Blur Part of a Video — Blur, Pixelate or Black Out

Censor Your Video Free

No masks, no keyframes — describe what to hide and review the preview. No credit card required.

How to Blur Part of a Video with Valmera

1. 1
   Upload your video

   Drag in a file up to 2GB and 3 hours. Valmera analyzes it once — building a word-level transcript and detecting shots — with visible progress on screen.
2. 2
   Type what to hide

   Try: "blur the license plate at 0:42", "pixelate the email address in the top-right of the screen recording", or "black out the bottom third of the frame from 5:10 to 5:30". Name the style, the place, and the time range.
3. 3
   Preview and export

   Check the covered area in the preview, refine with a follow-up ("make the blur region a bit bigger"), then export a full-quality H.264 MP4 rendered from your original upload.

The one-time upload analysis runs with visible progress — longer videos take longer. After that, every censor is a chat message.

Blur a Face, a Username, or a License Plate by Describing It

Censoring in a normal editor means drawing a mask, keyframing it, and hoping nothing moves. In Valmera you describe the thing: a bystander's face in the background, a customer name in a CRM screen recording, a house number in a walkthrough, an email address in a demo, a plate in a car vlog. The agent finds that part of the frame and covers it.

Because it is a conversation, correcting it is a sentence too: "a bit lower", "make the region wider", "keep it up for the rest of the video". Each adjusts the existing censor rather than restarting the edit, and the reply states exactly what changed.

Blur, Pixelate, or Black Out — Three Censoring Styles

The same region can be hidden three ways, and the choice matters more than people expect. Blur is the softest and looks the most natural in documentary-style footage — the frame still reads as a whole. Pixelate is the familiar mosaic; it signals clearly to viewers that something was deliberately hidden, which is often what you want for a name or a face. Black out is a solid block: the most absolute and the most obviously intentional, ideal for redacting text in a screen recording.

Ask for a style by name and switch it later if it reads wrong — "make that a pixelate instead of a blur" is one message. Multiple regions can be covered in the same video, each with its own style and its own time window.

Censoring That Survives Your Cuts and Reframes

This is the part manual editors get wrong. Censors in Valmera are placed in source-frame coordinates — they belong to your original footage, not to the current output frame. Reframe a 16:9 video to 9:16 for Reels and Shorts and the covered area stays covered, with nothing to re-apply and no risk of the blur sliding off the thing it was hiding.

Cuts behave sensibly too. An always-on censor survives later trims, so removing silences or filler words afterwards can't accidentally expose a plate. A censor limited to a time window follows the footage it covers — if that footage is cut, the censor is trimmed with it. Censor first and edit after, or edit first and censor after; the order doesn't trap you.

The Honest Limit: No Motion Tracking (and What to Do Instead)

Valmera's censors cover a fixed region of the frame. They do not motion-track a subject moving across the shot, and they do not follow camera motion. "Blur his face as he walks across the room" is the one request this tool cannot honor, and the agent says so instead of shipping a blur that drifts off the face — that is the honesty layer working as designed.

Three practical workarounds cover almost every real case:

• Size the region to the whole path. If the subject moves within a corner of the frame, one generous region covering that corner hides them for the entire span.
• Split it into time windows. "Blur the left third from 0:10 to 0:18, then the center from 0:18 to 0:26" steps the coverage along with the movement — a manual track, but a two-sentence one.
• Cut around it. If a bystander only crosses frame for a second, the fastest fix is often to remove that second: describe the moment and the agent cuts it.

Static targets — screen-recording text, a fixed plate, someone seated in the background, a logo on a wall — need none of this. They are a single request and they stay covered.

Who Needs to Censor Part of a Video?

Anyone shipping a product demo with real customer data on screen. Support and onboarding teams turning a call recording into a help video without exposing an account name. Vloggers who filmed a street and now need a plate or a stranger's face gone. Educators sharing a student's work anonymously. The common thread: you don't want to re-record, and you don't want to learn masking to fix a ten-second problem.

More Than a Video Blur Tool

Censoring is one instruction inside a full agentic editor. The same chat can resize the video to 9:16, 1:1 or 4:5, cut any moment you describe, color grade the footage, add word-accurate captions, and mix music smoothly under your voice — so a raw screen recording becomes a finished, publishable video without you opening a timeline.

Browse all AI video editing tools, read the effects documentation for exactly how censors behave, or see plans and pricing when you outgrow the free tier.

Frequently Asked Questions

Hide It in One Message

Upload your video, say what to blur, review the preview. Free plan — no credit card required.