Lewis Thompson

Exactly. One of the reasons I came up with the idea originally was because whenever I do a playtest and I go and buy weapons, I always get annoyed that I have to buy a weapon and I have to scroll down and find the exact mag for the gun that I want and stuff. One of the versions we want to do is a weapons case that comes with some attachments in it, some magazines and stuff. Like a full service kit.

Exactly. Just give me, you know, we call it a first aid kit. That's legitimately what we're looking at. Like a whole kit, you would just get this item and you would have everything that you need to go out for.

Yeah, because in real life, you wouldn't go and buy each bandage individually from the shop and go to the counter and buy one bandage at a time, you'd buy a full kit with everything you needed to get started. So that's kind of the idea that started this stuff. But its applications, there's a lot. So this one in particular will be a medical one, but there's loads of things that it could be used for.

Yeah, for sure. I know the core fundamentals of 3D, so learning Blender isn't as hard as it would have been if it was the first program I ever learnt, but the operations you do in them are a lot different. We have this in 3ds Max, which is called the stack, and basically it lets us put different modifiers on. Blender has a stack as well, but I don't think it's quite as versatile. The other cool thing about Blender as well is that it's free. So, say for example, I wanted to make something that I wanted to sell online, I think I could do it in Blender and you wouldn't have to pay the prices that you have to pay for 3ds Max. That being said, there's a lot of functionality that we use that comes with 3ds Max for production, which is why concept artists will work in Blender to create things, because they're not going directly into the game, they're kind of just being used to show us what we can do.

The last show of these that we just did had Alberto Pedronio on and he spent the entire hour working in Blender. Generally, we have a magnificent tools team here who spent a good portion of the earlier years of the project creating an entire suite of tools that allow this program to connect to that program, connect to that program. And then when our friends at Montreal came on board, they took it even a step further with Mighty Bridge, which was a whole internal suite of tools, again, designed to make that bridge between a whole bunch of our different softwares and our StarEngine. If you watched the show two weeks ago, you saw how blazingly fast Alberto Pedronio was in Blender. This is a human being who merged with Blender at some point in his young existence. And everybody I talked to afterwards, even internally, we're just watching like, how does he know all the commands that fast? He knows them all. But there are people like that with 3ds Max and there are people like that with Substance Designer and stuff like that. So we try whenever possible to provide the tools and the pipelines necessary so that our artists can work in whatever software they're most efficient and most effective in. It can't always be accommodated, like you're learning Blender right now, but we do try where we can.

To be fair, I prefer 3ds Max, I'm faster in 3ds Max. Blender, I still have the training wheels on, I'm still googling every time I do an operation to work out where everything lives, because obviously once you've learned it in Max, you know where it lives in Max, but elsewhere it becomes an issue.

Believe it or not, A-Robs, one of the big jokes that I have internally is that before my time here is done, I'm going to do every single department that ever contributes to the making of a video game. And one of the last ones probably being HR. But we actually recently did one of these shows just a few weeks ago, maybe seven, eight weeks ago, maybe 10, with one of the members of our talent acquisition team. And we talked about what they look for in hiring applicants and all the kind of prerequisites and stuff. So we actually did do a version of that show. You can check it out in the Star Citizen Live playlist up on our YouTube channel. I don't remember what we called it, but it was probably something whimsical.

If you're looking to get into art, like I said, Blender is a really good place to start, completely free. You don't need to go and pay for any of it. And you can learn all the skills that we use every day at work. As long as you've got a good foundation in that and some good work to display, then yeah, you're in. I don't even think you need a uni course or anything anymore. Obviously it helps to have that time to learn, but...

I would say it's not one of those things where... going to college and getting a degree is helpful if you learn and you grow from the experience. But there are plenty of people I've known in my life, I'm almost 50 years old at this point, I've met plenty of people who have a degree and it was very clear it was purchased. They did what they had to do to get through the schooling and get their degree, but it's not a true signifier of their experience and what they know and what they can do. One of the things I think happens a lot in the games industry is you can test. You can give them projects and say, hey, go and do up this or show me a version of this and show me your process, so we can evaluate those things. So the paper degree is less useful. The knowledge you can get from it is valuable. But if you have the knowledge and you got the knowledge some other way, we don't discriminate. I don't know a lot of game studios that do.

