21 November 2024 | Deon van Heerden
Hi there! I’m Deon van Heerden, a freelance composer and sound designer from Cape Town, South Africa. Over the past decade I’ve scored around 30 video games, countless trailers and ads and, recently, my first animated feature. I run this small company called PolyForge Media. Formed in 2020 in collaboration with my writing partner, Kristi Boonzaaier, PolyForge Media is a small group of trusted professionals working in game, film and documentary audio.
The Hidden Art: Why Do We Know So Little About Game Music Creation?
Every month, hundreds of games are released on Steam. The overwhelming majority of these ship with their own soundtrack, many of which are created uniquely for the games they accompany.
Game music has become a legitimate genre in its own right, with everything from cover bands to live orchestral programmes proliferating around the world.
Despite all this, though, the process behind the creation of video game soundtracks remains almost totally opaque.
Why don’t we know how soundtracks are made? Well, I have some guesses. Firstly, soundtrack production has been romanticized, and this idea is a potent marketing tool. We’re all familiar with the image:
The scoring stage, the earnest musicians, the conductor, the control room. All serve to lend a game an air of “Hollywood prestige.”
But, as emotionally engaging and occasionally transcendent as music can be, its actual creation is often fairly technical and decidedly unsexy.
The kind of recording I mentioned is increasingly rare, especially outside of the AAA space, and happens towards the end of the production process. The more common reality for most games simply isn’t as appealing. Bigger studios who do have access to visually pleasing productions and the means of documenting their processes often have a greater interest in producing shorter fluff pieces and sizzle reels than they do in educating indie studios and prospective composers.
This is also why, in many cases, it’s difficult to find practical advice and insights online. YouTube skews towards the flashy and the controversial, not the practical.
Furthermore, music production at an indie scale tends to be an insular process, undertaken by small teams under considerable time pressure. After all, in many cases, music is the final link in the chain. This leaves little room for documenting and publishing processes and best practices.
Fortunately, the tide does seem to be turning, with Steam-adjacent events like the Save and Sound festival gaining increased popularity, and giving composers a platform to start sharing their stories. But there are still many aspects of the process left largely undocumented.
What this means is that any indie studio commissioning their first soundtrack - or young composers booking their first gig - have access to precious little information on best practices.
How do you pick a composer, and agree on a rate structure?
How do you choose a direction and scope the soundtrack?
How do you develop a shared language, iterate and ultimately create something that fits your game like a glove?
Today, I hope to draw back the curtain a little for you, to show the nitty gritty of how a game soundtrack is conceptualized and produced, giving some of the advice I wish I’d had when starting out. I believe that having access to these processes can help empower indie studios, as well as composers tackling their first commissions. Throughout, I’ll be using the example of my work on Warhammer 40,000: Shootas, Blood and Teef, developed by the excellent independent Belgian studio, Rogueside.
Prelude: Before You Hire a Composer
Before you even start hunting for a composer, it’s a good idea, as a team, to get at least a sense of the kind of soundtrack you’d like your game to have. Other games in the genre are, of course, good starting points.
But I propose that the question shouldn’t necessarily be “what music does Game X have,” but “why does Game X have the music it does?” How does its music highlight its themes, how does it engage the player? As with art and mechanics, context is crucial to music. A soundtrack says a lot about a game’s identity. Asking this question from the get-go, will allow your team to think beyond the obvious and generic, and to consider music that will highlight your game’s unique identity, distinguishing it from the competition.
Remember:
Diablo's music makes players go "woah!" But music that sounds like Diablo's music in other games makes players go "this sounds like Diablo." Be the Diablo in your genre.
Nothing you decide here needs to be written in stone. But having some sort of internal consensus on a possible direction will make the process of sourcing a composer, and your initial chats, far easier.
Rogueside knew what they wanted the music to achieve and they knew roughly what style could achieve that. This also made it easier to identify someone who was likely to be able to achieve that vision. It would have been easy for a Warhammer game to default to dark, gothic orchestral music. Rather, they knew what they wanted their players to feel, identified points of reference outside of games, and remained focused on that vision.
The Process Begins: Hiring a Composer
This is always a challenge, and I don’t think I can be entirely objective...
I warned you.
You could hire someone based on their existing body of work, especially if it aligns with your vision for the game. Or, if you’re unsure of what you need, you could be hiring someone who you know to be experienced and flexible to help guide the development of your soundtrack’s identity from scratch. When in doubt, or trying to decide between multiple options, schedule some Zoom calls and commission a paid test from your prospective contractors. Provide a brief as best you can, based on your initial internal discussions.
During this process, there are a few things to keep in mind:
Are they able to accurately fulfill the brief?
How are their communication skills, both during the meeting, over email and when discussing revisions?
Is their personality compatible with your team?
