3d artist Robert Hodri described the way he produces environments and high-quality 3d assets for AAA-games.
3d artist Robert Hodri described the way he produces environments and high-quality 3d assets for AAA-games.
My name is Robert Hodri and I’m a self-taught 3D/Environment Artist with over seven years of professional experience in the video game industry. Prior to that, I spent almost six years in the Quake 3 and Doom 3 modding community making various multiplayer and single-player maps or working on personal 3D projects.
I started my professional career in 2009 at Crytek in Frankfurt/Germany as a full-time Environment Artist and worked on projects like Crysis 2, Crysis 3, Ryse, Warface and Homefront: The Revolution. Later I took a job offer from id Software and moved to Richardson, Texas where I’ve been working on DOOM, including all three multiplayer DLCs.
Tasks of an Environment Artist
I would say that can depend on the project and company you’re working for. Are you working on a real time strategy game, first person shooter, open world game with coop and online elements, virtual reality, mobile games etc.? There are so many different kinds of games these days, from low budget indie ones to high budget AAA blockbuster projects. Your role and tasks can heavily vary, depending on the team size, where you work, and what your position is (junior/mid-level/senior). I worked mainly on first person shooters, so that’s why I try to answer the question based on my personal experience. Normally when you start out in the industry as a junior environment artist, you work on things that can have absolutely nothing to do with environments or art in general. That can be for example, cleaning up the perforce folder structure, doing LODs and proxies for props, setting up breakable objects, gathering references for your lead, etc. Once you gain more experience, your responsibilities are also growing, of course. You’ll be working on smaller areas in the level and bigger props. A senior environment artist normally takes care of a whole multiplayer or single player level, not only from an art perspective. Because there are so many different people from different departments involved in creating a level, you also have to manage and organize a lot of people involved in the creation process.
From an art perspective, my tasks are mainly blocking out areas, doing a first lighting/particle pass, creating materials and textures or level art beautification like placing props and decals. Thinking about things like “environmental story telling”, giving the spaces you work on meaning. Finding good shapes, silhouettes, colors and creating a good and interesting mood for your environment. Normally you’ll get a block-out from a level designer and depending on how big your concept art department is, you’ll be provided with concepts and paint-overs. You then start detailing the area based on those approved concepts and try to stick to them as close as possible, while of course trying to improve on them as best as possible.
Often you have managing tasks like requesting assets from prop artists, doing quick overpaints or requesting them from a concept artist, talk about mood and colors with the lighting artist, communicate with designers to make sure nothing blocks gameplay, schedule tasks with your producer, make sure deadlines like alpha, beta, and content lock can get done on time and so on. It’s not only about getting all the art in and making things look beautiful. Later on when finishing a level you normally have to analyze problem areas for performance. So you need to have a good understanding of the engine and the tech of your game in general. A lot of things can have an impact on performance. The triangle count of an environment, too many expansive props, particles, dynamic lights or decals are the type of things you have to find and work out with different departments. And of course fixing bugs is another main task. Stretched UVWs, gaps in meshes, collision problems, all these issues need to get fixed before a level can be finished.
Because there are so many different persons involved in creating a level, being able to iterate quickly and take constructive feedback is very important for an environment artist. Normally levels get reviewed on a weekly basis, so you can’t take it personally when you get feedback from your lead, director or other artists and designer that some things may not work visually or gameplay wise and you need to change them.
The smaller your art team is, the more responsibilities you’ll probably have. In general I would say the main task of an environment artist is to make the game/levels look as beautiful as you can. When making games there’s a simple rule: gameplay before visuals. You can go crazy as an artist and add a lot of details everywhere but player path readability in a level is important for gameplay. That’s something you as an environment artist always have to keep in mind.
Hard Surface Work
There’s always this discussion about functionality versus design, especially for sci-fi assets. To me, when I design something I try to make it as functional as possible but I don’t stress myself too much because I don’t want to be limited artistically. Of course it’s easier for simple static objects like crates or doors. As long as it looks cool I’d say it’s alright. Once something gets animated and is more complex like weapons, robots, vehicles etc. it’s important to think more about functionality and how different parts of the mesh are built and move because they have to get rigged and animated.
