Meta Patents a Headset System That Mutes Audio When Others Are Listening
Meta has filed a patent for a system that can detect when multiple people are tuned into the same audio stream and automatically mute the sensitive parts. It pairs that trick with a separate ultrasonic sensing method that could sharpen how wearables "feel" their environment.
What Meta's listener-detection audio muting actually does
Imagine you're wearing a pair of AR glasses on a crowded train, listening to a private voice message. Someone nearby plugs into the same stream, or the device notices sound is leaking to another listener. Instead of playing on regardless, the device detects the situation and mutes the parts of the audio that shouldn't be shared.
That's one half of what Meta describes in this patent. A system spots when more than one person is receiving an audio stream at the same time, then automatically silences the portion that was meant to be private. The other listener hears a gap rather than your private content.
The same filing also covers a separate sensing technique. An array of tiny sensors fires a sound pulse, records the echo, reverses it in time, and fires it again. That process refines the echo into a cleaner signal, which could help a device map surfaces, detect motion, or sense what's around it with greater accuracy.
… a method for detecting, by a system, a plurality of users concurrently listening to an audio stream, muting, by the system, at least a portion of the audio stream, and transmitting the audio stream comprising the muted portion of the audio stream.
Translation: The system detects when multiple people are listening to the same audio and automatically mutes parts of it.
How the sensor array catches and reverses sound pulses
The patent covers two distinct methods bundled into a single filing.
Method one: listener detection and muting. A system monitors an audio stream for signs that multiple users are receiving it at the same time. When it detects concurrent listeners, it identifies a portion of the stream as private and mutes it before transmission continues. The muted audio is still sent as a placeholder, so the stream doesn't break, but the content is withheld.
Method two: time-reversed ultrasonic sensing. This is the more technically involved piece. An array of sensors fires a pitch-catch waveform (essentially a sound pulse sent from one point and caught at another). The system then:
- Records that waveform
- Reverses it in time (plays it backwards, a technique used in acoustics to focus signals)
- Emits the reversed version
- Records the echo that bounces back (pulse-echo measurement)
- Combines the echo with the reversed original to produce a cleaner final signal
Time-reversal in acoustics is a known way to sharpen a signal: imperfections in the original transmission tend to cancel out when the wave is played back in reverse and the echo is captured. The result is a more precise measurement of whatever the sound reflected off.
What this means for private audio on AR glasses and headsets
For anyone using Meta's Ray-Ban glasses or a future headset in public, automatic audio privacy detection would be a meaningful quality-of-life feature. Right now, if audio leaks or is intercepted, there's no automatic backstop. A device that can notice and react to that situation in real time addresses a real privacy gap.
The ultrasonic sensing method is harder to tie to a specific product today. Time-reversal signal processing is used in medical imaging and structural testing, but in a wearable context it could improve gesture detection, surface mapping, or proximity sensing. How far Meta is from fitting that hardware into a consumer device is the open question.
Meta's 70th filing we've tracked in the AR glasses race since May adds to a picture that includes shared AR objects fixed in place and lenses built into the frame.
The listener-detection and muting concept is the more immediately legible half of this filing, and it does address something real. Audio privacy on wearables is a mostly unsolved problem, and a software-layer solution that detects concurrent listeners doesn't necessarily require new hardware.
The time-reversal ultrasonic sensing method is a different story. Signal processing of this kind requires a sensor array capable of emitting and capturing ultrasonic pulses with enough precision that the reversal step actually improves the result. That's specialized hardware, and there's no evidence it exists in Meta's current consumer lineup. The shortest path to a product here runs through a hardware design cycle, not a firmware update.
Taken together, this reads like a single filing that bundles two loosely related ideas under one roof. One could ship as a software feature relatively quickly; the other is early-stage enough that it belongs in a research roadmap, not a product spec. a growing pile of Meta sensing and audio filings suggests the company is staking out broad IP territory across wearable perception, but broad territory and near-term products are different things.
There are more where this came from
We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.
The drawings
24 drawing sheets from US 2026/0301726 A1 · click any drawing to enlarge
Want this weekly breakdown for a company we don't cover? Patentlyze Pro →
Be the first to weigh in