Star Citizen Live: Planetary Art

Michel Kooper · Lead Environment Artist

I'll give a very simplified answer and then Sebastian can give it a bit of technical stuffing. We've essentially removed as much paperwork from the whole development process as possible. What we had in the earlier version of our Planet Tech, and it's something that I had as one of the slides in the presentation, is basically this file structure that we had with all these separate files that kept the settings. This was really slowing us down. One of the most important things for making things faster. So by getting rid of that space, getting rid of that time, essentially we free up more time to work on the actual artistic part of it. So creating more content, more library items, more assets, more textures.

Sebastian Schroeder · Senior Environment Artist

Yeah, I mean, that kind of sums it up, right? In essence, I'm not going to put in numbers, but it's gone from weeks to days for simple planets, just with how the process changed and how much more artist-focused it is now. It's just better all around, I'd say.

Jared Huckaby · Host, Community

Part of the challenge going forward, and this applies to many things, not just Planet Tech V4: we have now refined the tech that was used to build the planets that we've built, the moons that we've built. But as we go forward, we'll be building things that haven't been built before, different planet types, different biome types and stuff like that. It's hard to completely estimate how much faster it will take us to build something that hasn't been built yet. It's always a challenge in all game development. But I can say from watching you guys convert the existing planets and moons from V3 to V4, and not just the quality increase, but the incredibly, relatively short time span that it's taking you guys to do that, it's exciting what Planet Tech V4 can potentially do for us going forward.

Sebastian Schroeder · Senior Environment Artist

And to add to that, we were doing it while building the tools and learning how to use them. So it's only going to get faster.

Michel Kooper · Lead Environment Artist

I think the tech is really solidifying. And I think already from going through this process, reworking everything, there's a lot of best practices that come out of that, that anything that comes after will benefit from.

Sebastian Schroeder · Senior Environment Artist

I mean, it depends what people consider to be unique environment landmarks, I guess. The fact that we are no longer tying a biome to a specific height map allows us to do a lot more diverse things. So we could, for example, have a giant desert and just put one single oasis somewhere in there, which was not a thing we could do before without sacrificing an entire set of height map data just for that. So that's always exciting, and you will see it from space if it's big enough.

Michel Kooper · Lead Environment Artist

To add to that, I think with how the system works now, like Seb already mentioned, we're no longer repeating tiles that have a specific combination of this height shape, this terrain shape with these asset types and these colors. So all the elements are completely blended or mixed together in a unique situation everywhere. So depending on the temperature, depending on how this mountain range happens to be in that region where it's warm or hot or cold, what types of assets and all this stuff. So every area is in its own way more unique. And if we talk about landmark things, the goal of the tool is to do 80% super procedural and not have to worry about it. And that frees up more time for us to do more specific crafted areas. We have outposts, which are more purposefully built. It's a point of interest for the players to go to. On the organic side, we would like to do similar aspects to that. I think caves are a good example of organic points of interest for players to go to. And I could imagine having a massive rock or something big that is set up in the same way like we would do a cave or an outpost, that has similar things that would draw a player to it, but it's not necessarily manmade. This is something that we'll continue to explore. I think caves are a good first step.

Michel Kooper · Lead Environment Artist

So water remains a difficult thing. We have an ocean that is on Stanton 1. But as players know, you cannot go underneath the water surface, you cannot really interact with it that much. We have a frozen ocean, which you can actually walk over and all that stuff. So it's an ongoing process. One of the things that we cannot really do yet, but we are really looking forward to exploring and doing more work towards on the R&D side, is all the water-related stuff, because one thing that I've always wanted to do personally is have little small lakes or areas that are not necessarily connected to the ocean. I think that would be really cool. Same with running rivers and, when you get elevation changes, waterfalls that happen as well. But it's quite complex to tackle. So we don't have the answers yet for that.

