Uniform Solves for Volumetric Video
The promise of high quality volumetric video to create lifelike digital humans has Wild Capture excited about the creative possibilities of capturing real human performance with lifelike accuracy. This is the amazing power and promise of volumetric video!
But the reality is that when the technical artists import volumetric video files into their desired software, everything tends to bog down. The files can be extremely bulky, and when reduced for streamlined output they can fall apart. We at Wild Capture have many solutions to this industry-wide issue known as a “Uniform Solve”.
Volumetric video basically exists in raw form as a moving human shaped water blob in 3D environments. Nothing can touch it, or interact with it. For artists working with volumetric video in 3D environments the main challenge is one of boundaries. When working with traditional file types there is no boundary for the character to interact with its environment. No way for it to splash water, walk through smoke or sit down somewhere to relax.
Traditionally, volumetric video begins with a bunch of 3D snapshots over time that have nothing to connect one another. There's no easy way to translate the real world to that data. It exists essentially as a ghost or a hologram. The texture color on the 3D object generally changes very frequently with little consistency.
This data lacks certain physical elements that most 3D animated characters possess. Because those characters have consistency throughout the animation from the first frame to the last, the object is deformed over time.
Traditional characters are puppeted to life whereas Wild Capture is taking real life footage and manipulating it with an animation layer on top of the physical performance.
The actual performance from the data set that we generate provides all the features of the traditional character, starting with realistic velocity to interact without any CG effects or objects. We provide ways to uniform the topology to the structure of the character in a smart way that allows for the editing of the texture. That texture output and that uniform mesh is optimized across the entire 3D asset.
We can do a lot more with this sorted data than with the original raw data.
The original data is agnostic volumetric 3D mesh and texture sequences. That agnostic data is common format for OBJ, ABC and PLY files for the 3D mesh.
Wild Capture offers many solutions for this problem. Here is a demo from one of our uniform solve boundaries in action. This shot is from artist Kaelyn Kastle’s music video for her single “Sunsets Suck”. She turned to Wild Capture to bring this project to life in volumetric video.
Wild Capture is committed to developing bleeding edge digital human technologies. Our use of uniform solves, USD pipelines and our Cohort crowd and fashion toolkits, combined with the collaborative efforts of the XR Foundation and Hyperconstruct, Wild Capture is leading the way. These solutions show how using webXR to stream volumetric video inside live 3D environments can offer creators realistic digital humans for use in spatial media messaging, and more.