Star Citizen: Reverse the Verse 2.11 - DE

When I started at Foundry 42 it was about a year ago, last September. Back then we weren't really working on the procedural stuff. It was just an idea and a long-term goal. So I did some space station stuff. I supported the guys in the UK because there weren't a lot of environment artists here at the time.

You were telling me the other day, what was the very first asset?

Ian, the art director from the UK, came to me and gave me this white box mesh, it was really rough. He said, okay, dude, you have to make it look nice, that's your job. See how you do with the art style, everything is new. So I did that and I went through it, I finished it, I gave it to Ian, he's like, yeah, this is cool. And normally when you make a game, you wait for a long time till it's in the game. But in this case, a couple of weeks later it's in the game, and it's this console where every player goes to select his ship.

In the beginning it was kind of tricky because it's like two worlds clashing if you have art and code working together. So it takes some time to get used to how they talk. One dude thinks in pictures and the other thinks in numbers. So it took a bit of time to get adjusted, but basically Marco and the guys provide the code. I'm kind of the bridge for the artist who's in the end going to work with the tools. So I provide all the stuff that's needed like height maps and textures and materials, and just generally give art feedback: if I realize, okay, we want to achieve this quality, I know, okay, we need this and that. So I can talk to Marco: look, this is the visual fidelity we want to achieve. Marco himself obviously has very high standards as well. And together we try to find a way to solve all these issues.

The blending of different height maps in general, if you have mountains and then you have sand next to it, works. Technology-wise, it can be done really well, but in some cases, it just doesn't make sense. For example, if you have a pitch black mountain and next to it you have white sand, it's obviously going to look a bit strange. But this comes from the art direction. They tell you, okay, this doesn't look good, we just have to change something. But basically, in general, we get concept art from the UK. Our concept artist sets up all these biomes, these ecosystems, what they're going to look like. He kind of takes them apart and also shows you how it's going to look on a global scale, also on a detailed scale, what kind of vegetation there is and rocks.

So when you guys are jumping into it, you have a really good idea at least what your target is. And then it's you working with tech to go, cool, how do we achieve that?

Because sometimes we hit a certain limit. What we did so far was very organic. Earth-like planets, we can get them looking pretty good. But now we're thinking of stuff like ArcCorp that's very mechanical. It's a huge cityscape basically. So we had to figure out certain ways on how we achieve that. So before we go into production, we always make sure that we have the tech there, that it can be used by the artists. Because in the end, we always need more than one dude.

Looks nice. Looks good.

We're excited. We had Hannes working on that, primarily, right?

Yeah, he started a couple of months back and that was his first, his baby, basically. He's taking care of the ocean. So at the moment, visually, it's in a good place. We have all these tessellated three-dimensional waves. Interaction-wise, there's not much going on at the moment. Because we also have to take care of what's going on when you go underneath the water. And you need stuff like buoyancy, what happens if you want to land with your ship, and this type of stuff. As an artist, I don't really know if there's any plans for that yet.

I do know on the design side, we're not going to block people. It's not like you're going to get to the water's edge and hit an invisible wall. We know we're not going to do that, but honestly, I don't know what the design team has come up with or what that roadmap is, but I know there's been discussions of, okay, what would we do underwater and how would it happen? So we'll see.

We actually just started this week to prototype ArcCorp, figure out ideas. Can we do it with the planet terrain? Because if you do something like architecture, it's very different. If it's organic, you have these random shapes and stuff, and in architecture, everything is very straight and angular. And with the tessellation of the terrain, you can get away with certain looks, but you will never get this convincing city of ArcCorp with massive towers and skylines and this type of stuff. So what I would like to do in the future is prototype something that uses the object scattering, because all the objects on the planet will be scattered procedurally as well. So we will do something similar, just with really, really big objects, kind of scatter them over the planet. I don't know how happy Marco is with that. We'll have to figure out, does it work performance-wise?

But then on top of that, that's when level design comes in. They know where the player path is and where they want the gameplay to exist.

Well, so far we've gone with Earth-like stuff. We've been looking at the latest pictures from NASA, from Mars, and they have a lot of cool, inspiring stuff. But of course, I'm looking forward mostly to really alien planets where you just land and you're like... You can go really nuts with things like vegetation. A different atmosphere, there's more oxygen, things grow really big.

In a sense, it gives you guys more creativity. It's also a bit more of a challenge because of the unknown: you still need to create all those elements that still feel like they fit within an environment. Although that environment is make-believe, you still want those things to all seem like they're going to come from that same place.

