Prototyping in VR: 4 things I’ve learned.
I’m Liam Ferris, a software engineer working in Virtual Reality @KainosSoftware. At SXSW this year we launched “retne”, a VR experience built for the HTC Vive. Retne is an immersive adventure produced by artist @deepamann_kler.
Designing for VR is hard. Our intuition can’t be trusted. Our intuition is built from things like film, UX and games design; which are all very different than VR. There’s only one way to know if an idea works in VR — you have to try it.
So, try stuff. Learn what works and what doesn’t. It’ll make your VR application better, and it will improve your VR intuition. Here are 4 things I’ve learned about VR prototyping.
1. Stand on the shoulders of giants
Lots of smart people are building stuff, and lots already have. If it’s available to you, why not use it?
Look for VR UX research
For example Owlchemy Labs, creators of Job Simulator coined the term ‘Tomato presence’ in their research. It’s the process of making the controller/hand disappear when you pick something up. This feels natural in VR, it reinforces that the controller is currently tracking an object. We used this principle in retne.
Play good VR, try to identify what makes it good.
‘Henry’, the Oculus VR movie uses eyes to create an emotional connection with a character. This fed into our decisions while designing Lil’ Mo, retne’s main character.
During SXSW, the best VR experience I saw was Chocolate; a fantastic, incredibly weird cat-themed music experience. During Chocolate, to the rhythm of the techno music, flying cats turn and make eye contact with you… and it felt awesome. This was nice validation that eye contact in VR is effective.
Try to find open source projects, check the asset store. Use anything that could speed up prototyping.
Keijiro takahashi is a unity developer, he also has more open source github projects than there are atoms in the universe. He’s built really cool motion effects, particle effects and post processing effects. We used a bunch of these to prototype in VR.
2. Criticise
After trying stuff the next step is feedback.
VR is so novel that feedback from anyone not saturated in it will give false positives. If you get feedback from a VR noob, they are going to tell you it’s awesome. But what they really mean is VR is awesome.
Tunnel vision is also a problem. Sometimes you can be so in love with an idea that you can’t see that it doesn’t work. For example, in retne there’s a level where you are in space on an asteroid.
When we were prototyping this we came up with an idea we all loved; a Mario Galaxy style playing space, where the player could walk freely around the asteroid.
We built it and thought it was excellent, you can even do some really cool non-euclidean geometry puzzles. Although when we got play-testers to try it out the feedback was that it made everyone feel sick. So we cut it. ¯\_(ツ)_/¯
A good VR play-tester has the following characteristics:
- Has VR experience. If no one like this is available to you, find some people and show them good VR.
- They have to be outside of the day-to-day team.
- They have to be ready to hurt your feelings.
3. Affordance
Clues about how objects can be used.
From watching people in VR it’s surprising how goal driven we are. People won’t explore unless you explicitly or contextually make it their goal. So the challenge for VR design is that at any moment you have to be able to answer the question “What do I do to move forward?”. Confusion is the enemy.
You are in control of the player’s sight, sound and touch (still waiting on you, Smell-O-Vision) and because of this you are obligated to use all of them.
Visual affordance
Making objects glow, sparkle and vibrate will attract the users attention.
Audio affordance
Sound effects show the user that something is going on. Or after an interaction it confirms that they have done something. 3D audio helps users spatialise and can direct them to something outside of their field of view.
Haptic affordance
Make objects tactile. It confirms to the user that they can do something with an object.
4. The fidelity contract
Meeting natural and set expectations.
The fidelity contract is a set of unspoken rules that defines how every interaction works. This doesn’t necessarily mean we have to only design realistic interactions, we can still create engaging experiences that are unnatural as long as it’s perceived to be consistent in the world we’ve created.
An example of a natural expectation
When the user drops an object it should fall to the floor.
An example of creating a new expectation
When the player drops an object it falls up.
Both examples are perfectly good ideas for a VR experience. However, if some objects fall down and some fall up, in a way the user couldn’t determine you would be in violation of the fidelity contract. That’s how you get confused users.
You need to consider if you are honouring this contract in every decision. This extends to creating an interactivity base line. For example if you only want the player to interact with a single object, preventing interaction with everything else would break the contract.
If you are watching someone play-test your VR experience and you find yourself thinking “They are doing this wrong”, what you should be thinking is “I designed this wrong”.
Conclusion
VR design is hard. You’re going to be wrong more than you’re right. But that’s okay — be clever, be creative, be curious.
More about retne // @liamkferris
