A Brief History of Harmonix (1995-2007)

No Caption Provided

"We had this high-level directive that the founders, Alex and Eran, had concocted to which we’ve remained true to this day, and that was to use technology to allow non-musicians to experience music. We believe music is this incredible human joy which is denied to most people because they don’t have a decade or more to put into mastering a conventional instrument. So we wanted to use technology to cut out the huge learning curve and plug people right in to the awesome experience". - Greg LoPiccolo, Project Lead on Rock Band.

From 2007 to 2010 we saw rhythm game studios release twenty different band-based experiences for home consoles across five franchises. The plastic instrument genre of the late 00s became so ubiquitous that it's hard to imagine a time when producing one could have been a risk. However, when Harmonix developed the original Rock Band, they were putting their necks on the line and overcoming tragically bad luck during their stint in the industry so far. By the admission of the creators of the company, it was failing for its first ten years of existence, and for about half of that, they weren't posting a profit.

Harmonix Music Systems grew out of such idyllic and quirky circumstances, you'd have thought someone had made it up. Its two founders were MIT graduates who had experimented with accessible song-creation software within the institute. Eran Egozy was a computer engineer who played music on the side while Alex Rigopulos was a music major who dabbled in programming. They set up their company with the goal of making musical performance approachable to the masses, which they first tried to do with a product called The Axe, an instrumental improvisation program which had poor longevity as a creative platform and only sold about 300 copies. But they did go on to receive a glowing reception for a motion control music experience they developed for a Disney theme park. Their starry-eyed hope on the back of this was to do the same thing for learning and recreational spaces across America, like arcades and museums. Then they worked out the eternity it takes to establish a deal with a company like Dave & Busters and realised they, once again, had to reinvent themselves. Egozy and Rigopulos then tried to apply their compositional software to the karaoke industry but found no one willing to buy.

It's at this point that you see a fundamental shift in Harmonix's creative approach. All of Egozy and Rigopulos's software up to this point had involved putting a blank sheet in front of the player and asking them to create the sound, but you're threading a tiny needle by developing a product that lets players do this in real-time. Make the tool highly manipulable with many options for the player, and you will find that it requires some learning and that not everything a user can create with it will sound pleasant; this is why we must practice instruments to get consonant, structured sound. However, if you simplify the tool so much that you eliminate the learning curve, you've probably left your audience in a very claustrophobic creative space without the freedom to stretch and pitch sound however they want. From the founders' description, we can safely assume the latter is what happened to The Axe.

Even if a company did manage to pass down a magical Yamaha that let us sound out any tune in our head with the push of a few keys, only so many of us are going to have music worth recording. I'm not writing a screed against accessible music creation software and instruments; in my mind, there's no doubt they should exist, but in general, the public gravitates towards other peoples' music instead of writing their own. For most of us, composing music would be a frustrating job that would create, at best, listenable songs, but other people have already created beautiful symphonies and sensational pop that we can pull up with trivial effort. It's also the case that once music is widely available, we can congregate around it as a culture. Everyone knows The Rolling Stones' Gimme Shelter, but very few people know that guitar riff you came up with one Sunday in your garden shed. Pop culture follows the path of least resistance.

The balancing act of making your music creation suite low-entry but manipulable, and the public's predilection for popular music over amateur compositions were likely stumbling blocks for Harmonix in their infancy. This is why their karaoke project ran aground: The hobby was never about writing new tracks; it is about imitating ones that already exist. Rigopulos and Egozy were trying to give people a mixing desk when all they wanted was a microphone. This revelation reshaped Harmonix, and they would make almost every one of their future creations about the user responding to existing conceptions of how a piece of music should sound instead of providing their own suggestions on its content. And this is the mentality we usually associate with video games: The player doesn't decide which inputs are "correct"; instead, a developer sets out an objective scheme for the correct player actions and the player tries to determine what the scheme is and to conform to it. Once you have a set of rules for which moves are correct or incorrect, the player may then be scored, rewarded, and punished based on how close they got to the desired player behaviour. Rhythm games tend to do this with transparency and rigidity. There is a visual representation of the correct sequence of inputs, and we must match that sequence as closely as possible. So, that's the example that Harmonix follows.

