Ever since Oculus launched its Kickstarter, hopeful fans of VR technology have looked to the company to provide a cutting-edge gaming solution. Since then, we’ve seen a number of companies announce their own VR efforts, from theVive (Valve and HTC) the Gear VR (Samsung and Oculus), Razr’s OSVR, the new StarVR, to Sony’s Project Morpheus, there’s currently a race on to see who can provide the best hardware for virtual reality gaming. Unfortunately, as a new reddit thread illustrates, that future may be considerably more fragmented than users would prefer.
Oculus CEO Palmer Lucky took to the PCMasterRace subreddit earlier this week to answer questions around Oculus‘ decision to fund the creation of roughly 24 games that highlight the capabilities of VR technology. Oculus is building more than just a headset — it’s planning to build an entire platform, including motion controllers. Users took to the subreddit to ask how much support Oculus would be offering for other headsets or devices, and what consumers can expect in terms of hardware support.
Palmer notes that Oculus is not a closed platform, meaning the company does not require game developers to submit requests for approval to build titles for the Rift and does not collect a fee for adding Oculus support for games. At the same time, however, the company is investing in its own ecosystem, and it’s not prioritizing projects for other headsets. When asked if it would be easy to add such support, Lucky implied that it wasn’t, saying: “Extending VR support to multiple headsets is not as simple as a patch, it requires pretty deep integration into the code of the game, integration that the developers themselves have to spend a lot of time integrating and updating. This is especially true for games that rely on our SDK features like timewarp, direct mode, late latching, and layered compositor to get a good experience. We can’t possibly make any promises about support through external patches, and we won’t commit to supporting people who want to use our store to buy games for headsets that our store and software don’t currently support.”
Right now, there is no common standard for would-be VR leaders to adopt in the first place. Valve is pushing OpenVR for its SteamVR solution. Razr is pushing OSVR (Open-Source VR), which it claims can unite disparate standards on a common platform. Obviously a company like Oculus, which has invested in its own custom hardware and pioneered software approaches, isn’t going to be keen on just jumping onboard someone else’s platform — not when hardware sales are directly tied to its own financial success. We can safely assume that Sony will have its own SDK and solutions as well. Palmer notes that Oculus Rift hardware support “is currently (and has generally been) pretty much broken when it comes to Rift support. When it does work, support through SteamVR is far behind our own SDK. It is pretty clear that they have been prioritising Vive, and that is fine. They are working hard to launch a product as well, and it makes a more sense for them to focus on improving their own side of things than to try keeping up with every update we make.”
The situation is even more complicated when you consider the positions of the two leading GPU manufacturers. Nvidia has already announced that its VR technology will be folded into GameWorks, with all the preferential treatment that implies, while AMD’s LiquidVR is designed to take advantage of specific capabilities of AMD GPUs, like asynchronous shaders, that Nvidia’s Maxwell doesn’t support in the same fashion.
Sorting out the reality of VR support
The more I’ve talked to some of the developers working on VR, the more I’ve realised that some of the intrinsic assumptions gamers have made about how VR technology will integrate with existing titles are simply inaccurate. The long-term goal may be to take home any title and hook up any VR device, but that’s not going to be what happens when VR titles finally launch. In reality, most VR headsets will be tied to specific titles — and given the difficulty of optimising each title for each headset, initial compatibility is likely to be rather poor.
With the exception of Eve: Valkyrie, none of the games or demos we’ve seen demonstrated in VR actually approach the depth and scope of AAA titles. This may not be an accident. Most developers at AMD’s E3 event talked about integrating VR for short game segments or specific areas, and the full games on display that leveraged VR were either cockpit-based like Eve or simpler, smaller FPS multiplayer titles.
Some of the readers in the original thread accused Palmer of promoting console-style tactics and lockouts to prioritise Oculus’ own solution, but having talked about VR tech with both developers and hardware vendors, I’m not sure that’s true. Building good VR titles and hardware is difficult. It requires far higher frame rates than your typical PC or console title, and frame latencies must be kept to an absolute minimum. Because VR screens sit so close to the eye, the entire question of what to display and where to draw the player’s focus is different in VR than when gaming on a standard monitor.
There’s an increasing body of evidence that suggests designing a title for VR may mean making very different choices than when optimising it for a standard 16:9 monitor, and many of the techniques used to maximise performance are going to depend on the capabilities of the headset and the GPU that the player is gaming on. Bad VR ports also tend to induce nausea in players and there are few more reliable ways of ruining your reviews than by making gamers throw up.
What all of this means, collectively, is that the first set of VR games and experiences are going to be specialised and demonstrative of where the technology could go, not its final iteration. We still don’t know how cross-compatible solutions will be — it’s entirely possible that the “best” solutions will vary from game to game. It wouldn’t be surprising if certain titles favoured AMD + Oculus while others might play best on SteamVR + Nvidia (or vice-versa), and thanks to the complete lack of any formal standard or common platform, the experience is going to be erratic. We suspect a number of gamers may simply decide to wait things out rather than jump in, and despite having had an amazing experience with Eve: Valkyrie, we can’t blame them.
The thing to keep in mind is that this isn’t really Oculus’ fault. Until VR solutions are in-market, there’s no way to bring everyone together beforehand to agree upon a standard, particularly when different players are going to want different capabilities. I’ve loved VR when I’ve gotten to play with it, but the entire category is increasingly looking like a hot mess, and I’ve dialed back what I actually expect any of the companies to actually provide accordingly.