It's the final result that matters, and putting in the effort to get there. It's a very fun job, so definitely it's worth the effort.

How long anything takes in video games is incredibly subjective. We're not telling any tales out of school there. Not only does the complexity of the item itself factor in, but the relative importance of it. If it's what we call a hero asset versus just something that's going to be seen in the background, that'll mean the fidelity has to be higher. It means there are probably more review gates involved. The point at which you say, yeah, that's it, is probably easier to get to on a background item than it is on a big hero asset. This applies to anything from characters to props to flora and fauna and spaceships and everything like that. We used to talk to Paul Jones about this quite a bit, about how long this process took. And it's always a question of how soon you zero in on the idea. Because I always like to say this is game development, not game construction. You don't always just know what the thing you're making is when you start. A lot of it is about exploring a process and looking for a feeling. And as David can attest, David's my producer, even when making videos, I will sometimes be very, very prescriptive. I want this and do it like this and put this here and put this there and then let me see it like this. And then the editors will go out and they'll do it and come back, and I go, oh, that's terrible, don't do any of that, do something new. Even after doing this for almost a decade, I don't know if it's going to work until I see it. And that applies to just about anything, especially during the concept phase. It's one of the reasons the concept phase exists, because the idea is that you've answered most of those questions during the concept phase. But as is often the case, you discover things when you get to 3D art. You discover that, oh, we didn't look at the back, or the metric for this was wrong, and stuff like that. We're looking at the concept on the screen right now.

Yeah, I don't know how the hinge works. That's something that hasn't been answered for me, so that's something I'd have to think about a bit more. Working from a concept is really nice because it becomes more about problem solving to create forms as opposed to problem solving to create functionality and the entire object. You've got a starting point, which is really, really cool. You can see these are latches. There's a handle that you can carry it with. It's got branding on it. You can tell it's a medical crate from the branding, maybe the colours, obviously the text. Whereas if I came on this livestream and I didn't have a concept, I probably would spend most of the hour just spinning.

Concept's great for a lot of reasons, but I think chief amongst them all is somebody else goes through a certain amount of hell instead of you. Even when Alberto was doing his two weeks ago, two of them had very striking silhouettes. One of them was pretty much kind of what you expected. And one of them was a disgusting homunculus that he should still feel shame about. And he knows which one that was.

I watched it and I was very jealous how fast he works. He's an absolute machine.

But see, regardless of how fast he is, a concept artist is still attempting to hit something that's in the mind of multiple art directors. Whether it's Elwyn in Austin, or it's Ben Curtis here, or it's our new person in Montreal whose name is escaping me right now, or CR, or John Crew, or maybe Ian's got a real strong opinion on this one because it plays into a gameplay that he's... So you're trying to synthesize all these ideas. Even as fast as you can be or as fast as he can be, it always comes down to how soon you get to that consensus, that you realize that thing that somebody has in their head, or sometimes help them realize that thing in their head was wrong, and go, oh, I thought that would look much better.

Absolutely. The idea for these is that, yes, they can be everything from mission rewards or mission goals. You need to deliver X number of medical supplies to this outpost on such and such. It would be pretty silly to be carrying individual med guns, just dropping off a pile of med guns and a pile of the goo vials and stuff. So yeah, crates like these can be the objects of missions. They can also be a shorthand for purchasing in game. If you know you want to stack a couple of these in different cities, in different ships, why not just utilize some of the cargo grid as storage in case you have a ship that's got a rather small internal storage. Basically, the uses of these are plenty. They also can be used, as most props are, to simply decorate locations. When you go into the med center or any place where you expect to see these things, these things can be on shelves and decorating around. Or you get to a scene where some kind of trauma triage is happening, the NPCs would have these open with the guns removed and stuff like this. So like all props, they can be used to dress and decorate a scene. They can be used in shopping to speed up that whole pipeline. And they can be used for the mission system as both deliverables and retrievals. Possibly even rewards, I suppose.

