What does a material artist do?

The Liquid Swords Art Team is one of the biggest in the studio because we’re lucky enough to work with many artists dedicated to specific areas. We sat down with one of our material artists for a closer look at the different materials and textures that will bring the objects in Game1 to life!

My name is Mickis Lindén, and I work on the Art Team, primarily with materials and textures. For the past seven years, I’ve been working with environments and hard surfaces, so it’s really fun to explore a new area.

When you’re developing a game, you start greyboxing to get an idea of what the environment or assets should look like. Material is the composition of the normal map, including the diffuse, the specular, or the roughness, depending on the settings of your game engine. At Liquid Swords, we work in Unreal Engine 5 with PBR [physically based materials], so I put a lot of thought into metal and wood materials.

Most maps are either metallic or not metallic. If it is metallic, there are a lot of light properties you need to account for. For example, pure steel is pure metallic. But then you have rust, which is also a metal, but it’s not metallic; it’s dialectic. Wood doesn’t have any metallic features, so the difference between those types of material is the porosity or roughness values. Pure metal has a very low roughness. It’s basically a black texture mix. Driftwood, on the other hand, is very dry so it has a very white roughness, putting it on the high side of the scale.

Some games don’t use this map, so the developers have chosen cel shading instead, which gives you different materials. Cel shading doesn’t use any textures and doesn’t fuse; you work with the model’s own colors and lighting. Depending on where you've set up your light source, there’s more of a gradient from one color to the next. It’s the same principle that comic books use – the gradient is sharp and you only use a few colors that correspond to the lighting in the environment. Some cel shading uses and outline to pop out the models, you can see an example of this in Chants of Sennaar. To get what you want, you need to work with polygons.

The difference between them is the final result, so it really depends on the production and the style you want your game to have. Legend of Zelda: Wind Waker uses cel shading, while The Last of Us uses PBR. It’s filled with high-res and very detailed textures, making it a heavy game.

I don’t make a lot of the new material from scratch. Typically, we’ll find what we need by scanning the internet, checking out Quixel, or looking at different textures. I usually try to match it with a scanned asset to make sure I’ve gotten the values right. It can be a bit hit-and-miss. Sometimes, you just have to trust your eyes. Both scanned and procedural textures have their own pros and cons, at the end of the day, you make the choice depending on the game, the team, and time.

My next step is combining them with a very basic template, before adding them in Unreal and setting it up with our own shader. This allows other people to edit, so if they want something to be a certain color, they have the power to change it. I try to set up as many of these workflows as possible so everybody has what they need.

Recently, I’ve been making wallpaper, and that process starts with research. I went back in time and learned so much from looking at the wallpaper alone. The 1940s had terrible wallpaper; it was a depressing period, so that makes sense. The 1970s are my favorite. They had fantastic walls. But whichever wallpaper you end up making, you have to remember that no wallpaper in real life is perfect. There are lines that separate each roll from the next. It’s easier to spot them if there’s a pattern on the wallpaper, but even the solid colors will have a line here and there. The perfectionist in me wants to avoid this, but I know I have to fight that feeling and add these imperfections, along with air bubbles and tears.

I like metal; it has set values, so it’s one of the easier materials to work with. But I find wood to be the most interesting. There are so many things you need to consider. Has it been polished, or is it untreated? Even driftwood has a sheen to it, even if it’s been lying in the sand for a long time. The tricky thing is finding the correct values; otherwise, it will end up looking like plastic.

In my opinion, glass is the most challenging material. Making it look good from different angles in the engine is difficult. You can make it look great in a rendered cutscene, but there are so many more light values to account for in-game, along with the motion or where the player is positioned. It’s refractive, so it has to look believable. And it breaks, so chances are you’ll find yourself with small pieces of glass. Don’t get me started on mirrors! Both glass and mirrors are common in the real world, so they’re difficult to avoid in-game. It’s very expensive, both in terms of time consumption and for the engine. So you really have to decide if it’s worth having a copy of the character in a mirror or if a blurred shadow is enough. Luckily, most players are very forgiving when it comes to glass and mirrors.

I mainly work with the character and technical artists. They’ll come to me with a request to use a certain type of material, and I’ll either send them one we have or create a new one. I don’t work with the design team directly that much unless they want dynamic or damaged versions of assets. Sometimes, we have to push back on their ideas, though. They’re always fun but not always practical. Take a landline – a phone with a cord. I agree that it would be really fun to be able to pick up the phone and smash it into the ground. But we then have to consider whether you can smash every other phone in the game. Can you throw them? But the main thing is the cord. It will have to be dynamic, so now there’s physics involved. It’s going to go haywire, and it’s going to propel. So we have to figure out a compromise together.