Crucially, are they able (and willing!) to implement your feedback accurately?
What does your gut say?
These are some crucial points in determining whether they’ll be a good long-term fit for your team.
For Shootas, Blood and Teef, I was fortunate in that my existing body of work aligned broadly with what they wanted. There wasn’t any friction during our initial calls, and it was clear that we were a match, both professionally and personally. This is the best-case scenario, of course, and it can sometimes take some back-and-forth before a studio decides a composer is right for their project. Again, trust your gut, and don’t rush the process unless you have to.
Pricing and Payment Structure
This is a delicate topic, and one that’s rarely discussed candidly, much to the detriment of pretty much everyone. So here are some guidelines.
Make your life easier by deciding in advance what your budget’s limits are. Be up front and don’t try to fleece your composer.
A relationship of honesty and mutual trust will serve you far better than a few hundred dollars saved.
But how to structure the payments? You generally have four options here:
Per Minute, Project Fee, Per Track, or Hours Worked
Paying per minute of approved music is a popular approach, but one that’s never sat right with me, especially in the context of indie games. It lends itself to awkward conversations. If I believe that the music loops are too short, but accounting wants justification for spending more money on extending them, I’m wasting time that could be spent improving the music. We could have an idea for excellent credits music, but it wasn’t budgeted for. In some cases, music may be culled from one part of the game to finance music in another part. In indie games, where everyone on the project has a vested interest in ensuring their piece of the puzzle is as excellent and comprehensive as it can be, I feel that we can do better.
Deciding on a project fee is my preferred approach, based on the client’s available budget, generally divided into broad milestone payments. This means that I can grow the soundtrack alongside the game, ensuring it has everything it needs. While it requires a measure of mutual trust and respect, and so may not be appropriate in all contexts, it has always served me well.
Billing per track is a fairly safe middle ground between these two. If you’re bringing on a composer late in the process, and have temp music in place to give you an accurate idea of how many tracks you’ll need, this is a good bet in many instances.
Hours worked. This is my least favorite. I could finish your track in 3 hours, or I could finish your track in 20 hours, and you’d be unlikely to know the difference. To me, the value of my music is in what it contributes to your game, not how long it took me to produce.
If your budget is constrained, you could consider things like profit share, the granting of certain rights and sharing soundtrack sale revenues to sweeten the pot for your prospective composer.
On Shootas, Blood and Teef, the team had done their prep work and knew what their music budget was. They were upfront about what they could afford, and a project fee was quickly agreed on. What that meant was that, as the game developed, I could identify areas that could do with unique music, extend tracks as I felt appropriate and generally have a blast without ever having to justify my additions (aside from having them approved creatively, of course). I wanted to be ambitious and over-engineer things, and I was free to do so. And when, later in the project, they requested promotional music that was clearly outside the scope of the project, we agreed on a fair, proportional additional fee without fuss.
No haggling, no penny pinching, just clear, open communication.
I’m aware that any and all of these measures require some degree of mutual trust and respect so, if in doubt, put mutually agreed minimums and maximums in place to protect everyone involved.
Now, with the business aspects taken care of, we can start focusing on what’s really important: the music, and its relationship to the game.
Taking Stock
After an initial discussion, this was my impression of the game’s musical needs and limitations:
The soundtrack was likely to be expansive, including cut-scenes, but was to be produced within a fairly strict time frame.
No audio middle-ware. The music would be implemented directly in Unity. This immediately told me that we were looking at a functional, but uncomplicated, implementation.
The music would be dense heavy metal, a production-intensive style, which could present scheduling challenges. Therefore establishing scope and timelines was crucial.
It was a very ambitious project with little room for error due to it tying in to an existing IP with an established aesthetic and vast lore. The music would need to slot seamlessly into this.
However, all of the above seemed totally manageable, because the team already had a clear vision in place, and communication was clear and proactive.
We were on the right track.
Finding a Shared Language
Music is inherently a subjective thing, and therefore difficult to discuss.
Furthermore, musical terminology tends to be nuanced and widely misunderstood. Clients often feel a degree of pressure to communicate with composers “in their own language.” This frequently results in frustration, misunderstandings and confusion. But there is a better way. Any major musical undertaking requires the client and the composer to find a shared musical language, and musical landmarks that can guide discussions and revisions.
The way I like to start is with an exchange and discussion of reference music. What you hear as "heavy", "epic" or "emotional" is very different from what someone else hears. It’s a composer’s responsibility to pay close attention to their client’s likes and dislikes, to the terms they use to describe certain aspects of music and to be able to quickly and accurately translate a client’s descriptions into actionable musical directions for themselves.
Having a cache of references for each party to refer to to illustrate certain points removes the need for specific terminology.
