The Rise and Rise of Video Game Music

There can be no doubt that interest in video game music has grown considerably in recent years. It is notable that this passion is not reserved only for the latest releases, but extends back to the earliest days of the form, with as much praise heaped upon 1980s Commodore 64 or 1990s Amiga 500 music as on the high-profile symphonic soundtrack du jour. Contemporary game developers, such as Terry Cavanagh, underscore games like vvvvvv (2010) and Super Hexagon (2012) with soundtracks that draw directly on the early home computing and console instrumentation and form. In doing so, they demonstrate their own fandom, as well as a desire to draw on the distinctive aesthetics of early gaming and their association with an era of putatively simple yet unabashedly complex gameplay. We further note exhaustive online archival collections such as the HVSC (High Voltage SID Collection) or VGMrips that gather together music files extracted from original game code. For their part, enthusiastic programmers expend extraordinary effort on the production of software applications such as VGMPlay and Audio Overload, which are solely dedicated to the task of playing back these music files with no care or support for reproduction of graphics or gameplay.

The influence and impact of video game music is felt beyond the community of game fans: contemporary artists such as Tokyo Machine are fully fluent in the language of early video game music, while Grandaddy’s ‘A.M. 180’ (1998) and Beck’s ‘Girl’ (2005) demonstrate a bilingual mix of indie rock guitar instrumentation and 1980s video game consoles. In addition, a litany of tracks directly sample from titles as diverse as David Wise’s 1995 Super Nintendo Donkey Kong Country 2 (on Drake’s 2015 ‘6 God’) and David Whittaker’s 1984 Commodore 64 Lazy Jones (Zombie Nation’s 1999 ‘Kernkraft 400’).

To ease the process of accessing this sonic territory, contemporary synthesizer and sampler manufacturers draw on the distinctive sounds of early video game music in their banks of presets, programs and raw waveform data. Plogue’s Chipsounds software instrument offers painstakingly detailed emulations and recreations of a host of home and arcade-game sound systems, while Korg’s Kamata instrument explicitly references Namco’s 1980s game music, even going as far as integrating presets designed by legendary composer and sound designer Junko Ozawa, as it ‘reconstructs the C30 custom sound engine that swept the world in the 1980s … [and] lets you play the classic video game sounds of the past’.Footnote ¹

To palpably extend the impact of early video game sound into the present, some manufacturers build new machines explicitly influenced by the architectures and sonic fingerprints of specific systems, such as Mutable Instruments’ Edges module that mimics the Nintendo NES, or Noise Engineering’s Ataraxic Translatron that nods to the Atari VCS console. Some go further still in creating instruments that eschew emulation or simulation and instead are built around the physically extracted innards of old home computers and gaming consoles. For instance, Elektron’s SIDStation and ALM Busy Circuits’ SID GUTS repurpose the Commodore 64’s Sound Interface Device (or SID chip) sound chip, while Twisted Electrons’ AY3 module has at its heart two of General Instrument’s AY-3–8912 chips, variants of which provided the sonic foundation of countless computers, consoles and arcade systems, including the Spectrum 128, Mattel Intellivision and Capcom’s 1942 arcade game (1984).

But notwithstanding the enormity and creativity of the labour that continues to see these sounds celebrated and embedded into contemporary electronic music and gaming cultures, there remains little detailed historical work on early video game music. The few popular and scholarly histories of early game music that do exist are frequently based around discursive formulations and conceptualizations of technology that reinforce totalizing teleological narratives rather than add nuance to the picture.

This chapter suggests how we might arrive at a nuanced discussion of the role and function of technology in relation to video game sound and music. In doing this, the chapter will address some of the limitations of extant approaches, and outline some alternative ways of thinking that reconfigure the relationships between hardware, software and, crucially, the creative endeavour of composers, coders and sound designers.

Early Video Game Sound: When or What?

Before we proceed with this analysis, it is essential to address what we mean by ‘early’ game sound and music. Surprisingly, this is a difficult task, with no universally agreed definition. While there are distinctive practices, tools, technologies and techniques that are characteristic of game music making and reproduction throughout the 1970s, 1980s and 1990s, these ultimately continue to be evident, and are in use to this day, albeit in different contexts. For instance, central to the definition of early video game music is the sound chip. Whether they took the form of a TV-connected console or handheld device, a general-purpose home computer or a dedicated arcade cabinet, gaming systems created in the first three decades of the medium’s existence all generated their sound in real time via a sound chip. These specially designed pieces of silicon were effectively synthesizers capable of generating waveforms played back at specific pitches, with timing and durations specified by code created in advance by a composer and programmer. Just as a graphics chip takes data about colour, x-y coordinates and scrolling and processes them in real time to create images on screen, so too the sound chip responds to data in order to generate the game’s aural components.

The references to data are important here, and the methods and working practices for dealing with sound chips are particularly noteworthy. Contemporary video game composers likely expect to work with a Digital Audio Workstation for composition and sound design, might have vast libraries of hardware or virtual instruments and effects and use a plethora of software tools like Wwise to facilitate the integration of sound, music and interactivity. Each of these tools typically offers a graphical user interface, or a physical set of controls in the case of hardware, which are typically assessed in terms of their ‘user friendliness’, as Pinch and Trocco note.Footnote ² Composers working with 1980s sound chips found themselves in an altogether different situation. Crudely speaking, there were no tools save for those that they created themselves.

Commenting on the situation in the early 1980s when he began composing for the Commodore 64, Rob Hubbard observes,

Consumer-facing music-making hardware and software did exist, such as the Commodore Music Maker (1982), but these were too inefficient for use in game development, meaning that working with code remained the only viable solution. The experience of writing for the NES was very similar, as Manami Matsumae, who composed the soundtrack for the first Mega Man game, explains.

As Collins notes, the lack of any readily available or suitable tools, and the necessity to work either directly with machine code or via a code-oriented interface,

Even when such tools were later developed, they still demanded a combination of computing and musical literacy. The ‘Tracker’ to which Hubbard refers, for instance, is a sequencing application that takes its name from Karsten Obarski’s The Ultimate Soundtracker, developed in 1987 at German developer Rainbow Arts. The Tracker arose from the frustrations Obarski felt writing music for the Commodore Amiga, and derived its visual interface from the vertical piano roll.Footnote ⁶ Yet music and sound information was still entered as a series of hexadecimal values rather than as traditional musical notation and the Tracker’s resemblance to a code listing was unavoidable. Of perhaps even greater significance than the interface was Obarski’s implementation of the ‘Module’ or ‘MOD’ file format, which defined a collection of ‘instruments’ in the form of constituent samples, patterns that specified when and how the samples were played, and a list that set out the order in which those patterns would be played.

We will return to the importance of the interface between composer and sound chip later, but here we should note that the programming of sound chips to replay music in real time sits in stark contrast to systems such as the PlayStation 4 and Xbox One, for instance, which stream music from a hard disk, CD or from an online source. In these systems, music is not performed in real time by a sound chip responding to coded instructions but rather is prerecorded and played back in much the same manner as it might be played back on a CD player or portable music player. Accordingly, the scope of the musical creation is not defined by the capabilities of the sound chip, and might make use of any form of instrumentation available to the modern musician, composer, producer or recording engineer.

The utilization of prerecorded music in video games had been greatly precipitated by Sony’s 1995 PlayStation console, which brought the first commercially successful mainstream implementation of audio streaming from a built-in CD drive. In addition to creating and commissioning original music, CD playback brought the possibility of licensing pre-existing tracks. Psygnosis’s Wipeout (1995) led the charge, and serenaded players with the sounds of the Chemical Brothers and Orbital. Beyond this, some games even allowed players to swap out the game CD and its supplied soundtrack and replace it with a disc of their own, as with NanaOnSha’s Vib-Ribbon (1999), in which the game’s landscape was derived from an audio analysis of the CD audio content. The disc swapping possible with games like Namco’s Ridge Racer (1993), and central to Vib-Ribbon, arose because the entirety of the game’s program data could be loaded into the PlayStation’s memory (rendering the disc is no longer necessary), and because the PlayStation console was able to decode standard audio CDs. Indeed, the PlayStation game discs were often supplied as ‘Mixed Mode’ discs, meaning there was one partition with game data and one with CD audio, making it also possible to playback game music in a consumer CD player. It was this adoption of the CD audio format, or what is technically termed the ‘Red Book’ audio standard, that represents perhaps the greatest shift in game music technology. It also serves as a clear pivot point for ‘early’ game music, as prior to this moment, music was generated and created in a markedly different manner, with different working practices and tools and audibly different results.