We expect our players to try everything they can think of in the game, so we don’t want to give them things we know will break in the wrong way. If we fill a room with glass bottles, they could try to shoot at every single one, assuming they’ll shatter. That’s what glass does. This can quickly become very expensive, but you can find smart solutions like adding decals where possible or showing the glass demolishing effect. However, each of these fixes adds another node to this big web, so we have to think ten steps ahead. Otherwise, the web will get too big, and we’ll end up causing problems that we need to fix down the line.

We’re still in the middle of developing Game1, but if I look back at where we were a year ago, I remember we had this idea for a specific type of fabric. We had the material set up for architecture, and the mock-up worked well, but it would have ended up being too expensive. So we had to remove it, which meant that I spent two or three weeks going through the fabric folder and deleting it.

When you delete a material, the object transforms into a magnificent magenta color. It’s glaringly pink so that everyone can see that this is wrong. This asset isn’t meant to look like this, and because the color is so loud, we won’t forget to update it. Now, that fabric has been replaced with a cloth that works much better, and those assets are no longer pink. One of the final and most important things you need to do is update the documentation so everyone else understands how they can work with the new material.

Quick fixes. If you don’t do a thorough job or take the time to update your documentation, quick fixes often blend in and become a huge problem in the future. It will nestle in with all the other connecting parts and become a crucial element, so you end up having to unravel much more than you started with. They remind me of that screensaver in Windows 95, with all the pipes that just keep growing and connecting.

Try to stay focused on the right details. You’re going to look at everything as an artist, but the player isn’t going to be. They’re not going to be looking up, for example. I never look up when I play games. When was the last time you did? But when I’m working, I’ll notice all these small flaws and have to remind myself to stay focused and kill some perfectionist darlings.

Another big thing is not communicating properly. Anyone at the studio can create a material and add it to the library, but if they don’t let me know it exists, I can’t approve it before they start using it, and that’s when things start to spiral. So don’t underestimate the value of a clearly written Jira ticket! Another communication issue is naming conventions. I try to avoid temporary names, especially since so many other people need access.

You also need to remember that if you change a name, you need to make sure the new name is updated everywhere. I learned this the hard way when I accidentally removed the floor in an entire bank. I had changed the name, so I had to fix the redirector because it was still connected to the old name. The thing is, when you fix a redirector, it checks out all the material connected to it. So I checked out all the marble material that made up the floor for this bank, and poof! The second floor was gone because some of the assets hadn’t been checked out.

It’s much more technical than meets the eye. People usually assume my job involves drawing and painting or that I primarily work on a WACOM tablet. I spend most of my time working in different programs, such as Substance Designer, Painter, Blender, ZBrush, and some Photoshop. Sometimes I photograph or photoscan materials as well.

A general misconception about the entire field is that people think game developers sit around and goof off all day. We play many games in the studio, but most of it’s research. Players outside the field tend to follow the pipeline and do what the game tells them. The rest of us will run around in circles to check out the run cycle. How does the character stop? Does it jump? Now it’s learning, and so are we!

But honestly, we will play a lot of Mario Kart and Smash Bros. just to have fun and blow off some steam. That assumption is valid.

Do fun projects. Going to school is excellent, and they’re one of the best ways to land an internship, which is a good way to join the industry. That said, I think creating something that you care about shows potential employers what you can do and who you are. Game development is very collaborative, so your personality can help you. What you think is fun about your project says a lot about who you are as a person. You can always learn new programs and skills on the job, but your personality doesn’t change much, so show them who you are from the start.

Finally, mistakes happen, so let them happen. Mistakes can be good, and they help you learn. I’ll never delete an entire floor again, at least not by accident.

Thank you for telling us about your work as a material artist, Mickis!

The next time you find yourself in a video game world, take a closer look at the objects around you, the ones you pick up, and especially the ones you’re about to destroy. How do they fit into the world, and what makes you want to interact with them? Hopefully, this new appreciation and insight into the process will make taking a bat to a row of glass bottles even more satisfying.

If you want to get to know more teams at Liquid Swords, check out our interview with the UI/UX team or our Technical Artist!