Ideally, as mentioned, this task will be spearheaded by whoever on the team is the most enthusiastic about music, and in a position to help determine the direction of the soundtrack.
So a shared musical language, based on references, forms the bedrock of your future communications, but how do you then go about briefing a composer on specific pieces of music? "I want the music to be powerful!" and "I want the music to be unexpected!" are noble sentiments, but essentially useless. Similarly, “I want the music to be orchestral” or “I want the music to be super heavy” leave a lot to be desired.
The most useful question you can ask yourself, as a developer briefing a composer, is:
“How do I want the music to make the player feel here?”
Speaking in terms of emotions and narrative impact gives your composer your preferred outcome, without you needing to provide the details of how it should be achieved. “I want the player to feel relieved”, "I want the player to feel unsettled", "I want the player to feel like this boss fight is unwinnable, the enemy is just too powerful" are good starting points. Cross-referencing these desired outcomes with your reference list should help you hone in on the right approach, within the right genre.
And if your composer is worth their salt, having a genre and emotional outcome locked in is really all the initial input that should be necessary. Then, if any further changes are required, you have your store of references to refer back to.
So what did this process look like in practice on Shootas, Blood and Teef?
We mapped real-world music genres and bands onto the lore of Warhammer. Here we had the advantage of an excellent music director, Baptiste, who was also well-versed in Warhammer lore, and a shared background in the desired genre.
Orks are anarchic and wild, the Astra Miliratum are imperial, grandiose and masculine, the Genestealers are alien and cunning.
For each, we identified the subgenre of metal most appropriate to them. Power metal, thrash, doom and sludge all came to exist side-by-side. The Orkish comedy was then offset by self-serious orchestral pomp for the cut-scenes. We had our vision.
Committees are where art goes to die
Some years ago, I was contracted to score a game where the brief called for a very specific style of EDM. My music supervisor and I settled on references and I wrote an initial batch of music. Confident that we’d fulfilled the brief, the music went to the team, only to be rejected. “Why?” we asked. Because one of the leads hates EDM, came the reply.
Needless to say, it was a singularly unpleasant project. "Can we make it a little heavier," “can we get some guitars in there” and “can we make it less electronic” may be legitimate notes in another context, but tied to a game that's 100% aimed at a specific demographic, with an aesthetic built around a specific style of music, it's a catastrophe. The more you try to compromise, squeeze in a little bit of everything everyone likes, the worse your product will generally become.
A whole team’s worth of contradictory notes can be disorientating for a composer and, ultimately, it's not about individual preferences, but about what is best for the game.
Compare this to my experience working with Rogueside.
I had one point of contact who was familiar with the IP, and with the kind of music needed for the soundtrack. They were a good communicator, and were empowered by the team to present their shared vision and steer the musical ship. I had clear goalposts, could get immediate, meaningful feedback and, once the first batch of music had been approved by Games Workshop without notes, my focus became "how excellent can I make this" rather than "how can I tweak this enough to sneak it past Jerry in accounting?"
Scoping the soundtrack
Every video game needs to balance ambition with pragmatic reality.
That's not to say a game made on a smaller budget or during a shorter time frame will somehow be inferior, but understanding scope is key to making a game where year-long extensions to development time are simply not on the table.
Determining the scope of your soundtrack is difficult, but making an early attempt is essential. From the get-go, I knew roughly how many levels we’d have to deal with, I knew that we’d have a few - or possibly many - cut-scenes, and I knew the rough length of the levels. I also knew that we’d have several boss fights. This was all very useful. But I had one concern: the soundtrack’s genre.
High energy soundtracks are a tricky business, and this one would be a veritable rocket ship. If the soundtrack were simply to blast away for hours on end - during combat and exploration - we’d leave the listener fatigued and irritated. What works for a cathartic gunfight doesn't work when you're lost and trying to find the door out of a level. But committing to a complex music scripting system was clearly outside the scope of the project. Similarly, a system with distinct exploration and combat music wouldn’t work. If the same combat track played every time combat started, the player would hear the opening bars of that track dozens, or hundreds, of times in a playthrough.
We therefore opted for the same approach I’d used when scoring Broforce, a two-layered multi-intensity system.
Multi-intensity
Multi-intensity soundtracks, while compositionally challenging when done right, provide incredible bang-for-your-buck on a technical level. Simply put, two tracks play simultaneously. They’re two interpretations of the same track, one high intensity, one low. When the game state changes - i.e. when enemies are alerted, or when all enemies are dead - they crossfade, allowing the music to seamlessly become heavier or more laid back as needed.
The technical challenge was that, whereas Broforce’s level music had mostly been percussive, here we had to contend with tonal elements - riffs, lead guitars and orchestrations. I found, in time, that the key was in the basslines. Reinterpreting a 16th note bassline to give it an 8th or quarter note feel, while matching its contour and tempo, allowed the player’s ear to keep track of the tonal movements within the piece, maintaining a sense of coherence and continuity as we crossfaded between intensities.
