How to Anonymize Multiple Faces, Collaborators, and Bystanders

To anonymize multiple faces, list everyone who appears, choose a treatment for each person, and follow every visible appearance through the photo or video. In moving footage, check each mask when people cross, turn away, leave, or come back into view. Review the finished export one person at a time, including anyone visible in a reflection or at the edge.

You don't need to give everyone the same appearance. One collaborator may want their face visible, another may prefer an opaque cover, and a passerby may be easiest to remove with a crop. The useful starting point is a clear agreement about who can appear and a way to check that the edit follows it.

Agree on each person's appearance before editing

For a planned collaboration, show everyone what you intend to publish. A two-person tutorial, a behind-the-scenes photo, and a short promotional clip can expose different things even when they come from the same shoot. Discuss the actual files and where you'll post them.

Ask each collaborator which face treatment they're comfortable with, whether their natural voice can remain, and whether you can use a name or tag. Include any detail they want hidden, such as a work badge or tattoo. A message saying “you can use the clip with my face covered and no name tag” gives the editor more useful direction than “looks fine.”

Discuss uploads as well. If an editing service needs the original video, tell collaborators that the service will receive the unedited video. If that doesn't suit the group, use an appropriate local editing workflow or film a version that keeps those faces out of the source.

For people who happen to be in the background, leave them out of your post where possible. The EFF's guidance on photographing other people recommends seeking consent when possible and obscuring identifiable people when consent isn't available. It also calls attention to identifying clothing and tattoos. For creator footage, that means checking more than the face before deciding someone is adequately hidden.

If a bystander says they don't want to be included, choose another take, crop them out, or remove the relevant passage. Don't use a realistic face edit to include someone who asked to be left out. Keep permission to participate separate from the technical ability to edit a picture.

Write the agreed rules somewhere the editor and reviewers can find. Keep names and contact details out of public captions and filenames unless their publication is part of the agreement. Check any platform requirements for participants separately; covering a face doesn't settle those requirements.

Make a list of everyone visible in each shot

Start by watching the unedited sequence privately without adding effects. Give each person a short label, such as A, B, and C. Note when they first appear, when they leave, and any place where they become hard to follow. Add a brief visual description if it helps distinguish them while editing.

Count appearances as well as people. Your collaborator and their reflection are one person but two areas to cover. A face on a phone screen, a printed photograph behind the desk, or a second camera angle may need a separate treatment. Don't expect a mask on the main face to address those other appearances.

For a still photo, scan the frame in sections: foreground, background, edges, then reflective surfaces. For video, repeat that scan through the shot. A person who is hidden behind your shoulder at the beginning may be fully visible when you lean forward.

Here's a simple working list for an imaginary workshop clip:

• A, host: visible throughout; approved to appear with their real face.
• B, collaborator: visible from 00:02 to 00:18; face must be covered, voice approved.
• B, window reflection: visible from 00:07 to 00:10; needs a separate cover.
• C, passerby: crosses the doorway from 00:11 to 00:13; remove that background area or cut the passage.

Keep the list short enough to use while editing. Its purpose is to prevent you from finishing the obvious foreground face and forgetting a second appearance. If a scene contains too many overlapping people to follow confidently, simplify the shot before spending time on individual masks.

When you can still change the shoot, place collaborators farther from reflective surfaces and keep the camera away from busy doorways. The face-free filming workflow covers planning those shots. Here, use the list to decide what the footage you already have needs.

Choose a treatment for each person

Choose according to what the viewer needs to see. A collaborator demonstrating a hand technique may not need a visible head. A person passing behind them probably doesn't need to remain in the frame at all. Preserving a face-shaped appearance is only useful when it serves the scene and the person agrees.

An opaque cover can be a plain rectangle, a solid oval, or a graphic that fits your video. Keep the portion over the face fully opaque and large enough for movement. A translucent sticker, a hole between graphic elements, or a soft edge can leave part of the face visible.

