Export a frame, blur the face, and re-use the still — all processing stays in your browser.
Video redaction usually starts with a single frame: pause the clip, export or screenshot the moment, then obscure the face before the still travels to Slack, a newsroom, or a parent group chat. HideShot handles that frame locally with blur, pixelate, or black-box modes.
Or click to browse · Paste with Ctrl+V also works
Drop your image in or paste from clipboard.
Black Box, Blur, or Pixelate.
Rectangle, oval, or freehand lasso ΓÇö then hide what you selected.
Hit Download PNG. Done.
Video clips shared on social media, in HR investigations, or with family often contain one face that should not be identifiable — a bystander at an event, a minor in the background, or a witness whose identity must stay protected. HideShot does not re-encode video files; instead it redacts the still frames people actually share: exported PNGs from an editor, screenshots of a paused player, or frames grabbed before posting to Instagram or TikTok.
The people who land here are usually editing a single problematic moment rather than a full film. A teacher blurring a student behind a presenter in a recorded assembly. A neighborhood group removing a passerby from doorbell footage shared on Nextdoor. A journalist preparing a still for publication when re-shooting is impossible. Each needs the face gone from the image file that will be attached to email or posted publicly — and needs that work done without sending raw footage to a cloud editor.
Priority targets in a paused video frame are the full oval of the face (forehead through chin), distinctive hair silhouettes, profile outlines when the subject is turned partly away, and reflections that duplicate the face in windows or phone screens. Video compression softens detail but facial geometry often survives — especially in 1080p doorbell or conference recordings.
Bystander faces in viral clips create consent and safety issues that still images inherit frame by frame. A face that is identifiable in one exported still can be reverse-searched or recognized by acquaintances even when the original MP4 is never published. Redacting before the still leaves your machine closes that path.
Conference and classroom recordings add policy layers: minors, employees under investigation, and patients accidentally visible in telehealth backgrounds all appear in video frames that get screenshot and circulated. Local redaction keeps those frames off third-party servers during editing.
Scenario A — Doorbell footage: A homeowner wants to share a package theft clip but blur the neighbor who walked into frame. They pause the clip, screenshot, blur the neighbor's face in HideShot, and attach the still to a police report.
Scenario B — Zoom recording: An HR team exports a frame showing inappropriate background content plus a bystander's face. They pixelate both regions locally before adding the still to an internal case file.
Scenario C — Youth sports: A parent clips a goal celebration for Facebook but must obscure non-team children. Frame export, per-face blur, then re-import to a mobile editor for the final reel.
Using a thin blur on heavily compressed video frames. Compression already preserves structure; weak blur leaves recognizable geometry. Prefer pixelate at 12px blocks or a solid black box for high-stakes subjects.
Exporting a tiny thumbnail frame. Low-resolution exports make redaction boxes imprecise. Export at the highest resolution your player allows before editing.
Forgetting faces in mirrors, TV screens, or picture-in-picture tiles. Scan the whole frame — video calls often show the same person twice.
Raw video frames can be more identifying than finished posts because they are pulled from contexts viewers recognize — your home layout, a school logo, a clinic waiting room. Sending those frames to a server-based editor creates a copy on someone else's disk before you have obscured the face. HideShot keeps the sensitive still on your machine: local load, local canvas, local download. That matches how most people actually 'redact video' today — one shared frame at a time — without introducing a cloud vendor into the chain.