Before embarking on their next project, the studio plucked some promising new employees from Boston's thriving game dev scene. This included various former devs of Looking Glass Studios, the minds behind Thief and System Shock. One of their more notable acquisitions from Looking Glass was Greg LoPiccolo who would go on to become project leader on Guitar Hero. With this team, Harmonix developed Frequency, a rhythm game in which notes flow down a track towards the screen, and the player must hit corresponding buttons in time to play those notes. Publishers were broadly uninterested in the title, but after calling the front switchboard of Sony Computer Entertainment America, Harmonix managed to get put directly through to Shuhei Yoshida, then head of studios. He patiently heard out their pitch, and in late 2001, Sony published Frequency for the PlayStation 2. Frequency is about as straight up-and-down as rhythm games get, but you might ask why an ailing Harmonix didn't turn to the genre sooner. You have to understand that when Rigopulos and Egozy founded the company, there was no music game format to speak of. Rigopulos explains to GameCritics.com:

"Something very significant happened around 1997. Music gaming, which previously didn't exist, came out of nowhere and exploded in Japan, starting with Parappa the Rapper, created by Matsura and published by Sony for the Playstation One. It was followed shortly after by Konami's Beatmania and Dance Dance Revolution games. The category went from not existing to something gigantic and mass-market in Japan almost overnight, and has sustained itself for more than five years as a major entertainment category over there. When we saw that happen, it really struck us that videogaming was the mass-market interactive medium, and it was the medium through which we wanted to achieve our mission of bringing the music-making experience to people who are non-musicians".

Under this philosophy, a Harmonix project turned a profit for the first time in six years, but Frequency still dwelled in the shadows of obscurity. Rigopulos put this down to the game being ugly, intimidating, and having too sheer a difficulty curve, so, Harmonix tried to filter these issues out of the mix with a sequel, Amplitude, which they released in 2003. While we wouldn't label either of these games subversive by modern standards, the studio made a subtle advancement through Frequency and Amplitude's UIs. Most early rhythm games had notes scrolling across the screen on the x or y-axis to eventually meet the area in which the player could hit them, see: Parappa, Dance Dance Revolution, and BeatMania, but Frequency and Amplitude had notes crawling along the z-axis. Their interfaces were that of the 1990 block puzzle game Klax but applied to a music system: targets were spawning in the distance and moving towards the player instead of spawning at the top of the frame and moving towards the bottom, spawning on the right and moving towards the left, or doing the inverse of either.

In the traditional rhythm game UIs, all prompts have equal screen space dedicated to them, but under Harmonix's UIs, the notes closest to the player are afforded the most visual real estate, making them easy to notice and allowing the player to judge the distance between, and to them, at a glance. At the same time, notes that are still a second or two away exist closer to the horizon, taking up less screen space so that the player knows roughly the pattern they'll have to play but aren't reading fine cues for notes which aren't immediately relevant. There's also something about the gems flying right at you which makes you feel a part of the action. Harmonix dubbed this presentation element "the note highway", and it would later facilitate the creation of their most popular games, but they were still missing a hook for the experience. The note highway is an innovative, helpful projector for desired inputs, but in Frequency and Amplitude, what you do once those inputs reach you isn't all that compelling: it's a mundane button tap. These experiences also lacked the popular music that audiences tended to rally around and were filled out with niche artists like Freezepop and Meat Beat Manifesto. In focus testing, players gave Amplitude the lowest "intent to play" rating Sony had ever seen, but after demoing it, they gave it the highest "intent to play" rating the publisher had recorded. In short, once people played the game, they were ravenous for another helping, but they didn't know they wanted to play it until they'd already done so; it was a marketer's nightmare.