However, while this might seem like a neat delineation point, the launch of the PlayStation, Sega Saturn and subsequent home, portable and arcade systems employing prerecorded audio streaming did not mark the end of sound-chip-based music making or reproduction. Countless handheld systems such as Nintendo’s Game Boy Advance and DS series as well as their Nintendo 64 home console continued to offer no optical disc streaming and, therefore, relied still on the use of sound chips not merely as an aesthetic choice but as a technical necessity.

If we use the widespread adoption of prerecorded music as a historical watershed, the definition might seem to be based as much around what it isn’t as what it is; how it isn’t produced, and which technologies and creative practices aren’t deployed in its production, distribution and reception. Perhaps more accurately and more positively, however, we can define the era of ‘early’ game music as the period during which the use of real-time sound chips was the only way to create and hear game music, rather than the Mixed Mode discs and mixed economy of forms of streaming and sound chips that we identify from the mid-1990s onwards.

‘Chiptunes’ and the Flattening of Early Video Game Sound

One of the consequences of the increased popularity of early video game music has been its contemporary designation as ‘chiptunes’. As McAlpine notes, the term appeared in the early 1990s, has grown in popularity since, and is now often used for any music created with sound chips (or the sounds of sound chips), whether produced today or thirty years ago.Footnote ⁷ Given the slippery nature of definitions based around temporality, perhaps this chronological ambiguity might be useful. However, even though ‘chiptunes’ has the distinct advantage of being in popular parlance, it brings with it some significant issues. From the perspective of the historian, the term ‘chiptunes’ has an unhelpful flattening effect. Quite simply, there is such a wide variety of ‘chips’ with which video game music is, has been, and can be made, that such a designation is helpful only in distinguishing it from music from disk or online streaming. ‘Chiptunes’ tells us nothing of the differences between the affordances of the Commodore 64, NES or Space Invaders arcade cabinet, or how composers and sound designers might investigate, harness and extend their potentialities.

This flattening effect is particularly troublesome given the tendency of current accounts of early video game music to deploy the language of ‘bleeps’ and ‘blips’ in their often pun-heavy titles (e.g., BBC Radio 4’s While My Guitar Gently Bleeps, 2017).Footnote ⁸ The origin of this construction of early video game music is understandable as many of the very earliest home and arcade sound chips utilized in 1970s systems offered few of the sound-shaping tools such as filters, or fine-grained control over pitch and amplitude, that would become commonplace from the early 1980s. As such, raw waveforms were typical, and particularly square waves which, requiring only binary on–off instructions to generate, are the simplest to create. Certainly, such sounds are inescapably electronic, frequently associated with the sounds of sci-fi control panels burbling, and are unlikely to be heard outside the most experimental and avant-garde academic music-research labs. Indeed, if we take the Atari 2600 (1977) as an example, we find that the sound capabilities of the market-leading home video game console were decidedly challenging.Footnote ⁹ The system’s TIA (Television Interface Adapter) made use of a 5-bit frequency divider, which generated a limited number of mathematically related, but often musically unrelated, pitches. As the opening lines of the Atari 2600 Music And Sound Programming Guide have it, ‘It is difficult to do music on the Atari 2600 due to the limited pitch and two voices … many of the pitch values are not in-tune with others.’Footnote ¹⁰ Were we only to consider examples such as the 1984 Atari 2600 Gyruss soundtrack, we might well be forgiven for building a view of game music as bleeps and bloops (and fairly discordant ones at that). To compensate for the TIA’s esoteric tuning, Garry Kitchen, the developer of Activision’s 1983 Pressure Cooker, ‘determined a set of pitches that the Atari TIA could reliably reproduce. He then hired a professional jingle writer to compose theme music using only those available pitches’.Footnote ¹¹ Similarly, PC-compatibles and home computers like the ZX Spectrum were equipped with just a ‘beeper’ circuit primarily intended for system-alert sound effects. Yet, music and sound design of some sophistication and complexity was coaxed from the Spectrum, which speaks to the ingenuity of composers and programmers, and further complicates the idea of bleeping and blooping.Footnote ¹²

Yet, so pervasive is the idea that early video game music is no more than a series of bleeps, it has become necessary to tackle it head on. Question 3 on the HVSC’s FAQ list is telling: ‘Isn’t Commodore C64 music just silly beep-blop music?’ The answer, unsurprisingly, is an emphatic ‘no!’ Surely the palpable sonic differences between the TIA, ZX Spectrum, NES and Commodore 64 demand a distinction between their ‘bleeps’? In discussing the absence of fine-grained analysis, Altice notes that,

One possible solution entails adopting an approach that focuses on the specificities of the particular sound chips in the Commodore 64, NES and Atari 2600 and conceiving of them not merely as musical instruments or constituents of gaming systems but as distinctive platforms for music and sound design in and of themselves. A ‘platform studies’ approach recognizes the significance and character of the base technological systems and the contribution their underlying design and capabilities make in defining the sound of the platform. As Arcangel et al. put it, ‘Computers have personalities, shapes and architectures like a canvas that influence what we make.’Footnote ¹⁴

Conceptualizing early video game music as ‘chiptunes’ conflates and smooths out differences between platforms, but a game converted from NES to Commodore 64 will sound as immediately and profoundly different as it will look and feel different. The NES’s sound chip is recognizable, characterful and, most importantly, unlike that of the Commodore 64. The platforms contrast sonically in precisely the same way that Nintendo’s joypad is unlike Commodore’s keyboard and the NES’s colour palette and display resolution are unlike the C64’s. Indeed, if we look at the specification of the NES and Commodore 64 sound chips (see Table 1.1), we immediately begin to appreciate how these differences come to be so pronounced and why an approach that recognizes and foregrounds the specificity of design is so essential.

Examining the specification in this way, as in Table 1.1, we begin to reveal not only that there are differences between the capabilities of sound chips but also that the differences can be of great significance. For example, the SID chip offered variable waveform oscillators, Multimode Filter (buggy and unpredictable as it was), Ring Modulation and per-voice ADSR amplitude envelopes unavailable on the NES’s RP2A0X chip. On the other hand, the NES had more voices and a particularly characteristic and ‘lo-fi’ triangle waveform, so it would be unproductive to designate one platform better than another. It makes no sense to discuss these sound chips or sound-generation techniques as being better than one another, any more than it would make sense to have that debate about a guitar, flute or acoustic drum kit. That these instruments are markedly and audibly different is the key to understanding the call to disentangle the chips from the tunes. Examining them in this way, we begin to reveal not only some of the significant ways in which the gaming sound chips differ from one another but also how these differences materially affect the sonic fingerprint of the games created for the host system.

‘Chiptunes’ and the Oversimplification of the Sound of Sound Chips

While it is essential to recognize the distinctive nature of different sound chips and move away from the monolithic nature of chiptunes, it is important that we do not overstate the role of the chip in affecting the sound of a given gaming system. Of course, the SID chip’s oscillators sound different to those of the NES. For that matter, different revisions of the SID chip sound different to one another, with the filters varying depending on the month of manufacture, just as the different iterations of the Game Boy and Mega Drive vary, with one version of the latter being labelled ‘The Stinker’!Footnote ¹⁵ However, although the SID Chip, RP2A0X, YM2612 and Paula all shape the sonic landscape of C64, NES, Mega Drive and Amiga respectively, the chips do not write the tunes. And many of the compositional, stylistic and sound-design techniques most commonly associated with chip music, or even with specific sound chips, are really better understood as interactions and improvisations between person and technology.

By way of example, the ability of the Amiga’s sound chip, Paula, to play back samples, is certainly a distinctive, if not wholly unique feature of the platform. Paula’s ability to replay samples certainly affords a different kind of compositional and sound-design style, but it is only one aspect of the characteristic sound of the Amiga. Those snippets of audio are replayed at a specific sample rate and bit depth, which gives them a characteristic, grainy, subjectively ‘lo-fi’ quality, especially when they are pitched away from the root note. This is also affected by the quality of the computer’s DAC (Digital Audio Converter), which is responsible for outputting the sound in a form capable of being replayed to a TV set, speakers or headphones (indeed, it is here that the audiophile might take exception with Altice’s point about CD players not having a distinct sound). Each step of the audio pathway imparts its own distortions and colouration, some pleasing, some intended, but all present.

