The musical road from Halo to Starbloom

A former Halo producer went indie to make music games, and found it wasn't quite what he expected.

Jump to

The plates are cleared. The iPads come out.

It's March 2012, the week of the Game Developers Conference. Matthew Burns sits at a long table at an Indian restaurant in the heart of San Francisco. He's surrounded by game critics, writers, bloggers and developers, all animatedly chatting about games they've seen, developers they've met and talks they've attended. If there's ever a time to show off a game, now is that time.

Developers at the table take out their iPads and show off rough builds of their games. Burns watches as they give impromptu tutorials to anyone who will listen. After a while, he takes out his own iPad and places it on the dining table.

"Not all the sound has been put in," he says as he fires up an early build of his music game, Starbloom. Burns doesn't say or do much to sell his game. He's quiet and reserved, unsure if he should tell people how the game should be played. He pushes his iPad across the table, inviting another developer to try it out.

Little planets appear on the iPad screen and begin circling a galaxy. As they hit certain points, bursts of energy emanate from the planets. The player uses their index finger to tap at the explosion before quickly realizing they're meant to trace them as they appear, producing different sounds. As more planets join the galaxy and more explosions erupt, their tracing becomes quicker and a musical score begins to play.

Burns watches intently. He's worked on games like Halo 3, Halo 3: ODST, Halo Reach, Halo 4 and Call of Duty — all of which have been extensively playtested. He's used to receiving feedback on big projects. But this is different. This is his game, his idea. He nervously watches the player and the game, hoping for a response — any response.

"Hey," the player says, conducting a little synth orchestra with his finger. "This is pretty neat!"

The tastiest cheeseburger

Like many game developers, Matthew Burns entered the industry with the intention to make his own things. He began playing the piano at the age of five, and as a child he composed his own music and wrote extensively, even writing his own little HyperCard stacks. He was attracted to games not because they were "fun," but because they sat at the intersection of so many different domains of knowledge — visual art, music, programming, technology — all things he had an interest in, all things he wanted to make and merge and experiment with.

He wanted to make his own games. He spent the next decade making other people's games. For the most part, he was OK with this.

"We're basically making the best cheeseburger you can possibly have."

"I appreciated the chance to work on games like Call of Duty and Halo because they're interesting, I think, from a logistical perspective in terms of how they get put together," Burns says. "But creatively speaking, they aren't what get me out of bed in the morning."

It was a job. It was a job he found interesting enough, a job he enjoyed. But after spending almost 10 years working as a producer on AAA titles and shipping a big-budget shooter almost every year since 2004, that urge to create something of his own began to bubble. As he worked on bigger and bigger projects at Bungie, where he was one of 15 producers managing up to 300 developers, he couldn't help but feel jealous of small indie teams of three or four or five people. They had complete control over their projects. They didn't have to sit in meeting after meeting explaining to higher-ups who was doing what and why. He wanted what they were having.

Matthew Burns
Matthew Burns doing, as he calls it, his "Jonathan Blow impersonation."
Matthew Burns
Burns shows what working from home looks like.

"There's a sense that indie developers can push the limits of the medium in ways that people hadn't thought of before, whereas with Halo, one of the designers at Bungie used to call Halo a cheeseburger — it was his metaphor for it," he says. "Basically it's going to be really good and a lot of people are going to like it, but we're basically making the best cheeseburger you can possibly have. And that's great — there's nothing wrong with a cheeseburger — but you are making a cheeseburger.

"Whereas an indie game developer might be able to experiment with something like, I don't know, some gastropub crazy thing."

In 2009, at the end of production for Halo: Reach, Burns decided to go indie. He handed his resignation letter to Bungie. He was finally going to go full-time making the gastropub-equivalent of video games. He had no idea what he was doing.

Planck — the game that was never released


Burns thought he knew how games were made. He knew production procedures like the back of his hand. If a new weapon needs to be introduced to a game, he can tell you exactly which teams need to be involved, what steps need to be taken and how much it will cost. If a new element of design needs to be implemented, he can tell you how long it will take, what might need to be cut and the pros and cons of its inclusion. He thought this knowledge and experience would help him as an indie developer. It didn't.

"It turned out a lot of what I had learned on a game the size of 150 people is not really applicable to a little indie game," he says. "Saying it hurt me is maybe a little too much, but it didn't help me as much [as you might expect]. I forget who but somebody did a talk at GDC that was like 'Former AAA Devs Go Indie' and the takeaway was everything learned on AAA games was so invaluable when they went indie, and ... I'm not so sure."

Burns says that many of the lessons learned working at a AAA studio were incredibly general and could almost apply to anything, such as believing in your project, learning to not be too invested and knowing when to cut things. But his time as a producer left him with many large gaps in his knowledge. He had no idea how to market a game — all the studios he worked at had multi-million dollar marketing departments that helped grow the games into the cultural phenomena that they were. He knew nothing of interacting with fans, customers or those interested in his game — every studio he'd worked at had dedicated community managers to handle that. Having complete control over his first project also introduced conundrums he'd never had to face while working on Halo or Call of Duty.

Shortly after his departure from Bungie he formed his own indie studio, Shadegrown Games, with four other developers who wanted an outlet for their own ideas. Teaming up with Brenton Woodrow, Justin Kimball, Kyle Murphy and Chris McCarthy, he began working on Planck (named after the physicist Max Planck), an exploratory, shoot-'em-up music game.

