Author: Breanna Lemons

Posted on 12/12/2025

I am supposed to be doing AI for this project, but we haven't gotten to the point where it needs to be implemented. Therefore, I volunteered for a task that we haven't been shown how to implement - Audio. I do have some background in music and some basic understanding of mixing but the only game related audio I've done so far was in Unity. There also doesn't seem to be a lot of tutorials or explanations that fit our current project setup, so I've been trying different things to attempt creating a system where functionality can be added later without a lot of work. 

===========================================================

The Problem:

At first, as I was doing research and thinking about how this should be built, I was pretty confident in making a subsystem so that the audio would prevail over level changes... but that proved to be more difficult than I anticipated. I am sure I am close to full implementation but there seems to be a lack of understanding somewhere.

I did start off in code, because I knew how I would like the system to be set up. I used the GameInstanceSubsystem and overrode the initialization and deinitialization functions. 

At first, this was enough, and things were working fine. But as more content and functionality started being put into the project, there started to be issues and crashes. After some more research I created an initialization class to try to mitigate this:

and it did work for a little while. It created a line inside the DefaultGame.ini file just like the tutorial said. Unfortunately, crashes started happening again dealing with garbage data and caching... since I didn't yet understand what the problem was and I needed to get the audio in quickly without crashing or hindering the other developers, I decided to switch tactics until I could figure it out. 

===========================================================

The Solution:

In the interest of time, and since I haven't been able to get ahold of anyone knowledgeable on the subject to ask, I decided to implement a second approach. Instead of the subsystem, I created an actor component that I attached to the game state.

Since the project currently has a game mode for each map, I had to make sure that it was added to both. This resulted in fixing the crashes between levels. 

I was able to create a blueprint to handle the spawning of the audio depending on whether or not the Main Menu was loaded:

  I know this is not ideal for the setup that was originally intended but it does get the music into the game quickly for now and plays the appropriate music on the right map with no crashes. 

For the pause menu, I made the function that would handle both cases on whether the pause music was properly created or not, in lieu of the previous issues. This does not seem to have any issues.

The bugs related to this new system have to do with pause and resume. I am sure this is also due to my limited understanding on types and how unreal is handling things. If I had more time, I could work this out properly. 

Music Pause and Music Resume are also wired to account for any null pointers or missing information, however, there seems to be an issue with the gameplay music properly pausing and resuming...resulting in it restarting instead of resuming as intended. This didn't seem to be an issue before, so more research and testing is needed to figure out what is going on. 

===========================================================

Both techniques are called at the same places in the blueprint, they are just different approaches to the same idea. 

The ultimate goal was to get the music to play when it needed to play, which the alternate method does do. It just isn't built as I would like it to. I have a vision in my head as to the intent but not a clear path of knowledge or the experience to get from point A to point B quickly in the time frame allotted. Audio Design will have to be a focus for future studies and projects to ensure I fully understand how to implement what is intended in different situations. 

The other audio aspects have not been a problem. Meta Sounds, Sound Cues, adding 3D spatial like birds and SFX like footsteps - have all been implemented fairly quickly. It's just the music and management that seems to be more complex than it first appears. 

I believe both issues above may have to do with my understanding of Audio Components and how to manage them. I will need to do more work in this area to figure that out.