But even this extension of the ‘platform’ to encompass the audio format conversions fails to recognize the brute fact that Paula doesn’t create music. Paula’s ability to replay samples might strongly imply a compositional workflow based around small snippets of sound acquired from original recordings, sound libraries or from other pieces of music, and this workflow might have much in common with the cut-up techniques of Steve Reich or the burgeoning hip-hop scene. However, whether Paula plays back the sound of a barking dog or a series of samples derived from Roland’s (1987) D-50 synthesizer (as in the case of Andrew Barnabas’s SWIV soundtrack, 1991) is the result of a complex series of aesthetic and technical decisions arrived at through the dialogue between composer, programmer and chip. The compositional decisions, solutions and compromises are negotiated with the affordances of the sound chip, the architecture of the system as a whole, and the amount of memory and processing resources afforded to the sound program by the development team. Interestingly, the availability of samples gives the video game composer access not only to a broader sonic palette but, as we hear with SWIV’s ‘Decimation’ theme, access to a sound palette immediately familiar from popular music. Laced with the sounds of the Roland D-50, SWIV is audibly connected to the industrial soundscapes of Jean-Michel Jarre’s influential 1988 Revolutions album, for instance.

Yet, there is considerably more to explore than spec sheets and more to chiptunes than chips. Not only does each chip have an identifiable musical/technical ‘character’, but also each composer/programmer bestows upon it a unique and equally identifiable personality born of specific compositional, sound design and technological interactions and intentions. In discussing her work on early Namco arcade games, Junko Ozawa notes how foundational this personalization can be, by pointing to her own library of hand-drawn waveforms meticulously laid out on squared paper.Footnote ¹⁶ Like the Amiga composer’s bricolage of samples drawn from countless sources, these waveforms operate in dialogue with the sound chip’s playback capabilities to define the sonic territory available to the composer. Ozawa’s stamp of individuality is evident in her personalized waveforms.

Other composers forged their own distinct paths by rejecting the typical approaches to writing music for a particular chip. The NES’s triangle waveform is often used to take on bassline duties, as we hear in the case of the overworld theme in Super Mario Bros. (1985). The triangle wave is suited to this function because of a lack of control over its amplitude (thereby offering fewer opportunities for dynamics) and its pitch range extends down to 27.3 Hz, unlike the two pulse voices that bottom out at 54.6 Hz. Yet, some composers challenge convention by utilising the triangle wave in a wholly different way. By rapidly modulating the pitch of the triangle, the single voice can do double duty: it can alternate between a pitched bassline note and a kick drum. When the drum note is reinforced with a burst from the chip’s noise channel, the result is heard as combination of bass, kick and percussion. The effect was used extensively by Neil Baldwin (Hero Quest, 1991) and Tim Follin (Silver Surfer, 1990). Composers devised their own innovative ways of writing for the technology, resulting in a diversity of approaches to composing for even one type of chip.

Some composers took this approach even further. Some SID chip composers, such as Rob Hubbard, would use one chip channel to play multiple voices, and/or blend chip channels to create new hybrid sounds. For example, in Commando (1985), to create his percussion sounds, Hubbard changes the waveform of a channel every fiftieth of a second to create a single sound that combines a burst of noise followed by a swooping pitched ‘tom’ effect. Even though the SID chip only has three channels, Hubbard’s Commando score does not limit itself to three discrete musical parts. Where the listener may hear the rhythm track as a single ‘part’, it is actually written and performed across multiple voices. Sometimes voice two performs a ‘drum’ hit and elsewhere the same sound is played on voice three, for example. This happens because, depending on the nature of the composition, voice two is not always available and may be ‘busy’ playing the bassline or countermelody. It is clear from analysing the Commando theme that it is not a piece of music simply transcribed for the SID chip but, rather, is written with the specific capabilities and affordances of the chip in mind.Footnote ¹⁷ The piece’s thick sonic texture is further enhanced by the frequent use of ornamentation characteristic of Hubbard’s work. Commando only uses three channels, but by continually shifting the musical material across the three voices, making space to pack in other musical elements, it gives the listener the impression of more than three voices being present.

Martin Galway’s soundtrack to Wizball (1987), by contrast, uses similar processes for different ends. Here, Galway doubles and offsets melodies across voices to simulate a delay or echo effect (common in recording-studio outboard racks but absent from the SID’s specification).Footnote ¹⁸ What we hear in Galway and Hubbard’s work, then, is as much a function of the SID chip as it is a reflection of their individual compositional characters. And so, when Guay and Arsenault note a tendency in ‘chiptunes’ towards busy ‘baroque’ ornamentation, this holds true only for some composers of early video game music.Footnote ¹⁹

If we look to the NES, we see similar variations in musical style and form that are best read as reflections of differing artistic and aesthetic intentions and sensibilities. Koji Kondo’s work on the Super Mario Bros. and The Legend of Zelda series takes a very particular approach to the RP2A0X sound chip. Unlike the work of Hubbard, whose extreme multiplexing sought to create the impression of a greater number of voices than were actually sounding, Kondo’s NES themes are stripped back. Not bare or minimalistic, by any means, but neither disguising nor shying away from the presentation of a four-part composition of bassline (triangle), melody and countermelody (pulses) and percussion (noise). For McAlpine, Kondo’s technique in Super Mario Bros. has its antecedents in the ‘shell voicing’ technique used by Art Tatum and Bill Evans that implies complex harmonies with just a few notes, and the syncopations and polyrhythms that are not entirely dissimilar to the rhythmic devices employed in Debussy’s first Arabesque.Footnote ²⁰

Of course, music is typically not the only sonic output of a game, but has to co-exist with sound effects in the game’s soundscape. One question we might raise in relation to Kondo’s Super Mario Bros. theme is why the jumping sound effect, triggered as Mario takes flight and heard with some frequency, should cut out the main melody line. The technical answer is simple. With only four channels available for music and sound effects, something has to give. Kondo’s was but one of a number of strategies deployed throughout the 1980s and 1990s to combine music and ‘interactive non-diegetic sound’ as Collins puts it.Footnote ²¹ Some games reserved voices on the chip exclusively for sound effects (The Human Race, Mastertronic, Commodore 64, 1985), while others asked players to choose between music and sound effects (Delta, Thalamus, Commodore 64, 1987). But why replace the lead line rather than the countermelody? One answer, favoured by McAlpine, connects the pitch of the jump sound effect with the key of the game’s soundtrack.Footnote ²² As such, the harmonic integrity of the soundscape (if not the melody) is preserved and the sound effect is integrated into the music, joining the visual, sonic, haptic and ludic aspects of the game into one experiential whole.

What we see in these examples is not simply the result of different musicians imprinting themselves on an instrument through their differing sensibilities. Rather, the different approaches become encoded in the distinct software routines that are used to create and craft music.

Although it is tempting to view the 1980s or 1990s console, home computer or arcade-board sound chip as a synthesizer, if we consider the interfaces by which these sonic potentialities were accessed, they have little in common with the commercially available instruments used by musicians on stage and in studios at the time. The SID chip, for instance, presents itself as a series of data registers that can be read and written to. There are no inviting faders, sliders, potentiometers, or pitch and modulation wheels to be seen. All sound design is undertaken by accessing these data registers in hexadecimal. Similarly, in the 1980s, no visual compositional tools existed to help write, arrange or edit sequences of game music. Any tools had to be created for the task. Some musicians created their own bespoke sound design, composition and playback software routines – or ‘drivers’ as they are typically called – while others relied on the services of coders. Either way, these sound drivers play an immense role in defining the sound of the sound chip. These routines reveal, or make available, facets of the sound chip’s capability.Footnote ²³ As Hubbard notes of his Monty on the Run (1985) composition, ‘The middle section was an excuse to use the new pitch bend code that I wrote for this project.’Footnote ²⁴ Similarly, Martin Galway, in-house composer at Ocean Software, describes how he was ‘mastering the C64’ as he was developing his drivers.Footnote ²⁵

Technical capabilities are transformed into musical utilities by the software driver. For instance, the SID chip’s technical capability to stably control the pitch of its oscillators over a wide frequency range or to continuously vary the duty cycle of its pulse waves become the musical operation of portamento. A driver includes and omits musical and sound design features in accordance with the aesthetic judgement and predilections of the composer and sound designer. As new affordances are revealed, these features can be added to the driver to present them in a musically useful manner. Sometimes these are features unintended and unanticipated even by the creators of the sound chips themselves, as in the case of the SID chip’s sample playback capability.

To speak of the sonic characteristic of a given sound chip is to tell only part of the story. Without sound driver routines written in software and acting as a mediating interface between the composer and silicon, the sound chip’s features remain musically inaccessible data points on a specification sheet. As Chris Abbott observes, ‘Rob Hubbard sounded very different from Martin Galway because they had to write their own synthesizer engine, as well as the music.’Footnote ²⁶

Perhaps the most persuasive example of the significance and personality of the sound driver in mediating and shaping the interface between musician and chip is found in the rapidly arpeggiating pseudochord. This musical device must surely rank as one of the most instantly recognizable features of ‘chiptunes’. It consists of an arpeggio of two or more notes played at a speed so high that the individual tones almost appear to meld into a single chord. Almost, but not quite, which is what gives the figure a warbling or chirping effect a little like a mobile telephone ring, as the individual tones and their transients are still identifiable.

