Interview: Tobias Johansson, Head of VR development at Stunlock Studios, talks about Battlerite’s unique use of VR

Tobias Johansson is a programmer and Head of VR at Stunlock Studios, the developers of the rapidly-growing arena brawler Battlerite.  Whilst his focus was officially related to AI, he has also been tinkering with VR in his own time, developing a system for filming Battlerite matches using a virtual camera that can be positioned to record the best shots of the action.  He’s now officially Head of VR at the studio, and we caught up with him to talk about his vision.  If you haven’t already, be sure to check out this video which introduces the idea:

Ben (TVR): How did you get into working on VR for Battlerite?

Tobias: We don’t want to downplay our enthusiasm for VR, because most of my colleagues are really loving VR in general.  But the thing is, we were on a super tight budget with Battlerite – most of my time back then had nothing to do with VR because, I mean, I was a programmer, I did the AI, and all that other stuff.  So, my boss was like, “Yeah, I mean, VR is super cool, I support you, but I can’t really put you on that kind of thing right now”. That was just the right call back then.

But I’m the kind of guy that lives for his work, so…  I stayed after work a bunch of times tinkering.  I’ve been working on VR since the first DK1.  So I thought: VR observer mode?  Obviously! So back in early June, I ended up developing the VR camera system in order to shoot our gameplay trailer. That was when I started working on VR for real, instead of just a side project. Since then I’ve worked almost full-time on VR for the game!

You can have the characters bigger than you, or you can be like a god and see it as a small Warhammer tabletop.

Ben (TVR): Did the inspiration partly come from what Valve has done with Dota’s VR observer mode?

Tobias: I was working on this way before anything was announced!   The first prototype for observing was with the DK2, I think it’s over a year old.

In Dota, when you go into the map, you have these 3 different scale modes:  you can be a god, or a bit smaller, or one-to-one with the characters.   I tried that a year ago, and was like, I don’t like it!

Ben (TVR):  I guess the major difference, and the really novel aspect of what you’re doing, is putting VR cameras into the world that can record footage the same way a regular camera does.  Where did that idea come from?

Tobias:  A lot of this was actually me and my old roommate, a guy called Sebastian, he actually came up with the camera idea.  He was zooming around inside a replay, and was like, “Damn, I want to have a camera!”.  We both knew it was a good idea.  I was living in the office for the next 2 months!

Ben (TVR): With Battlerite, the arena is kind of the perfect size to be a tabletop that you can peer into.  How is the sense of scale:  does it feel like you’re floating over a real battle, or watching toy soldiers duke it out?

Tobias:  You’ve seen the 3d-pinch-to-zoom, right?  That makes it so you can be whatever you want.   You can have the characters bigger than you, or you can be like a god and see it as a small Warhammer tabletop.

Ben (TVR): Games like BlazeRush where it’s like Micro Machines racing around, you can lean in but they are still small scale.   3d-pinch-to-zoom actually changes the scale itself?

Tobias:  It’s completely seamless – I have capped the scale from 100x bigger than the characters to 0.1x, so between those sizes you can scale it seamlessly, however you want.

Ben (TVR): Some people have problems with motion sickness in VR.  If you’re using pinch-to-zoom at the tabletop scale it must just feel like you’re manipulating a small object, but if you use it to spin the whole world around you, do you think there’s any possibility of motion sickness?

Tobias:  I take motion sickness pretty seriously when I work in VR.  I’ve had – just a guesstimate here – less than 1% actually feeling sick in the game.   The way we handle it is by, as soon as you grip the grip button that makes you able to move the world, we fade-in a grid:  a transparent grid that is in your room-scale, physical space.   So that never moves relative to you.   That solves the problem for most people.

Ben (TVR): A point of reference with the real, physical world.

Tobias: Yeah exactly – so the point is you should feel like you are moving the world around you, but you are stationary.   That’s the aim of it.

Ben (TVR): There have definitely been many different approaches to solving the motion sickness problem.  Are there any disadvantages with pinch-to-zoom?

Tobias:  So, if you’re playing a game where you’re actually supposed to be immersed in the game, I think it’s sort of impossible to use it because it obviously breaks the immersion to be able to spin the world around.   But if it’s like this – a tool – or if you are just observing, then I think it’s the ultimate way.  The best way I’ve come across yet, anyway.  And I’ve tried a bunch of different ways for observer moding.

Ben (TVR): Do you think 2D is going to remain the medium of choice for spectating e-sports in general, in pubs or bars or whatever, and that VR is going to be used more by the cameramen who are filming the best shots?

Tobias: With VR you still have the problem of UI and the social aspect if you want to watch the game with people.   Or, you know, I don’t wanna have this big thing on my head, I just want to be able to relax, have my drink or whatever, and watch a nice game of Battlerite.   Then obviously it’s not going to be for you, at that moment.   But, I have to say, I watched a bunch of games in internal tournaments in VR, at the office.  And I prefer to watch the game in VR.