Jared Huckaby · Host, Community

Now, when you say it's complex to tackle, that's not the same as saying it's not going to happen. That's not the same as saying we're not going to do it or pursue it. It's just that when you start any endeavor, you start with the things that are accomplishable now. That's what we do. It's what V0, V1, V2... You start with the easier wins. You start with the immediate wins, and then you build from there.

Michel Kooper · Lead Environment Artist

Exactly. I think considering that we had to rethink how we do terrain because it's on a sphere, it's on a planet, there's a lot of groundwork that needed to be done before you can get all the other elements in there as well. But I don't think it's impossible. It just needs dedicated time to spend on it. Considering the changes we've been able to do in the last year and all the effort that went into that and the people that we have working on this, I think we will get to that point at some point.

Sebastian Schroeder · Senior Environment Artist

The repeated textures are obviously an issue. They have been, at least in my personal opinion, better in V3. But that being said, V4 is still very much in a groundwork phase almost. I mean, it's pretty good already, but it's like the foundation of things to come. And the fact that the textures are tiling so much is a compounded issue coming from technical issues about precision. Then we have to go into the way we apply textures to the terrain. It's very manual right now, which we definitely want to improve in the future with more procedural things to help the artists do a quicker, better job at this. And then there's also the issue that, I'm not sure if I should say this, but we've added tech to V4 in the last two weeks. So it's still an ongoing thing. And there are definitely plans to improve this from a purely technical procedural approach as well, without any artist input. So it's not something we're going to fix in 3.8, but it's definitely on the list of things to do, hopefully very shortly.

Michel Kooper · Lead Environment Artist

Well, I wouldn't say it's final. I think there's a foundation now that we can continue to build on, and it's one that we put a lot of effort into thinking through and making sure that it is something that we could continue to build on. I don't think there are changes that would warrant a rename to V5, because I think everything that goes on top of this is an iteration on what, in essence, we have now, if that makes sense, from my point of view at least.

Sebastian Schroeder · Senior Environment Artist

Yeah, that's totally fair, and I think a very solid point. I'd say it's good in all aspects; some better than V3, some still a bit behind, but we have plans for everything to improve them. When and how are we going to get around to do that? We'll see. But nothing should warrant a complete rework or a major number change, I guess.

Jared Huckaby · Host, Community

We're never going to say never. Let's put it that way. We don't want to sit here and say, oh, there's never going to be a Planet Tech V5, take that to the bank. That's not what we're saying here. It's just that with V4, we believe we've got a foundation that's pretty solid that we can continue to build on. And we're hopeful that this gets us to where we need to go and will continue to support us for all the time to follow.

Sebastian Schroeder · Senior Environment Artist

I mean, Planet Tech V4 doesn't really change anything about that, to be perfectly honest. That being said, there are some improvements in terms of what data the objects are spawned on. That's more stable and doesn't jump around as much as it LODs out. That's one improvement. But it's also not necessarily connected to the LODs. And then there's also the fact that we continuously work on the assets and try to improve them as best as we can. But it's also disconnected from Planet Tech V4, really. It's just the thing we try to improve our pipelines all the time. So it should be better than before, but it's not necessarily completely different.

Jared Huckaby · Host, Community

From a layman's perspective, it certainly looks smoother to me. And I don't know if that's just me imagining it, but I see more fades than just pops.

Sebastian Schroeder · Senior Environment Artist

Yeah. I mean, that's part of the whole data being more stable thing, really. Before, a lot of the moving of assets was not so much that the LODs were badly made, but just that the data they are spawned on moved around with the mips. When you zoom out, the quality of data changes. And this is something that thankfully the engineers were able to improve a lot. So it should be much better than before in this regard. But the underlying loading process is still the same.

Michel Kooper · Lead Environment Artist

No. In the sense that we will not have the terrain dynamically change based on water flow, or create this playground where people could direct the water flow and make things happen on the terrain. I don't think it will be that dynamic, but it will more likely be the opposite, where we have terrain shapes that, especially with the simulated data that's behind it, will more likely create opportunities for water to appear or rivers to flow when we get to that point, when we get that part of it working.

