Postmortem

My early concept for this project was to make  a goal-oriented automata-based sound toy. Specifically, I wanted to expand on my early prototype Gerridae by imposing some explicit goals onto the sound toy system, with the idea that goals would both give the game musical direction and complicate the automata's behaviors. At some point thinking about this concept, however, I got stuck on the idea of a tower-defense game, which took me away from expanding Gerridae as I felt its ML and movement systems would be too difficult to fit into a tower-defense game. Instead I chose to work with a grid-based movement system and simple rules-based automata and ended up with this project.

I also had some process-oriented goals for this project, specifically building up a prototyping kit for rhythm and music games and exploring options for generative music systems in Unity.

What went well?

This was a challenging project in terms of complexity with a lot of moving parts and timing sensitive systems. I think I was successful in getting those systems working at a basic level to a point where I can see interesting behaviors emerging in the intersection of the game space and music generation, and I'm generally happy with the music, visuals, and gameplay so far.

My goals of building up a prototyping kit and finding some good tools for generative music in Unity went very well. As I worked, I was careful to implement things in a portable way and now have some solid scripts for implement rhythm-based behavior that I can easily use in other projects. I also finally tried out Audio Helm, which I had a great time with. I found it to be very easy to use and more performant than I was expecting. I also like that with the exception of making presets, it lives entirely within the Unity project. I'll probably use it in future projects, especially if I learn the sampler and sequencers features. 

What could be improved?

For the game itself, as much as I am happy with everything mentioned above so far, I think it all has room for improvement. I spread myself a bit thin on this project and didn't give quite as much time as I'd like to each part. With another pass, I think  the unit behaviors could be more dynamic, there could be more interaction between units, e.g. collision, the music and sounds could be tightened up and have more variation, the visuals could be more cohesive, there could be more feedback to indicate game events and states, and of course the actual tower-defense systems could be more balanced and varied, which would likely benefit the flow of the music as well.

In terms of process, I was bit lost when making this in that I didn't have a plan on what to work on when. Instead I worked very moment-to-moment, taking whatever feature was most concrete and clear in my mind and implementing it. This was a good way to work quickly, but also led me to leaving important things like the music for end when I had less time. I'm not sure this could have been much different during this early prototyping stage, but going forward now with a concrete prototype to work from, I could definitely benefit from planning ahead more.