Of course, with a little audio middleware, even more can be done. Tempo-synced intros, outros and stingers can be incorporated. But again,
a pragmatic assessment of our time, budget, developer manpower and needs let us settle on an approach early on.
A quick word on feedback and iteration
It’s a fact of life that you listen to music that you’ve commissioned differently to music that you encounter in games and films you aren’t involved with. Being critical is good - and this isn’t to diminish the importance of giving feedback - but there’s an irrationality that can occur, where
an idealized idea of other music negatively colours your perception of the track you’ve been given.
The snare needs to be bigger, the kick needs to be bigger, the guitars need to be bigger, the bass needs to be bigger…
So how do we deal with this? With a simple trick I’ve been using for years.
Spend a little time listening to music you know well, but pretend it’s yours. Convince yourself that this is the music that’s been submitted for your game. Very soon, you’ll start second guessing aspects of it. Actually, the guitars are a little bright aren’t they? The orchestra’s too far in the background. The drums sound fake. When we look for fault, we find it, especially under stress.
Now, pretend the music your composer has submitted isn’t yours. Pretend it’s just something that came on the radio. Pretend it’s from another game. This is a powerful hack for perspective. And always be sure to test in context: put the music in the game!
Music Production
For the demoing phase, speed is of the essence. Programmed drums and bass and software synths accompanied the guitar riffs. Any missteps could be quickly corrected. But once the demoing phase is complete, there are several things that can be done to ensure the final product shines, without breaking the bank.
Live drums are an essential part of my soundtracks. Programmed drums are replaced with real drums, and programmed percussion is augmented with real performances. The cost to quality ratio is extremely good.
Similarly, layering in live strings is a common trick used to bring life and realism to orchestral mockups. A string quartet, or simply a cellist recording a few layers, is enough to give the character of a live performance.
A proper mix is critical. There are very few composers who score and mix equally well. As a composer, a mix engineer’s services should be factored into your budget. I’m fortunate enough to have been working with the same mix engineer, Jo Ellis, for close to 15 years. With a great mix engineer in your corner, you can have the confidence that your score is going to sound its best, both in the game and on Spotify.
Music as a marketing tool
One of the most puzzling omissions in the game industry, from my perspective, is the lack of focus on audio branding.
Music is the quickest and easiest way to develop an emotional connection with your audience.
The ultimate expression of this is, for my money, the theme song. Something that resonates thematically with the game, that captures its spirit.
The Broforce theme song had been a hit, and the promotional track The Ballad of Rambro, won a Golden Trailer award. It therefore made sense that we agreed to create a theme song for Shootas, Blood and Teef called Six Shootas.
The track captured the punk energy of the Orks, and was sung by death metal vocalist Adri Jordaan in a cockney accent. The lyrics not only told the story of the game from the perspective of the protagonist, but included countless references and in-jokes for the Warhammer faithful. And, importantly, the chorus contained the game’s name. The track, and some of the music from the soundtrack, formed the basis of the game’s marketing and trailers, creating a coherent identity for prospective players to latch onto.
By contrast, the approach many publishers and trailer houses seem to take is to license an unrelated, inoffensively generic piece of music from a music licensing site, and to build a trailer around that. That’s a catastrophic own-goal, if you ask me.
So there we have it, from before the beginning to after the end.
In summary
Preparation
Before reaching out to a composer, take the time to consider the kind of soundtrack you're looking for. What is the range of emotions you're trying to evoke? Does the composer have a demonstrated range of musical output, with consistent quality? If you're considering more than one composer, don't be afraid to do a paid test. Also be sure to remember that, while the initial response to the brief is important, iteration based on feedback is just as important, if not more so. Have a good idea of what you can pay for music, and what rights or incentives you can offer if your budget is constrained.
Developing a shared language
Exchange references, discuss them, and let your composer develop a feel for how you respond to music. Discuss music in terms of outcomes, emotions and references. Don't feel pressured to use musical terminology you're not comfortable with.
Scope realistically and pragmatically, leave room for the unexpected
Does your game need a complicated music implementation? Do you have the developer manpower and time to achieve this? How can you maximize the resources you have at your disposal?
Iteration and Feedback
Maintain perspective, and test music in context, remembering that this is a collaborative effort.
Respect music as a marketing tool
Music is the quickest and easiest way to develop an emotional connection with your audience. Use your game’s music - or music based on it - in your marketing!
Still reading? That's impressive. Here's a little gem for your trouble: the Making of the Soundtrack for Warhammer 40,000: Shootas, Blood and Teef.
Enjoy!
