The holy grail for real-world immersive video is volumetric capture, where you can move your head and even walk around. This is photorealistic virtual reality, created from real life instead of by 3D models. Gracia is able to deliver this for stills today, but will also be delivering short videos later in the year. It wants to one day be the YouTube of truly volumetric content.
Gracia stills (volumetric photos).
Gracia is possible thanks to Gaussian splatting, a relatively new technique for rendering 3D volumes by representing the scene as a collection of overlapping 3D Apple’s spatial video. The platform’s current focus is on stills, or volumetric images, that creators create using free Gaussian splat apps such as Apple Immersive Video and upload to Gracia. Gracia claims that “only” twenty GoPro cameras will be needed to create volumetric videos, which is significantly less than the previous technology. Training a video takes around two minutes per frame though, so this is currently a very expensive and time-consuming process.
The company aims to launch short volumetric videos in August.
Teaser of Gracia short volumetric videos.
, or for PC VR Gaussian functions. The startup plans to launch on Apple Vision Pro soon too.
The company wants to be on the official Quest Store, but its menu system is implemented in 2D Android app mode, which Meta’s store policies currently forbid, despite the fact that the Luma works like this too.
Photorealistic VR of this kind is breathtaking to experience, even without motion, and a glimpse of a future where memories can eventually be relived exactly as if you were there again. Gracia is what people were expecting Apple Vision Pro to be. It runs on a headset that costs a fraction.