Blur keeps some of the scene's colors and shapes. Judge the result at the size you'll export, including paused frames. If you can still make out facial details you intended to hide, increase coverage or choose an opaque cover or crop. There's no universal blur setting that you should copy across every face size and resolution.

If you want an anonymized appearance for your own separate shot, the NeoFace walkthrough describes selecting a synthetic face, uploading a photo or prerecorded clip, comparing the processed result with the source, and downloading it for review. A short single-person test can help you assess that workflow for your part of a collaboration.

That walkthrough doesn't explain how to handle several faces at once. For the shared scene, use a method that lets you select and review everyone who needs treatment; don't assume one processed face means the whole group is handled. Keep bystanders and reflections on your list even if the foreground output looks convincing.

For a broader editor comparison, use the guide to apps that hide faces in video. The job here is to finish the whole group scene, including corrections, with whichever suitable editor you choose.

How to hide several faces in a photo

Work on a copy and settle the final composition first. If cropping removes two unnecessary people while keeping the subject clear, do that before drawing individual covers. Then return to your list and deal with every remaining face that needs treatment.

For each face, inspect the forehead, cheeks, chin, and side view. Cover the full area you intend to hide, including any portion visible around glasses or beside a hand. If hair or clothing is also identifying, decide whether the cover needs to include more of the person. An eye bar alone leaves much of the face visible.

Use separate shapes when people are spaced apart. That makes it easier to move one cover without changing the others. Where several bystanders overlap in the background, one larger covered area may be clearer and easier to inspect than a cluster of tiny circles. Check that the treatment doesn't hide the action the photo is meant to show.

Export a finished image, then open that file outside the editor. If you're using Photoshop, its Save a Copy workflow supports creating a flattened JPEG or PNG while retaining the original working document. Share the reviewed finished image for publication; keep the editable source separate.

Review every image in a carousel individually. A successful cover on one photo doesn't tell you where the same person appears in the next. Also inspect any tighter crop you make for a cover or profile preview. Finish with the file you will upload, rather than a screenshot of the editor with selection outlines still visible.

How to track several faces in prerecorded video

In an editor, a mask is a selected area where an effect applies. Tracking moves that area through the clip. A keyframe saves a position or another setting at a particular moment, so you can adjust the movement over time.

Give each moving face its own coverage. Adobe explains how to use several masks with independent tracking. It also notes that an unassigned mask has no effect by itself. Check the visible result with the blur or other intended effect enabled, rather than treating an outline as completed concealment.

Use this sequence for a manageable shot:

1. Finish the rough cut. Remove unwanted passages and choose the framing before detailed tracking.
2. Choose one person. Find a clear frame where their face is easy to locate and size the coverage around it.
3. Track the visible interval. Follow that appearance toward its beginning and end using the editor's controls.
4. Inspect the result. Correct position, size, or shape wherever the coverage slips.
5. Handle separate appearances. Add coverage for a reflection or a later re-entry as needed.
6. Repeat for the next person. Keep the first person's treatment enabled while you review everyone together.

Adobe's mask-tracking instructions include forward, backward, and one-frame tracking controls, followed by review and adjustment of masks or keyframes. Use those smaller corrections where a long automatic track stops following the intended face.

YouTube offers another option for a video already in Studio. As of September 2026, its blur instructions describe selecting faces after processing, plus Custom blur with adjustable boxes, timing, and a Track object setting. Check the selected faces against your own list and add custom coverage where needed.

Studio editing also means providing the video to YouTube. If your group doesn't want the unedited source uploaded there, finish the edit elsewhere. Keep a separate working copy before saving Studio changes: YouTube says its editor no longer offers Revert to original after saving. Review the resulting video before using it as your publication copy.

Whatever editor you choose, label masks or effects clearly where the software allows it. “B face” and “B reflection” are easier to review than several indistinguishable effects. Keep a short correction note for moments that need special attention, so the next reviewer knows where to look.