Yeah, so I can show you now it's not just a mesh. This one is something I've been developing for a little while. So the animation is obviously super, super basic right now because it is just me bashing it out in one hour on a stream. But this is its most basic form, right? You've got a box, you can pick it up, you can carry it. It's actually quite big. So the metric obviously hasn't been decided on for these yet. We're still testing it and stuff. But we'd adjust the grips. We'd probably make it a bit smaller. But then you can take it on your ship. You can open it. Opens up and then you've got your stuff inside. I can't even pick them up for some reason. Oh, I know why. Basically we have physics proxies on these, and the collision for this box at the moment is currently blocking my ability to get an accurate hit onto the items. So all I need to do is take the top of this, I'll lower it down temporarily and I'll make a more accurate one later. And that's what controls stuff like when you're getting shot at and there's a gap somewhere that you're trying to shoot through, but then the bullets just keep bouncing off the gap because the physics proxy isn't accurate. So we've got our med gun and then we can stow it away again. Same with, you've got two refills, and we can customize these. When missions gets hold of these, I'm sure they'll create, maybe a mission like Siege of Orison, maybe they have a staging area with lots of different items in crates and maybe you can have a support person running supplies back and forth, that kind of thing.

Yeah.

You have to watch out for loose components like this. I've told this story a few times, so I apologize if somebody's been watching me for nine years now and you've heard it before, but there was a time back in 2014, this was still in the Arena Commander days, people were flying a Hornet in Arena Commander and it would just randomly explode. Even in Free Flight, with nobody else shooting, in a completely open map with nothing to collide with, just randomly as it flew, it would explode. And all bug finding and fixing is based on reproduction, how often you can repro it. And they just couldn't repro it. It seemingly happened at random. Some people would fly their ship for an hour and a half and it would be fine. And some people would fly for 15 minutes and the ship would explode. And what it ultimately ended up being was that on the original flight suit, the white and red one where everybody was bald, there was a flashlight that dangled on the belt. The flashlight was physicalized. And depending on how you flew, how many G-forces you did, that flashlight would dink against the side of the Hornet for one point of damage, dink, one point of damage, dink, one point of damage, just rolling around inside the Hornet until eventually it exploded the Hornet. So that took forever to figure out. So whenever I see loose things in a box like that, that's what I always think about. You're going to need to find some way to secure those in the box.

Yeah, luckily I think with item ports we're safe with that kind of thing. This is why I'm developing this, because before that we could just physicalise everything in the box but it would be an absolute nightmare, you wouldn't be able to do anything. I think we had a problem fairly recently with something to do with the chessboard. When a ship was blown up, the chessboard inside the Mercury Star Runner... obviously now cargo kind of scatters, right? When the ship is, I think it's when it's soft-dead, all the cargo comes off the cargo grid and it kind of scatters. That includes the chess pieces on the chessboard. So whenever that chessboard was there, all the pieces of that chessboard would have to break off and it was causing some performance problems. We were trying to work out what it was.

It's always the tiniest little things. There was a version of the Freelancer, the second version of the Freelancer. Freelancer has been made three times, so remade twice. The first time it was remade, that version had 100, 200, I don't have the exact number, little rivets all around the exterior. And it was built by an outsourcer at the time because the company was so small and early. And the outsourcer took the rivet from the concept mesh and just Control-C, Control-V and put it into the in-game model. So the person did not make the rivet at a production scale. He took it from the concept. And concepts are terribly non-performant. And this individual screw had as many polys as the entire ship. So by the time the ship was done, the ship had a resource footprint bigger than the Idris. The Freelancer. And so then, during the old hangar module, people would have six, seven, eight, nine, ten of these things spawn in their hangar, and their hangars would just crash and be really, really slow. It was all because of a stupid little rivet.

