7 Tips for Creating Virtual Reality Worlds for Oculus Rift

Matthew Warneford
3
June
2014

We love our Oculus Rift! Six years ago, going from a Blackberry to the first iPhone was an incredible leap forward, today going from a screen to the Oculus Rift virtual reality headset feels just as important. Where iPhone was a huge jump forward, a user experience stepchange that ushered in the modern cellphone market, the Oculus Rift is doing the same for user immersion. There is no experience that’s more immersive. I believe going from a screen to virtual reality is the next big leap forward, and a huge wave of apps, games, and new experiences are going to follow. Others think so too, Facebook payed $2bn for Oculus VR in March, Sony announced Project Morpheus at GDC, and Chuck E. Cheese are adding virtual reality to their restaurants.

I’m not normally a fanboy. I didn’t believe the early hype, I’d dismissed it as just another 3D TV gimmick. But when our Oculus Rift dev kit arrived in January the guys in the studio twisted my arm. I was convinced immediately.

Since January we’ve jumped into VR with both feet:

- Dubit’s research team are studying children’s reactions to virtual reality. The report will be published soon in partnership with Sheffield University.
- We’ve launched the biggest portal for virtual reality games, WeArVR
- Developed a VR world called Fairy Forest.

I’ll be blogging more about virtual reality in the coming weeks (and sharing our VR report). In this first VR post I’m going to share the lessons we learned developing our first VR world, Fairy Forest.

To checkout Fairy Forest you can download the game from WeArVR.com or if you have not got a Oculus Rift headset there’s a YouTube video.

1.    We found the optimum frame rate to be 60fps in order to make the experience feel real – anything less makes the user aware that what they are experiencing is on a screen.

2.    Maintain the realistic feeling by ensuring the user can enjoy the world at a realistic pace; it’s more comfortable as well as being realistic – so no Mario-like invincibility stars or Sonic-style speed runs.

3.    Avoid zooming in and out and other tricks that could motion sickness or alter the user’s sense of stability in the environment.

4.    Interactions help the user feel immersed in their environment, so the more of the world they can interact with the better. It could be as simple as seeing the grass move as they walk to being able to operate machinery or in our case, cave words into the trees.

5.    It’s important to tackle latency as it can really affect the player’s immersion, as we mentioned with regards to frame rates, but it also extends to the engine. Initially we tried using the Unreal engine but we incurred problems from vertical tearing and lag where the frames would be behind the movement in the real world. This was a shame as Unreal is a great engine. We’re sure this will change but for the time being Unity was our best bet.

6.    Be sure to choose the right control system. The first option we looked at was for the player to move towards waypoints by looking at them, but it proved fiddly for the user and required a lot of neck movements. We also looked at using a control pad. The problem here is that it can take the player out of the experience, but the analogue sticks worked well and felt more natural than a keyboard. As the initial design was for the world to be experienced at the conferences we were exhibiting we decided to put the game on rails with the player controlling where they looked but not where they moved to. However, it allowed us to control the player’s position and make the most of trigger effects like the bats coming out of cave. As we move forward with the development we will probably re-introduce the hand-held control method.

7. Lastly, as we were using the 720p development kit and because of the distance from the screen to the user’s eyes you could pixels in the display. This meant that even with anti-aliasing on things tend to look a little jaggy. As this is dependent on the silhouettes of your art assets ‘thin’ objects such as grass, foliage and tree leaves, which our demo features quite a lot of, can look bad. Blocky stylised art would probably have looked better. We’re sure this will change with future development kits and the consumer release.

Other Articles