How an Audio Team Works with Video Games

Whether with a symphony orchestra or white noise, audio plays a crucial role in your video game experience. Audio cues warn you of impending danger, and voice actors bring the characters to life, and even the ambient sounds you might not pick up keep you immersed in the action as you watch it unfold. Sounds help make the digital world believable, no matter how far away it is from your reality.

We visited David Schlein-Andersen and Grégoire Iwaniec in our bespoke sound studio to learn more about their process of bringing a world to life through sound.

David Schlein-Andersen: I’m a Voiceover Designer at Liquid Swords. VO is everything that has to do with a character making a sound with their vocal cords or vocal folds as they're called nowadays. This usually means lines of dialogue, but they can also extend to grunts, sighs, and other noises that characters make in the game. If you're fighting, there are a bunch of grunts. Running can also extend to their efforts and the noises that they make. I am responsible for the characters' sounds and act as a bridge between the audio and narrative departments.

Grégoire Iwaniec: My name is Grégoire Iwaniec. I am a sound designer at Liquid Swords. My role captures the design, production, and implementation of everything that isn’t voice and music in the game. Everything that makes a sound, all the effects, the ambiance, everything you can interact with in the game. The idea is to record as much source material as possible and make it fit into the game. We also create a set of rules that will cause it to play back in the most natural and believable way while conveying the sonic environment we want the player to feel during the game.

David: There's surprisingly little overlap. Most of the stuff we've been doing together over the past couple of months has been building and planning our studio space in the office and getting our pipelines working because we use many of the same tools. Still, we don't necessarily use them the same way or with the same content.

Greg: We have the same end goal and work in parallel toward the same creative direction using the tools we developed together. It's mostly parallel because, as David mentioned earlier, his main job will be with narrative as well, as the connection between narrative and sound. That's where VO comes in and shines. VO will make our characters and narrative come to life, whereas audio is more of a realm of its own that creates the city's soundscape.

David: Just not directly touching the same content.

Greg: It will remain that way throughout most of the project. However, towards the end, when you mix a project, you want everything to work together, and there’s overlap. But right now, when we're in the early stages of the game and figuring out how our pipelines are supposed to work, I'd say it's a little bit more working together. Then, when we get into the actual production of working with the content, it will be less collaborative because we have our pipelines and our content. It's not until the final stages, when we begin mixing, that it converges more.

Greg: We use Wwise as a middleware, creating all the logic for the sound’s playback and then integrating that over in Unreal Engine 5. That is where the programmers will feed us all the parameters we need to make the sound dynamic. Typically, on a day when I work on cars and vehicles, I spend my time driving around in the different cars in the game, making sure that the wheels' sound reacts correctly to the parameters we feed them in Wwise.

Greg: It depends on what area of sound you focus on and what production stage you’re in. Let's say we split the audio area into vehicles, weapons, and ambiance. You're going to be recording and designing assets that you will use in the game later on, but this comes at a different point of production. You can spend an entire week planning a recording session and then executing that recording session. Editing it will also take a large amount of time. Once you have all of these elements, then you can spend more time fiddling with them.

David: Where sound design touches all the objects in the world that make sounds, voiceover involves the characters, and that usually means you need a person to make those sounds by reading lines. Or AI.

David: Currently, there is a big debate about using AI to voice characters, which is a very interesting issue. I understand and agree to a certain extent that it is, perhaps not worrying, but there are different roads and routes to take. The use of AI could go for better or for worse, and what people are scared of, myself included, is that human beings will become obsolete. That AI would just replace all VO.

There are plenty of really good AI voice-over tools today, but they can only get you so far when it comes to directing. You can't really do the same thing as you can with a human being where you can give them the context and subtext of the scene. “This is the line you’re delivering, but this is what you’re really saying.” We’re wired to pick up on these things as humans, and maybe AI will be able to do that one day, but I also think that by then, our society will treat that practice with respect toward humans.

Greg: The closer you get to the actual emotions that you try to deliver with VO, the better people can picture how the game will turn out. We start with subtitles, and then we generate placeholder voices through text-to-speech AI synthesis, which gives a certain amount of emotional output in order to further iterate on the writing, narrative, and pacing before we start recording people. All of these degrees of improvement make it a lot easier for us to picture how the final product will sound and how it will feel. In that sense, AI is super useful.

David: Right now, we’re using AI as an iterative tool to show very quickly how certain things might feel or sound when we're creating a game. Once we know what we want, we get actual human beings behind microphones.

Greg: We aim for the highest fidelity of sound, which means the highest quality possible. We’re going to record as many versions of the sound we're going to use as possible or use professionals who know how to record specific sounds, such as vehicles. Vehicles need a lot of microphones, and you need to know what you're doing to be able to record them correctly. We work with professionals who specialize in that area to record this type of sound, but mainly, at this stage, we use libraries and record a little bit on our own. Later in the production, we will be recording our own material as much as possible. You can record with intention when you know what you want to achieve. Not all recordings are equal, such as the choice of microphone, the microphone placement, and what material you're going to record. All of that has an impact on the end product. So, if you know what your end goal is, then it's easier to record it correctly.

David: Since we're a small studio without a proper recording booth in our office, we'll be using a lot of outsourcing in order to find and record our voice talent for the game. We're aiming for the highest possible quality in regards to VO as well, so we'll need some outside help in order to be able to record the amount of voices necessary to populate a living open world that we're creating.

Sounds like the first game will have more than one high note! While you wait for it to launch, learn more about what we’re up to on our News page!