We had something similar about a year ago. We have a cinder block in the game somewhere, I think it's an environment asset, and it was the highest poly mesh ever. A cinder block is a square, right? But I swear it was around 50, 60k. It was mental. And we put it in our backlog to basically work on it at some point, and I think it was Joe that did it, he went and fixed it. We crowned him as a hero that day for fixing the cinder block. But yeah, there's lots of stuff like that. You'd think it would be the big stuff, but more often than not it's the smaller bits and pieces.

Yeah. I mean, we're at 1,200, 1,300 people in five studios around the world, and maybe not all of us, but most of us are here because of our shared love of sci-fi video games. So there's the chats that go on. Actually just today, I was trying to schedule a D&D game for one of the groups, and they're like, no, Starfield comes out that day, I can't do D&D that week. And I was just like, alright, understandable. We've been on record before. We don't see it as competition. We see it as the rising tide raises all ships. Every space game out there just brings more people into the genre. And you play that game and finish that game and then go looking for the next space game. So the more the merrier in that regard. And there are plenty of people here that I know are excited for just about every game you see out there. I was just talking to somebody who's major into that ship breaker... Hardspace. I love that game. There's a sequel for Deliver Us Mars that's coming. I can't remember what it's called right now, I'm really bad with names, but I want to play that sequel.

Yeah, you gotta love sci-fi. Especially when you don't have a concept. A lot of the time, the concept team doesn't have time to do, for example, just a trash bin. We'd probably design it ourselves. So just having an eye on the kind of stuff out there really helps, especially with the design language.

That's a very good question. And I think it's something that's going to be hashed out over the coming months a bit more. But at the moment, as it stands, it's basically a props team 2.0. But the goal of this team is to push the props a little bit further. Whereas we've just done the art before, the Interactables team is about interactables. So obviously more complex props with more functionality, being able to take on a bit more of the technical setup and stuff like that, and be our own entity really. So usually we're just supporting environment team, AI team, mission feature team. Different teams we provide support, but this is going to be more of a team where we can start shaping our own game features and stuff ourselves, which is really, really exciting. Yeah, it's like a props team, but a more in-depth technical version.

Yeah, so basically, our team's brand new, and we're just starting to find our feet and get the ball rolling on the stuff we have to work on. But previous to this, before we got started, as we were ramping down our work with the previous props team, we took a bit of time to start looking at what kind of cool new features we could develop, what we're able to do with the skills we currently have, and then maybe what we're able to progress into later on. So to that end, we sat down and thought, okay, what can we do with props? So we thought it'd be awesome to create a cool set of props that players would be able to maybe pick up, take with them to places and place wherever they'd like. Within reason. So we sat around and thought, oh, it would be really cool making some form of camping set, something like that, where players can load up their ship with some equipment, take them to anywhere in the verse, land, and place them and have little camps and stuff.

Make little tableaus, make little campsites of their own and stuff.

Yeah, we went wild with the thought process. We're just like, imagine when all the tech comes online and we have lots of players on a server, they'll all come and set up tents and stuff. We can have our own version of Glastonbury, Star Citizen.

And it's not the only initiative that the new Interactables team is tackling. This is just the one that we've chosen to highlight and explore today. Not everything that we make in Star Citizen Live Game Dev goes into the game, but whenever possible, we do try to piggyback on work that you're already doing so that we can see the process of making stuff that has a better chance of someday making it into the game. The team is already scoping out camping gear. We thought, let's make some camping gear today.

So that would be a good starting point. So part of this was I did go back and review some of the artwork that was already made to be fit for purpose. But the problem is a lot of them aren't built with the animation in mind. For example, we have a bunch of camping beds already in the game, but they're not built in a way which means they can easily be deployed. So part of what we want to do is create some art from scratch that will actually work from the get-go.