Jared Huckaby · Host, Community

Now, as artists, we will go through and we will create rivers and anything that seems to have carved its way through rocks and stuff. But we're saying that it's not going to be something that procedurally happens live in the game. We're not talking about deformable terrain where the water will slowly wear away the terrain or anything. But with our tools, we will go through and we will create the impression that these lakes and rivers and streams have been here for years and have worked their effect on rocks and stuff like that.

Sebastian Schroeder · Senior Environment Artist

I mean, I assume they're talking about cold and dead lava planets, which you can totally do, no problem at all. No, in seriousness, I guess they want liquid lava and stuff, but it's the same issue as the rivers, really. It's not something that Planet Tech V4 can do right now. We'll hopefully get to that.

Michel Kooper · Lead Environment Artist

Personally, I think volcanoes and lava are really, really awesome. I would love to have them. But I think whatever we do next, whether it's more advanced water stuff or lava stuff, we want to make sure that it's a complete experience and not just a visual representation of what lava is doing on the planet, and then you walk up to it and you cannot interact with it or anything. So if we add it, we want to make sure that it's a meaningful addition beyond just the visual representation of something that looks cool. We want to make sure that gameplay is involved with it and that we deliver a complete package of, oh, this is what it's like to have lava in the game and to interact with it.

Michel Kooper · Lead Environment Artist

Steep cliffs. Overhangs. But this is more of an asset issue than something we realistically would be able to do with the terrain anyway. And all the smaller bodies of water, all the cool stuff you could do with that, like a lake in a mountain area or flowing water, waterfalls.

Sebastian Schroeder · Senior Environment Artist

That being said, with the way that Planet Tech V4 works, with the whole climate data informing the visuals and every asset being able to know what kind of climate data it's sitting on, it will make it much easier for us to build assets that nicely integrate with the terrain, so we can make essentially terrain extensions as meshes, and they will hopefully look the same as the terrain or at least have everything the terrain has around it. And that is definitely something we want to work on very soon.

Michel Kooper · Lead Environment Artist

Technically, or theoretically, there's nothing holding us back from having the mountain peaks stick through the clouds. However, at the moment, the clouds, as people who do their landings know, are not really that volumetric yet. It's something that's high on our to-do list and on our wish list that's being worked on. We want to make sure that if we allow the player to travel through the clouds on a mountain peak, that experience is what you would expect it to be. So you want to traverse through a thick layer of clouds and just be in the middle of it, just like you would if you fly with an airplane, for example. So that's one of the reasons why we felt like it's not quite the right moment for it yet, because right now it would not be that experience.

Just to hook onto that thread that Ali responded to and just the general questions about steepness and terrain height: I think Ali did a really good job of explaining it and trying to explain why there's a limitation or what the issues are that we're running into. Actually, I have a little video prepared just to help me with that, because I've been reading that thread, so I sort of expected this would come. I just recorded this little example of what actually happens with the terrain and how terrain is basically functioning. So we have a patch of mesh, polygons, and we have our height map and it's displaced. So it's stretching up and down, right? This is based on just a height map of a piece of terrain that goes in and it's pushed out. And what you'll see is sometimes with these triangles or polygons, they get stretched. So the steeper it is, the more they will get stretched. And some of the issues that we'll run into, and I prepared this one as a bit more stylized version of it. This is basically a grid with polygons doing the same thing, a very simple height map that will displace it. And these are basically our textures applied. And what you'll see happen is if there's a smoothness, these textures will remain quite nice and smooth, and they will not distort or displace too much. But what you'll see on the steeper areas is that these textures will get really, really stretched. And it's a very old-school problem to have. But it's one that is also quite hard to fix, especially if you're dealing with a whole planet's worth of height content that is doing all kinds of things. Some games fix this by covering up a lot of these surfaces with additional geometry. That's what we also do. But the steeper it is... and this is just the textures problem. But also, if you look at the triangles or the polygons, it's the same issue. We distribute our assets based on these points, these vertices. And if an area is super stretched, the more stretched it is, there's actually no points to attach these assets to. So they will find a spot to spawn on and it will look very jarring. So if there's not enough geometry to support it, it will cause issues. The steepness is a bit of a technical limitation, and that's the reason why it's hard to do and get results that are very visually satisfying as well, because we can do higher stuff, especially on the large scale. I think Ali said it as well: if you want a very high mountain, you would have to do a bit of a pyramid shape. The top maximum height might be more than six, seven, 10 kilometers. But in order to get there, you would have to do it very gradually. And I think that's what happens in most situations on Earth and most large-scale mountain ranges as well. Yes, the maximum height is very high, but in order to get there, you're traveling kilometers and kilometers on the horizontal axis as well before you hit that highest point. And we sort of had to make a compromise saying, well, in a patch of four by four kilometers, how much would you expect the maximum height difference to be? How steep would it be? And we sort of settled on a maximum of two kilometers between the lowest and highest point. And if you combine that with the height map on the planet itself, which doesn't really have a limitation, we could go way higher. But then it would be less steep, but the maximum height and lowest point would be more than one or two kilometers apart. I hope that sort of makes it a bit clearer.

