Q&A: Making music with former Bungie composer Marty O'Donnell

Composer Marty O'Donnell is perhaps best known in our industry for his work on the iconic score of Bungie's Halo series, though he's been working with the company since helping to produce the music for Myth: The Fallen Lords back in 1997. Before that he worked on commercial jingles for products like Mr. Clean cleansers and Flintstones Vitamins, and in the wake of his departure from Bungie earlier this year he's now sorting out what the third act of his career will look like. I spoke with O'Donnell via phone recently in an effort to learn more about how he works, and to unpack some of the audio production experience locked up inside his brain after so many years in the industry. The nuts and bolts of producing audio for games is often given short shrift in industry circles, and I hope our interview with O'Donnell can help shed some light on how the sound design teams at AAA studios like Bungie work. What follows is an edited transcript of our conversation, which touches on both O'Donnell's personal experiences across more than a decade at Bungie and his recommendations for developers seeking to improve the quality of their sound design work.

You spent years working as a composer and audio director for Bungie. What does that entail, on a day-to-day basis?

O'Donnell: In my mind, as an audio director I'm responsible for everything that comes out of the speakers in a game. Anything that makes noise is my purview, though that doesn’t mean I do it all -- especially at Bungie, I had quite a large team to do a lot of the work and I was essentially a director. I directed a lot of the processes and creation, oversight, and how we’re approaching audio in a game. And I work with the other disciplines too, so I’d work with the art director, the story director, the overall creative director, to fulfill their vision of what audio needs to be in the game. So that’s a pretty general statement, I know, and on a day-to-day basis it can lead me to all sorts of things. In preproduction I may be sitting at a table with a story director and a writer, talking about characters, and as scripts are being written I’ll often do a table read and read the script aloud to the writers in the room so they can actually hear their dialogue. Sometimes by reading it out loud, especially if you aren’t the writer, and the writer’s listening to someone act out the part, they can get a sense of what is and isn’t working. Sometimes syllables that are spoken into the air that look good on paper, don’t sound good when spoken. These sessions also give me a good idea of how to cast the actors to bring the parts to life.

Do you miss spending time in the studio, making music?

I don’t miss it because I still do it. I’m not the kind of composer that wants to sit 8-10 hours a day and do nothing but compose; I enjoy having a specific task, a reason to compose. Sometimes that’s what I do for a week, y’know; I cancel all my meetings and make it clear to

people that I need to compose for a while. But I do it in chunks; especially in the Halo series and the Destiny series, there was so much music that I really didn’t miss out on composing at all. But I really enjoy having my finger in more pies than just the music, because I feel like it makes the overall product better. If I can understand why the music is important in a particular scene or encounter, or what mood is appropriate, I’m able to help make all those decisions in how the player responds to the audio of the game. So I’m not just sitting outside composing music and then "throwing it over the wall" to the developers, where they have someone else implement it; there’s an implementation stage for music and audio that I still really enjoy.

What does that process look like, in your experience?

It’s a staggered process, and I’ve done it differently on different projects. It’s hard for me, personally, to have a sense of what the music needs to be during a game's preproduction phase, when we’re still trying to figure out whether it’s gonna be, you know, a medieval mythic fantasy or a futuristic sci-fi setting. If that hasn’t been decided, it’s pretty hard to start composing appropriate pieces of music! But once we have some characters and a sense of the universe, then some music will start to flow. And just like artists will do concept art, sometimes we create what I like to call concept music. Those pieces can be used in internal meetings to inspire the team, and vice versa; I often get inspired by concept art that I see. So the process of making music for games happens in chunks. Theme from the original Halo To get really specific, I think it’s better to have the game up on its feet and some of the encounters actually working so you can get a sense of its flow and timing. Music is all about flow and timing, and if you can better understand how those elements play out in the game I believe you will write better music. So a lot of those elements come together near the end of a project, which usually happens five or six months prior to a AAA game’s release. See, just like with a movie, you usually wait until you’re in post-production, when editing is happening and scenes are being cut together; that’s when music is usually composed. I think music composition and production, for the majority of big games, happens toward the end of the production process.

So what most often goes wrong for you while scoring a game, and how can other devs avoid it?

I would say it’s just a matter of scale, usually. If you’re working on a team that’s 25-30 people, communication problems are minimized because you know everyone personally. You don’t really think about communication being a systemic problem. Once you start getting projects that last for multiple years with hundreds of collaborators, you realize how easy it is to just accidentally leave someone out of the equation that’s essential. Like, an entire team might not be aware of something it needs to be involved in working on, and you’re not even aware of it. So you’re just blithely proceeding down a path that you and several other teams have all decided is really cool, and weeks later you realize that some other essential team needed to be involved, and so the whole thing isn’t going to work. I’d say that as the teams get bigger, the management of communication between the teams becomes a thing in and of itself that has to be managed. I don’t know if I have an easy answer for how to do that; it’s certainly a production issue, and I work with the producers and managers, and I’ve seen it happen anyway. I’ve been on very small teams, and I’ve been on the team at Bungie, which grew and grew and grew, and the communication issue is one of the hardest problems to solve.

