How to blur a face in iMovie

How to cover a face in iMovie

This is a workaround, not a built-in feature — iMovie has no blur or pixelate filter to reach for. It works, but it's slow.

1. Add the clip to your timeline and split it wherever the person moves noticeably. iMovie has no motion tracking, so you'll be repositioning things by hand at every cut.
2. Duplicate the clip (or export a still frame) onto the row above as a cutaway. Select the clip, choose Picture-in-Picture, and drop the copy directly above the original at the same point in time.
3. Crop the overlay tightly to the face. Use the crop handles on the picture-in-picture window to shrink it down to roughly head size.
4. Cover the face rather than blur it. iMovie has no blur or pixelate effect, so the overlay itself needs to already be obscured before it goes in — fill it with a solid-colour clip, or pre-blur a still frame in another app and bring that image in as the overlay instead.
5. Reposition the overlay for every split. The picture-in-picture window doesn't move on its own, so repeat the resize-and-place step for each segment, nudging the box to track where the face is in that portion of the clip.
6. Play the whole thing back at normal speed. Gaps between cuts are where the cover visibly lags behind the face — that's usually where you need another split.
7. Share or export once every segment lines up.

Where this breaks down

The core problem is that iMovie was never built to obscure part of a frame — Picture-in-Picture exists for cutaways and reaction shots, not redaction. There's no blur filter to fall back on, so the overlay has to already be hidden before it's placed, which means a trip to another app for a pre-blurred still if you want anything softer than a flat colour block.

Because the overlay window is a static rectangle with no tracking, a face that moves at all means splitting the clip into short segments and repositioning the box for each one. A few seconds of someone walking can easily be a dozen splits. Multiple faces in the same shot means a second overlay track, cropped and repositioned independently of the first — and the two have to be checked together so they don't overlap or drift apart.

It's also a rectangle, not a face shape, so a tight crop on a moving head either clips the edges or leaves a visible margin of unblurred skin as the person turns.

None of this is destructive in the useful sense, either. Because the cover is a separate clip sitting on top rather than an effect baked into the pixels underneath, anyone who gets the original project file, or who screen-records around the overlay's edges, still has the unobscured footage sitting one layer down.

What tends to go wrong

A few specific problems come up often with this workaround:

• The overlay lags behind the face. Because repositioning only happens at a cut, any movement inside a segment goes uncovered until the next split.
• A pre-blurred still doesn't match the lighting or motion of the video, so the overlay reads as an obvious patch rather than a blend.
• Audio doubles up if the overlay clip isn't muted, since duplicating video usually duplicates its sound too.
• Multiple faces need entirely separate overlay tracks, stacked and timed independently, which gets hard to manage in a short clip with more than one or two people.

iMovie vs. Deface

The faster way

Deface exists for exactly the gap iMovie leaves open. Import the same photo or video and it detects every face on device in about half a second, tracking each one automatically as it moves — no splitting, no repositioning a picture-in-picture window by hand.

Choose blur, pixelate, a solid block or an emoji, and it's applied to every detected face at once, switchable afterwards without re-processing. Each face has its own toggle if you want to leave someone visible, and a manual region tool covers anything the detector isn't meant to find, like a licence plate or a street sign. Export strips the photo's metadata too, which iMovie leaves untouched.