they appear, your own tag intact if you asked for that, and the action, the score and the timer untouched. The blurred regions move with the tags rather than sitting as permanent panels over the picture.

When does this not work well?

Some footage cannot be cleaned to a publishable standard. Names baked into a busy centre-screen scoreboard, tiny text in a compressed stream recording, or a game whose chat shows full names in a scrolling panel may leave something readable at full resolution.

• Heavily compressed recordings. Small text is mush, but still legible enough.
• Names inside a scoreboard. Covering them covers the scores.
• Chat panels. Better cropped out than blurred.
• Very long recordings. Every kind has to be tracked throughout.
• Names on skins or in-world signage. Not part of the interface at all.

How do you fix a result that came back wrong?

Give a timestamp and say which kind of name was missed at it. Because Vizard Agent handles each kind of overlay separately rather than as one blurred region, it can extend one of them without enlarging the others or covering any more of the picture than before.

• "A name is readable at 1:40." That kind is re-checked across the whole clip.
• "The blur covers the action." The moving mask is tightened to the tag.
• "You blurred my own name." Your tag is excluded and the rest kept.
• "The kill feed still shows names." It is tracked as a scrolling region.

How does Vizard Agent compare to doing it yourself?

By hand you put a blur box over the top of the frame and call it done. It covers the HUD, it covers a quarter of the picture, and it does nothing at all about the tag floating over the teammate in the middle of the screen — which is the one a viewer can actually read.

Common questions

Will it blur my own name too? Only if you want. Tell Vizard Agent to keep your tag visible.

Does it work for any game? Yes. Vizard Agent maps that game's own overlays rather than assuming a layout.

Can it black them out instead of blurring? Yes. Ask Vizard Agent for a solid box, which reads as deliberate.

Will the score still be readable? Yes, if you name it to Vizard Agent as something to keep.

What about names in the chat? Vizard Agent can blur them, though cropping the panel out is often cleaner.

Can it do a whole stream recording? Yes, though the tracking Vizard Agent does scales with the length.

Is blurring enough for privacy? Strong blur or a solid box is. Vizard Agent checks the result at full resolution.

What if two tags overlap? Vizard Agent handles that case separately so neither name becomes readable.

Can it keep the kill feed but hide the names? Yes. Vizard Agent keeps the feed and hides only the names inside it.

Does it re-encode the whole video? It renders a new file; your original recording is untouched.

Can it also remove the interface completely? That is a different job, and usually a reframe rather than a blur.

Will it slow down the video? No. Vizard Agent keeps the frame rate and the length as they are.

How do I check it myself? Pause at full resolution on a kill-feed moment; that is where names survive.

Can it handle a spectator view? Yes. Death cards and spectator names are one of the kinds it looks for.