Level & Environmental Design Principles | Collaborative Exercise

Postmortem:
This project was a collaborative exercise focusing on both level & environmental design. The main objective of this project was to implement level design principles common within the industry, such as leading lines, shape language, & breadcrumbing. In this postmortem, I will be sharing & going over three things that went right & three things that went wrong throughout the development process.

What Went Right:
1) [Teamwork] ~
First & foremost, regarding what went right, I have to bring up the teamwork. As soon as our team had finished forming, we created a Discord channel for us to communicate & contact each other easily. Oftentimes, during development, the entire team would be in our Discord server’s voice channel collaborating & actively contributing to the project.

2) [Accurately portrayed my level (concerning the initial concept, storyboards, & mood boards created during the planning stage of development)] ~
Another thing that went right was my ability to go from Alpha (mental concept + mood board + initial blockmesh) -> Beta (source control + playtesting + replacing blockmesh assets) -> Gold (polished level + commentary video), all while maintaining (& iterating upon) the integrity of my level’s initial conceptualized overall look & feel.

3) [Optimizing] -
During the final days of collaboration, our team decided the time had come to optimize our project. These optimizations include (but are not limited to) adding Post-Processing Volumes & Lightmass Importance Volumes to our levels (individually), increasing the texture streaming pool, & adjusting the LODs on many of our static meshes & textures.

What Went Wrong:
1) [Assets/Materials not showing for everyone] -
Shortly after the team began migrating assets from Unreal Engine’s marketplace into the project, we noticed that some of each other’s assets & materials weren’t getting added to the source control software we had been using - Perforce. As we came to find out, only some of the assets, materials, & textures were being ‘marked for add’ upon importing them into our project & pushing them to our server repository; for those that weren’t, we had to manually ‘mark for add’ ourselves - that way, all of the assets, materials, & textures we had imported were successfully committed & uploaded to the repository - correcting this issue.

2) [Changes not being seen by everyone] -
This somewhat goes hand-in-hand with what was previously mentioned. Some of the changes to our sections weren’t visible in everyone’s project despite verifying that our changes were successfully submitted to our server depot on Perforce. Sometimes, only the individual who submitted those changes would see all the changes, whereas the rest of us would only see some of the changes. As we realized, this was mainly due to not having all of the assets, materials, & textures we migrated into the project being ‘marked for add’ upon importing. To prevent this issue from continuing, we each had to manually review what we would migrate into the project to ensure everything that required being sent to the server was (actually) going to be sent.

3) [Texture Streaming Pool Over Budget] -
All of our optimizations were relatively straightforward, aside from dealing with our project going over the texture streaming pool budget. When handling this, one would think to just type the appropriate console command within the editor to adjust the budget, thus solving the problem - right? Wrong. The console command only adjusts the texture streaming pool budget while inside the editor & only for that session. As we discovered, to modify this value permanently, one must delve into the DefaultEngine.ini file & change the allotted budget value within that file. Adjusting our texture streaming pool budget using this method enabled us to package our build with the altered budget value intact, finally solving this issue.

In conclusion, this collaborative project had ups & downs, strengthening my level design, environmental design, teamwork, & problem-solving skills. At the end of it all, my teammates & I are proud of what we could create within a slim timeline (25 days).