That's a good thing, because it's still artist-driven. It's not completely procedural. So you still have the control: you can say, okay, this is how a tree looks in our world and I'm just going to modify it, but it still keeps that essence of a tree. You can say, okay, it grows like this, and it still has this believability to it.

I think one big part would be the actual narrative PCAP that feeds into all the cinematic scenes for Squadron 42, that we actually go to shoot at Imaginarium. So I would accompany Chris as a kind of co-director, basically directing the actors, while I'm making sure that the staging is correct and the lighting looks good. I'm basically like a virtual director of photography.

So you were there directing with Chris basically every speaking role bit for Squadron, right?

Yeah. And we had the main shoot last year and then another pickup shoot in May, July this year.

But ultimately, as director of cinematics, primarily you are responsible for all of the cutscenes for Squadron 42.

I'm leading the team here, and we are making sure that the PCAP gets integrated properly into our scenes, and it's a bit tricky with our game because it's not just pre-rendered cinematics. We want to achieve most of them in real time. So it's a constant back and forth and handover with the designers that actually make it happen, so that we can trigger it from AI and back and forth.

I had a full range of emotions. It went from living the dream to 'I can't believe I'm still stuck here', because it was so long. I had everything going on. But it was amazing. Especially, my favorite was, I think, both Marks were great, and Gillian, of course. And just seeing them putting on this suit, that humbling suit, which is totally not comfortable, and just playing these roles that we had in the scripts for so long on these pages, and actually getting into these roles in a proper way. Gary Oldman was actually asking things about how would he speak and what is the whole background.

I do still remember the first time Oldman stepped up on the podium for his very first shot on his very first day. He was just rehearsing, but he was in full character. And everybody was running around, there's 40 people on set moving everything around, and he gave his thing in full volume and everybody stopped. The place was just perfectly quiet and he ran through the entire thing. It was awesome.

Then we were like, did we actually record this?

I think one goal was we wanted to see how mature our planet tools are when they are actually touched upon by artists, and they actually create something substantial with it that is actually playable. So this is where the idea for Homestead came from. And then we basically went, okay, do we really want full jungle already? And we wanted to keep this for later, because also our asset pool was basically not full jungle already. So we were like, first let's do lots of jungle and trees, and now let's move this to later. And then we were like, okay, what can we really do now that looks really great already, which would be an arid desert planet, but also savannah highlands. So we chose kind of like if you would go to Tanzania, a Kilimanjaro-style landscape. So we picked that, and from then it was basically working with the engineers, Carsten, Marco, Chris Bolte and Sascha, every day, making sure that the newest features are stable and make sense for what the artists need to do.

Yes. The only difference is always: what is your field of view? How close can you get to the guy? Is depth of field on? What is the lighting? These are the same assets. We are not switching to another type of asset if you're in-game.

I will say that that's fairly rare, at least in my experience: hey, here's a cinematic, now we have a different LOD for the head that's crazy high, or here's the special cutscene head that we're swapping in. But he's absolutely right. Our faces are the faces. And they need to be able to hold up from a distance, and then you need to be able to get close and see them as well.

They're actually so detailed in some cases that our current accessories for the faces are lower detail than the actual faces. So they are very detailed, and in-game you need to really bump into them, really close, to enjoy the detail that they can offer, which is great.

It's definitely the character department and the animation department, I would say. Constant back and forth with the character department on, look here, some penetration on the armor, or how can we fix the Snoopy cap, and what is with the hair, and all kinds of things. Animation, of course, because it's actually the raw material of what we're using in the cinematics. Without animation, we wouldn't have anything. We could just show ships drifting around and the characters would be static. And then, of course, there is a certain part that is kind of tool-side R&D, because, for example, our ships need to move in a proper way in cinematics that is kind of puppeteered, almost like remote control. So there's this part of the equation as well.

And we have a lot of those guys. When characters are created and so on, things get funneled here, and Hannes is working with those various studios, and a lot of things end up getting assembled here. With you and with our lead cinematic designer and the animators here in Frankfurt, you guys end up cramming it all together and throwing cameras in and lighting it and the whole bit, yeah?

We actually did lots of R&D as well on the animation side when it came to what we showed at Gamescom with Miles Eckhart. This was also done by our animation department here that is attached to cinematics.

