3D Workflow for Game Development
November 11th, 2024My journey with the senior capstone project so far has taken the direction of honing my skills in using Blender 3D for game design. The “isometric” perspective lends itself well to our proposed farming game. While we could build a 2.5D title like Stardew Valley, we have decided that our specific game will benefit from a few different advantages of 3D. We are building a farming game centered within a post-apocalyptic world. Thousands of years after the collapse of humanity, the player is tasked with restoring the environment via agriculture. Two key levels of this game will be underground bunkers and the surface world. By building these levels in 3D, the meticulous process of managing transitions and depth within a large and varied environment is avoided. Another deciding factor for 3D development is that multiple teammates have experience with 3D asset creation and processes. Our experience will allow rapid development of a unique aesthetic that will help differentiate this title from other farming titles out there.
The above image is the most recent blockout I made of our v0.0.1 bunker level. While it still needs to be fully textured, shaded, and populated with assets, this provides a great starting point for the development process.
Something I am learning, working with my teammates, is that an initial concept may not convey the final vision very well. We are all eager to rush to the stage of development where concept becomes polished design, and ideas are fully visualized on the screen. It can be hard to meter expectations during early stages of design, so I have opted to finish one asset (the hallway) in order to show what something close to the final product may look like.
After a few stages of back-and-forth, I sought notes on the hallway asset. I found that my teammates are looking for a toon shader aesthetic with hard outlines. Notes also brought out that my teammates prefer contrast between bright and dark tones. This process of design and review is likely a good simulation of being a 3D artist. It is similar to the process of requirements gathering, except the clients are my teammates. Knowing this information moving forward, I have a concrete idea of what is needed for each asset. For hard outlines, sharp corners are favored over beveled geometry. Toon shaders simplify the interaction of materials and put the focus more on color value rather than material properties. I am glad we were able to describe these aspects so early on in development.
My prior experience with 3D modeling in Blender largely involves creating designs for my 3D printing business. 3D printing has a unique set of requirements, such as keeping models manifold, and notably, there is no major cap on polygon count. Game design, on the other hand, heavily relies on efficient topology with low poly counts. Because of this, I have held myself to a higher standard of efficiency when building concept models. Initially, this YouTube series on game asset creation in Blender by LMHPOLY was extremely helpful. It outlines modeling, texturing, and exporting a 3D asset from Blender to Unity while maintaining best practices. Adapting my previous workflow to this new set of standards, I now feel I am experienced enough to create any asset we might need for the game, given enough time.
My next focus will be modeling and rigging 3D characters. In our post-apocalyptic setting, there will be organic life, robots, and cyborgs. Each type of character represents a different combination of modeling techniques. For organic life, I will use subdivision modeling and rigs with weight painting. For mechanical AI, I will use hard surface modeling and rigid rigs via vertex groups. Cyborgs will use a combination of these methods. Below is a block-out of a cyborg Llamma trader that would likely pull carts of goods to be traded with the player.
When creating more detailed 3D art, a concept sketch is invaluable for referencing proportions, etc. Ideally, you can use sketches from multiple perspectives to facilitate the creation of 3D shapes against the reference. (As a side note, I have been working to further my 2D art skills. Inspired by my teammates, I think a more general understanding of art will improve my 3D work)
As with earlier stages of the project, I predict that further creation of concept art will be a great boon to development. With each new design, I learn new methods of simplifying the modeling process or improving the quality of assets for our game. Feedback from my teammates helps us hone the game’s aesthetic and design philosophies. In conjunction, these realizations will ensure smooth development.
Read the post...