Viz Engine 5.1 release features Reality Connect powered by Viz AI
Vizrt has released Viz Engine 5.1, which includes several advances to live graphics workflows.
Building on the debut of Viz Engine 5, this version introduces Reality Connect – powered by Viz AI – marking a new era of AI-driven 3D talent immersion. Among its many anticipated uses, Reality Connect uses advanced AI algorithms to enable hyper-realistic talent immersion, including shadow casting and reflections within virtual studios and XR sets.
“Reality Connect solves the problem of people appearing to float in a virtual studio by enabling realistic shadow casting and reflections,” said Gerhard Lang, CTO at Vizrt. “Reality Connect anchors the talent within the virtual studio by utilising virtual 3D models of the talent, which are continuously updated using Vizrt’s AI pose estimation. Our AI pose estimation operates on the same GPU and utilises the live video feed used for keying.”
Saving on cost, Reality Connect does not require additional hardware. Presently, Reality Connect is exclusively compatible with the Viz Engine 5 native render pipeline and works in conjunction with the Precision Keyer. Reality Connect also offers the flexibility to seamlessly integrate data from any motion capture system, as an alternative option.
Vizrt’s invention of the Virtual Window transformed the utilisation of video walls as far back as 2011, marking a significant industry shift and expanding their capabilities. Viz Engine 5.1 evolves the way Virtual Windows are created when using Viz Engine’s native render pipeline or Unreal’s render blade. The resulting solution takes advantage of the increased sensitivity of both new renderers to screen resolutions, and supports Vizrt’s mandate to enhance usability, particularly when working with intricate video wall configurations, such as curved surfaces and floors.
“With the updated version of the Virtual Window effect, no compromises in scene complexity and effects must be made. All shapes and forms of video walls, floors and ceiling are covered by our new method,” added Lang.
The combined output resolution of Vizrt’s new Virtual Window effect, video walls, ceilings and floors can be up to four times UHD, allowing for large setups to be managed by a single Viz Engine.
Previously, one of the challenges users faced when using an XR sets was the lack of a proper preview. Vizrt works with Roe Visual to combine Viz Engine 5 and GhostFrame to introduce a better experience that is more comfortable and natural for talent.
GhostFrame enables up to four content feeds into an LED screen – which can be individually captured by a camera. Users can select which feeds are visible for the presenter and production crew and which will be the live broadcast feed. The talent can truly interact with an AR object for example, without the hazard for virtual collisions, or look directly at the video wall for markers, teleprompting cues, or when conducting a remote interview; all of which are not visible in the live broadcast feed.
GhostFrame can also incorporate four different brandings or implement multiple video wall layouts using Viz Multiplay. Roe Visual will continue collaborating on the numerous creative workflows and possibilities achievable by uniting these innovative technologies.
Another Viz Engine 5.1 advancement is an extension of Adaptive Graphics – which now includes user-defined aspect-ratio support for video walls. Now, Adaptive Graphics can be intelligently adjusted to fit any video wall size, shape and resolution.
By extending Adaptive Graphics’ faster and more efficient graphics production line to studio video walls, customers ensure a consistent look and unified brand identity across all platforms and displays.