So you're saying that in situations like this, it would be more time and cost effective to adapt the art to the animation, than try to take something that was already created and completely redo the animation.

Exactly, yeah. Because sometimes when you're doing modeling, you can take shortcuts. So for example, if you have a folding mechanism, sometimes you'll just model it as it is folded out and then optimize the geometry behind where it folds so that if you actually tried to fold it via the animation, it would not work. Just because where you've tried to optimize it, maybe the mesh isn't clean on the other side, or it just looks a bit jank, or some faces have been deleted to save on polygons for optimization. So going forward, what we usually do is we will set up a template. And then that template has the functionality that is required. And then the artist will go on to make the actual final art.

Well, the idea behind this is that this is the kind of chair that can be deployed anywhere out in the open world. That's the whole idea. So it's not one particular or specific scene. It should be able to go out to microTech or out into the marshes of Lorville or wherever. And chair, bed, table, be able to set yourself up a little scene.

It would be absolutely amazing to see, players already organize stuff, right? So for example, the Daymar Rally, can you imagine if players could bring their own camping chairs and set up and watch alongside the track and stuff? Or later on, we're looking at potentially getting a programmer on our team and being able to create features and stuff like that. So imagine cooking mechanics maybe or something like that, a deployable stove where you can cook things for your mates, tents. One of the ideas we really, really liked the sound of was football goals. So bringing a ball and a football goal that deploys along with you. Thought that would be a really cool idea. But again, this here was a task just to learn some of the more technical stuff in regards to the setup because I know a lot of the art stuff, but in terms of setting it up in DataForge and basically making it actually work, I don't know it that well.

For lack of a better word, interactable.

Stuff like when I did the med bed, that was already preset up by a different team. So all I had to do was place the points, animate it, and then that was done. Yeah, essentially, this was a task to learn the setup behind creating deployables, really.

It's pain. It's drop-down pain.

DataForge is pain. This is DataForge here. DataForge is basically the place where we keep all of the connections between every individual item that makes up the universe and set up the rules and the guidelines in how that interacts with one another. The crazy, insane database at the heart of the Persistent Universe. I'm sure some folks who have been around here for a while might be able to remember links to old ATVs where DataForge was being developed originally. We did some segments on it way back when, if you're really curious to learn more. But it's a totally internally developed way to store and retrieve the data that is necessary to make the Persistent Universe work. And one of the biggest things it did was it got us away from CryEngine's XML system for a lot of stuff. Not everything, for being completely honest, but for most things.

I don't even really know what else you can compare it to, really. It's really, really cool. Essentially what I did when setting up the first deployable that I ever did, it was essentially mashing two DataForge records together and learning how they both worked. And it's a really cool modular system that can actually do stuff like that. It's really awesome.

There are other things like it from other developers; the issue is that other developers don't share them. They don't talk about them. They don't expose them. So the references wouldn't make any sense. But it's our version of a common necessity for MMOs.

Yeah, so really all this is just some demonstration geometry. The main thing is the actual record. So I was talking about DataForge earlier. This is the record for that. And they can essentially do any kind of animation they want as long as they stick to those metrics. So the chair metrics so that the player always sits in the same position, they're able to rest. And as long as it folds up so that they can carry it in the same way, there's no reason they can't change the animation up. Even if it's a different size when it's folded down, they've got the ability to go into this record and actually change down here, I think. They've got stuff like the grip data, so they can change what orientations it can be picked up from, the shape of the grip. So maybe it's a flat grip, maybe it's a rounded grip, the dimension of the hands. So there's all sorts of features. You could spend all day in this just looking it up, but there's a lot of customization to be done through this. And this is essentially just going to be the base version. And then you can see down here, there's all sorts of different...

C_Chair 890, C_Chair Constantine Hurston, all the different...

Exactly. So every chair in the game is sort of based from the same original chair, right? It's always going to be the same sort of height. And that just means that people can be more flexible in their approach when making stuff. All the tech travels forever and yeah, it's cool beans.