Harmonix's next game would be a more direct attempt to bring interactive Japanese music media to the west and would provide tracks that people would fondly recognise on the back of a box. Getting into the music software game early had paid off: When Konami was looking for a studio to develop a karaoke video game, they picked Harmonix, as it was the only saloon in town as far as western rhythm game developers went. Harmonix would get another stab at that karaoke project, and this time, they'd bring their newfound philosophy to the task, using the software to help players imitate well-known vocalists instead of becoming composers themselves. Unlike in Frequency and Amplitude, the soundtrack for this karaoke title was not predominantly underground musicians but was graced by chart names like Michael Jackson, Aretha Franklin, and Van Halen. Although, yes, Freezepop was still there.

Karaoke Revolution, released in North America in 2003, was a money maker, and the developer would revise and retool it countless times over the following years. Other companies like Sony and Microsoft picked up and ran with the concept in titles like SingStar and Lips. SingStar arguably got one over on Karaoke Revolution, featuring some master tracks and music videos for songs while Karaoke Revolution stuck with covers of songs and virtual performers. But Harmonix's sing-a-long experience marks the first time that they lived up to their mission statement of sharing music beyond the world of musicians. They had people who never would have sung vocals or weren't the typical audience for video games picking up a microphone and belting out the classics. It's the scope of the audience and the attraction of a non-traditional demographic that are the achievements here.

As Harmonix had gotten their foot in the door with Sony, the publisher then assigned them their first explosive commercial success, but one that would complicate their vision enormously. In 2004, Harmonix and Sony released EyeToy: Antigrav, a hoverboard action game that made pioneering use of motion control but which critics regarded as little more than a novelty. In the founders' minds, this had dire implications. For almost a decade, they'd been trying to bring interactive music experiences to people, and every time they'd failed or achieved only modest progress. The one time they strayed from their company mission and made a piece of non-musical software, it had turned out to be a cash cow. So, what now? The evidence hinted that continuing to navigate by the stars of the rhythm game industry would result in, at best, barely commercially viable games that would never reach that wide an audience. But down the other path lay games that would betray their core mission and bore critics. Quoting Rigopulos in CNN Money:

"We got really gloomy. We started to wonder if everything we were trying to do was just a fool's errand. When it came to making music games, we couldn't make any money or even a return for our investors. We were paying the bills, but to go on with the business would have been a departure from the founding premise of the company. It would have been an emotionally and psychologically crushing defeat".

But their darkest night came before their brightest day. While Konami had no idea they'd done it, they had set off a chain reaction which helped Harmonix become a godly force in western rhythm games. Konami had developed and published Dance Dance Revolution, and North Californian peripheral manufacturer RedOctane got their start developing third-party dance mats for it. However, RedOctane eventually realised that their operation could be wiped out in one fell swoop if Konami stopped selling in North America, and so, they entered the publishing business.

Harmonix often gets the credit for creating Guitar Hero, but it was RedOctane that approached them with the concept of a game aping GuitarFreaks for players in Europe and the states. They wanted to ferry a title in the image of Konami's guitar simulator to the west, the same way they'd helped bring over DDR, but Harmonix was initially sceptical of their pitch. RedOctane was a company with modest resources who were proposing taking on an expensive, high volume manufacturing effort, and the developer wasn't sure people were clamouring for a rock soundtrack in the mid-00s. Remember, one of the reasons Frequency and Amplitude failed to capture mainstream attention was that their music selections were too obscure, and one of the reasons Karaoke Revolution did better was that it was full of pop tracks. So, it's logical that a game where players would be headbanging to Deep Purple and Judas Priest might have felt like a step in the wrong direction in that day and age.

Whatever doubts may have crossed their mind, Harmonix agreed. RedOctane needed someone who knew software; Harmonix needed someone who knew hardware, not to mention a publisher; it was a perfect match. This time Harmonix wouldn't just be escorting Japanese rhythm trends westward, they'd do it while finding an application for the note highway. Guitar Hero provided the exciting input task that Frequency and Amplitude lacked: when a note reached the live zone of the track, the player would hold down a fret button and strum their guitar which was miles more stimulating than just pressing Triangle. Or, looked at from the other end, the note highway took the tactile satisfaction of Konami's rhythm games and paired it with a far more readable board. It also helps that the notes in Guitar Hero were chunkier, and therefore more perceptible, than those in GuitarFreaks.

