A new patent from Sony outlines a frame generation system that combines motion vectors from the game engine with controller inputs to predict player actions, promising to reduce perceived latency on future PlayStation consoles. This technology may be applied to the PS6, but this has not yet been confirmed. The article also raises concerns, such as visual artifacts and developers relying on generated frames instead of optimizing performance.
For better or worse, AI image enhancement and frame generation technologies are the future of the gaming industry. NVIDIA's DLSS, AMD's FSR, and Intel's XeSS image enhancement technologies have been used in PC gaming for over five years, while their frame generation technologies have been in use for three years. Image enhancement technologies are now also available on game consoles, with PlayStation 5 Pro's PSSR image enhancement technology launching two years ago, and Xbox Series consoles supporting FSR technology since 2022.
Following PSSR, frame generation technology is clearly the next target for Sony's PlayStation consoles. This was confirmed by Mark Cerny, Chief Architect of PS5 and PS5 Pro and Senior Vice President of Computing and Graphics at AMD, in an interview with IGN. Now, Sony has revealed its plans to implement frame generation in the next-generation PlayStation console (tentatively named PS6) in a new patent application.
"Generation Supersampling" (US-20260320740-A1) is a new frame generation technology patent filed by Sony for its game consoles, which we may see for the first time on the PS6. While the core concept remains unchanged, with the model analyzing rendered frames and generating a duplicate frame, PlayStation has made significant improvements to how it generates this duplicate frame.
During DLSS frame generation, the model analyzes rendered frames and uses motion vectors from the game engine. However, in addition to using motion vectors from the game engine, PlayStation also aims to incorporate controller input data to more accurately predict player movement in each frame.
As described in the patent, this method aims to reduce input latency, or at least create the illusion of extremely low latency. The patent also does not rule out the possibility of multi-frame generation, stating: "In some embodiments, prediction is not limited to the next image in the image stream. The decoder system can be trained to predict any number of images before or after the input image."
What this could mean for console gaming
Currently, most people agree that console games do not need frame generation technology, especially since the PS5 uses AMD hardware and FSR is still refining its frame generation solution. All current-generation consoles support refresh rates up to 120Hz, so frame rates exceeding 120Hz would reduce visual quality by introducing visual artifacts.
Furthermore, frame generation technology may continue the trend of poorly optimized PC game ports to console platforms, with developers potentially relying on this technology to compensate for lack of optimization. If frame generation technology is adopted, it should be an optional toggle rather than a mandatory requirement.