How to Approach Environment Production?

We’re happy to present our full interview with Liz Kirby – a very talented environment artist, who’s recently finished her work on critically acclaimed first person shooter Halo 5: Guardians. Liz was kind enough to talk about her experience building game environments and 3d model. She also gave some recommendations as to her choice of tools for beginner artists.

Introduction Liz Kirby

I’m Liz, currently working as an environment artist for 343 Industries, having worked on Halo 4 and Halo 5: Guardians. Before that, I was doing medical animations at Arkitek Studios as an intern. At the tail end of my internship, a friend that I met at school recommended me for an integration role at 343, which involved taking already-made assets and getting them to work in-game, or adding additional functionality to more dynamic assets. I did this for awhile, until eventually I settled into content creation. Now I build props and environments!

I’m originally from Washington state, which is where I remain today – though more outside the country and closer to cities now, haha. Growing up, I didn’t realize Washington was a hub for game companies, so I can’t say that was my reason for wanting to become a game developer, but it definitely helped later on, both education and career-wise. I played games pretty regularly as a kid, starting with your usual Nintendo lineup and a lot of Sierra games, but it was when I discovered RPG Maker 2000 through my brother, (and countless months making terrible games through it), that I decided that yes, this is what I want to do, and I can do this as an artist. (I always loved drawing, and I had the most fun making art assets for my bad RPGs.) I thought I wanted to be a concept artist for a very long time, but 3D grabbed my attention in school, and I’ve stuck to it ever since.

What’s it Like Working on A Big Project for an Artist?

I wholly recommend beginner artists to not shy away from bigger companies and to seize the opportunity to work for one if given the chance. I almost didn’t go for the job I have now, thinking I was too inexperienced and would get drowned out immediately. I always had it in my head that I would be ready for a big studio in maybe tens years time, but then came the unexpected opportunity to start at one immediately. I figured this was a rare chance to hit the ground running, and hit the ground running, I did! If you can keep up with the pace, there is so, so, so much to learn. It will be incredibly overwhelming at first, but you’ll find yourself improving at a remarkable pace. There’s just such a wealth of experience to learn from; I was constantly, and still am, learning something new every day.

There is some room for creative exploration at large companies, but how much is somewhat inherently tied to a person’s role. A 3D prop artist, for example, is usually at the end of an asset’s development – the prop has already been designed, tested, and given artistic direction with concept art. The 3D artist will have to interpret that concept, (never expect a hyper-detailed orthographic), and here is where creativity can shine, but for the most part, the direction of the asset is already decided. A world artist, on the other hand, may have more creative input in fleshing out design block-ins, though they, too, usually work within concept, an established palette, and need to adhere to gameplay design. I guess what I’m getting at is: creative exploration exists to varying degrees, but at the end of the day, a production requires constant collaboration, and isn’t a space for complete independence. A team needs to work as a team.

Approach to 3D Modeling

Whenever I start modeling, there’s a few guidelines that I stick to, but form and function are definitely at the top of my priorities. Detail shouldn’t be added just for the sake of adding detail, there needs to be consideration on how everything fits together.

It should look functional. For example, say you were given something like a power core to model. If this power core existed in the real world, would it appear as though an electrician could fix it if it were to break? Do the panels and wires running across it make sense? How would they begin to disassemble it if they needed to? Even with alien structures, you always want to sell functionality. This doesn’t mean it HAS to be functional, just enough to convince the viewer. If an object has a large bevel, it should have a reason to other than “it looks sci-fi.” We have different ideas of what “looks like sci-fi,” but we can all look at something and innately determine if it has a function, even if we’re not sure what that function is.

This even applies to foliage, or anything organic. How does it grow, what kinds of animals eat it – is it even edible? If it’s not edible, is it carnivorous or poisonous, is it colored in a way that denotes this? Maybe it is edible, but it’s spiky and hostile to protect itself. How does it grow, how does it die, what does it grow out from? If you consider functionality, more interesting shapes will follow suit. Form follows function.

At a larger scale, the same applies. Don’t just add things every which way to fill up a scene, try to consider how it all exists together. There is a story to every environment, environments are shaped by what’s in and around them. If you’re going to build a sci-fi hallway, consider everything about that hallway – give it a story. When I build environments, I like to think of characters that might inhabit them and insert their personalities into the scene. (Which is why my cockpit cross-section has sticky notes everywhere.) If you have narrative to work off of, all the better.

Composition of the Environment

Composition is very important, even in games where the composition is constantly changing and you can’t pin the player down. You’ll never be able to create an environment where every shot is a masterpiece, even in the most beautiful games, the player can still manage to find unflattering angles, or maybe stick their face right up against a wall, but the scene should still try its best to hold together and guide the player when looking at it overall.

