VR projections
Posted: Fri Aug 06, 2021 4:09 pm
So according to current trajectories, standalone 8K per eye VR resolution will be pretty much mainstream by the end of the decade. But does anyone know if those 8K standalone VR headsets will have the GPU processing power to support 8K content?
Right now, most VR content is rendered in 1080p (a couple of games are supersampled) even though we have 2K and 4K per eye VR headsets. Rendering in lower resolutions is probably the reason why most, if not all VR games are cartoonish. Do you think the lack of polygons for rendering prevented many developers from making true triple AAA VR games?
When I saw Ready Player One and its rendered worlds and characters, it didn't look too real BUT it also didn't look too cartoonish like current VR content. According to John Carmack, from a technology standpoint, the VR resolution in RPO is most likely 8k per eye. It's crazy to think how most VR content in the 2010s and up to now, all look cartoonish (limited by 1080p to 2K resolution). Then somewhere in the late (or mid) 2020s to mid-2030s, the VR content will look a lot like RPO (6K to 8K). THEN somewhere in the late 2030s and thereafter, the VR content is pretty much indistinguishable from reality (12K to 16K). I'm a little bit saddened by this projection and how it'll take two decades for VR to be the same resolution as reality.
However, It's pretty cool how VR will achieve RPO type resolution in this decade but I'm not sure if it'll have the GPU power to support 8K content.
Will eye tracking/foveated.rendering really help in reducing computational cost? Will it really be dynamic as our eye movements?
(I'm new to VR and I just bought a Quest 2. It's amazing and looks promising. Just wearing the device and looking around, I could feel its potential once we get to 8K. However, the screendoor effect is noticeable; I won't be buying another VR headset unless its 8K and the prevalence of 8K content)
Right now, most VR content is rendered in 1080p (a couple of games are supersampled) even though we have 2K and 4K per eye VR headsets. Rendering in lower resolutions is probably the reason why most, if not all VR games are cartoonish. Do you think the lack of polygons for rendering prevented many developers from making true triple AAA VR games?
When I saw Ready Player One and its rendered worlds and characters, it didn't look too real BUT it also didn't look too cartoonish like current VR content. According to John Carmack, from a technology standpoint, the VR resolution in RPO is most likely 8k per eye. It's crazy to think how most VR content in the 2010s and up to now, all look cartoonish (limited by 1080p to 2K resolution). Then somewhere in the late (or mid) 2020s to mid-2030s, the VR content will look a lot like RPO (6K to 8K). THEN somewhere in the late 2030s and thereafter, the VR content is pretty much indistinguishable from reality (12K to 16K). I'm a little bit saddened by this projection and how it'll take two decades for VR to be the same resolution as reality.
However, It's pretty cool how VR will achieve RPO type resolution in this decade but I'm not sure if it'll have the GPU power to support 8K content.
Will eye tracking/foveated.rendering really help in reducing computational cost? Will it really be dynamic as our eye movements?
(I'm new to VR and I just bought a Quest 2. It's amazing and looks promising. Just wearing the device and looking around, I could feel its potential once we get to 8K. However, the screendoor effect is noticeable; I won't be buying another VR headset unless its 8K and the prevalence of 8K content)