Making DM-Reverence, Part 2: The Mentality of Art Design in UnrealEd

Hello again! For anyone who remains unaware, you should know that this article is the second half of a two-part series I’ve written about the design and development of an Unreal Tournament 3 Deathmatch map, “DM-Reverence.” The first half of the article (which can be found here, if you’ve yet to read it) deals exclusively with the design of the map — how I took a pencil-and-paper sketch and transformed it into a working in-game prototype. This second half is all about the process I went through to bring art assets in and stop the map from looking like a giant, multi-dimensional chess board.

As I’ve mentioned before, this is most definitely not a tutorial. If you need help figuring out how to bring models you’ve made in a different program into UnrealEd (or, for that matter, how to do just about anything in UnrealEd), I have to recommend Hourences.com. That’s how I started learning, and I’d recommend his stuff to anyone.

Instead, this article is an overview of the mentality that should go into a level’s artistic design. Making a level look good takes a whole hell of a lot more than tossing in a few meshes and hoping for the best. Art has a lot of effect on gameplay, and is certainly what sells a map to an audience.

There’s a hitch, though: I did not find total success in the artistic design of my map. While I am very happy with how the Beta version of DM-Reverence ended up, the fact of the matter is that time constraints kept me from being able to go where I wanted to go. I make no claim to being a master level designer, but I have certainly learned a whole bunch in trying, and I don’t think it would hurt to tell you about it. If you can save time by understanding some core concepts up front, you’ll have more chances to do what you really want: transform your design into an artistic masterpiece.

Ready to get started? Allrighty then, cue the lights.

Part 1: The Skybox

This is really important.

Well, okay, it was a heck of a lot more important for me than it might be for you. After all, I made a space map: there’s skybox in every direction. If you’re making a map that’s entirely underground, you won’t need a skybox at all. But in most cases, you’ll need to create a non-playable ambiance around your map, and that’s what the skybox is for.

To be clear, UnrealEd doesn’t actually have skyboxes. It has sky domes (big half-circles you can map sky textures onto), but I don’t even mean those. When I talk about skyboxes, I’m talking about anything — absolutely anything — that exists beyond the playable boundaries of your map. Maybe its a cityscape. Maybe it’s a mountain range. Maybe it has a skydome as part of it. But whatever it is, it’s there to be seen, not touched.

This is key, because in many maps, the skybox is there to help players orient themselves within the level. DM-Reverence is basically a bunch of floating platforms in a void, and because I wanted to make it symmetrical, a lot of its shapes ended up looking the same. To combat this, I built a skybox with a really big planet that’s down and to one side, and a sun popping out from behind said planet. The result: a fairly intense distinction between one half of the map and the other, at least in terms of the skybox:

See? There’s a “sun-side” of the map and a “space-side,” and the players can more easily figure out where they are every time they spawn. Obviously you can’t just create a skybox and call it a day, nor can you do so without some concept of your map’s theme, but it’s an important thing to keep in mind as we move on to…

Part 2: The Style

I’m going to suggest something here, and I’m going to ask you all to bear with me when I say it: Don’t make your own art assets.

At least, don’t do it the first time around. I understand that you might be an awesome, gung-ho artist. You might be able to twirl Photoshop around your pinkie with metaphorical pizazz, and make Autodesk Maya bend to your will. I am very impressed, I promise. But don’t start there. Instead, recognize that your first map is, very likely, something that you’re not going to want to release to the ardent hordes who crave the scent of fresh level in the morning. It doesn’t have to be a portfolio piece. It is, in fact, there to teach you how to stretch the boundaries of the engine; how to work with junk and turn it into something awesome.

See all those static meshes in the Generic Browser? Yeah, that’s a LOT of static meshes. Holy crap. So use them. Practice with them. Then, when you’re ready, you can fire up 3D Studio Max.

You see, I skipped that step, and it hurt. Once I’d completed my art-less, Alpha version of DM-Reverence, I told myself I was going to create all the art from scratch — the models, the textures, the animations. I spent weeks toying around with different artistic styles, modeling cool animated jump pads, testing out cel shader effects, and trying my hand at TF2-style textures, before I realized I could have spent that time learning how to pull all those pieces together. If you’ll recall, DM-Reverence is the product of a school project, which meant I was getting graded on the outcome and time was a factor.

