How to Blur Faces in Video — Free Tools vs. Automatic Services, and How to Choose

When you post a video to social media or YouTube, you often want to hide the faces of passersby or children who happened to be on camera. In this article, we compare the main ways to blur faces in video — based on official sources — and lay out how to choose the method that fits your needs.

Key takeaways

• The four main ways to hide faces are YouTube’s built-in blur, manual editing software, mobile apps, and automatic services.
• The more people, camera movement, and runtime your video has, the more the manual effort explodes.
• YouTube’s blur only works on videos uploaded to YouTube, and since June 2025 it carries restrictions such as edits that can’t be undone (YouTube Help).
• Video sharing is now everywhere across all age groups (in Japan, LINE 91.1% and YouTube 80.8%, per the Japanese Ministry of Internal Affairs and Communications, FY2024 survey), so being mindful of incidental faces matters more than ever.

Why face mosaicking matters

Video sharing has reached every age group. According to a survey by Japan’s Ministry of Internal Affairs and Communications (MIC), the all-ages usage rate is 91.1% for LINE and 80.8% for YouTube (MIC, “Survey Report on Information and Communications Media Usage Time and Information Behavior, FY2024,” published June 2025). In an era when anyone can film and publish video at a moment’s notice, it’s important to be mindful of other people’s faces caught in the frame.

The main ways to blur faces (compared)

There are four common approaches, each with its own strengths and weaknesses. Let’s start with the big picture in a table.

Method	Auto-tracks faces	Many people / long videos	Where it works	Effort
YouTube blur	Yes (auto-detect)	Limited	YouTube uploads only	Low
Editing software (manual)	Needs tracking setup	Limited (work scales per person)	Anything	High
Mobile app	Depends on the app	Limited	Anything	Medium
Automatic mosaic service	Excellent (all faces auto)	Excellent	Anything	Low

1. YouTube’s blur feature

YouTube has an official “blur” feature with two modes: face blur (auto-detect) and custom blur. You can choose a rectangular or elliptical shape, and have it automatically track a moving subject (YouTube Help, “Blur parts of your video”).

There are caveats, though. According to the same help page, edits may not save on videos with high view counts, and since June 2025, blur edits can no longer be reverted with “undo.” And of course it’s only for videos uploaded to YouTube — you can’t use it on video destined for other platforms.

2. Manual mosaicking in editing software

Editing software like DaVinci Resolve and Premiere Pro includes a mosaic (blur) effect plus tracking to follow a subject. In DaVinci Resolve, you combine a Power Window (mask) with the Tracker to keep up with a face (Blackmagic Design official manual). Premiere Pro does the same by tracking a mask on the Mosaic effect (Adobe Learn).

The finished result is fully customizable, but you have to adjust the position face by face and cut by cut — and with many people or a moving camera, the effort spikes fast.

3. Mobile apps

Apps are convenient, but auto-tracking accuracy depends on the app, and they can miss profiles or small faces. Apps without dedicated auto-tracking fall back to manual keyframes, which can’t keep up with large movements (CyberLink blog). They’re a poor fit for long or high-resolution video.

4. Automatic mosaic services

You just pick a file, and AI detects every face and applies the mosaic automatically. The strength is that it can process all faces regardless of how many people or how much the camera moves. Mosaic Post uses this approach, fading the boundary smoothly along an ellipse that follows the shape of the face — so you don’t get an unnatural square box.

Services like this normally send your video to a server. Mosaic Post instead runs the whole thing inside your browser, so footage containing other people’s faces never has to be handed to anyone.

How it works with Mosaic Post (3 steps)

1. Pick your video on the home page (no account, no upload).
2. Take any faces you want left visible out of the detected list (or just hide everyone).
3. Save the file once the export finishes.

You can choose from three concealment styles: mosaic (pixelation), blur, or solid fill. It works on videos up to 3 minutes and costs nothing. Large files such as 4K are scaled down to Full HD on export, and processing takes roughly 1.3 times the video’s length on a computer, several times that on a phone.

Choosing the right method

• A few cuts with few faces → your existing editing software can handle it.
• Many people / moving camera / long runtime → an automatic service saves overwhelmingly more effort.
• Need it for platforms beyond YouTube → a platform-independent automatic service.

For use-case examples, see Faces of passersby in vlogs and street-walking videos. For the thinking around portrait rights, see How to hide other people’s faces (the basics of portrait rights and privacy).

Sources

• YouTube Help, "Blur parts of your video" — support.google.com
• Ministry of Internal Affairs and Communications (Japan), "Survey Report on Information and Communications Media Usage Time and Information Behavior, FY2024" (published June 27, 2025) — soumu.go.jp
• Blackmagic Design, "DaVinci Resolve 18 Reference Manual" — documents.blackmagicdesign.com
• Adobe, "Premiere Pro: blur a face with masking and tracking" — adobe.com

Blur faces in your browser → (no upload, free)