Michel Kooper · Lead Environment Artist

In the future, we want to get more stuff, road-related stuff. But we would need a tool for that. It had to take a bit of a backseat in favor of properly doing the Planet Tech.

Michel Kooper · Lead Environment Artist

Not yet. But this is something that we are continuously talking about how to improve, and we want to improve it, because I think now that this foundation is set up, we can actually look at the details more. Which means that, for example, the cacti: if you have your rover or your Cyclone, the beast of a machine, and you're just blasting through the wasteland, and then you run into those cacti and you're standing still, it doesn't really match your expectations. When you're driving around in a vehicle like that and you're hitting what I would expect to be a softer object, or at least I would expect it to break when it gets hit. So stuff like that is stuff that we are still looking at on the to-do list.

Michel Kooper · Lead Environment Artist

Yeah, I think that ties into the height map, or the technical limitations of how the terrain works and is displaced. Any type of crack would have to have some sort of angle. I think Ali mentioned that it works best up to 70 degrees. So if you want something super steep, actually, we would come down to assets along those edges to sort of get more vertical faces, which is also something that we're looking into.

Sebastian Schroeder · Senior Environment Artist

ArcCorp is a very special beast in its own right, I'd say. Even in V3, we had to put a lot of extra effort into adjusting the tech just for ArcCorp. And we have done the same for V4 as well. That being said, since we had to redo all planets, we didn't go all out rebuilding ArcCorp utilizing everything that V4 has to offer in terms of improvements. So we just tried to get it close to or slightly better than V3 ArcCorp, which I think we hopefully did. It benefits from some of the improvements, like the space-to-ground transition being more consistent, but it's not a very different planet, I'd say.

Michel Kooper · Lead Environment Artist

Yeah, I think everything that improved in terms of the consistency from orbit to ground also is the case for ArcCorp. But a lot of the improvements that happened for V4 are very much based on the organic side of how a planet would work. So temperature and all this stuff, and what happens with terrain and what happens with assets and how stuff sort of mixes together. And ArcCorp is a very unnatural place because it's basically completely covered, right? So it's not benefiting as much from it as the organic locations.

Sebastian Schroeder · Senior Environment Artist

For now. We'll see when we hopefully get around to it.

Jared Huckaby · Host, Community

Eventually, we've stated before in the past, ArcCorp will need to be revisited to add things like its oceans and stuff like that. So I would imagine that whenever that time comes, there will probably be a good chance to look at the other stuff.

Sebastian Schroeder · Senior Environment Artist

Look at what it can actually benefit from in Planet Tech V4 properly, with a proper rework of content.

Jared Huckaby · Host, Community

There's a couple things on the bucket list, I think. The big butterfly seam at the equator.