The key here is that the design and operation of the sound driver for a video game are intimately related to those of the system processor, game program and TV standards. The sound driver is just one of the processes requiring access to the finite computing resources of the system. If we take the case of the typical 1980s console or home computer game, as graphics need to be drawn on screen fifty or sixty times a second (depending on the exact system and international TV specification) to create the impression of continuous, smooth motion, other elements of the program, including the sound driver, can run only when the processor is not busy undertaking these tasks. With the driver written to issue new instructions to the sound chip fifty times a second, this sets the maximum rate of change for the opening or closing of a filter, the adjustment of a waveform duty cycle or the pitch of a note. The effect of this is that the rapidity of the rapid arpeggiations that comprise the pseudochord is dictated not by any musically significant or even musically derived division, but rather by the driver’s maximum speed of operation. Importantly, the maximum speed of arpeggiation does not necessarily reflect the extent of a sound chip’s capability, and serves to underscore the significance of the driver in shaping the available features.

Yet, while arpeggiation has become almost emblematic of an entire genre of music, it is not endemic to any particular chip nor, indeed, is its use a product of or limited to sound chips. Rapid arpeggiation has been evident in monophonic music for many centuries, but here it is accelerated beyond the ability of even the most dexterous human player. More importantly, because this rapid changing of pitch is not a feature of the sound chip that is simply enabled or disabled, but rather is a function of the sound driver routine, its design and implementation is as personal and distinctive as any other sound-design or compositional device we might identify. So, where some composers use three- or even four-note arpeggios, some, such as Hubbard, often use just two. Indeed, where a composer writes just music with no need for graphics, game logic or user input, a sound driver may be written to run at a higher rate thereby offering increased resolution of sound manipulation and note retriggering. As such, though the pseudochord is characteristic of early game music, its implementation is always coloured by the driver and composer in question.

In addition to its flattening effect, then, the chiptune soubriquet is also problematic because it goes beyond setting focus on technology to giving it primacy. As such, the interactions and creative endeavours of the people who design the chips and the sound drivers and those who ultimately shape the technology are silenced. A focus on the machinery of music making is key to helping us avoid the tendency to generalization rightly identified by Altice. However, we must also exercise caution by ensuring that we do not separate the sound chip from the contexts of its design and use and inadvertently create a deterministic account that is deaf to the ways in which technologies are shaped. Similarly, by focusing on the ways in which the particular combinations of design decisions, technical and musical opportunities, quirks and affordances are revealed, harnessed and exploited in different ways by different composers and programmers, we offer ourselves an opportunity to account for the often chaotic and haphazard manner in which these investigations, innovations and musical inventions are arrived at.

Teleology, Linearity and the Tyranny of Limitations

Both popular and scholarly accounts of game music history display a tendency to centre on the inevitability of technological progress or evolution. These teleological approaches are based upon two interrelated principles. First, that there is an implied perfect or ideal state which this inexorable forward motion is heading towards. Second, that less advanced stages along this journey are characterized by limitations, which are progressively eradicated and rendered immaterial.Footnote ²⁷

Histories of early video game music typically present an evolutionary narrative that is both teleological and demonstrably written from a backward-facing perspective in which the limitations are evident given the bounty of what is now known to have followed. As such, rather than offer a study of early video game music and sound chips in terms of the ways in which foundational practices of design, composition and technique were established, they typically serve the purpose of a (brief) contextual ‘prehistory’ for contemporary practices. Popular accounts are more explicit in their conceptualization of progression in their titles. Both BBC Radio 3 (2018) and NPR (2008) have produced programmes entitled The Evolution of Video Game Music, for instance.

The orthodoxy of the technological timeline begins with the silence of early video games such as Steve Russell’s Spacewar! (1962) before indefatigably moving along the pathway to the present, noting the series of significant milestones along the way. Early milestones typically include games such as Taito’s coin-operated Space Invaders (1978), noted for including the first continuously playing soundtrack. This consisted of an ostinato pattern of four descending notes played in the bass register, and the tempo increased as the player cleared more invaders, making it the first example of interactive, or perhaps more accurately reactive, audio. For the first instance of a continuous melodic soundtrack, our attention is drawn to Namco’s 1980 Rally-X, while Konami’s 1981 Frogger added introductory and ‘game over’ themes linking music to the game state. The evolution of arcade sound is often articulated in terms of new chips (including the introduction of the numerous variants based around Yamaha FM or Frequency Modulation synthesis method), or the inclusion of greater numbers of sound chips working in parallel to provide increased polyphony (Konami’s 1983 Gyruss included five of General Instrument’s AY-3–8910 chips, for instance).

For home gaming, the evolutionary narrative is typically mapped to the successive ‘generations’ of consoles which move from the discordance of the Atari 2600’s TIA, through the SID chip’s synthesis, FM (e.g., Sega Mega Drive/Genesis which actually used a combination of Yamaha YM2612 along with the Texas Instruments SN76489 from the earlier Master System console), and sample playback (the Nintendo SNES/Super Famicom’s Sony-designed S-SMP allowed 8 channels of 16-bit samples).

In the domain of general-purpose home computers, while it is ironic to note that the Atari ST was a fixture of commercial music recording studios,Footnote ²⁸ it is the Commodore Amiga that is more commonly identified as key in the evolution of game music. This accolade arises from the development of the ‘Tracker’ software sequencers we noted above and the associated ‘MOD’ file format, which brought some degree of compatibility and portability.

In contrast, the PC-compatible games marketplace was characterized by a variety of competing sound cards such as the AdLib, Creative’s Sound Blaster and peripherals such as the Roland MT32, offering differing playback quality, sound libraries and available effects (the MT32 included reverb, for instance). Where consoles provided a stable platform, the variety of potential PC-compatible sound cards created a situation whereby music might sound markedly different from one system to the next. In the case of a game like LucasArts’ Monkey Island (1990), although the PC speaker version is recognizably the same melody, when this is compared with its playback on a Roland SCC-1 card, it is rather easier to note the differences than the similarities.

Aside from the inevitable magnitude of the task, one issue we might immediately have with the timeline approach arises from the fact that, while games and systems unquestionably provide useful sites for investigation, their milestones entangle aesthetic, creative and technological endeavour and innovation. Nonetheless, such narratives remain powerfully seductive, not least because they chime neatly with the dominant discourses identifiable elsewhere in writing on video game history and, indeed, as deployed by the video games industry through its marketing and advertising messages.

As I have suggested elsewhere,Footnote ²⁹ this narrative is not merely one that implies the existence of a continuum with its imaginary end point of gaming perfection, but is also one predicated on the supersession of successive generations of hardware. Simply put, each platform generation embarks on a journey beginning with its construction as mythical object of speculation, anticipation and desire through to an ultimate reconstruction as a series of weaknesses, limitations and failings that are rectified by its replacement. It becomes the baseline from which the next generation’s performance is measured. Given the prevalence of this narrative and the ways in which new platforms and games are seen to improve upon their predecessors and render them obsolete, it is not surprising to find so much history of game music reproducing these formulations.

However, it is worth probing the efficacy of the conceit of limitations, particularly as its use in relation to early video game music is so prevalent that it seems almost to have become an uncontested truism. Much of what we have seen above, whether it be the use of the distinctive pseudochord or the waveform manipulation/multichannel drum design, can be explained as a response to the limitations of the SID chip and RP2A0X polyphony, just as Mario’s occupation as a plumber can be explained as a response to the limitations of the NES colour palette and sprite resolution and the decision to clad him in overalls. On the other hand, these examples can equally be framed as manifest examples of the creativity and inventiveness of composers, programmers and character artists working within specific resource windows and working in dialogue with the technologies at their disposal to shape their outcomes. That their creations, whether auditory or visual, appear to confound and exceed the boundaries of the ‘limitations’ is a testament to their ingenuity.

The incidence of looping is another area worthy of study in this regard. The construction of early video game music either as a single looping phrase, or as a series of nested loops, is for many commentators a defining feature of the form.Footnote ³⁰ The reliance on looping is typically explained as a necessary compositional response to the limited availability of memory within which music data and driver code had to be stored. However, the tendency towards heavily looping sequences might also be read as a result of the iterative nature of compositional/coding practices.

