Launching today
Syntonym Blur Studio
Detect, draw, or describe what to blur in images & videos
17 followers
Detect, draw, or describe what to blur in images & videos
17 followers
Syntonym Blur Studio anonymizes faces, people, vehicles and license plates in images and video. Detect them automatically, draw a region by hand, or just describe what to hide. No install, runs in the browser, exports at original resolution.
Free Options
Launch tags:Privacy
Launch Team
Render Sandboxes: a computer for every agentGive every agent its own machine. Starts in under a second
Promoted
@Syntonym Blur Studio
Hey Product Hunt 👋
I'm Uslu from Syntonym. We've spent the last few years building anonymization tech for automotive and robotics teams who need to share camera data without exposing people's identities.
Along the way we kept seeing the same problem at a smaller scale: creators, journalists, and small teams blurring faces and plates frame by frame in video editors. Hours of work for a 2-minute clip.
So we built Syntonym Blur, with three ways to hide what shouldn't be seen:
🔍 Detect: faces, people, vehicles and license plates are found automatically
✏️ Draw: mark any region by hand and it's tracked through every frame
💬 Describe: type what to hide in plain words ("the laptop screen", "the logo on his shirt") and it finds it for you
You can then pick blur, pixelate or blackout per target. It works on images and video, runs in the browser with nothing to install, and exports at original resolution.
🎁 Launch offer for the PH community: use code HUNT20 for 20% off any plan (one-time discount).
We'd love your honest feedback and support. Tell us where it fails, that's how we make it better.
Love it, but the tool keeps the original?
@thomas_wazza
Great question, and a crucial one for a privacy tool. Short answer: your original is kept only as long as you want it to be.
Here’s how it works:
Analyzed once: your file is scanned a single time to find what needs blurring, and it isn’t reused for anything else. All the editing and blurring after that happens right in your browser.
Where, never who: our AI finds where a face or plate is, not whose it is. No face recognition, no plate reading, no identity matching.
Export & Delete: open the dropdown next to Export and pick “Export & Delete”. It’s the fastest way to remove the original once you have your result.
Auto Delete: under Profile → Data Management, turn on Auto Delete and choose a retention period (e.g. 24 hours). Uploads are then removed automatically, even if you forget.
And on our roadmap: a version that works entirely on local files, which has been part of our core design from the start. We’ll open it up in the browser soon.
Thanks for asking!
Combining automatic detection with a hand-drawn region sounds useful for the things a detector misses. Flagging the frames where a target disappears or tracking becomes uncertain would help someone review the export before sharing it.
@oleksii_sekundant
Thanks, this is exactly the use case we had in mind! Detection gets you most of the way, and drawing covers the long tail a detector misses.
You're also spot on about review. For a privacy tool, the frame where a target slips out of its region matters more than the 99% that went fine. Flagging low-confidence and lost-track frames on the timeline, so you can jump straight to them before exporting, is on our roadmap.