I'd even say that Guitar Hero was more effective than Karaoke Revolution at realising Harmonix's mission statement. While Karaoke Revolution provided a compelling way for people to interface with the music they loved, everyone had sung along to a song before, even if they couldn't see the pitch visualised in front of them or weren't scored on it. But there wasn't any analogue for imitating the guitar parts of the songs we listened to; not unless you were one of the microscopic circle of westerners who discovered GuitarFreaks. Guitar Hero was a scintillating new way to interface with music, and a powerful demonstration of the wholly original interactions rhythm games enable.

The tactile experiences we have with games are determined in no small part by our controller. With the number of different ways you could design a controller, there are potentially more ways we could talk with video games than we could experience in our lifetime, but for reasons including practical design and cost, we get one controller per console (generally speaking). It's not an approach that we favour in software; successful designers tailor the mechanics of their games to suit the circumstances and imagine if they could do the same thing with the input hardware. Many rhythm games offer the rare chance to play a title where not only do you get bespoke software designed for a specific game experience, but you get bespoke hardware towards the same end. Basic inputs that would be unremarkable in another title become part of the fun in rhythm games, usually through physical exertion and roleplay, as we get to tap a touch-sensitive pad like a dancer or hit a fret button like a guitarist. And where Karaoke Revolution let people pretend that they were hitting the high note as Diana Ross or Madonna, Guitar Hero and its peripherals enable you to live out your dream of firing off face-melting solos as Slash or Jimmy Page.

You'll also notice that conventional controllers have to keep all buttons roughly in reach of your fingers at all times, for the sake of not being a pain to operate, but rhythm game controllers often function on the principle of rudimentary input being challenging. Someone talented enough with the instrument can quickly press all buttons, but learning to perform those inputs with a split-second's notice is half the game. Guitar Hero takes this concept further than GuitarFreaks by giving us five fret buttons where GuitarFreaks used three. You have four fingers on a hand but five buttons, so, on higher difficulties, you must move your hand, or at least a finger, up and down the guitar as new notes come into view. Through this, Guitar Hero emulates a whole activity in guitar playing that GuitarFreaks could not: Changing your fingering as the song progresses.

Harmonix also does wonders for the presentation of the guitar game. GuitarFreaks looks like someone's Winamp plugin while Guitar Hero has been sprayed and stamped with a heavy metal authenticity. It's everywhere from the graffitied note highway to the metal album fonts to the industrial combo meter. And there's a sense of an actual person behind the music. GuitarFreaks showed its note stream next to a window with video in it while Guitar Hero has its note highway fading inward from an animation of our chosen character on stage, thrashing out the song. You're playing a track both on the fretboard where you're inputting the right notes and on stage where you're shredding in front of a baying crowd.

Harmonix developed this experience in just eight months with a puny budget of $1.7 million, but when it released for the PlayStation 2 in 2005, it became a household name overnight. It took the developer a decade and countless failed experiments, but they'd finally ended up on top of the music game pyramid, and they'd done it without having to abandon their mission of placing music in the hands of regular people. Yet, as strongly as we may associate the Harmonix name with Guitar Hero, the company only got in one mainline sequel to the game before the industry noose was back around their neck. In 2006, the year that RedOctane and Harmonix would wow fans with Guitar Hero II, RedOctane was bought up by Activision and the Guitar Hero IP went with them. Meanwhile, MTV Games, an underling of Viacom, purchased Harmonix. Harmonix was now adrift in the industry; their competition was the best game they'd ever been able to create, but now with the megabucks of one of the industry's most prolific publishers behind it. Plus, they would have to compete against that rhythm genre juggernaut without their hardware manufacturer. But wait, because it gets so much worse.

