Well, I never thought I would change my position on VR but here we are. If this works it will make VR viable.
It’s so disappointing to remember the hype in 2012 and then look at VR in 2018 only to see an anemic library of titles, little adoption, and not much improvement of the actual headsets. It’s pathetic.
I’ve heard of some headsets in the pipes that are supposed to actually be good. Extremely high resolution and full field of view coverage are absolutely required. There is no technological limit that is stopping a headset like that from existing, and yet it doesn’t. Pimax maybe. Am I missing one?
If you control motion sickness and have a headset like I described above, then you have a minimum viable system. Going beyond that, we need non-discreet light fields and perfect body motion capturing.
VR ‘cured’ my motion sickness. I worked on a VR project in 2017. Before that I'd get carsick unless I was driving. At the start any artificial motion in VR made me queasy, but I got used to it. Now I'm fine.
> Extremely high resolution and full field of view coverage are absolutely required. There is no technological limit that is stopping a headset like that from existing, and yet it doesn’t.
GPUs are the technological limit.
The Vive and Rift push 2160 x 1200 pixels at 90Hz. 90Hz is the bare minimum speed for VR. Nvidia's new RTX 2080 Ti (retail $1,200) cannot maintain 90Hz at 3840x2160 in most games. Two in SLI probably could, but that is much more expensive than the original Vive/Rift CV1 GPU requirements were. We are several generations of GPUs away from being able to double the VR resolution while maintaining 90Hz for less than $600 worth of hardware.
We're getting very close to high quality and inexpensive eye-tracking technology that will make it cheap and easy to implement foveated rendering in VR headsets (basically, only rendering high quality where your eye is looking in a completely unnoticeable manner). That will really drop the hardware requirements and make high field of view/high resolution headsets totally doable.
You are, of course, expecting a certain level of detail in games to have this kind of limitation. I would expect a game like Rez (PS2) to be easily runnable in super high resolutions. Obviously games like RE7 and Fallout 4 wouldn't run without significant changes and engine optimisations, but I could live with simpler looking games.
Just because a headset has high resolution and therefore complete fov coverage and no screen door, does not mean that it has to be utilized. If someone has a weak computer they can render a lower resolution and smaller area and scale it. Every headset should be capable of high resolution because anything less than that is not compelling vr. Multiple gpus is fine.
I agree that in general it is a very important improvement for VR to allow motion-sensitive people to enjoy it. However I do not agree this by itself will make VR much more viable than it is now. Motion sickness is just one problem, but without addressing others (high price, low resolution, narrow FOV, bulky headset), it will not get much more viable.
And I say it as someone who considers current state of VR to be amazing and enjoys it daily. But personally for me relatively low resolution would be #1 priority to improve.
It’s so disappointing to remember the hype in 2012 and then look at VR in 2018 only to see an anemic library of titles, little adoption, and not much improvement of the actual headsets. It’s pathetic.
I’ve heard of some headsets in the pipes that are supposed to actually be good. Extremely high resolution and full field of view coverage are absolutely required. There is no technological limit that is stopping a headset like that from existing, and yet it doesn’t. Pimax maybe. Am I missing one?
If you control motion sickness and have a headset like I described above, then you have a minimum viable system. Going beyond that, we need non-discreet light fields and perfect body motion capturing.