Today, you can spend about $15 and buy a Cardboard-type phone attachment to get simple VR with your phone. You can also spend $1500 and in a few months you will get a nice computer and what would be considered today a high-performance consumer VR experience, touch controller not included. As I write this, the price for an HTC Vive has not been published - perhaps it will be even higher.
$1500 is a lot of money. Maybe I just don't have $1500 available to spend. Maybe I spent $900 a year ago on a nice computer and I don't want to buy a new one. Maybe I'm running a classroom and we just can't afford a $1500 machine on every desk? Maybe I worry that VR is moving so fast that the $1500 VR system I buy in 2016 will be obsolete in 2017?
$15. $1500. Excluding Gear VR ($100 if you already have a fairly new Samsung phone), is there any good solution in-between? How can we bridge the gap to allow "Everyman's VR"?
A few things could - and perhaps should - happen:
- HMD vendors would select different trade-offs geared towards different use cases (media viewing, casual gaming) and offer them at lower price points.
- Software platforms, like OSVR, would continue to allow games to modulate the required computing power. For instance, the Unreal Showdown demo on OSVR runs nicely on an NVIDIA 6xx card.
- Benchmark utilities would allow users to determine how fast a game can run on their PC and have the game adjust to it. Can't run Elite: Dangerous at 90 FPS? Run it at 45 FPS with time-warping or reduce the resolution and image quality.
- Eye trackers embedded inside goggles allow using foveated rendering, which is the process of doing high-quality rendering at the portion of the image you are directly looking at and lower quality rendering elsewhere. This would save GPU resources and allow running on lower-end PCs. Similarly, VR driver optimizations from NVIDIA/AMD/Intel could make better use of GPU resources.