Q&A: Developing games for the new Samsung Gear VR

Last week saw Samsung debut its Gear VR headset, which was developed in close partnership with the folks at Oculus VR and is powered by the new Oculus Mobile SDK that John Carmack (pictured) has been scrambling to finish. When the Gear VR developer kit (known as the "Innovator Edition") is released later this year, developers will have an opportunity to start mucking around with mobile VR, and selling VR games in the busy mobile app market. But more than a few developers are already working with early versions of the hardware to build Gear VR games in advance of the headset's launch later this year. Some of them have already blogged about their experience on Gamasutra -- check out Dreadhalls developer Sergio Hidalgo's post about the challenges of porting to Gear VR and E McNeill's technical breakdown of what it's like to create a Gear VR launch game. In addition, Gamasutra reached out to both Darknet developer E McNeill and Fireproof Games cofounder Barry Meade (The Room) for more information on why they started mucking around with Oculus' new Mobile SDK and what developers need to know about developing for Gear VR.

What is the most complicated or challenging aspect of developing for the Gear VR?

Meade: Well it's still a gaming device so many of the issues are the same as developing on any platform - performance, memory, frame rate etc. but where VR is different is in the importance of these things. On any game for any platform you can drop 1-30 fps and all it means is annoyance for the player. In VR, it's a potential barfing incident. You absolutely cannot drop below 60fps if you want the proper experience, which in reality means your game has to run at 70fps+. Combine that frame rate with a huge screen res (2560x1440) plus our ridiculous idea of setting Omega Agent in an open world - with all this is running from a mobile phone to boot - and we are talking an eye-watering project in optimization. McNeill: Performance is probably the biggest challenge, but it's definitely possible to get a good-looking game running fast. Input might also be tough for some types of games. I elected to try to make my game playable with just the built-in controls (a small touchpad and back button), but you could require a bluetooth gamepad as a fallback option. (Developers will also need to deal with all the usual challenges and opportunities of VR design, but those aren't unique to Gear VR.)

How did you end up working with Samsung to develop for the Gear VR, and why?

Meade: Fireproof are just very interested in VR in general. VR is truly a new platform, with all the properties of a blank canvas that suggests. It has tremendous strengths to build on but also obvious weaknesses to overcome. So it's a challenge first, and that's great fun to attack, but we also see VR as an open field, and the kind of gameplay it will support and experiences it can give are entirely unknown - it will lead not just to new games but new genre's too and that's a supremely interesting space to play in for any developer. But as a small team with limited resources we're kinda locked out of VR on PC/console - Gear VR allows us a more lightweight, limited but tighter type of game that suits how we work. Gear VR is a great piece of kit. The speed of the response, the lack of lag is stunning. But really we were attracted to its consumer-focused setup. VR is amazing but only if you get to try it out personally and with the top spec stuff requiring $300-$400 and a beefy PC, VR was not really going to be accessible to the mainstream audience who are now being turned on to gaming through mobile. With Gear VR we hoped Samsung and Oculus will change that. They can bring these small, lightweight, low cost machines to any shop in any mall in the world and just let people have a go and perhaps buy it on the spot. That's a more realistic, dynamic approach to VR. And more practically, The Room requires months of design work before we can put the whole team on it. So when we were approached by Oculus & Samsung we were in the position where we could split our team and make something new as long as the projects remained a manageable size. And making something completely different than our bread-winner series is a great pallet cleanser. VR is entirely new ground, we know nothing about it and the market is all to play for: that's like catnip to us. McNeill: I've been in contact with Oculus ever since they sponsored the VR Jam last summer. Around GDC, they let me try out an early dev kit of Gear VR, and that was enough to convince me to develop for it. My game is built in Unity and has simple, stylized graphics, so it was a good fit for mobile VR, and porting it to Android was relatively straightforward?

How does the Gear VR differ from the Oculus DK2 (or DK1) from a dev perspective?

Meade: Gear and Rift are of course different, with Rift being more fully featured than Gear VR, but the astonishing reality of presence that Oculus gives remains just as real on both kits. McNeill: I'm working in Unity, and just like the DK1 and DK2, Gear VR has an excellent Unity SDK available from Oculus. Getting the game running in Android was a cinch. Gear VR doesn't have positional tracking, so it feels like the DK1 in that respect, but in every other way, it's as good as the DK2 or better. The resolution, latency, and comfort are pretty fantastic. In most ways, I feel that VR is VR; once you're wearing the headset, you don't much care whether it's powered by a PC or a phone. Developing for Gear VR feels very similar to developing for the DK2, except that I try to be more careful about efficiency and performance.