As a 3D artist, you don’t always work on the most important hero assets of your game. Barrels, crates, trashcans, rocks are not the most exciting assets to work on, but pretty much every video game needs them. You’ll probably end up working on a lot of those smaller props in your career. Just like an environment where you try to tell a story, you can also have the same approach with simple objects like those. It doesn’t matter if the visual style is realistic or cartoony, just try to make them as interesting as possible. And even if you’re working on yet another crate, try to make the latest one always better than you’re previous one. Same should go with all your other art pieces. Try not to reuse ideas too often and come up with new things. That keeps you being creative and helps improving your skills.
I’d say modularity really depends on what you’re working on. It can be a huge time saver but it can also block you in your designs. For things like concrete/brick buildings it’s kind of essential to have a very modular workflow. Good examples are open world games that are set in a big real world city. It would be a waste of time to make unique assets for all the buildings. It also would make no visual difference because in the end they are just some blocky brick and concrete buildings. You keep those things very modular and easy to use because you can give them a unique look later on in the polishing phase by adding props and decals on top of them.
A non-modular workflow also limits the designers when they want to move things around for gameplay reasons. Level designers work most of the time in the level editor, so it’s important for them to have smaller objects that allow them to be flexible when they need to change things. Depending on the company and workflow, designers are often not too familiar with 3D programs like 3ds Max, Maya or Modo. So whenever they have big level sections as one giant model they are very limited to make any changes and have to rely on the artist who worked on that object to make any changes when needed.
When it comes to sci-fi environments and if the project I’m working on allows for it, I personally like to stay a bit away from a too modular workflow. I normally have some elements that I like and try to reuse and modify them. I specially tend to make building facades unique. Smaller generic corridors on the other hand, I like to keep them a bit more modular.
For level optimization, I think it depends on your tools and engine if you want to use a very modular approach with your environments or just export huge areas as one mesh.
Color and Light
At the blocking out and first detailing stage I don’t think about light and colors at all. Normally I work with just a couple of different colors that also kind of represent different material types like metal, stone, wood, chrome etc. Or I simply use different default materials to separate floors, walls, ceiling and trims. I work with those for some time, especially when it comes to futuristic sci-fi designs. Once everything is detailed out and I’m happy with the main shapes and silhouettes of a scene, I think about the colors and do a first texturing pass, not caring about UVWs too much because I know the geo can still change. Once there is a lighting pass for the level, I add props to the scene and look how the color, light, mood, meshes etc. work together with the rest of the environment. At this stage I constantly check the area I’m working on from every angle, from close up and from far away, always trying to make sure it works from everywhere. Because it happens quite often that you work on a small section of a level and add a lot of different things and then the next day you look at the whole picture and it just doesn’t work together.
You don’t want to give everything the same color everywhere. I like to separate the ground from the walls with different textures and colors. It helps visually and makes gameplay spaces easier to read. I also try not to use too many different textures for them. Of course it also depends on the kind of environment if it’s more organic and natural or hard-surface. For the latter I tend to use more different textures.
Whenever it comes to colors it’s also important to communicate with the designers about which colors are reserved for gameplay purpose. Best example would be red barrels or pipes. They’ve been in video games for such a long time that every player knows when you shoot them, they’ll explode. Or ledge grabs often have a specific color to them that the player can easily read and identify throughout the game.
In multiplayer, gameplay and player path readability is even more important than in single-player levels. Different color themes can be important to give each area a distinct look that helps players to know where they are. Big visible saturated color stripes for each area or bright lights at doorways are good ways to show the player path better.
Communicate with your lighting artist and level designer and make sure the hero asset is placed in a spot where the player will notice it or walk by quite often. You don’t want to spend a month on a prop where the player just walks by and sees it for five seconds or even misses it.
Put your effort and details where it matters and the player can actually see it. It’s also important to know how it will be placed in the level. Will the player be able to come close to it? Will it be covered with vegetation? Or is it more in the background? Depending on the size of your hero asset, you also have to think about how many textures you’ll use for it and how to split up the mesh. Does it make sense to use a couple of unique 4K textures or use tileable textures for some parts? Sometimes you also do those hero assets in a very modular way so you can reuse parts later in other areas.
Designing Pretty, Believable and Realistic Levels
I think it also depends on the game you’re developing. Open world games I would say normally shouldn’t have any barriers at all. Whatever the player sees, he should also be able to reach it. Most linear action games block the player by simply having props placed in the way or destroyed elements. After a while the player will understand the gameplay mechanics and know whenever he can jump over obstacles or open doors and realize where to go or not to go.
If you want the player to purely focus on certain things in a level there are a lot of different things you can do like lighting, prop placement, enemy spawns, or small cinematic moments. All that can help to let the player focus on certain moments.