In fact, in this final point there is a hint that, like Kondo’s integration of the jump sound effect into the melodic top line, repetition in early video game music might be read as aesthetically and functionally matched to the often repetitive, iterative nature of the gameplay it supported. The four notes of Space Invaders are undoubtedly efficient, but they also suit the relentlessly cyclical nature of the gameplay just as Super Mario Bros.’ repeating phrases suit gameplay and levels comprising repeating graphical and structural elements.

More than this, however, we might also re-evaluate whether some of what are presented as limitations truly were so, or have ever been truly remedied. Returning to drum design, achieving this through the rapid manipulation and alteration of an oscillators’ waveform is a remarkable feat of ingenuity in sound design and creative thinking. It is also made possible because the SID chip’s design allows the waveform of its oscillators to be altered at audio rate. It is possible to achieve this because the sound driver software provides an interface to that data register and enables the composer and sound designer to manipulate the SID’s output accordingly. Is this example best understood as limited polyphony, with voices and waveforms accessed through an unforgiving interface? We might well argue that the SID chip was far from limited. Indeed, while modern replications and emulations of sound chips (such as the SID Guts, AY3 and Edges) provide altogether more user-friendly and tactile interfaces to the underlying hexadecimal data registers, the nature of these interfaces and protocols means that they actually offer considerably less precise control. The sound driver’s interface may have lacked the tactility of the contemporary hardware synthesizer, but it did enable the composer, programmer and sound designer to effectively write directly to the silicon. If for no other reason than this, our discussion of the role of technology would greatly benefit from a radical rethinking of the centrality of limitations and the teleological, evolutionary process that implicitly frames early video game sound as a prototype or underdeveloped iteration of what would inevitably later come.

In this chapter, I hope to have demonstrated the importance of keeping an awareness of technology at the heart of our consideration of early video game music while simultaneously guarding against the rehearsal of the deterministic, teleological discourses so dominant within scholarly, popular and industry accounts. In particular, I hope to have foregrounded the ways in which contemporary analytical frames such as the centrality of overcoming or working with ‘limitations’ or the construction of ‘chiptunes’ actually serve to hinder our ability to create nuanced accounts of early video game sound and music.

Introduction

Chiptune: for players of a certain age – and, as a child of the 1970s, I certainly count myself as one of them – it is the aural embodiment of video games. There is something about that raw, geometric sound that captures classic video gaming in its most immediate form, a distillation of pure gameplay. It represents a period of gaming in which technical and musical creativity combined in the most exquisite way, as video game programmer-composers ingeniously coaxed the primitive hardware – primitive, at least by today’s standards – into feats of musicality that it had never been designed to achieve.Footnote ¹

That period of game audio, however, was relatively short-lived. The programmable sound generators (PSGs) that served as the voice for the 8-bit machines – Atari’s TIA and POKEY and Commodore’s SID, for example – used simple digital sound synthesis.Footnote ² While this gave those 8-bit soundtracks a unique and very characteristic sound, it was almost impossible to make those PSGs sound anything other than ‘blippy’.

By the mid-1980s, in a drive towards greater sonic range and fidelity, the PSGs had largely been superseded and were beginning to be replaced by dedicated sample-based hardware, such as the Paula chip that provided the four-channel stereo soundtracks of Commodore’s Amiga,Footnote ³ and FM and wavetable soundcards in IBM PCs and compatibles,Footnote ⁴ but it was the arrival of the CD-ROM drive, and particularly that of the Sony PlayStation, that created a fundamental shift in what players could expect from their video game soundtracks.

For some, it heralded the end of an era. Mark Knight, an industry veteran who got his break writing the Wing Commander (1990) soundtrack for the Commodore Amiga explains:

Just as an industry drive towards filmic realism and shifting audience expectations normalized colour cinema in the 1940s and ’50s,Footnote ⁶ bringing an end to the era of black-and-white film and its brooding unreality, so too PSG music disappeared from video game soundtracks,Footnote ⁷ to be replaced by MIDI arrangements, sampled loops and licensed commercial tracks on CD-ROM.

It was a shift that saw video game music take on a more polished and commercial edge. The PlayStation racer Wipeout (1995), for example, featured a high-octane electronic soundtrack that was mostly written by composer Tim Wright with some tracks licensed from Leftfield, the Chemical Brothers and Orbital. Sony also licensed music from some non-mainstream acts to create an original soundtrack album that was released to promote the game at launch (Columbia Records, 1995).Footnote ⁸

Colin Anderson, who, in the 1990s, was Head of Audio at DMA Designs, the company that created both Lemmings (1991) and Grand Theft Auto (GTA, 1997), described how that shift away from sound chips to full production music changed how he approached game audio.

One of the principal soundtrack innovations introduced by the GTA franchise was its in-game radio stations, a feature that created a sense of pervasiveness for its diegetic world, making it seem broader, richer and more multifaceted than the player’s direct experience of it.

But, as Anderson continues,

That shift in end-user expectation proved to be a spur for further innovation, both in the way that music was utilized in games – the move back towards video games with real-time adaptive soundtracks, for example, was as much an industry response to that ‘so what’ factor as it was a desire to create tightly integrated interactive audiovisual experiences – and in how that music was acquired by developers.

If established artists were prepared to lend the weight of their brand and fan base to a game franchise, and perhaps even pay for the privilege, it is hardly surprising that developers and publishers would embrace that new commercial model, particularly in the high-budget blockbuster development space, where development times and budgets are often huge, raising significantly the overall cost and risk of production.Footnote ¹²

The age of PSG video game music, then, was brought to an end as much by the commercial realities of video game production as it was by the increasing technical capacity of home consoles. However, while chiptune might have disappeared from games, reports of its demise were greatly exaggerated. Chiptune was about to develop an edge, one that would set it in direct opposition to the corporate world of professional game development.

Going Underground

Since the earliest days of gaming, software piracy had been a problem for publishers.Footnote ¹³ In the late 1970s and early 1980s, the issue was most pressing on tape- and disc-based machines like the Apple II, the Commodore C64 and later the Commodore Amiga and Atari ST, these media lending themselves more easily to analogue or direct digital duplication than did cartridges.

The industry responded directly and technologically to the threat, by combining a range of sophisticated copy protection routines with the threat of legal action against those who circumvented the copy protection to distribute the games.Footnote ¹⁴

That stance created two parallel but overlapping worlds; on the one side the corporate world of the games industry, and on the other, the world of crackers, skilled coders whose self-appointed role it was to strip the copy protection from games and release neutered versions within days – and sometimes hours – of their official release.

Removing copy protection was a complex process, akin to surgically removing a non-vital organ that, nevertheless, forms part of a complex biosystem. The more complex the copy protection, the greater the surgical skill required to remove it. For the cracker this was the primary motivating force, not the resale value of the software or its functionality; they wanted to be able to demonstrate that they were nimbler and more skilled than those who designed the copy protection and all of the other crackers who were scrubbing up in countless other bedroom operating theatres.

The first organized groups of crackers, or cracking crews, came out of Western Europe, specifically West Germany (JEDI) and the Netherlands (ABC Crackings) around 1983,Footnote ¹⁶ but by the mid-to-late 1980s crews were working across international borders to produce not only cracks, but sophisticated digital calling cards – crack intros, or cracktros – that were displayed onscreen as the games loaded. These combined scrolling text, algorithmically generated plasma and 3-D effects, and music to mark the technical achievements of the crack (see Figure 2.1).

Figure 2.1 The Melon Dezign cracktro from their crack of Thalamus’s Creatures (1993). The cracktro features the music of Mark Knight, credited as TDK. Note the use of the lyrics to Boney M’s Ma Baker, an ironic nod towards the illicit practice of cracking. Cracking groups often riffed off pop culture references

The code to execute cracktros had to be compact and efficient to fit in the boot sectors of floppy disks, so that the cracked game could be uploaded and downloaded easily from bulletin board services via dial-up and rewritten to new floppies. The simple waveforms and sequences of PSG music, which could be stored as space-efficient single-cycle samples and tracker sequences for playback on sample-based systems, lent itself perfectly to this end. Chiptune became the sound of the digital underground.

Over time, the competition to demonstrate both coding virtuosity and graphical and musical creativity became more important than the cracks themselves. End users would actively seek out cracked software for the cracktros, rendering the game an almost insignificant by-product of the cracking process.

The production and sharing of cracktros became an end in itself, and evolved into the demoscene, a distributed online community of digital arts practice dedicated to the production of complex real-time audiovisual displays.Footnote ¹⁷ That combination of anti-commercialism, a distinctive sense of community and a culture of sharing marks a definite point of departure of chiptune, as a constituent part of the crackscene and demoscene, from the increasingly professionalized and corporate approach to video game music production.

