Several iPhones. One camera.
VoxaVerse Beacon turns the iPhones you already own into a synchronized, multi-camera volumetric rig. Set them around the scene, press record once, and every device captures colour, LiDAR depth and audio against one shared rig clock — so the footage lines up in post by itself.
Works standalone. Record a full volumetric take with nothing but iPhones and take the footage anywhere — standard video, with depth, camera intrinsics, lens distortion and exact per-frame capture time recorded beside it. A single device works on its own; two or more make a rig.
A capture rig that fits in your pockets.
One device leads; the others join over your local Wi-Fi network. No timecode hardware, no cabling between cameras, and no cloud in the middle of a take.
Every phone stamps every frame against one clock the rig agrees on, and the app shows the alignment it is actually achieving rather than a number from a datasheet. Exact per-frame time is what lets a solver place each observation at its true instant.
On top of the clock, the rig steers each camera’s exposure window toward a common grid, in software, so the cameras are not merely timestamped together but exposing together.
Where the hardware supports it, a compatible USB-C sync accessory drives readout electrically — the one path that does not depend on a network at all. Genlock and shutter-phase alignment are different mechanisms, and the app never calls one the other.
Depth is recorded losslessly alongside every video frame on devices that have the sensor. Fused capture resolves thin geometry and reflective surfaces that a photo-only rig misses.
Every shutter press delivers a phase-locked sync frame with its depth, plus a full-sensor detail shot — each carrying honest per-still timing metadata rather than an assumed capture instant.
No account, no cloud, no upload. Footage never leaves your devices unless you move it, and the optional subject matte runs on-device too.
Every technical readout in the app is a measurement taken during your take, not a promise made in advance. Sync quality depends on the devices, the network and how warm the phones are — which is exactly why it is on screen.
Not a mockup. The actual app.
Straight off the devices — a live two-iPhone rig, shot on the phones themselves.
The rig starts on a single screen — lead it, or join one as a camera. What this device can contribute is stated before anything records.
Anything on the same local network appears by itself. Here an iPhone 13 Pro is about to join an iPhone 17 Pro’s rig; each device shows its own capability badges, so nobody discovers a missing sensor mid-take.
Connected nodes, the current take’s timecode, and the shared rig clock’s live alignment readout — always on screen, always a measurement.
ISO, shutter, white balance, tint, focus and exposure bias set once and pushed to every camera, so the whole rig locks to one interpretation of the light for a take.
Take number, elapsed time, thermal state and remaining storage stay visible during the take — the rig warns before a take cannot finish rather than after.
Every take is a numbered roll on the device with its own sidecar data. Copy it off with the Files app or a USB-C drive — nothing is locked inside a proprietary viewer.
Screens are from a pre-release build — the interface and the feature set may still change before release.
A take is a package, not a pile of files.
What comes off the rig is what a reconstruction actually needs, recorded rather than reconstructed afterwards.
HEVC, with Apple Log on supported devices — readable by anything, not a private container.
Recorded frame by frame beside the colour on devices with the sensor.
Per device, recorded from the device rather than estimated later.
The single fact that makes a moving scene solvable instead of merely filmed.
Drop that session into VoxaVerse Prism and it imports as itself — no conversion step, no reformatting. Bringing footage from a different rig? The open Capture Package is the route in, and it is public.
Privacy isn't a feature. It's the architecture.
The volumetric category runs on cloud upload. Beacon doesn't — there is nowhere for a take to go.
Recording is crash-safe: an interrupted take recovers to its last complete frame instead of becoming an unreadable file. The rig is thermal- and storage-aware, and says so before a take starts rather than after it fails.
Be holding it on day one.
Beacon is in active development alongside Prism, and the viewer their output streams through is already live on this site.