Sebastian Schroeder · Senior Environment Artist

That's still there.

Michel Kooper · Lead Environment Artist

I would definitely say any footage that comes straight out of an editor, like you mentioned, is flying around at ridiculous speed. So the popping or the loading of these assets is not meant to keep up with that speed. So it will be much, much less for players just playing the game and flying around.

Sebastian Schroeder · Senior Environment Artist

Eventually, we will. But for 3.8, it's just redoing what was there before and getting the whole set of planets out in time.

Michel Kooper · Lead Environment Artist

We do have mud textures. We do have some mud textures already on some of the planets. I think you'll find some mud on Stanton 1 in and around the trash. But I think the question is more about will we have slippery mud and will we see deformation, footsteps. I think it ties into what you would expect to see with a more complex simulation of snow on the ground as well. Those effects of stepping into deeper stuff, that's on the wish list that we'll still have to look at. So we have mud textures, but not slippery mud that will change your physics drastically. You won't be able to create these massive mud slipstreams yet.

Sebastian Schroeder · Senior Environment Artist

We don't have deformation, but I do think we can make the mud slippery for vehicles and stuff. I think that's possible. I mean, the frozen ocean will also be slippery, so I think the mud should be the same.

Michel Kooper · Lead Environment Artist

No, we have some basic friction, but it will not be visually very involved.

Michel Kooper · Lead Environment Artist

I think in a lot of ways, the environments that we're building are not landscaped by humans. First of all, we've tried to really improve our scattering methods in 3.8. So with the terrain being much more an element that feeds into the scattering system and where we distribute assets, that should already create a way more natural feeling of, okay, this is a dense area here, it's actually a clearing that you can sort of follow, and it feels like a natural path. But due to the procedural nature, you will still find areas that are super dense and less dense. In some areas the algorithm just gives you maximum rocks. So that can happen. I don't think that will ever really change. But we've really tried to balance that out wherever an artist is involved and is looking at an area.

Sebastian Schroeder · Senior Environment Artist

But nature isn't always nicely traversable either.

Jared Huckaby · Host, Community

When you say things like, I don't think that'll ever really change, we're talking about there will always be difficult areas. When you're building an entire planet, there will always be places where it's difficult to traverse. It's meant to be challenging for you to get through there. But I really like what you said before about how we are constantly iterating on the algorithm, on the way that these things are placed, on their densities and stuff like that. We've worked on it in 3.8, but that's not the only patch where we've improved it. We've actually improved it in previous patches as well. And it's something we'll continue to work on. I know many of our developers watch the YouTube videos. They watch the Twitch streams. In addition to playing the game ourselves, we watch other people play. We see their experience. And we do want this to be a fun but challenging experience in a lot of ways. So we will continue to watch, we will continue to evaluate, and we will continue to adjust, as we have been, based on the experience that we see players have.

Michel Kooper · Lead Environment Artist

Just to add to that, in some of the sci-fi novels that you read, like The Martian, one of the things that I found really, really cool about it is he had to plan out this route of kilometers to get from point A to B and plan the whole trip. And there were so many dangers; even basic rocks were very dangerous. Like, this boulder field is going to be a real issue. I personally think it's really cool. But I also appreciate the desire to have a good, smooth-running rally across the moon's surface or do racing. So we're not doing it to make life difficult. We want to make sure that we can facilitate both, basically.

Sebastian Schroeder · Senior Environment Artist

Hopefully. I mean, the fact that we have everything based on temperature and humidity now, which is almost climate, will allow us to do much more than we could before. But it's still something we need to put more work in and build things specifically for. So hopefully we will be able to have seasons soon, or eventually. Who knows? But it would be nice to have it. And we're already preparing some of the assets in a way so we can actually change the color, like have the leaves change color and stuff. So we are slowly preparing for it, but I'm not sure if there are specific plans for when that is going to be a thing.