Ben (TVR): I can imagine, especially if caster audio and a solid UI can be implemented.  Any plans for that?

Tobias:  Yeah, the UI for the whole thing here is actually what I’m putting my focus on right now.  It’s super challenging.  So the HUD – just displaying basic information about the game – I think I have sort of cracked.   I don’t wanna spoil exactly how it’s going to work, but it’s not going to be…  In Dota, they just project the stuff into the camera, and that doesn’t look too good, I think.   And I bet they’re doing improvements there.    But I am going with a different approach.

But then, the UI – the actual like, “I wanna do something” – the controls – I have prototyped the button for leaving the game for 2 days now!   Not my actual workday, but you know.  A bunch of time.   Just a little thing like that:  How does it feel to leave a game naturally?

You know SteamVR spawns some 2D projected surface that you laser-point at.   And I mean, that works!  Obviously.  But that’s not what the best UI’s gonna look like.  It’s just boilerplate.

Ben (TVR): For VR developers there’s still very few established ways to do things, so it’s the perfect time to get creative.

…being able to direct multiple cameras in VR, and then be able to export it in-game, to share it…  That’s the end-goal, so to speak.

Tobias:  Exactly.  So on the subject of the leave button.  I can spoil how I think it’s going to work, because it’s quite interesting I think.    You know I try to get inspired by real life, because 3D is real life!  So, just physical buttons.  You know these buttons you have in the movies for launching a nuclear weapon?  A big red button with a flap on it – that’s the kind of thing I’m interested in, because then you have a two-step thing that you need to do, so you can’t press exit by mistake.   But it still feels like, when you look at it, you know exactly how it works, because “oh, I opened a flap and I pressed a button”.   Integrating that with sound, the visuals, vibration.

Ben (TVR): So you can almost imagine having this big control room with all sorts of buttons and levers, each being tied to a different aspect of the UI.  

Tobias: Exactly.  And then you get to use some of the things that you automatically have, like magical drawers – pulling a drawer out of nothing and it’s just a bunch of buttons.  That’s a good idea to use to your advantage as well.

Ben (TVR): On a similar note, you talked about having player interaction of some kind, like throwing roses out into the arena and such. 

Tobias: I would love to talk about it but, we don’t really know exactly how we want to do it – we’ve got to prototype it.  There’s actually some technical difficulties involved in, like real technical difficulties with the back-end.  We want to do something in the future.

Ben (TVR): The “Odeum” in-game lets you choose segments of your replays to post to a news feed that your friends can like or share.  It’s already a really cool idea, but do you think at some point there will be integration with the VR observer mode for saving “directed” replays that have been filmed using the VR camera?

There’s so many different options.  Handheld, static, curve-controlled, player-following, a bunch of mixed-reality tools, all at your fingertips in VR.

Tobias:  Now I’m walking a fine line between what I really want to talk about and what I don’t want to promise!  But you have the right idea, the right vision, I hear that you think like me.  And yeah, obviously that is something we’d love to do.  I don’t know how far we’re going to be able to go with it, but yeah, being able to direct multiple cameras in VR, and then be able to export it in-game, to share it…  That’s the end-goal, so to speak.

Ben (TVR): I guess there’s two sides to it – recording from old replays, and also live e-sports camerawork.  

Tobias: Live would work similar to broadcasting in Dota 2 – you register yourself as a cameraman, and then someone – as a watcher of the game on a 2D screen – can choose to watch between my directed camera perspective instead of the regular game view.

Ben (TVR): I think for me, that’s the part that’s really fascinating and that I haven’t seen before.   And when you think about it, it’s obvious!  This is how shots are filmed in the real world, so it makes perfect sense as a tool in VR.

The other side of things is the social aspect of watching a game with your friends in VR, the more “traditional” application of VR observing.  Is that not as much of a priority?

Tobias:  We don’t have in-game voice yet, and I don’t really know our stance on that right now because, you know, TeamSpeak and Skype and everything does the job for us right now.   Maybe it’s just a bad idea and it would become toxic and stuff, but it’s not on my table at all, to choose how we’re going to proceed with that.   But if we don’t have a voice function inside of the game, then it would obviously be kind of a big deal to put it in just for the VR mode.

Ben (TVR): So it sounds like the idea is more focused on a sort of cameraman/director tool, rather than a social experience.

…you’ve gotta imagine: the biggest, baddest, most expensive camera kit from Hollywood… I can just replicate it in VR because it’s just code, and it can be even better.