So I went back to the drawing board, and I came up with an art style that was entirely modular: a series of geometric shapes that could fit together to form everything I needed in DM-Reverence. It was not, by any means, a perfect solution, but it allowed me to pick a theme and stick with it. That’s critical, folks; you need to know, going into art development, what sort of feel you’re trying to pull off. And remember, location and theming aren’t the same thing. You can have an Egyptian crypt that’s creepy, bloody, stark, or cheery. You can use a rich color palette or a drab one. You can go for a cel shaded art style, an illustrative style, a realistic style, or something else. The fact is, you have a LOT of options, so pick one, but for the love of all things holy, stick with it.

For my part, I started with a basic hexagon shape and worked off of it to create interlocking parts that could combine to form almost any shape I wanted:

Each shape followed the same basic structure, with a lit groove around the edges and a slightly bumped central panel, but more importantly, each shape fit to the grid.

I cannot possibly emphasise how important it is to build to the grid. In UnrealEd, it’s not even feasible to build off the grid, since you have such precise control over the grid level you want to work at. But if you’ll notice, those grid levels are always at powers of 2. If you want to find success with your own static meshes, build them in powers of 2!

My result: a series of interconnected shapes that I could put to almost any use:

Part 3: Light and Color

You might notice that my final art style makes use of bright, primary colors. This was intentional: I wanted to use light and color to identify pieces of the map, so that players know exactly what they’re working with.

Examine item placement in stock UT3 maps and you’ll probably notice that every major pickup is placed in a fairly prominent position. This is because, as I’ve noted in the past, UT3 is a game of resource management: Everything else being equal, the more items you can stock yourself up with, the better your odds are. Players should be encouraged to grab armor and health as often as possible, without making it a hassle. That means helping players find those pickups.

As a level designer, you must remember to make items visually prominent in the map, and one of the best ways to do that is with light and color. In a dark room, a well-lit corner with a rocket launcher is going to be instantly visible. In a room that’s dripping green slime, your players aren’t going to notice the Biorifle unless it’s set apart with a different color, like red or bright yellow.

In DM-Reverence, I based my tile design around color. The borders along the edges of any of my tiles could be set to glow any color, allowing me to highlight pieces of the ground. Then, by layering those colors on a monochromatic level, I used the border lights to guide players towards vital resources.

Note how color becomes the primary means of identification in this shot. Red equates to weapons; if you run over a red tile, you’ll get either a weapon or ammo for it. Yellow equates to armor, while green tiles have health vials. I also used color to differentiate jump pads from surrounding geometry, making them blue and their destination tiles purple. And, if you’ll notice, additional lights are used to highlight the most important tiles — the ones that have the actual weapon or the big health pickup, rather than the ammo or vials.

The result, while admittedly a little gaudy, remains effective: You always know what you’re getting into, wherever you go.

Part 4: Wrapping it Up

Am I happy with how the DM-Reverence Beta release came out? Absolutely. I came up with and stuck to a consistent art style, used light and color in meaningful ways, and created the sort of bizarre, otherworldly environment I wanted to create. But am I finished with the map? Hell no.

Thusly do I deliver my final tip: Don’t stop polishing until you’re out of time. There’s always something that can be done better, some piece of the level that could be made more fun, interesting, or visually dramatic. Playtesting in the art phase is just as important as it is in the design phase, because art can impact your players’ perceptions of the level, and thusly their interaction with it.

There’s a lot I still want to accomplish in DM-Reverence. I feel that the catacombs blend in too well with the dais because of their identical color schemes, but am still working on how to mix light and color together to solve the problem. I’d like to provide a bit more visual variety to the platforms while pulling back on their chaotic feel — again, a challenge I continue to muddle through. You’ll always have a deadline for your map, be it for a grade or for submission to Epic’s Make Something Unreal competition. So start early and work hard, because you can pretty much guarantee that you’ll want more time than you have. After all, mapping is fun! And we all wish we had more time to have fun.

That’s all from me, folks. Thanks for reading, and good luck in your own mapping endeavors!