Michel Kooper · Lead Environment Artist

It's a very large-scale dynamic thing that would need to happen. So you can sort of predict or imagine how we would set up assets for it, but how all those components would tie in together...

Jared Huckaby · Host, Community

The point is it's not coming next week, if it comes. One of the really cool things about Planet Tech V4, and we talk about it quite a bit in the ISC that went out yesterday with the VFX team, is what you just mentioned, Sebastian: how we now have temperature information in the surface types. And we can pull so much from that. That's what the VFX team is tying into to inform the type of storm that is produced over each area. So it's definitely conceivable, at the very least. It seems theoretically possible to tie into that same information to flock trees in the winter or something like that. But at this point, I'm not aware of any direct plans or intentions to do that. But who knows what the future holds.

Sebastian Schroeder · Senior Environment Artist

Did Ali talk about the snow depositing on the suit of the player? That was a thing in ISC, right? So that's also going to tie into the whole environment as well. So we will be able to have snow deposit on everything in the environment eventually. So there are things coming together that will hopefully give us seasons eventually. But we're not there yet.

Michel Kooper · Lead Environment Artist

Yeah, it's a lot of building blocks being put into place.

Jared Huckaby · Host, Community

Now, the snow buildup affecting the player outfits, if I remember correctly, that's not in 3.8. That's coming later, 3.9, I think, because we have to go back and set up all the different armor types and character types to do that. So the tech is there. We just need to implement it across everything at this point.

Michel Kooper · Lead Environment Artist

Not for 3.8. And this has to do with, if we think about caves on microTech and its moons, I think we would really like to see a new archetype, which would be frozen caves, ice caves. So we really want to do it justice and do it properly. And if we would put in caves now, it would probably be the archetype that we already have, simply due to time limitations. But when we do caves on microTech, we want to make sure that they are proper ice caves and they fulfill that ice cave feel and look as cool as they can be. So not for 3.8, but in the near future.

Sebastian Schroeder · Senior Environment Artist

Yeah, Crusader is even more detached from Planet Tech V4 than ArcCorp, because ArcCorp is at least a solid thing. But Crusader needs its very own tech. Maybe we will be able to use some of this tech and repurpose it for Crusader. I don't know about that. That's probably for Ali to answer, I guess. But it's not going to have any impact on Crusader.

Michel Kooper · Lead Environment Artist

Yeah, there might be some logic or just general methodology that we can sort of take into whatever we do for Crusader. But Crusader will be its own piece of tech. And that would involve all the clouds and having those look and feel correct. It's not even a physical or solid thing that you would fly to. It's basically all types of clouds and layers. It's a very different system.

Sebastian Schroeder · Senior Environment Artist

For me personally, it's definitely the things we can do with it in the future, just because it's so enabling in all the different kinds of ways that we always wanted improvements to have, but we were kind of limited with the groundwork that we had. And now, like I said before, we have plans for essentially every part of it that can be improved. And that's what excites me the most, just so we can get better-looking planets much faster in the future. And a lot of things we couldn't do before. So that's what excites me the most.

Michel Kooper · Lead Environment Artist

Yeah, I think that's pretty much my answer as well. Thinking back on earlier Planet Tech, a lot of the stuff that we wanted to do had to be an exception to the ruleset that we have. The more complex stuff got, the more we found ourselves adding exceptions in order to do stuff. And in the end, you'll have this very complex system of all these rules and exceptions and checkboxes that you need to make sure that you enable for it to work, and then you forget one. And now we're at a new system which fundamentally has a very grounded logic to it, and you can imagine doing stuff and you can imagine, okay, well, this requires temperature, or I understand that I can plug this into the flow maps and the humidity in order to achieve a certain thing. And so with that data, almost everything you could imagine, I wouldn't say everything, but a lot of the stuff that we would want to do, you can sort of plug into the logic that is there now. And with that foundation, we can actually look forward to, okay, now we can start building again on top of it instead of adding exceptions and little additional checkboxes just to enable certain things.