Abstract:
A new Sony patent outlines a frame generation system that combines motion vectors from the game engine and controller input to predict player actions, potentially reducing perceived latency on future PlayStation consoles. This technology may be coming to PS6, but it's not confirmed yet. The article also raises concerns such as visual artifacts and developers relying on generated frames rather than optimizing.
For better or worse, AI image enhancement and frame generation technology are the future of the gaming industry. NVIDIA's DLSS, AMD's FSR and Intel's XeSS image enhancement technology have been used in PC gaming for more than 5 years, and their frame generation technology has also been used for 3 years. Image enhancement technology has now also been applied to game consoles. PlayStation 5 Pro’s PSSR image enhancement technology was launched two years ago, and Xbox’s Xbox Series consoles will support FSR technology from 2022.
After PSSR, frame generation technology is obviously the next target for the development of Sony PlayStation consoles. IGN confirmed this in an interview with Mark Cerny, chief architect of PS5 and PS5 Pro and senior vice president of computing and graphics at AMD. Now, Sony has revealed its plans to implement frame generation in its next-generation PlayStation console, tentatively called PS6, in a new patent application.
"Generation Super Sampling" (US-20260320740-A1) is a new frame generation technology patent applied for by Sony for its game console. We may see this technology for the first time on the PS6 console. While the core concept remains the same, the model analyzes the rendered frame and generates a replica frame. However, PlayStation has made significant improvements to the way it generates this replica frame.
During the DLSS frame generation process, the model analyzes the rendered frame and uses motion vectors from the game engine. But in addition to using motion vectors from the game engine, PlayStation aims to take controller input data into account to more accurately predict how players will move in each frame.
As described in the patent, this approach is intended to reduce input latency, or at least create the illusion of extremely low latency. The patent also doesn't rule out the possibility of multi-frame generation, stating: "In some embodiments, the 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
At present, most people would agree that frame generation technology is not needed for console games, especially when PS5 uses AMD hardware and FSR is still perfecting its frame generation solution. All current-generation consoles support refresh rates up to 120Hz, so frame rates above 120Hz degrade visual quality by introducing artifacts.
In addition, frame generation technology may continue the trend of unoptimized porting of PC games to console platforms, and developers may rely on this technology to make up for the lack of game optimization. If frame generation technology is adopted, it should be a toggleable option, not a requirement.
Comments