It also points to a difference in mindsets. On one side sits the corporate perspective, which recognizes that there is value – and cost – in the production of professional content, be that music or software, and that it is therefore justifiable for a company to protect its investment by using a combination of digital rights management (DRM) and litigation to ensure that only legitimate copies are in circulation.

Set against this, the Hacker Ethic, the core philosophy of hacking culture, which originated from Massachusetts Institute of Technology in the 1950s and 1960s,Footnote ¹⁸ sets out the intellectual counterpoint to this enterprise-driven process of making intellectual property out of everything, namely the ‘belief that information sharing is a powerful good and that it is an ethical duty … to share … expertise by writing free software and facilitating access to information … whenever possible’.Footnote ¹⁹

But while that intellectual tension is most often framed and discussed in terms of the lone hacker against the multinational corporation, in practice, its impact can often be felt on a much smaller scale. As Mark Knight says:

I would argue that it’s not that people don’t ‘get it’, it’s that these perspectives lie at opposite ends of a continuum on which we all sit, and our position on it shifts, depending on context and whether we are predominantly creating or consuming.Footnote ²¹ It also points to a fundamental shift in how we collectively value intangible products, be they musical works or software.

An Open Letter to Hobbyists

Early in 1975, a young Bill Gates and his friend Paul Allen picked up the January copy of the news-stand magazine, Popular Electronics. On the cover was an Altair 8800. Manufactured by Micro Instrumentation and Telemetry Systems (MITS), the Altair, almost overnight, would become a commercial success: its designer, Ed Roberts, worked out that he needed to sell 200 machines to break even; within three months he had a backlog of 4,000 orders.Footnote ²²

Gates, then a Harvard undergraduate, had been following closely the growing phenomenon of personal computing, and had come to the conclusion that there was value in software as an indispensable counterpart to hardware. Sensing that the Altair represented a breakthrough moment, Gates and Allen called Roberts and offered to demonstrate a BASIC interpreter for the machine, hoping to contract with MITS as a key supplier. In fact, the pair didn’t even have an Altair, let alone the BASIC interpreter that they were offering.Footnote ²³

Roberts agreed to meet them, and in the space of just a few weeks, Gates and Allen had developed an Altair emulator that ran on Harvard’s PDP-10 mainframe, and then the BASIC interpreter. The first ‘Micro-soft’ agreement was sealed in April; Gates and Allen received US$3,000 immediately, with royalties of US$30 per copy of 4 K BASIC, and US$35 for 8 K BASIC for each subsequent sale.Footnote ²⁴

As they were about to discover, however, the early adopters of personal computing had a very different perspective from Micro-soft on the value of software. At one time, computer software was not something that was bought and sold. As computing hardware trickled out from research labs in the early 1950s, most end users wrote programs themselves, largely because none of the hardware manufacturers provided any for them to use. IBM’s first production computer, for example, the 701, came with little more than a user manual.Footnote ²⁵

Developing applications was a major undertaking that required specialist support. Even relatively mundane programs required thousands of lines of code. They were difficult to debug and needed continual modification and improvement in response to the demands of a changing business environment.Footnote ²⁶ Most companies maintained a team of programmers to service a single mainframe machine, a significant portion of the overall cost of maintaining and running a computer.

IBM, which, even in the 1950s was an old and well-established company, recognized that if that continued, ‘the cost of programming would rise to the point where users would have difficulty in justifying the total cost of computing.’Footnote ²⁷

In response IBM created SHARE, a community of makers and consumers whose key mission was to share information and programs, thereby reducing the overall cost of computing, and in turn making IBM’s machines a more attractive and cost-effective option.

From this group came many of the standardized notions of operational computing that continue through to the present day,Footnote ²⁸ but so too did the idea that software could – and should – be something that was freely distributable. It was a commodity whose main value was in making the leasing of hardware more attractive, rather than as something that had value in its own right.

By the late 1960s, IBM had become the dominant player in mainframe systems. In much the same way as Microsoft achieved with its operating systems throughout the 1980s and 1990s,Footnote ²⁹ IBM, by power of ubiquity, had created a de facto standard. By bundling software, IBM was able to provide users with a tight-knit group of products that would work seamlessly together, and that presented a problem: when the choice was to go with an IBM system, complete with training and support, or to try and bring together several applications from different suppliers that had not been proven to work together and which might receive uncoordinated updates and fixes from their individual manufacturers, most customers did not consider it a choice at all.

In 1967, the Antitrust Division of the US Department of Justice began an investigation of IBM, citing IBM’s practice of bundling as evidence of the company’s anti-competitive practice, and so, on 6 December 1968, IBM announced that it would unbundle the five major services – system engineering, education and training, field engineering, programming services and software packages – that it had previously included free with its hardware, and charge separately for them.Footnote ³⁰

And so it was, fresh in the wake of IBM’s unbundling initiative, and with the concept of software as a saleable product still a relatively novel and untested idea, that Gates found himself colliding head-on with the established mindset that software should be free.

Gates had embarked on a national roadshow to demo the Altair and Microsoft’s BASIC interpreter. At one event in Paolo Alto, Gates presented to a hotel packed with members of the Homebrew computing club, many of whom had already built an Altair and were waiting for MITS to release BASIC.Footnote ³¹ When they saw that the Altairs on display were all running BASIC off punched paper tape, one unnamed member ‘borrowed’ the tape and ran off a few copies. At the next Homebrew club meeting, there was a cardboard box filled with dozens of BASIC tapes for members to take, with just one condition: you had to make a couple of copies for each one you took.Footnote ³²

Gates was furious. He wrote an emotionally charged open letter, which set out both the tone and the agenda for the debate around intellectual property that has raged since (Figure 2.2).

Figure 2.2 Bill Gates’s Open Letter to Hobbyists. This letter sets out clearly the opposing perspectives of the hobbyist community, which had legitimately come to think of software as a freely shareable resource, and Gates’s more corporate viewpoint, which sought to commoditize its value. This debate has, if anything, intensified, and has grown to incorporate all manner of intangible commodities, including music recordings

History has demonstrated beyond doubt that Bill Gates and Paul Allen were right about the commercial potential of software as a commodity, but the continued growth of the underground warez scene, and the legitimate adoption of freeware, Creative Commons and open source as models for publishing and distribution suggest that attitudes around the sharing of digital content remain as strong as ever.

The Incentive to Share

While the technology of sharing is now different – chunks of data shared as torrent files as opposed to cardboard boxes of punched paper tapes – the nature and themes of the debate have remained remarkably consistent as they have played out across different domains and distribution media and at different points in time. Although there are some notable differences between software development and music publication and distribution, there are also some quite striking parallels. Both, for example, have forced legislators and the public to deal with new technologies that have challenged our fundamental assumptions of what constitutes publication and ownership.

At the turn of the century, piano rolls, for example, were considered to be part of the machinery of a player piano, like the mechanism of a music box, and so not subject to copyright law, despite the fact that the punched paper rolls, unlike a music box, were distinct from the playing mechanism and could easily be swapped for other rolls, which contained all of the detail of the original music manuscript, albeit in a mechanically encoded form.Footnote ³³

In 1978, novelist John Hershey, a member of the National Commission on the New Technological Uses of Copyrighted Works, argued, in a similar vein, that computer code is dramatically different from other copyright works because the ones and zeroes in a computer program are designed to have

In some respects, chiptune represents both of these key characteristics: the machine code routines and data are no more than ones and zeroes designed to control electrical impulses in a machine, and, like the rolls of a player piano, they convey no meaning unless they are coupled with a PSG,Footnote ³⁵ which provides the machinery necessary to turn that code into sound, and yet few creative coders would challenge the idea that in writing sound drivers and music data, they are encoding both the musical score and the performance characteristics that will realize a musical work; they are not simply performing routine mechanical tasks.

However, such notions remain abstract until they are tested in law, and because of the anti-commercial sharing ethos that is prevalent in the chiptune community, chip musicians have generally been happy for others to appropriate, adapt and cover their work provided nobody makes any money from it.

In an interview, Ben Daglish,Footnote ³⁶ the prolific C64 composer, described how he felt about new generations of chip musicians rediscovering and using his video game themes:

There have, however, been instances where chiptunes have ended up in court: David Whittaker’s soundtrack to the classic C64 game, Lazy Jones (1984), for example, was reused commercially without permission by the German techno outfit, Zombie Nation, who used it as the central hook in their track ‘Kernkraft 400’ (1999), while in 2007 Timbaland used elements of the demotune ‘Acidjazzed Evening’ (2002) in the Nelly Furtado track, ‘Do It ’ (2007).

