How to pixelate a face
Pixelating a face breaks it into visible blocks of colour instead of smoothing it into a blur. It is the look most people associate with 'this was deliberately hidden' — here is when it is the right call and how to do it.
What pixelating a face actually does
Pixelation divides the face into a grid and replaces every block with the average colour of the pixels inside it, so fine detail — eyes, nose, mouth, expression — is discarded and replaced with a mosaic of flat squares. It is the same technique broadcast TV has used for decades to censor faces in news footage, which is why it reads instantly as 'anonymised' to anyone watching, in a way a soft blur sometimes does not.
To pixelate a face, you need something that finds the face first. Doing it by hand in a photo editor means selecting the region, applying a mosaic or pixelate filter, and — for video — repeating that for every frame the face appears in. Deface detects the face automatically and applies the pixelate style in about 0.4 seconds, with the option to switch to a different style afterwards without re-processing.
Pixelate vs blur vs a solid block
All three cover a face, but they say different things and remove different amounts of information.
Blur softens the face while keeping its general shape and colouring. It reads as a documentary or artistic choice more than a redaction, and a light blur can leave enough shape and colour information behind that the image still looks natural.
Pixelate breaks the face into blocks by averaging groups of pixels together, which discards more of the original detail per block than a soft blur does across the same area. It is also the more universally recognised 'this face was deliberately hidden' signal, which matters when you want a viewer to know a choice was made, not just that the photo is slightly out of focus.
Solid colour replaces the region with a flat block of colour and removes all facial pixel information. It is the only one of the three that leaves nothing of the original face in the exported file.
Why block size matters more than it looks like it should
A pixelated face is only as effective as its block size relative to the face itself. Large blocks over a small, distant face can end up averaging so much of the surrounding background into each square that the result looks blocky without actually obscuring much — the face was already low-detail before the effect was applied. The same block size over a large, close-up face leaves more structure recognisable between the blocks, because each square is covering a smaller share of the facial detail.
This is part of why doing it by hand in a general photo editor is fiddly: you're picking a mosaic size that has to be judged against the size of the face in that specific photo, then redone if you crop or resize the image afterwards. Deface applies pixelation at a block size tuned against the detected face's actual dimensions, so a small background face and a large close-up subject both get proportionate coverage without you tuning a slider for each one.
How to pixelate a face on iPhone with Deface
No manual selection and no frame-by-frame work — detection and the mosaic effect are applied automatically.
- Import a photo or video. Deface reads it from your library, up to 4K.
- Let it detect. Every face is found automatically, and tracked across frames if it is a video.
- Choose pixelate as the style. Switch between blur, pixelate, solid and emoji at any point before export, with no re-processing.
- Export. The result keeps the original resolution and, for video, the original frame rate.
Pixelation is more robust than a light blur, but it is not a guarantee. Because it discards more information per block, coarse pixelation is harder to undo with a simple sharpening filter than a light blur is — but that does not make it unbreakable, and how much it protects depends on how coarse the blocks are, the resolution of the original, and what an attacker already knows. If you need the option that removes all facial pixel information from the exported file with no ambiguity, use the solid colour style instead.
When to reach for pixelation specifically
Choose pixelate when you want the viewer to clearly register that a face was hidden on purpose — a news-style photo, a screenshot you are sharing publicly, a listing photo with a bystander in frame. It reads as intentional in a way a light blur does not always manage.
Choose a plain blur when you want the image to still feel natural, such as a background face in a travel photo. Choose solid colour when the priority is removing the face information completely rather than how the cover looks. And if what you are actually hiding is not a face at all — a license plate, a screen, a document — the same styles apply to a manually drawn region.
Frequently asked questions
What is the difference between pixelating and blurring a face?
Pixelation replaces blocks of the face with a single averaged colour per block, discarding more detail per area than a soft blur and giving the recognisable mosaic look. A blur softens the face while keeping more of its general shape and colour, which reads as a softer, more natural edit.
Is a pixelated face impossible to identify?
Not guaranteed. Coarse pixelation removes more information than a light blur and is harder to reverse with a basic sharpening filter, but the strength depends on how large the blocks are relative to the face and the resolution of the source image. The only style that removes all facial pixel data outright is a solid colour block.
Can I pixelate a face in a video, not just a photo?
Yes. Deface detects faces in video and tracks them across frames, applying the pixelate style consistently as the person moves, so you are not pixelating each frame by hand.
How do I pixelate just one face and leave others visible?
Every face Deface detects has its own toggle, so you can apply pixelate to some faces and leave others untouched, or switch individual faces to a different style entirely.
Does pixelating a face remove the photo's location data too?
Not on its own — pixelation only changes the pixels in the face region. Deface strips EXIF, GPS and timestamp data from every export separately, regardless of which censor style you use.
Can I control how coarse the pixelation is?
The pixelate style is applied at a consistent block size tuned for reliable coverage of a detected face. If you need a different look, blur and solid colour are available as alternate styles on the same detected face or region.
Pixelate every face automatically.
Free to try. One-time Pro unlock for unlimited, watermark-free exports.