When you listen to the more recent interviews with Harmonix's head honchos, there doesn't seem to have ever been a question that Rock Band would be the next game they'd make. The idea existed long before development started, and they considered making a game that simulates a full band to be the natural evolution of Guitar Hero. You can also see how Rock Band iterates on the formula behind Konami's Bemani series. Bemani was a line of instrument games that included play on vocals, guitar, and drums. Harmonix also wanted to cover all elements of the rock instrument pantheon, but unlike Konami, they were going to do it under the roof of a single game which is a hell of a task to set yourself.

With Guitar Hero, Harmonix only had to worry about the mechanics and note charts for one instrument (or two if you count bass), but Rock Band was going to drastically increase their workload by adding vocals and drums to the party, and Harmonix had never made a drum game. Additionally, they would have to manufacture both of those instruments despite never having made peripherals before, and they would have to make back the cost of all those peripherals at sale. On both the software and hardware end, they also couldn't play it too close to earlier blueprints; this was a new series, and they had to design the game and hardware from the ground up. And while Rock Band was an unknown property that a studio was building from scratch, Guitar Hero had the leg up of being an established brand, and a number of other advantages too.

Releasing in 2007, Rock Band would be competing against Activision and Neversoft's Guitar Hero III: Legends of Rock. Both were the kind of high ticket Autumn items financially dependent on people picking them up for the holiday season, but Guitar Hero III was getting a one month head start on Rock Band. Then there was the pricing. Anyone who had bought a previous Guitar Hero could pick up III for the price of any other AAA game because they already owned the guitar, but Rock Band was a new game that required new controllers and asked for an excruciating £180/$170 for the game/instrument bundle. Even if someone was choosing between buying Guitar Hero or Rock Band for the first time, Guitar Hero would run them about $100 less, require shorter setup, and wouldn't take up half their living room. From Reuters back in 2007:

"Game industry analysts almost uniformly predict that “Guitar Hero” will sell more copies [than Rock Band] as a result of its earlier release date and established franchise as well as its lower price (about $70; “Rock Band” costs $170) and broader availability".

When Harmonix's hardware team in China finally delivered a prototype of the Rock Band guitar to their HQ, the manufacturers were shocked to find that players would be pressing quickly and forcefully on the guitar and had not built its strum bar to endure such rigorous use. Clearly, they needed to make some big changes. Meanwhile, the folks in charge of content at Harmonix had their own problems. They would need to be pickier in their song selection for Rock Band than they were for Guitar Hero because, while you can sight read a song in Guitar Hero, Rock Band has a vocal component where players need to know songs beforehand to be able to sing along with them. As Rock Band was built to be a group experience, the studio also had to pick songs which would work well when performed alongside other players.

Expectations were high for the game after it debuted to much applause at EA's E3 press conference in 2007. It was EA doing the hosting because it would be EA distributing the game and its peripherals. The stage demo notably included a moment in which then EA Sports head Peter Moore, providing the guitar for The Hives' Main Offender, accidentally paused the game. As Rock Band approached its judgement day, EA told Harmonix that the game had more bugs in it than they'd seen in any game in late-stage development; the publisher was adamant that the developer would have to delay it. Additionally, Harmonix's content distribution model wherein players would be able to buy individual songs for their soundtrack was not something that the Xbox or PlayStation online stores could support. In the words of Egozy:

"At any moment in time, we were sort of like this close to catastrophic failure".

To get Rock Band out of the door, Harmonix's developers crunched hard, and their factories in China pumped out three million hardware bundles in the space of two months. Depending on which staff member you listen to, at the height of production, there were either 10,000 or 12,000 labourers building instruments during sixty-hour work weeks.[1] I want to pause for a second here to acknowledge the trying work conditions that the people constructing these devices endured. It's hard enough for American software developers to secure basic worker's rights, but anyone walking a Chinese factory floor has even less chance of being treated compassionately by the western electronics companies which rely on their labour.