Take a game that has a huge open-world space, for example. The player may be plopped in the middle of a sprawling field, but there are still ways to guide and orient them, like maybe a focal point in the vista. They can run around in an empty field all they want, but if that’s all there is, eventually they’ll find that object in the distance far more interesting. Capture your player’s attention.

If you don’t want them straying a certain way, leave that area relatively plain and make their objective more visually interesting. If your scene is inundated with detail, make the main path less detailed. Use color cues, negative space, set dress, (or don’t), in a way that makes the primary path obvious. You don’t want to take away a player’s freedom, but you do want to make their goals more visually appealing to convince them that’s the way they want and need to go.

Production of 3D environments and 3D models

The production of whole environments and separate 3D models go hand-in-hand, but when building environments, it’s more level assembly and set dressing than modeling. Pieces are being fit together rather than built. Sometimes there might be a need for a unique asset that wasn’t planned for, in which case it’s made by kitbashing the asset together, mapping the asset to trims and tiles, or just outright sending the piece out for modeling, but mostly the world builder should be living within an existing library of assets.

When modeling separate assets, these could also be kitbashed and mapped to trims and tiles to save on draw calls, but there are also unique bakes. If the asset is something like a wall, you’d want to build it to a grid and make it modular so that the world builder has an easier time building with them. If it’s unique, you should still be aware of its usage and the role it plays in a scene, but it might not have to adhere strictly to a grid. When building individual assets, you also need to make sure that they’ve been prepared correctly, such as creating LODs and collision for them. The world builder needs to be aware of this too, but they will be further downstream and shouldn’t be inheriting broken work.

While building environments, there’s also a need to be far more conscious of game design choices. The environment needs to look good, sure, but it also has to be fun to play on. Same goes for props, (ex. don’t build something the player will get stuck in or doesn’t fit within the visual language of the game space), but with environments, design really needs to be taken into consideration and coordination with the game designer must exist. You can make a fun game with bad art, but you can’t make a fun game with bad design.

Tools for a 3D Artist

For a 3D artist, I’d recommend Photoshop, xNormal, ZBrush, and the artist’s choice of MAX, Maya, Modo, or Blender as a good starting toolset. Autodesk software has free student licenses, xNormal is free, Photoshop can be bought on a monthly subscription basis, and while ZBrush has a higher price tag, I wholly recommend licensing ZBrush. You can model, sculpt, retopologize, texture, render – sometimes I rarely leave ZBrush during a project, that program is next level wizardry magic.

Maya and ZBrush are usually the two I complete models in, with Photoshop, ZBrush, and xNormal coming in to finish them off with textures. Sometimes I use Knald to generate additional maps, as well as the Quixel Suite for additional touch-ups. Lately I’ve been really interested in Substance Designer, too, which is all but synonymous with texture creation now. I’d recommend exploring these options, as well.
For personal work, if I’m working on a single game asset, I’ll bring it into Marmoset Toolbag for renders, but if it’s a scene, I’ll use Unreal Engine and do my assembly in that. If my work isn’t meant for games and is only going to exist as a high poly asset, however, I like to render everything in KeyShot. KeyShot is a great tool for beginners, since it’s all very intuitive and produces great results. Like ZBrush, it’s a bit more of an investment, but I think entirely worth it.

The Future of Game Development

I think we’ll definitely be seeing great things come out of VR, though whether or not that’s where gaming is directly headed, I’m unsure, as I haven’t had too much experience with VR, personally. From what I have seen, though, I think it’s fantastic technology and definitely has its place in the industry.

I think game development has already rapidly become more accessible. Ten years ago, I couldn’t do half of what is possible now. I had RPG Maker 2000 and edited sprites with a program called iDraw. (No relation to Apple.) Somewhere along the way, I also received a copy of Paint Shop Pro. (I think it came with a printer?) I used what I could scrounge up, really. Social media didn’t really exist back then, either, so if I got stuck, if my two or three web forums and fan sites contained no answers, I was on my own.

Now, there’s everything! You can contact companies and like-minded developers in so many different ways. CryEngine and Photoshop have affordable subscriptions, Unreal and Unity are free. Autodesk software is free to students, and there’s also Blender. Then you have games like Mario Maker and Minecraft, or Skyrim with its enormous modding community, which I’m sure will be the makings of future designers. The knowledge base is all there, too – you’re not limited to hitting F5 on a forum page or hoping your local community college has something to do with game art classes. You can learn directly from veterans in so many different ways.There’s also crowdfunding through sites like Kickstarter and Indiegogo, so not only have the tools and knowledge become more readily available, but now there’s also the potential to fund it.

I don’t think we’ll ever see the downfall of indies, not with all this at our disposal. Indies are agile, with untethered creativity and bottomless passion for what they do. There will always be someone with an idea in their head and the ambition to see it through.

Liz talks about her studies in DigiPen!

  Liz Kirby, Environment Artist, 343 Industries