Many people were surprised, about a year ago, when the news broke that Oculus' virtual reality (VR) headset would require a high-end, $1000+ PC. Today, as consumers finally unbox long-awaited high-end VR systems, the demanding PC requirements of Oculus and HTC Vive continue to draw discussion.
Even as someone who has worked with VR for over a decade the requirement was initially a head scratcher for me. VR had run for years on cheap PCs. I was running my Oculus DK2 at the time on a mid-grade iMac. Regular mobile phones could drive VR using Google Cardboard or Samsung Gear VR. Surely, a cheap PC should suffice.
These steep requirements are rooted in a decades-old technical limitation that could keep virtual reality headsets from ever reaching the cutting edge of visual experience. The gap between head-mounted and standard displays will likely be a permanent obstacle for the nascent VR headset industry. Here is why:
In order to understand the logic of headset manufacturer's decision to require high-end computers, we must first understand what I call the VR gap. VR presents separate images to each eye, which doubles the strain on a PC's graphical hardware. This is a factor of 2x. VR also needs to show about 90 frames per second to the viewer, as opposed to the 60 frames per second typical of a high-end computer game. This is an additional 50%. If we put these factors together, VR needs 3x as much graphical horsepower as a standard screen to display exactly the same image. Some sources--companies like Nvidia--claim this VR gap is 7x. Whatever the exact number is, it's undeniably significant. The intense graphical processing needs for VR make one thing clear: games played on a standard display will be richer and more detailed than anything available in VR for years to come.
This VR gap is nothing new. In the 90's, it was common to use a $100k+ supercomputer to drive VR because, well, that was what was considered needed at the time. However, then like now a supercomputer wasn't "needed". It was merely desired, because otherwise VR would look 3-5 years behind the times, in terms of visual fidelity.
When it comes to Oculus and Vive, the high-end PC requirements start to make sense with this VR gap in mind. Both companies are releasing high-end products for the PC Gaming Master Race. This audience is used to AAA games. Anything less than a top of the line PC would result in a steep decline in graphics quality when hopping to VR. Gamers complained about the the clarity and visual richness of early Oculus prototypes. Everything is on the line with a big product launch. Understandably Valve/HTC and Oculus/Facebook want to limit the experience to PCs that, even in VR, can match the visual fidelity of top PC games, not to mention the Xbox and PS4.
The pricing also makes sense when we compare it to other innovations like the iPhone. At launch, the $600 iPhone price-point was a shocker. However, rabid early adopters bought anyway, and the technology naturally became more affordable over time. Oculus and HTC may have similar motives. Apple's first iPhone had a graphical richness and responsiveness which matched average desktop PCs at the time. It was, in my experience, the first phone to do so. By the same token, these VR headsets are being built and sold to match today's desktop average.
The mobile gap, like the VR gap has not gone away. While advances in mobile like the Unreal Engine for iOS (anyone remember Infinity Blade?) upped the ante with gaming graphics on phones, consoles and PCs have gotten even better. Mobile, for the foreseeable future, simply won't catch up.
For similar reasons, we will be stuck behind this VR gap for the foreseeable future. The gulf seems solid. And it is not going away without a surprising innovation--an unknown unknown at this point in time.
Good VR should be good VR because it is good in VR. The benefits of the VR technologies (whichever they may be--I have perhaps the world's broadest definition of VR--the subject of a future post) must be apparent even if the resolution is low, and the tracking is poor. We in the VR community need to build VR that is good because of the nature of the medium itself (the subject of a past post).
As VR-ready PCs get cheaper, we see more signs of the VR gap playing its course. Today we see signs that Microsoft is releasing a new, more powerful Xbox console to handle VR. And there are rumors that Sony is developing a new Playstation 4 called NEO, because otherwise--well, VR would be "terrible".
This year, gamers will be making a choice. Play in a VR headset, or have higher visual fidelity on the flat screen. Let's hope that the VR-ness of the VR is enough to make the tradeoff worthwhile.