Can you offer any specific examples?

Oh well there can be things like, people on the writing team and the story team and the design team and the audio team all get together and decide a certain kind of dialogue for characters is going to be really, really cool. And you even get some AI engineers who do some programming to help with the dialogue system, and you feel like you have a really great plan, and then weeks later you realize you left out the networking team and none of your stuff will work because it doesn’t play nice with the networking code. You go ‘Oh crap, why didn’t we think of that?’ And often it’s just because you don’t have the experience at that level to realize that the last time you did this, your programmer was doing both the AI and the networking -- it was all in one guy. And we’ve had things in the past where a physics system changes or a particle system changes, and nobody really noticed that the way we implemented sounds for particles broke. That happened late in the process on...I think it was Halo 3? At one point I suddenly realized I wasn’t hearing the shotgun shell casings hitting the floor around me. And I was like, ‘...wait a minute,’ so I went to look into it and I saw the sounds were still properly hooked up, but we had changed the shotgun shell casings to be a kind of particle that just disappeared instantly if you weren’t looking at them. That made the sounds stop, and I had to go back to the team and say ‘guys, wait a minute...that’s not realistic! I know I’m not physically looking at these casings, but mentally they’re still in my head and I want to get the sounds of the casings hitting the floor around me.’ And we were able to fix it, even though it was late in development, but it’s just one of those things where sometimes, as your studio gets larger, the left hand doesn’t know what the right hand is doing.

Sounds like you've spent a good deal of time advocating for the music side of things.

When I say I’m responsible for everything that comes out of the speakers, I mean that I am totally willing to take the blame for what didn’t happen. I’d like to give the credit for successful things to the team that did it, and as the team gets better, sometimes it’s harder to give credit in interviews and have people actually hear what I’m saying: that the team who actually puts this together deserves all the credit for the successful parts. I feel like I should take the blame for the stuff that just doesn’t work out, because it’s my fault for really not ensuring that our vision gets implemented. But if the vision is implemented, there’s a huge, talented group of people that are responsible for that.

Fair enough, but what advice can you offer developers who don’t have Bungie-level scale or resources?

It’s been a long time since I was just starting out, so I would just say that creativity is always going to find a way to bust through limitations. The most important thing to focus on, besides just your basic chops -- education, practice, going to professional conferences -- is to focus on your relationships with other people around the industry. If you have a friend who’s making a film and they need audio or music, offer your services: be part of a small team of friends that makes something. That gives you experience collaborating with other disciplines, and especially now, with the amount of easily-accessible third-party game development software that’s out there, find people who are making indie games using Unity, or UE4, or CryEngine, or whatever it might be, and offer your services. You won’t have a big budget and you shouldn’t sell yourself too cheaply, but the experience of actually producing something is invaluable. And also there are now software solutions that used to not exist: you can build a home studio relatively inexpensively now, and there’s some really good software out there for sequencing and music production. From Pro Tools to Digital Performer, you can get demo versions of these things and even buying the full versions isn’t horribly expensive. And it’s really worth having tools so you can produce something from start to finish that sounds really good. And then as an audio person, you need to learn about microphones and recording sound effects, manipulating sound, that kind of thing. Finally, you need to learn to use tools like Fmod and Wwise, wonderful tools that let you get into audio production at a beginner level and ramp all the way up to the highest levels of game development. If you’re trying to break into the business and you’re unfamiliar with those kinds of products and techniques, you’re going to be beaten out by more competitive audio folks.

How do you feel about musicians like doseone and Paul McCartney making music for games?

I’m thrilled by it; to me, it shows the maturing of the industry. The music side of the game industry is now seen pretty much as legitimate a career path, and legitimate a creative outlet for composers and musicians, as film and television and concert music. We’re starting to be recognized as mature expressions of the art form, and not sort of as the kid brother. It’s great to see people like Hans Zimmer out there composing for games. It was amazing to have Steve Vai and Incubus and Breaking Benjamin and Nile Rodgers working with me ten years ago on Halo 2. That culminated for me personally with working with Paul McCartney, because he wanted to see what this industry is all about: what it’s like to do music for a game. Paul McCartney himself was actually interested in that, and that was just a total thrill for me. Overall, I think it’s a great indicator that music and audio for games has matured to the level of audio production in any other art form.

So where do you think we go from here, as an industry?

I’ve been seeing groups of what I would call ‘veteran indies’ -- people who have been in the industry for a long time -- getting together in smaller groups to make smaller expressions of games as an art form. They’re going to be really high-quality, but maybe not quite as huge in scope as some of their prior projects. We’ll see how successful that is, especially with all these new platforms like mobile and virtual reality. I think VR is going to be a really interesting platform, and I’m excited to see how audio enhances that experience. That’s gonna be a lot of fun.

Do you see yourself as such a ‘veteran indie’?

Well, I have nothing to announce about that, but maybe soon. I think the concept of a veteran indie company is actually very attractive. We’ll see!