Devon Fay is a 3d artist with a vast portfolio. He worked on matte-painted environments for Starcraft II, World of Warcraft: Cataclysm, Diablo III. After Blizzard he became the senior environment artist at Infinity Ward. If you played Call of Duty: Ghosts and Call of Duty: Advanced Warfare, you probably saw his levels. Apart from professional work, Devon also does a lot of teaching. He’s an online instructor at Gnomon School of VFX, where he teaches environment creation. All in all Devon is the guy to go to learn about good environment art. We were fortunate enough to meet with him and discuss the creation of high-quality virtual locations.
I’ve always been into art, even as a young kid. I drew a lot and appreciated all sorts of art, from fine paintings to comic books and album covers. I was also lucky to have an uncle, Ted Fay, who was in the movie industry on the visual effects side of things, so I knew about computers and CG at an early age.
As I grew up and graduated from high school, I continued to work on art and was self-taught on Maya, Photoshop, and other programs. Around that time, I’d heard about the Gnomon School of Visual Effects and knew that’s where I needed to go. So, I moved out to Los Angeles and went to school there.
After graduating, I got a job in Blizzard Entertainment’s cinematic department. I worked on projects like Starcraft II, Diablo III, and World of Warcraft. Since then, I’ve moved to in-game environments at Infinity Ward, where I’ve worked on Call of Duty: Ghosts and Call of Duty: Advanced Warfare.
Stages of Environment Design
Before I start a piece, I need to be inspired. Sci-fi, to me, is always a source of inspiration, so that’s why I tend to lean towards those types of scenes.
After the inspiration comes, it’s important to develop your idea so you have a clear path moving forward. Not everything needs to be figured out, but it’s important that you have a clear enough idea that you aren’t overwhelmed with questions.
The next important step would be locking down your composition as soon as possible. This, to me, usually involves locking down the block-out and camera angle for the shot. Once that is locked, I may move it a bit here and there, but usually I don’t move on until I am very happy with it.
Lastly, I also find it incredibly important to have some sort of story being told. Even if it’s just hinted at, a story can bring an otherwise sterile environment to life.
Saving Time During Production
My most recent scene Sci-Fi Alleyway did take a lot of time, probably too much time. When starting something large like this, it’s important to be inspired in the start, and keep with it to the end. By the end of a project like this, I usually just want it to be over, but to not finish it would be a massive waste of time.
There’s a lot of things you can do to save time though. Things like rendering at low quality, using quick automatic uvs when you can, using tileable materials, and fixing things in Photoshop on your final render. I go more into time saving techniques with my Gnomon Workshop.
When you are working at home, there’s no real deadline so you can tweak things endlessly. This is great to achieve an impressive, refined result. At work, working with large levels or complex shots, you don’t have the time to iterate forever. In both cases it’s important to understand which aspects are worth spending more time on, and which are good enough for the final look. Even at home, you got to finish it sometime!
Overflowing Scenes with Details
I don’t think it’s necessary to have so many props and objects to make an interesting environment. I don’t think what makes my environments interesting is just the amount of detail. I hope the composition, storytelling and overall read and feel of the environments are the most important pieces of the puzzle, the details are just the icing on the cake, to add a bit more interest.
Part of why I like to put so much detail, however, is because I can. At work we usually don’t have time to add so much detail. Not just in the time it takes creating it, but also it becomes very inefficient for rendering (realtime or otherwise). It’s fun to push the limits of what can be done, because every year programs and PCs get faster and faster. It’s nice to try and find the limit.
The Logic of Environment Design
‘Just make it look cool’ motto is viable, but that can only take you so far. I think a good environment should feel real, usable, and interesting. You should be able to see the logic in it, feel the people that lived there, and see how it has been used.
Just randomly placing cool looking objects around a poorly planned area will never be as interesting as an environment which is planned from the ground up, supported by reference, concepts, and research.
When working on a project for work, you have an overall story, setting, style guide and all of that. Those are great things you can use to make your environments interesting and cohesive. Even at home though, you can use that same approach: create rules, story and such to help guide the decisions you make in your scene.
For my sci-fi scenes for instance, I always think about if you removed everything futuristic about the scene, would it still feel like a real place? In most cases, yes, they do. They aren’t as interesting, but it would still feel real.
Creating Levels for Multiplayer Games
Multiplayer maps are really fun to make, in my opinion. They tend to be a bit smaller with less moving parts, so you have a bit more time to polish them.
Some of the biggest differences between single player and multiplayer maps is the replayability of the maps. Multiplayer maps are usually played over and over again, so they need to hold up visually for repeated playing. Single player maps usually are played once or twice, so you can be a little more focused where and how you add details.
The biggest mistake an environment artist can make, while working on multiplayers maps, is forgetting about the gameplay of the map. The gameplay should always be the most important aspect to the game, and the art should compliment that. For example, it’s easy to put WAY too many details on walls and floors of multiplayer maps. While this may look cool, it might make it very hard to see enemy players. This can lead to situations where the game doesn’t feel good, or that you have been cheated somehow. You should always be weighing your artistic decisions against the gameplay.
The Tools of Trade
Being a 3D artist is very complex. There are a LOT of tools that constantly change, evolve, and get added all the time. It’s really important to stay on top of changing trends and always try out new tools as they come.
Personally, my most important tools are all the same that most environment artists use: Maya (or any other high-end 3D package), ZBrush, Photoshop, and VRay. With just those 4 programs you can basically do anything. Taking away even one of those would be a massive loss.
I’m familiar with some of the more common game engines out there, but don’t have a lot of experience with them. At work we have our own engine that is really powerful, and at home I tend to do pre-rendered environments.
In-game resolution is really amazing, but when working on things for myself, I don’t really love having to do a high resolution mesh, low resolution mesh, unique UVs, normal maps, all the hoops it’s takes to make game ready assets. That said, with game engines like Unreal Engine getting faster and faster, I’m curious to try doing a more high resolution approach in it in the future, to see how it holds up.
Advice for 3d Designers
When starting in 3D there are a few things that you should pay attention to. The first is mastering the tools. As I said before, the tools are vast and complex and pipelines at work are very complicated. The more you can understand, the better and faster your work will be. It’s important, however, to remember that 3D art is still art. So, all the things that have been taught for centuries should still be studied as much as possible. Composition, design, color, storytelling, all of that is so important. Lastly, there is SO much information out there these days, you can read endless articles and watch endless tutorials, but there is no replacement for doing work. You have to put the hours into 3D to get good at it, no way around it.