Martian Observatory

For my Level Design class, I was tasked with creating an elimination map fit for a third-person shooter using the Project Lyra template in Unreal Engine 5 in 5 weeks. Playing the demo level that came with Project Lyra, I found that the shotguns were not very practical, as you would end up getting killed by the enemy AI from far distances. I was inspired by this limitation to challenge myself, by designing a level focused just around the shotgun and its close proximity combat.

In order to have an arena that encourages the use of the shotgun, I needed to make sure that there were not a ton of long-distance vantage points, and that all primary chokepoints could be navigated to in 3-5 seconds. With that in mind, I decided that making the arena have a small surface area with multiple levels, akin to the final fight scene in Star Wars: Episode I - The Phantom Menace, would be the most beneficial. I had a vivid image in my mind of a player dropping from the top, and hitting a player on the way down. To start the construction, I made a massive hollowed-out cylinder as the arena with multiple platforms, and stairs moving between the platforms, only to realize a couple of flaws when I playtested this idea. The first was obvious - it was extremely dark, even with the generous amount of lights that I placed in the scene. In order to resolve this, I made the exterior cylinder glass, which looked so pretty with Unreal’s lighting system that I decided to make glass and copper the primary elements of the level, mimicking arboretums and observatories, with a golden hour kind of hue inspired by games like Journey and Life is Strange.

The second flaw was more fundamental and took a lot of work to fix. The main complaint that I was getting repeatedly was that the arena was too small, but increasing the size of the the arena led to weapons like the shotgun proving pretty useless. My professor advised me to play Quake III Arena, which I marathoned that night with my friends, playing countless LAN Deathmatchs and CTFs on the lab computers. The main thing that I noticed about Quake was the way their labyrinthine levels were laid out. There were a lot more twists and turns to their levels, and every area had at least 2-3 exits, allowing for lots of different paths through the space. Inspired, I added two teleporters to the bottom area to allow for quick and easy access to the top area, in addition to creating symmetrical wings on the side, allowing the player to reach the second floor in a multitude of ways as opposed to just having to take the stairs. This relatively late decision completely changed the feel of my level, where during playtesting I was starting to see my testers encounter interesting combat scenarios, where the primary ways to win were reaction time and positioning. After that, I did a final lighting and post-processing pass, and submitted a video of my level with explanations of the various decisions that I made.

  • Having continuous testing and getting feedback based on those playtests was invaluable towards making a more enjoyable experience for players.

  • I turned an arena that was way too dark for players to navigate comfortably into a visually cohesive, incredibly pretty arena, in my opinion :). Learning the basics of Unreal Engine’s post-processing and lighting systems allowed me to create the golden-hour lighting that I was picturing in my mind.

  • Inspired by Quake III, I decided that my level needed to have multiple entrances/exits for each chokepoint. By adding side wings to my level, in addition to teleporters, I successfully accomplished that.

What I Did Right

  • At the time, I had no experience using Unreal Engine’s modeling tools, so I made plenty of mistakes creating the shapes for this level, which resulted in having to recreate the level several times. I didn’t learn about mesh cuts until halfway through the project, so I was using plane cuts in the meantime, which made the cylinder incredibly sloppy.

  • With the amount of time I spent on the level, I had to finish the video to turn in rather quickly. There are some issues with the mixing of my audio compared to the SFX and music, and I definitely would’ve swapped the black lines for white lines for my drawings, because they blended into the background too much. Also, I just hate my voice on recording.

  • Like with using the modeling tools, I only found out about the cube grid tool very late into production, and it took me a bit to get situated with it. This resulted in the cylinder not blending great with the multiple pathways (which were made with the grid tool). I would’ve laid out both the area the cylinder would take, in combination with the pathways next time.

What I Did Wrong

  • I liked the post-processing and lighting passes that I did, but I think I could’ve improved it with more time and knowledge. I spent a while trying to get volumetric god rays, but I couldn’t figure it out on time. There was a method that I saw with shell texturing to mimic the rays, and I would have loved to try that.

  • The landscape around the arena could’ve used another pass. I was trying to mimic a Mars look, but I don’t think I used the right textures for it. I could’ve added more dust and less rocks blanketing the surface, and I would’ve liked to experiment with the terrain height generation if I had the time.

  • I didn’t try to change any of the Project Lyra enemy AI, and I think it could’ve been worth looking into that. In the project, the AI couldn’t use the jump pads in order to get the weapon floating in the air, so getting them to do that could’ve resulted in there being more challenge.

What I Could Do Better