Q&A: What Insomniac has learned from working on VR brawler Feral Rites

It’s been a couple of months since Insomniac Games announced its virtual reality ambitions, following up the release of Edge of Nowhere with plans to launch two Oculus-exclusive games: The Unspoken and Feral Rites.

We got to see a bit of The Unspoken a few months ago, but Feral Rites wasn’t quite ready to play at the time. At E3 however, Insomniac's Marcus Smith gave a demo of the game, sharing some of Insomniac’s major lessons from working in virtual reality. 

For lead designer Smith, who’s worked on Sunset Overdrive and other Insomniac third-person action games, the biggest challenges have been trying to build level design and combat that’s similar to what Insomniac has done in the past, but now in the trials of virtual reality. 

First things first, what did you steal from Edge of Nowhere? I'll say this, I didn't get sick playing Feral Rites, because I walked backward in Edge of Nowhere and had a very bad day.

Yes. So that was one of the first things, they had done a lot of work into simulator sickness. Camera work in particular. We knew that we wanted to make a game that was a lot more open world. This demo is pretty linear, so it's hard to explain that.

Well, I did see it on the map that there was a larger path for me to take, I just took the more direct route.

Yeah, okay. If you scroll forward on the map you saw a little bit further, it opens up more. That was the first thing, we knew we wanted to make an adventure game that allowed you to explore and having a follow camera causes problems when you come back towards the camera. So with Edge of Nowhere they had to design the same, where you're mostly moving forward and trying to minimize that as much as possible.

For us we had done a little bit of experimentation, especially after playing games like Hero Bound, kind of going back to the old Zelda games where you would walk through a doorway and it would cut through to the other side. 

Then we had to figure out how to make the level design for those types of cameras. Finally, after we got all of that working it, we learned it looks a lot better when you're close to the player and the enemies. So now all those cameras that used to be a gridded system and a lot more procedural, now they're a lot more authored. We can bring the camera in closer for certain setups, we can bring it wider for larger setups where you need that spacial awareness and everything.

You discovered that you had to take time to put the camera in places instead of just procedurally tying it to certain kinds of rooms?

Yeah, exactly. 

What's changed from combat encounters in the last few months, that you've learned?

I think the big one that we're tackling now is, even in this demo, we have executes that are tied to combinations. They're just sort of ten hit combinations regardless of what they are. People are successfully button mashing their way through. I'm fine with people feeling like they're successful getting through the game. The problem is that it will ramp up in difficulty as soon as you start meeting other enemy types. So at the end, we just threw the mandril beast in there.

I noticed all the health bars just shooting up when I was in that room. 

Yeah, that's the thing. It feels like, because it's the end of a demo, that it's a boss. It's not a boss. That's a normal guy that you'll meet in the world a lot. But it changes things up when you meet him in combination with other guys. We're going to have a lot of other enemy types where the skill level needs to be higher to get past those guys. That's sort of our challenge. how do we encourage people to do the more action-packed skillful fighting, to string combinations together? Because that's what the game is designed around.

Has the fact that sometimes the player can be really far away and the animations are going to be smaller factored in at all?

That's where we really need to go in and author cameras. When we throw up that magic wall, (A wall of fire that erupts and locks the player and enemies into a fixed arena), we can do that closer in. That's easy, we don't need to adjust the camera for that. But  enemy AI can tend towards the camera. So that way you kind of have to come towards the fight, which means getting closer to the camera.

Another tricky thing that we've had is navigation. At one point we only had a north-facing camera, which caused its own set of problems, but at least you always knew which way you were going. Then we switched it to every time you cross a camera line, the camera will follow you. Which is cool because you can explore the world a lot more seamlessly. You don't feel like you're running backwards all the time. 

The drawback is, if you cross that line and then immediately come back, it will reverse your camera and it will look like you're at a brand new space, even though you came from that area. Which means people will see it and think that's the right way to go and they'll backtrack. 

That's interesting, yeah. 

MS: So we've had to do things like, we've added a timer so when you cross the line there's a five second timer right now. It's an arbitrary number, we'll adjust it as we test more people against it, but basically if you go over the line for under five seconds and you return, we'll pop you back to the camera you had before. So it will feel more like, "Oh, I'm where I was before." But if you spend more time there and come back then we'll reverse it, because we figure people are trying to explore. 

So I've noticed a lot at E3, this year, that a lot of the VR things being promoted are really short. Especially all the announcements coming from PlayStation. This is planning to be a longer game, though, isn't it?

It's a longer game. This is much more of a normal [length]. I think the big trick is, there's no book on how long people play, how long people feel comfortable. For me, personally, when I first started playing the game I could only play, putting on headsets, even if I'm playing a game that doesn't have a lot of sickness issues, I couldn't play that long without feeling a little weird. But having played a lot, I can probably spend all day in a game like this now. 

That's the biggest trick, we just don't know what a good time of play is. In a regular game you could say, "45 minutes is a good chunk." People will play that and make the judgment, "I'll do another 45 minutes," or, "I'll back out." In this we don't know what people are going to do. So we need to cater the experience to something that people can pop into, play a little bit, pop back out. I think that's the biggest trick: how do you do that?

Have there been any big new UI lessons you've learned since last time we talked?

Yeah, I think the problem that we used to have that we were trying to solve was players getting disoriented of where their character was because of third person and because you can look anywhere. We used to not have those blue lines that were the demarcation of where the camera cut will be and it was complete chaos. People would get sick, experience a camera cut they weren't ready for, not know where their character was -- Depending on what angle you were coming in and out of, you would lose your character a lot. 

That's why we did the follow cam, we brought the camera lower, so it's easier to find them on the horizon line. But that causes the problem you were saying, which is you can get really far away and be tiny. So the blue line helped a lot, but then also we've got the little head with the arrow, that shows you where your character is off screen, so it's a lot easier to come back and find them. A lot of those little UI issues have been our Achilles' heel. Also we use scale form, which is a third party vector based, flash based kind of interface. That's fine when you're making a 2D game, but it's all in 2D so everything is flat. 

So for VR we've been trying to figure out ways to give depth in interfaces that is built on a platform inherently made for 2D. That's tricky too. A lot of times we'll just do multi-layers to give depth in something that would normally just be a flat list. 

What other VR things have you been studying to learn about VR and, what are the non-game experiences you're using to learn things about VR?

It's funny because the first thing I thought of was Lucky's Tale was a game that I played a lot just to see third person. I think they do a really good job of having a true follow cam and feeling like you're in a world, and making it very much, "Oh, this is why this is in VR." It's a traditional game, but it's not a traditional VR experience. I think they did an outstanding job. 

For VR, I've been to game jams where people are making VR experiences. Sorry, I'm going to talk trash on somebody, but I went to a game jam where the whole thing was "Jenga." It was like a virtual Jenga, but they had a lag on it purposefully and they were trying to simulate how you do things when you're intoxicated. It was so disorienting. Basically it was like the Drunk Jenga Simulator.

I've been playing a lot of different VR experiences. I'm really excited about experiences where you have the helmet on but you're in a place where there's physical, tangible, objects representing things that are being rendered. I went to a USC thesis class where a guy had done a motion capture and he had built sets of everything you could see. In the headset it looked like I was in this medieval dungeon grabbing weapons off of tables and things, but you could feel the table. It felt tangible. That really makes the immersion so much greater.