Tobias: The plan for me right now, the release situation, is to release something that is excluding the camera, and just introducing the actual observer mode with a bunch of interesting UI concepts.  Because, the thing about the camera is that it’s quite a complicated system inside of VR:  there’s so many tools available.  So I need to get the UI to actually work well, so people can actually use it without adding a bunch of instructions.

There’s so many different options.  Handheld, static, curve-controlled, player-following, a bunch of mixed-reality tools, all at your fingertips in VR.  And, how you control that…  It works, and my colleges who have tried it, and myself, we can handle it, but if I just were to just put it up on the web, people would be unsure how it works.  And I don’t want that.

I don’t want to put out something 100% polished, because that’s just unreasonable to try, but I want to try to make it something people can just jump in, and say, “Oh, I get it!” and I think introducing some of my UI ideas, and trying it out in the bare-bones observer mode first, that’s my goal.  Get some feedback, like “this doesn’t make sense” or “this is a really good idea”, and then I can transfer that over to the camera UI.

Ben (TVR): So the challenge is about finding the right balance between power and simplicity.  If someone’s just spectating, say, a Dota game, they’re just using their mouse and keyboard: limited and very well-understood controls.

Tobias: Since however-many years we have been perfecting normal mouse-based 2D interactions and everybody knows how it works.

Ben (TVR):  It’s a solved problem.

Tobias: And it’s not at all solved for VR.  Not even remotely.  So it’s an uphill battle, for sure.

Ben (TVR):  The concept behind this may have more far-reaching applications.  Imagine an entirely CGI scene in a film, such as a Pixar film, and being able to have a professional cameraman actually in the scene filming with similar tools to how they would film in the real world.  Do you think that’s a potential outcome of this kind of development?

I could imagine that a lot of games in the future would have their gameplay trailer made in this way, because it’s so superior in every way.

Tobias: You’ve gotta think ahead!  And you’ve gotta imagine: the biggest, baddest, most expensive camera kit from Hollywood… I can just replicate it in VR because it’s just code, and it can be even better.  I’m thinking how to exploit that in the real world, try to bridge the gap and use it in other ways besides just the application we have now.  I mean, that’s for the future, that’s a long way ahead, but if you could use that somehow, it’s so powerful.

I’ve done some really interesting camerawork in a matter of minutes.  Did you see the Blackstone teaser? The whole thing was done in VR, less than a day of work for me.  And doing that kind of thing, being able to prototype it in just a few minutes, it’s so powerful.  And if you could do that, maybe with animated movies?

It’s so much easier when it’s all digital!

Ben (TVR):  The big questions with VR tend to be about its “killer app” – what will be its primary use – and the answers given tend to be consumer-focused: gaming, media, virtual tours.  What’s interesting is that you’re working on a VR application that is about providing a practical tool that previously did not exist.

I saw one movie being made in Battlerite with the traditional WASD/mouse type of camera, and I’m blown away by what people can do, and then I imagine if I put this in their hands, it’s just going to explode.

Tobias: I don’t want to say like, “I think this is going to be revolutionary for everything”, but I could imagine that a lot of games in the future would have their gameplay trailer made in this way, because it’s so superior in every way.

Ben (TVR): So we’ve got to ask, when can we get our hands on this for Battlerite?

Tobias: We’re still trying to prioritize everything, and I’ve got a bunch of other things in the game that I’m responsible for.

Ben (TVR): And I imagine, of the whole player-base, you’ve only got about 1% of people actually wanting the VR aspect at this stage.

Tobias: Exactly, but still, it’s important, and being able to do the live e-sports things we talked about, that’s super good PR for us, and most of all we really want to push e-sports to new levels, so it’s worth it.  Even though just a few people might be the ones using it directly.  And in the future, if just a small percentage of our content creators have Vives, or Rifts (because obviously I’m going to implement for both)… I saw one movie being made in Battlerite with the traditional WASD/mouse type of camera, and I’m blown away by what people can do, and then I imagine if I put this in their hands, it’s just going to explode.  And that’s just PR in itself, a good way of advertising the game.  “Look at this: it’s like a super-complex movie scene inside of a game!”.

Ben (TVR): Sounds like you need to get Steven Spielberg to direct your next trailer using VR cameras.

Tobias: Yeah, exactly!  We actually thought about trying to get some famous movie people and doing it, but I don’t have too much to announce on that subject yet <laughs>.  But it’s on our mind!

One thing, just to be clear here: this whole system is super-multiplied in usefulness by our replay system.  So the technology is super cool, super flexible but if you can’t replay a scene over and over and try different camera angles, that’s not as useful.  Our gameplay back-end people have done an amazing job with that, which really makes my job easier!

Ben (TVR):  Definitely can’t wait to try it out. Thanks so much for talking with us!

Battlerite is available to buy (Early Access) on Steam for £14.99.  With 95% of its 4,500 reviews being positive, you don’t need us to tell you how good it is.  Take a look!