FIEA Level Dev Blog

 This past week, I started blocking out the tutorial area and began reworking the landscape and architecture now that I’m in charge of the cathedral exterior. I’ve been working closely with our environment artist, who put together some guidelines for how the architecture and modular kit will work together. We’re also able to build proxy pieces outside of the modular kit when needed, and add them to our asset list. It’s been helpful having that flexibility while still keeping everything consistent with the overall art direction.

This past week was spring break at GDC. Even though I'm originally from San Francisco, this was actually my first time attending GDC. It was a cool experience getting to drive people from our cohort around the city and show them some of my favorite spots. I tried speed networking a few times, it was kind of hit or miss, but the most interesting conversations I had were with people adjacent to level design. Funny enough, I didn’t meet any other level designers during speed networking, but I did connect with a couple of environment artists and a tech designer, and we had some really good talks. The FIEA reunion was a highlight. I got to meet a lot of alumni and former faculty who were super welcoming and open to talking about breaking into the industry. One of the best moments was when we ended up at another bar after the reunion, where alumni Marchand introduced Greg, Cameron, and me to level designers from Obsidian some of whom were also FIEAns. They shared a lot of helpful advic...

 This past week, the biggest thing I worked on aside from finishing my level was reworking my portfolio and creating a web page for my level. Before that, I made a final push to complete the horde encounter that serves as the climax of my level. Getting the enemy AI to behave the way I wanted was a challenge. It was mostly about staging the encounter to work visually rather than having fully functional AI. Since the enemies were aggressive and difficult to fight, and I didn’t have a proper checkpoint system in place, I ended up setting their damage to 0. This way, playtesters could get through the level with the right amount of friction without constantly dying. For the portfolio, I focused on production value while keeping the styling consistent with both TLOU2 and my own portfolio’s style guide. I took inspiration from TLOU2 's Seattle map and designed my level’s top-down map to resemble a physical tourist map of San Francisco, then placed it on a mock-up of a creased piece of...

 This week, I continued working on the high-rise building and started using a hybrid approach for the blockout. Now, all the main structures like buildings and room layouts are built in Maya, while certain parts of the environment, like the landscape, are done in Unreal. I also realized that some elements, like props and cover, need to be placed directly in Unreal to iterate quickly. Being able to move things around in-engine makes it easier to test if they work in gameplay and adjust them on the fly. This setup gives me more control over structure while keeping iteration fast where it matters.

This week for my TLOU2 -inspired level blockout, I worked on several floors of the second building the player platforms onto. I continued using the process of first building the office as it originally was, then deconstructing it and repurposing the floor plans to create an interesting flow and rhythm as the player ascends. The biggest challenge I’m facing is making sure the floors don’t feel repetitive. Since things like elevators and stairwells have to stay consistent across multiple floors, it naturally limits how much variation I can introduce. I also have to remind myself that not every part of a floor needs to be accessible. Destroying sections and limiting space helps guide the player forward while keeping the environment believable.

 Initially, I started my block out in Unreal to playtest and set up the level, but I later switched entirely to Maya for building everything. This change gave me more control and speed when creating objects. Since my level is inspired by The Last of Us Part II, I wanted to emulate Naughty Dog’s process by using Maya. The downside is that constantly exporting to Unreal slows down my ability to test quickly. While ND likely doesn’t have this constraint, I still think it’s valuable practice to follow industry workflows. In the long run, this process is useful since it allows me to block out in Maya and export to any engine as needed. For my level, I’m referencing several real buildings and landmarks in San Francisco. The Bay Bridge serves as the main objective, acting as a weenie to guide players. While I’m basing the layout on real locations, I’m adjusting scale and placement to improve navigation. My approach is to build the environment first, then deconstruct it—tearing down walls,...