So what we have is basically two types of dressing currently with our planet. One is we can just populate everything with a certain rule set for rocks and trees and little bushes and everything. That's totally automatic. So if you go to a different location on the globe and jump into a forest, it looks populated already. The content currently needs to be hand-placed. If there's a crashed ship, we hand-place it in a certain area that we find suitable. There's a real sense of adventure going on. If you have this 1,000-kilometer diameter planet in front of you and you're actually scouting, you're actually flying around: this would be awesome as a landing site for pirates. We will basically have these bubbles of content, like a crashed Javelin or a crashed little Cutlass, whatever, and these get, currently at least, hand-placed. We are also thinking about distributing them logically, of course, because planets are so big, but this is ongoing, so they need to do lots of R&D for that, so that we basically don't place a Cutlass where it's floating in the air, or in a canyon that you can't reach. There needs to be a certain logic rule set for that, so that it actually works properly and reliably.

Yeah, I'm definitely confident. The number of scenes we have is crazy high.

But we knew that for a while.

So there's lots of things going on, but for example, we have lots of scenes that are comms calls, and these comms calls are kind of easy to set up. So there's a big chunk that is comms calls. Then there's lots of scenes that are kind of drifting into gameplay territory, where we are just helping out or supporting. And then, of course, there's a big bulk of scenes that are the social interaction, interstitial scenes during each chapter, which is a big part of the story. And then a small but very important part is the full-on filmic cinematics, which is the stuff that is the most fun.

Those are the ones you always get excited about the most, right?

If I have control over the camera, I'm in love. Because then I can get really close to the characters, I can change my lenses and everything. But our rule set for Squadron normally is: if we can do it, it should be coming from player agency. We still stay in the player's eyes. We have some cases where, if the player is not present in the scene, yes, we will cut to this, and then we can cut to it filmically.

That's interesting to note, that we're doing as much as we can to keep control in the player's hands. Even though there were some, as we were talking them through, where we kind of had to take control away from the player, we didn't want to completely, so we put them on rails. But again, there's so many of these that are interactive, if you will. Somebody approaches you and, hey, I can go this way and hear part of the story, I can go that way and come back to this later if I want, but you're making a choice, and whatever choice you make, something new clicks off and you run down that route.

I'm fighting a constant fight. Some gamers out there probably are like, oh god, this guy is a cinematic guy. He will fight for quick time events and framing dead bodies instead of actually discovering them, and stuff like this. But there's this back and forth between design and our department where we went through these scenes: does it make sense to take away the camera for a certain bit and only give you look control? Or do we want to give you full freedom, but then can we funnel the player into a certain view? It's an ongoing process, but I would say I'm happy about the balance that we have in the game. It's definitely not something that's constantly on rails, or complete freedom.

We kind of answered this, but basically the rule of thumb is: if the player is present and he's not knocked out or something, then it's from the player's point of view, and he at least has look freedom. Most of the time he has full freedom. When conversations happen, we will at maximum lock the player in a soft lock. He can still back away from the conversation. But for some important things, like you get a briefing with Mark Hamill in the briefing room, we actually sit you down in the chair to watch the briefing.

There's certain elements we need to give the player in order for them to move. Even if that's just a few pivotal ones, they need that information to know that they'd better do this thing or this bad thing is going to happen.

Our captains for the hub ships, at least we captured it and it's pretty cool: [if you behave] too much like a jackass on board and not like a military officer, then you land on the bridge and the guy is basically like, I had high expectations for you and you disappointed them. So there will definitely be the opportunity to also behave like a creep or a bunny-hopping idiot. But in some crucial scenes where we want to get the story across, we say no. We don't want to have key characters dying or something and the player is concerned.

The performance itself was pretty good, but we had lots of technical issues. I can name a few. I think some normals on his face were inverted. We had the wrong type of skinning. The hair was really not good, and his uniform was just the first pass. From an asset point of view, it was not really that great. So I would probably say facial performance maybe 85% or 80%, and what the asset actually looks like maybe 60%. So he looks much better now. Of course, at one point I need to go back and re-render this, because this will also be game content that you can actually view on TV screens. So it will definitely be re-rendered with the latest Bishop, much better.

The reason why I ask is I've seen things deforming a lot better, and new shaders on the eyes, and once you see things actually firing off the way that they need to as a one-to-one, some of the stuff I've seen has been nuts.

Our problem is always that we kind of make it look polished, but of course it's an ongoing process. There's lots of things that you still need to work out, technical things. For example, a month ago we got something really amazing from Jeff, one of the graphics programmers. Before that, in CryEngine the eye shader was always just giving these little dots, no matter how big the light source was. And now, if there's a big area light in front of the character, we actually see it reflected in the eye. Way better than before.

It's amazing to me how just that simple thing, when I walked in and saw that face on your screen, it's immediately noticeable. Sometimes it adds something where you're like, wait, what's different? You may not even know and recognize it.

The eyes especially, you always know that something's wrong with them, but you cannot really tell. The eyes are something that humans, since they are babies, can immediately recognize what's off there.