SVG Sit-Down: Vizrt's Nicholas Jameson on AI in Workflows, Pushing Boundaries With XR/AR

SVG Sit-Down: Vizrt's Nicholas Jameson on AI in Workflows, Pushing Boundaries With XR/AR Virtual assets may someday soon be changed by voice command By Ken Kerschbaumer

Thursday, May 8, 2025 - 7:00 am

Print This Story | Subscribe

Story Highlights

There was a time when Vizrt had to do a lot of evangelizing around the role of extended reality and augmented reality in live sports or studio programming. That is no longer the case: XR and AR were two of the buzziest topics at NAB 2025. Couple that with the growth of AI in data visualization, another core aspect of Vizrt systems, and Vizrt Senior Customer Success Manager Nicholas Jameson had plenty to discuss at the show.

What were some of the key messages for Vizrt at NAB 2025?

Obviously, a lot of the news this year is around the investment in AI and focusing on AI functionality, AI features, and AI models that can help end users with their workload as opposed to just creating crazy graphics. One of the results of that is using AI with Reality Connect and Virtual Studio to allow [for example] talent to interact with AR objects and move around them or even do object tracking or telestration by just moving their arm or hand.

At NAB 2025, Vizrt's Nicholas Jameson noted a focus on AI that can help end users with their workload as opposed to just creating crazy graphics.

For workflow purposes, we want to use AI to make things like Libero and Arena and making them a bit faster and more intuitive and bringing them to the masses. Libero Go, for example, has object tracking and detection using AI, thanks to Nvidia cards that are smaller, denser, and can speed up what were previously longer processes. Viz Arena 5.5 has downstream cut detection, which smooths wipes and dissolves in camera feeds. And there are things like making virtual ads easier and cleaner.

What are some of the trends you are hearing from customers?

The consistent theme, especially from some of those core users, is more people are getting involved with AR. If you had asked me five years ago how many sports teams were doing AR graphics on a regular basis, it would be less than 10. Now there is a market of 120 teams in the U.S. alone that are ready to do AR, and they have the budget and appetite as well.

The networks - Fox, ESPN, NBC, CBS - are pushing the boundaries and chasing after AR. All of them are increasing investments, looking to make AR repeatable and available to all their productions.

Now they are looking at virtual-set extensions and doing keying without the need for a green screen, even taking those workflows to remote locations and quickly spinning up a virtual and an AR studio with keyable assets.

Where do you see generative AI helping take that to the next level?

A lot of people are trying to figure out the best way to use AI, and customers are deciding what they want to do with it. We recently put together a data center that is kind of a backbone data distribution inside the Viz Engine and can do things like creating an AI data connection in Libero that can use tracking and stats data to do things like social-media crops and push out player-specific highlights much more quickly than in the past.

Obviously, at NAB 2025, you continue to highlight virtual sets and volumetric capture, and you've been a leader in those spaces. How are those operations improving?

The new Render Pipeline we introduced in Viz Engine 5 gives more photorealism and realistic graphics with things like casting shadows and reflections. Our integration with Unreal opens a huge market for us, as everybody is building in it and playing around with it. We continue to make sure our Viz Engine renderer supports Unreal, but we also support other 3D build and render platforms, like Cinema 4D and Blender.

The name of the game, ultimately, is making sure you're getting the most out of the hardware platform, and it's all about bringing immersion to a new level. We're seeing users move back to using the Viz Engine renderer as the core render platform, and we can use Unreal so our tracking data can create a smooth look for projects in stadiums or volumetric shoots with players and things like that.

It seems like, at some point, we'll be seeing voice commands to create a virtual-studio environment.

I think so. I've seen a couple of CAD design tools that can create a CAD with voice prompts. Right now, we have the ability for on-air talent to interact with a virtual touchscreen in the virtual studio without the need to buy a physical screen. It uses hand gestures, but we're probably closer than we think to being able to use voice commands to change virtual assets.

LINK:	https://www.sportsvideo.org/2025/05/08/svg-sit-down-vizrts-nicholas-ja...
	See more stories from svg

Most recent headlines