Moving back to Harmonix's legacy, I think that most people, if presented with the first couple of acts of this story, wouldn't guess that it ends with Rock Band being one of the most successful and acclaimed western rhythm games ever released. That Harmonix broke beyond their limitations is highly impressive. A combination of improving on Guitar Hero's ideas, a killer licensed soundtrack, and a penchant for unifying players made Rock Band 1 explode in popularity and stand as one of the most influential rhythm games ever released. It probably also helped that Harmonix had hoovered up a lot of employees who were in bands or had been in the past. The development of the game may have been hot with risk, but they probably had to take that risk to have a chance of holding their own against Guitar Hero. Yet, as much as I want to celebrate all the adversity Harmonix overcame to get to this point; it's not hard to simultaneously see their journey as a dispiriting story.

Harmonix had a commendable vision, not just wanting to make games that people would find fun, but going past that and trying to lend people the opportunity to create music and connect them with some of the most talented musicians who've ever lived. But more often than not they found themselves unable to realise that vision. Some of that barrier between Harmonix and their target audience was down to flawed premises in projects such as with The Axe or their karaoke plans, but even when they made the right choices as a developer, they were still often punished for them. Games like Amplitude and Guitar Hero were beaten back primarily by low marketability and competitive acquisitions, respectively, not because of anything that went wrong in the craft of these titles or because players familiar with the games didn't want them. And sure, Harmonix got their big break with Rock Band 1, but you might be surprised by how little long-term security that lent them.

We have this idea that all industries have built-in filtration mechanisms which ensure that the cream rises to the top while the dregs are sifted out. The theory goes that companies put their products and services onto the market, and if there's been no widespread adoption of them by the time they're a few items in, then it's proof that those companies have nothing to offer. They won't make a profit and will have to drop out of the industry, and so, resources are freed up for people who would make something of worth with them. If, on the other hand, the company can develop something consumers desire, then after a few years, they'll start posting a profit and remain part of the economy. But a history like Harmonix's highlights how naive this view is.

Harmonix didn't make a proper game until they'd be open for six years and they didn't make anything all that successful until they'd been in operation for almost a decade. By the time they got around to creating a work of entertainment that fulfilled their vision, they'd been open ten years. Harmonix is the kind of company that most of us would have written off in its early days; labelled as a derelict with no treasures to offer, but that couldn't have been further from the truth. And now I wonder about all the other companies which made a few mediocre to passable games and then went out of business. It feels all too possible that the industry is full of Harmonixes which didn't ever get to reach their full potential. Maybe we're just lucky this one did. Thanks for reading.

Sources

Harmonix Interview by Edge Staff (October 20, 2008), Edge Magazine.

How 'horrendous failure' led to Rock Band by Maggie Overfelt (September 3, 2009), CNN Money.

Interview with Alex Rigopulos – Part 1 by David Stone (March 30, 2004), Gamecritics.com.

PAX East 2014 Keynote - Story Time with Alex Rigopulos by Alex Rigopulos (April 22, 2014), YouTube.

How The Huang Brothers Bootstrapped Guitar Hero To A Billion Dollar Business by Derek Andersen (December 30, 2012), TechCrunch.

Dueling videogames "Guitar Hero," "Rock Band" face off by Antony Bruno (November 4, 2007), Reuters.

The Harmonix Interview: Alex Rigopulos and Eran Egozy by Harmonix Music Systems (Dec 1, 2011), YouTube.

The Harmonix Interview: Greg LoPiccolo by Harmonix Music Systems (Jan 2, 2012), YouTube.

Rock Band: Manufacturing the Hardware by Harmonix Music Systems (Nov 23, 2011), YouTube.

Rock Band Credits by Harmonix Music Systems (2007), Rock Band.

All other sources are linked at relevant points in the article.

Notes

1. Alex Rigopulos puts the number at 10,000, but COO Michael Dornbrook states the number is 12,000.

4 Comments