They wanted Planck to be an exploratory, free-form game about discovery. They wanted the game to be about the journey and not focus too heavily on goals. They were hoping to get funding, but publishers didn't understand what their game was trying to do. When they entered it into the Independent Games Festival, the judges didn't understand the point of the game. The publishers they spoke to wanted them to change it, make it smaller, add goals, add leaderboards, change the scope. As different publishers tugged in different directions, they realized Planck wasn't going to work in its current form. They needed to create something that they wanted to make but others would also want to play. They decided to reset the project and try to make something smaller.

After a year of experimentation and watching Planck go nowhere, Burns joined 343 Industries as a contractor on Halo 4. At the same time, Shadegrown Games began working on Starbloom.



The power and meaning of music

A little planet appears on the screen. It circles the galaxy at a controlled pace. Every time it hits a bump, a fantastic ring of energy explodes from its center, spreading outward until it disappears off the edge of the screen. Using a finger to trace the ring will eliminate it from the screen. In exchange, the player is rewarded with a sound. The first set of rings play a steady beat. When a new planet is introduced and a new set of rings explodes, tracing them adds an array of synthesized sounds. As more and more planets are added and the explosions come more frequently, players have to choose which rings to trace and which sounds to play. The player becomes the frantic conductor of an intergalactic orchestra.

For Burns, a large part of Starbloom is about performance. It's about creativity, about making things and doing something that makes you feel good.

"Games and music are actually very similar on a really granular level."

"When people talk about games like Rock Band, they talk about the fantasy of being a rock star and I think there's an element of that. There's also an element of just being good at an instrument," he says.

"People who are good at instruments, who just play them, it feels really good to just be good at playing an instrument and knowing that you're making that sound. In fact, you don't even have to be good at playing an instrument to enjoy playing it. A lot of people enjoy playing instruments badly."


That's certainly part of the reason why the two games he has made as an indie developer have been music-oriented. But he believes there's more to it than just merging two of his interests. Games can be merged with anything — name a subject matter or domain and chances are a game has been made about it — but Burns believes that there is something special in the relationship between games and music, something that translates into a powerful experience that people are drawn to.

"Games and music are actually very similar on a really granular level," he says. "I think [Kotaku writer] Kirk Hamilton has written about how improvisation in music can be likened to gameplay because musical improvisation is kind of feeling your way through a system of rules.

"When you improvise when you play music, there are still rules you have to follow in order for whatever you're doing — whether it be your jazz solo or Baroque ornamentation — to make musical sense."

Burns says that game experiences and music can also be broken down into discrete events, and when the two are tied together at that level, it is possible for gameplay events to also be musical events. The two are unified, giving musical meaning to gameplay events and vice versa.

One of the ways this is achieved is through quantizing, which occurs in many music-oriented games like Rez and Shadegrown's Planck and Starbloom. The process of quantization subdivides a space of time into, say, 16 pieces of time. Any event that happens in the game thus has to fall exactly on one of those 16 slices. In most games, if a player shoots at an enemy, that enemy would blow up instantaneously. In a quantized game like Planck, if the player shoots an enemy, the enemy will transition into a different state — it will begin falling away and smoke will appear — but it won't explode until the next beat hits. This then associates events in a game with musical sounds, allowing the developer to construct music through game events.

Starbloom to the future

Burns wants to finish Planck. He thinks he knows what he wants the game to be, and it's OK if it doesn't match the vision of publishers. In fact, Starbloom is, in a way, a response to the feedback the team received on Planck.

"What I would like to get across with Starbloom is that there are leaderboards in this game, but there are always ways to play the game where you are just pursuing the sounds you want to hear — and I would like to get across that that is also a legitimate way to play."

In the fifth level of Starbloom, if players chase after the highest score, then the music they get will also stop abruptly at times. If they don't pay too much attention to the score, they can create much more interesting-sounding music. This isn't made explicit in the game, although the point Burns wants to get across is seeking the highest score does not need to be the ultimate goal in a game.

"That's maybe the secret message of Starbloom," he says. "The feedback we got on Planck was you've got to have leaderboards, the game has to tell you if you're doing it right, and I had originally wanted Planck to be in response to games like Rock Band or even Audiosurf. I like those games, but with those games there's one right way to play the song and they reward you if you play the right way. If you reproduce the song as it already existed before, you win. If you don't follow the directions, you're wrong."

Burns says he wanted Planck to be more improvisational, performative and exploratory. He wanted people to just make sounds because they're cool. He admits that no one is going to pay his team to make a game like that, and a lot of players want someone to tell them what to do, but that's the kind of experience Shadegrown wants to create.

Burns understands that what he and Shadegrown want may not sustain them for quite some time. Every member on the team has a day job. Burns now teaches educational games at the University of Washington. He says that there are ways that the studio could probably become self-sustainable, but he isn't sure it's something the team wants to do. The group has been advised by industry colleagues to look into free-to-play models and monetization strategies, and perhaps this could pave the way for Shadegrown to start itself up as a business and pay the salaries of its developers, but none of the team members feel that this would be the right route to take.

"Shadegrown Games is kind of our outlet for the cool stuff we want to do, and if we start turning that into like, 'Oh now I have to design a free-to-play rewards ladder for virtual currency,' that would be like our day jobs anyway," Burns says.

The studio plans to continue making the kinds of games it wants to make. Maybe things will work out. Maybe they won't.

"I don't know where it's going to take us," he says. "The hope is eventually we might get somewhere with it." Babykayak