In both instances the legal challenge failed, and in part, that failure stemmed from the fact that it was difficult to prove ownership and establish the mode of publication. In the early days of the games industry, in Europe and North America at any rate, nobody gave much thought to the value that was present in the intellectual property that comprised the game. Video game music was not something that was imagined to have distinct value, and in most cases, those soundtracks were commissioned verbally – often by telephone – and so today there is simply not the paperwork to go back and prove who owns what.

In some respects, that legal ambiguity presents a real challenge for video game historians who seek to document and archive the ephemeral elements of early gaming culture. Video Game History Foundation founder Frank Cifaldi notes that ‘there is no alternative BUT piracy for, like, 99 per cent of video game history’ due to ‘the completely abysmal job the video game industry has done keeping its games available’.Footnote ³⁸

That rarity argument – the idea that if the industry is either unwilling or unable to maintain legitimate access to a back catalogue, then end users are justified in using whatever means are available to source the material that they seek – is discussed more fully by Steven Downing,Footnote ³⁹ and it represents another driver in the underground market for digital content, particularly when the boundaries between the legitimate and the illegitimate are fuzzy or uncontested.

The desire to accumulate and collect is a common feature in most fan communities,Footnote ⁴⁰ and the chip music community is no exception. The High Voltage SID Collection (HVSC), for example, is an expansive and in-depth online repository of community-ripped C64 SID music files that combine both the copyrighted music data and the music driver code required to play it. It was created specifically to meet the growing demand for the specific sound of classic 1980s video game music in its original form as gamers migrated to new platforms that offered a different – more ‘produced’ – musical experience.

Collections like the HVSC represent the latest manifestation of a culture of illicit supply and demand that goes back to the very beginnings of the recording industry, when bootleggers, playing a role similar to the fans who rip SID music files for the HVSC, stepped in to provide consumer content that was not available through legitimate channels.

Bootleggers and Mixtapes

The term bootlegging rose to prominence in Prohibition-era America,Footnote ⁴¹ and it first started to be associated with the practice of music recording and distribution in the late 1920s, just as the culture of record collecting started to emerge. An article in Variety in April 1929, for example, notes that ‘There is almost as big a market for bootleg disk records as there is for bootlegged books’.Footnote ⁴²

Many of the recordings that fans were interested in collecting had been produced in low numbers by small or unstable companies, and major labels like RCA Victor and Columbia were not interested in keeping obscure records in production. Collectors’ magazines sprouted up in the 1930s in response to the growing public interest, and this, in turn, boosted the collectors’ market, with the value of individual recordings being determined largely by their rarity. Bootleggers began supplying the demand, sourcing and reproducing rare and deleted works without incurring any legal reaction.

Collecting and bootlegging, from the outset, existed in paradoxical symbiosis: fan culture depended on bootlegging, and yet bootlegging undermined the rarity value of the recordings it supplied. The relationship highlighted the long-term commercial value of a back catalogue at a time when the music industry still treated recordings as products of the moment, aimed at contemporary markets and abandoned as consumer tastes shifted.

By the late 1960s, the availability of quality portable recording equipment and cassette tapes meant that an increasing number of unauthorized recordings of live events began to surface. Bootlegs became a valuable commodity in the shadow cultural economy of fan culture that sat – from the perspective of the industry, at least – uncomfortably alongside the more mainstream commercial channels of popular music. It was an economy that relied on an honour system, where those who received tapes from fellow traders and collectors made multiple copies to pass on to others within the community, echoing the sharing culture of many other anti-commercial groups and in particular, the hacker code that had so incensed Bill Gates.

‘Home Taping Is Killing Music’, cried the British Phonographic Industry (BPI) as the 1980s dawned and twin cassette decks and blank tapes became more affordable, which in turn domesticated music duplication.Footnote ⁴³ A few years earlier, in 1977, the BPI had estimated that the industry had suffered around £75 million in losses through lost revenue to home taping. A study released by CBS went further, blaming home taping for the loss of hundreds of millions of dollars of record sales, and industry commentators began to predict the death of music just as surely as Gates had predicted the death of professional software.Footnote ⁴⁴

It was a hard-hitting message that had little impact. It was at odds with consumer experience, who viewed home taping at worst as a victimless crime, but largely, thanks to the subversive DIY ethic of punk, primarily as an expressive and creative act. A whole culture and social infrastructure grew up around the mixtape,Footnote ⁴⁵ allowing music lovers to spread the word about what they liked, to make statements about themselves or to reinvent themselves to others, or, in the days before the complex personality-matching algorithms of internet dating, to tentatively sound out the personal qualities of a potential life partner. The counter-slogan of mixtape culture? ‘Home Taping is Skill in Music.’Footnote ⁴⁶

Home taping did not kill music, just as VHS did not kill the theatrical movie release. The emphasis of the rhetoric was wrong. It wasn’t music itself that was under threat, but the commercial framework that surrounded it, and here, amateur taping and informal distribution did, slowly, begin to change the way that commercial music was produced and distributed: streaming content, peer-to-peer file sharing, aggregators, online music collections like the HVSC and netlabels have all changed the way we access and consume music, and all have their roots, at least in part, in the digital underground.

Rather than move with the times, however, and embrace and adapt to disruptive technologies and changing public attitudes towards content, particularly as music became decoupled from physical media and labels could no longer justify charging consumers for simply accessing content, the industry reacted slowly and heavy-handedly, targeting the technology and the cultures of practice that grew up around them. As early as 1974, the BPI had threatened legal action against ‘hardware manufacturers whose advertising of tape equipment emphasises its potential for home-copying of copyrighted material such as recorded music’.Footnote ⁴⁷ Ten years later, the BPI made formal complaints to both the Advertising Standards Authority (ASA) and the Independent Broadcasting Authority about Amstrad’s advertising for their music centres, which highlighted the ease with which its two-in-one cassette deck could duplicate tapes. Their complaints were dismissed. The ASA pointed out that it was not unlawful ‘to advertise [the] features and capabilities of lawfully constructed appliances’.Footnote ⁴⁸

What the BPI was trying to do, of course, was protect the interests of its members, but it was doing it by demonizing the fans who were their bread and butter, in the process stoking mistrust and disenchantment that would further erode the sell-through music market. It was a counterproductive attack aimed in the wrong direction. A survey by tape manufacturer Maxell, for example, showed that ‘premium’ cassette users, those who were apparently killing music, actually bought twice as many records as non-tape-users,Footnote ⁴⁹ lending some credence to the notion that – amongst the fan community at any rate – collectors are likely to seek illicit copies for consumption to augment legitimate hardware and software that is bought for archival purposes.

Conclusion

Chiptune exhibits several significant links to and parallels with other, established areas of cultural practice. In particular, the increasing commercial pressures of video game development contributed – in part – to a schism between the production and consumption of video game music: this led, on the one hand, to an increasingly corporate and professionalized approach that has seen video game soundtracks evolve to become a tightly produced and interactive form of media music, and on the other, to chiptune becoming one element of a manifestation of a set of co-operative and anti-commercial community values that can trace its roots back through computer hacking to the bootleggers who supplied content for the early record-collecting community.

That anti-commercial ethos, however, and the pervasive culture of sharing, not just within the chipscene, but more broadly within the different subcultural groups that lurk beneath the increasingly corporate digital mainstream, certainly poses a challenge: as Bill Gates noted, who will create professional content – be that music or games or productivity software – if nobody is prepared to pay for it? History, however, suggests that content will still be produced both commercially and – to a very high standard – within deprofessionalized communities like the chip music scene.

That anti-commercial ethos, however, does impinge on the community itself, as musicians who invest heavily in the music they produce find themselves unable to recoup that investment by charging for product.

Perhaps more significantly still, the anti-commercial ethos provides the rationale for the continued adoption of stringent music DRM, which commercial publishers use to protect their intellectual property (IP). Again, however, history has demonstrated that while DRM may well reduce casual sharing, it seems to have little impact on piracy; instead, it primarily inconveniences legitimate customers by limiting what they can do with their legally purchased content.Footnote ⁵⁰ As Dinah Cohen-Vernik et al. discuss, since DRM-restricted content is only ever purchased by legitimate users, only they ‘pay the price and suffer from the restrictions … Illegal users are not affected because the pirated product does not have DRM restrictions’.Footnote ⁵¹

Contrary to conventional wisdom, then, it seems that because DRM restricts the legitimate buyer, thus making the product less valuable, and increases the cost of the product, the effect is that fewer people are willing to buy; instead they make an active decision to source their music illegally. That inconvenience may also play a role in driving new listeners to grass-roots subcultures like the chip music scene, where sharing is the default, and the boundaries between creation and consumption are less distinct.