Fix crossings, obstructions, and re-entry

A straight walk across an empty background is only one kind of movement. Group footage also contains people stepping in front of each other, turning into profile, bending down, and returning from outside the frame. Review those moments deliberately instead of judging the treatment from the clearest section.

When two people cross, follow each separately. Begin just before the overlap and step through until both are clear again. Confirm that each cover stays with its intended person. If one face should remain visible and the other hidden, inspect both after the crossing so you don't miss a treatment applied to the wrong person.

When a hand or object hides part of a face, check the reveal. Keep coverage over any facial area that remains visible. Inspect the first frame where the face emerges again; don't wait until the person is fully clear to restart the treatment. A short manually corrected section may be easier to manage than repeatedly rerunning the whole shot.

When someone leaves and returns, inspect both boundaries. End coverage only after their last visible facial detail disappears. Start the next section at their first visible return. Treat the re-entry as another appearance to check, even when you're confident it's the same person.

When the face changes size, resize the treatment. A cover that fits someone at the back of the room may be too small when they approach the camera. Check turns as well: a profile can extend outside the shape you drew around a front-facing head.

If a section keeps failing, choose a simpler repair. Widen the covered area, cut to a close-up of the task, or omit the difficult seconds. A short video with complete, reviewable coverage is more useful than a longer version containing a face you meant to hide.

Work through a shared scene from start to finish

Return to the imaginary workshop clip. A is explaining a craft technique, B is helping, and C crosses the doorway. B has agreed to keep their voice but wants their face hidden. The window also shows B's face for a few seconds.

First, try a tighter crop that keeps both pairs of hands and removes the doorway. Watch the entire shot after cropping. If you later reframe the video to follow B, keep the doorway outside that version too.

Next, give B's face an opaque cover and track it through the working area. At 00:08, B reaches across A. Review that crossing frame by frame and correct the cover until B's visible face remains covered throughout. Add a separate cover for the window reflection over its full visible interval.

Then watch the result for usefulness. Can viewers still see where the tool touches the material? If the cover blocks the demonstration, cut to a hands-only close-up for that section. You don't need a more elaborate face effect when another shot explains the technique better.

Finally, export the clip and ask B to review that version. Include the intended thumbnail and short teaser. If B spots a badge you overlooked, correct it before making the publication copies. For repairs elsewhere in the picture, the background-anonymization guide covers visible surroundings in more detail.

Review the result one person at a time

Open the finished export outside the editing project. Watch once while following A, then again for B, and again for any other person or separate appearance on the list. This gives you a concrete review task when several things are moving at once.

After those passes, scan the whole picture for anyone missing from the list. Include the first and last frames, transitions, and any added photos. Listen for spoken names or off-camera voices that conflict with what people approved. Face coverage won't remove a name from the soundtrack.

Use this checklist before sharing the publication files:

• Everyone visible has an agreed treatment or has been removed from the shot.
• Each reflection, screen image, and repeated appearance has been checked.
• Faces intended to remain visible have the correct treatment.
• Hidden faces stay covered during turns, crossings, and size changes.
• Coverage includes the first and last visible frames of each appearance.
• No cover leaves an unintended transparent area or exposed edge.
• Background people missed by automatic selection have been handled.
• Voices, names, tags, and identifying details follow the agreed plan.
• The edited scene still shows the action viewers need to understand.
• Each collaborator has reviewed the exact proposed files.
• Covers, teasers, and alternate crops have their own review.
• The upload files are clearly separated from unedited sources and earlier versions.

Use clear filenames such as workshop-v03-approved.mp4, and send collaborators the approved publication assets instead of asking them to recreate the edit from raw footage. If someone makes a new crop or teaser, bring that file back for review. Otherwise, the group can end up publishing several versions with different coverage.

For the wider check of audio, metadata, captions, and the upload package, use the full faceless-video privacy checklist. Once the group edit is ready, publish the version everyone reviewed and keep that version identifiable for future reposts.