Local multiplayer could be done via raytracing, strange as that sounds

Peter Thoman, more commonly known in gaming circles as “Durante,” is renowned for his work on the infamously horrid Dark Souls port from the early 2010s. Focusing on technical rendering solutions as of late, Durante has now set his sights on a new kind of local multiplayer gaming.

More specifically, Durante’s latest blog post, titled “A New Approach to Local Multiplayer / Splitscreen Perspective With Raytracing,” goes in-depth on Thoman’s idea of leveraging raytracing to develop a new kind of hotseat multiplayer screen-sharing experience. Historically, hotseat games have relied on one of several different types of split-screen implementations. A prominent example comes from racing games such as Gran Turismo, for one, where each player is allotted either one-half or one-quarter of the screen, depending on whether two or four players are participating.

“However,” says Thoman, “even with the most advanced hardware support, only up to 4 different views can be rendered at once without substantial performance implications – and even that requires quite a lot of development effort.” This is where his unorthodox reliance on hardware raytracing comes into the picture.

Leveraging hardware raytracing for local multiplayer camera

As the video featured above plainly demonstrates, Thoman has come up with a way to use raytracing to “dynamically adjust the [players’] camera(s) at a pixel level,” resulting in a trippy, albeit novel way of rendering that side-steps all of the hotseat games’ previous issues. The game itself is a technical demo that Thoman’s team had worked on in the years prior, Sphere Spectacle.

“The most interesting part is what happens in the “no-man’s land” between [the players’ screens],” explains Thoman. “where space appears squished as the camera parameters are interpolated.” While the solution is fairly similar to the usual suite of split-screen setups that hotseat multiplayer games have previously relied on, Thoman claims this solution is “fundamentally so much simpler than trying to do anything like it in a rasterization pipeline.”

Further, it doesn’t appear to be overly difficult to implement, either: “I ultimately only spent two working days hacking this demo into Sphere Spectacle – and that includes implementing support for two players!”

Naturally, none of what Durante has shown off here is done just yet, even on a conceptual level. Instead, this is a proof of concept for a simplification of local multiplayer rendering. As he himself admits, “Now we just need someone to build a full game around it,” which is bound to take a while. Yet, the feature is curious and has potential, and there’s no telling what might come of it down the line.