Metastage's 106 cameras build a mesh, not a free-standing image

What you see

Metastage's finished output is a moving 3D figure that must be played back through an AR or VR app or engine before anyone can see it; nothing is visible during the capture session itself. The studio's own volumetric video page describes the goal as letting an audience view a captured performance in virtual or augmented reality as if the person were right in front of them, from any angle the capture supports.

How it works

Metastage's specifications page states its stage uses 106 cameras and 8 shotgun microphones to record a performance from every possible angle in one take, generating about 10 gigabytes of raw data per capture. That multi-camera footage is converted first into point-cloud data, then into mesh data, before smoothing, compression, and encoding produce the final asset, directly derived from photographic and video data rather than reanimated or hand-modeled by an artist afterward. The company's team page describes this pipeline internally as Holocapture processing, and outputs can be delivered as compressed streaming video, as raw mesh-and-texture sequences, or through plugins for Unity, Unreal, Magic Leap, and Quest.

Viewing conditions and limits

Metastage's specs page states delivered assets range from 250,000 down to 5,000 polygons with textures from 4K down to 1K, chosen to fit anything from cinematic VR down to streaming mobile AR, and that a compressed 10-second clip can be as small as 5 megabytes. Typical delivery takes two to three weeks after a client selects frame ranges from dailies. As with Microsoft's comparable studios, none of this content is visible without a headset, phone, or screen playing the finished file; the studio's own figures for camera count, data volume, and output size are not independently audited.

What it is not

Metastage's own volumetric-video page uses the word holographic loosely, describing its work as part of immersive and holographic media, while its specifications page is precise that the underlying technique is volumetric video capture, a time-varying 3D mesh built from many synchronized cameras. This is not an interference hologram and not a light-field display; it is closer in kind to the output of Microsoft's Mixed Reality Capture Studios or Intel True View's reconstruction step, a captured 3D asset that must be rendered and displayed by something else before it is visible.

• Ask whether a claimed volumetric hologram was captured on a multi-camera stage or generated by other means entirely.
• Ask what polycount and texture resolution were actually delivered for a given project.
• Ask what device or app the audience needs to view the finished asset.

Metastage's own technical page is more careful than its marketing language, and that gap, precise specification sheet against looser promotional copy, is a pattern worth watching across nearly every volumetric capture company profiled in this guide.