But what of those scenes today? Interestingly, while both the chipscene and the demoscene evolved both conceptually and technically from video game hardware, it is the social and performative expressions of that hardware that have seen both scenes flourish into vibrant contemporary movements.

A hacked Nintendo Game Boy took the sound of chiptune from the desktop to the stage,Footnote ⁵² and created a new generation of chiptuners who brought with them new musical influences, particularly the sound of contemporary electronic dance music, giving the chip sound a harder, more aggressive edge.

In the intervening years, the scene has grown in scale – thanks largely to social media allowing geographically remote performers and audiences to form communities of practice – and in significance; the lo-fi sound of chip music has been adopted by a number of major commercial acts, including Beck, Kraftwerk and Jme.

In a similar way, interest in the demoscene has surged, reaching a peak of activity in the early 2010s, with the informal DIY parties of the late 1980s and early 1990s growing to become huge international stadium events attended by tens of thousands of people, all gripped by the spectacle of competitive creative coding.

Ultimately, however, while there are links in both chiptune and the demoscene with hacking, bootlegging, mixtapes and gaming, both groups exhibit a collective and very distinctive form of self-identity, and by staying true to their core ethos – of using technical constraint as a mechanism through which to explore creative expression – they demonstrate that there is value in creative ideas distilled down to their most fundamental form and expressed well.

Few members of the chipscene would disagree that a big part of chiptune’s appeal comes from it being unconventional and musically heterodox, both in terms of its production and its sound, and these are characteristics that I think demonstrate that these scenes are distinctive and well-established subcultures. Chiptuners and demosceners are quite happy to occupy that space. After all, as one unnamed blogger is quoted as saying in an article in the Pittsburgh Post-Gazette: ‘Hoping to god this genre never goes mainstream. It’s too [expletive] brilliant to get run over by the masses.’Footnote ⁵³

Introduction

Junko Ozawa was born in 1960 in Saitama prefecture. After attending the Musashino College of Music as a student in the Musicology Department and graduating with a major in Instrumental Piano, she joined the Namco Corporation in 1983 (now Bandai Namco Entertainment). The first game she worked on was Gaplus (1984, the name was later changed to Galaga 3 in the United States), and following this game, she was in charge of the music for The Tower of Druaga (1984, for which she also wrote the sound driver) and Rolling Thunder (1986), amongst several other games. She was also responsible for porting some of her game music from the arcade to the Famicom versions (e.g., The Tower of Druaga), and further created the music for games that Namco developed for other companies, including Nintendo’s critically acclaimed rhythm game Donkey Konga (for the Nintendo GameCube, 2003). She is credited alongside Toshio Kai, Nobuyuki Ohnogi, Yuriko Keino and Yuu Miyake for the song ‘Katamari On Namco’ on the Katamari Damacy – Touch My Katamari Original Sound Track 2 soundtrack release of the PlayStation Vita title Touch My Katamari (2011). Since leaving Namco in 2008, she has continued to compose music, alongside giving piano performances and doing a variety of musical activities.

The music she created for Gaplus and The Tower of Druaga was included in the second game music album ever made. Produced by Haruomi Hosono (famous member of the Japanese synthpop and computer music pioneers Yellow Magic Orchestra) and released in 1984, the album entitled Super Xevious contained music by Namco’s female composers: Ozawa-sensei and the other major composer, Yuriko Keino. The success of this album and its predecessor Video Game Music released earlier the same year started the game music boom in Japan.Footnote ¹ Streets of Rage composer Yuzo Koshiro, who is also often lauded for having an influence on 1990s electronic dance music, mentions Ozawa’s The Tower of Druaga soundtrack in a video posting as one of the major inspirations for his own work. In 2016, the Danish indie rock duo The Raveonettes dedicated the song ‘Junko Ozawa’ to her and her work. Ozawa-senseiFootnote ² herself appeared in Nick Dwyer’s Diggin’ in the Carts documentary series, in which she also talks about her work at Namco and presents her own wavetable library. While other major manufacturers worked with PSGs (programmable sound generators, as documented elsewhere in this book), Namco’s customized arcade game boards created sound with built-in wavesound generators (WSGs), including the C15, and the later model C30, which featured a dedicated 8-channel wavesound generator. This allowed for the use of customized waveforms that Ozawa-sensei and her fellow Namco composers used to create a distinct and recognizable Namco sound. As is the case for many other female Japanese composers (see Chapter 21 by Lemon and Rietveld in this book),Footnote ³ her musical impact and legacy is still to be fully acclaimed in academic research.

Note from the editors: The following interview with Ozawa-sensei was conducted in the Japanese language, then translated back into English and brought into the form of an article. This would not have been possible without the dedicated and outstanding help and translations of Lyman Gamberton, to whom the editors owe their deep gratitude.

Becoming a Video Game Music Composer

The first time I touched a piano was when I was two years old. My sister, who is four years older than me, had begun to learn piano and so my parents brought one home. Although I begged my parents to let me take lessons, because I wanted to learn the same way my sister did, I was still too little and wasn’t allowed to study properly. Imitating the pieces that my sister played, I played them in my ‘own style’; and from around the time I was four years old, I was able to study under the guidance of a music teacher.

When I was in high school, a friend invited me to join a band they had started. I was the keyboardist. Although I only played one concert during my high school years, I resumed band-related activities from the time I started at university, and gradually I began to compose my own melodies. My harmony classes at university were also really interesting, and it was there that I developed a further interest in composition.

Because I majored in music at university, I wanted to have a job where I could use my musical skills and knowledge. I found Namco Ltd. through my search for companies that employed music majors. I was already familiar with them via their Micromouse events.Footnote ⁴ I thought I would be able to do interesting work there and so I sent in a job application. At that point in time, I was willing to do anything as long as it involved music – I didn’t expect it to be video game music.

My first completed project was Galaga 3. While my senior (senpai)Footnote ⁵ was teaching me, I was able to compose slowly and by trial and error; there were barely any impediments. The method of entering musical data is different to the way a score is written, so I struggled a bit at the beginning – but every day was a fresh experience, and it was a lot of fun to, for example, create 60-second sound clips.

I think that recent video game music does not greatly differ from TV/film music and so on, but when I was just starting out as a composer, game music was decisively different. For game music at that time, the whole point was that the ‘performer’ was a computer. Computers’ strong point is that they can play music fast and accurately: the sound was completely different to what musical instruments had been able to do up until then. There is only a slight electronic tone to the sound. Because the number of output sounds is limited to three or eight, if a different sound occurs in the middle of the song, the vocal portion within the melody is completely cut out.Footnote ⁶ That is characteristic of this kind of game music.

The most difficult thing is always being under pressure to meet a deadline. The music is created after the character development has finished: although the last part of the game’s development period is busy no matter what, because the programmers and the art/design team’s appointed deadlines get delayed, I have to make the music in what little time (barely any) I have before my deadline. The thing that brings me greatest joy is that many people all over the world are hearing the music I’ve composed as they play.

I can’t really say if my musical style has been influenced by any particular musician, but an artist I do like is Stevie Wonder. If we’re talking about composing video game music specifically, I think the greatest influences for me were songs from the various anime series I often watched on TV as a child.

Starting to Work at Namco

At that time, the places where you could play video games (they were called ‘game centres’) had a very bad image. They were the kind of places that elementary and middle schoolers would be told by their teachers and parents to avoid because they were dangerous – and since my parents were both teachers, I received fierce opposition, especially from my mother. While this was going on, Namco was working as a company to improve the image of the games industry. At that time, I thought the creative environment inside Namco was something really amazing. The latest computer equipment was set out for us, and there were very expensive musical instruments that you could not buy as an individual collector. Working every day with these great instruments, being told that I could use whichever ones I wanted, was my dream.

At first it was ‘one composer per game’ in terms of composition, but as the games got bigger, the number of melodies also increased, so we eventually composed as a group. The first thing I considered was songs that suited the game. The game’s ‘image’ grew out of its time period, location, content etc. as I composed. When I first joined the company, I was hyper-aware of trying to make music that would not fall short of the expected standard, because the people who were already ‘senior’ composers were making really fantastic video game music.

Waveforms and Sound Driver

For the waveforms, I created software that added additional waveforms, or I analysed the waveform of the sound created by a synthesizer, using an oscilloscope or an FFT analyser.Footnote ⁷ Anyhow, amongst the many prototypes I’d created, I had about twenty that looked potentially useful. As for the sound driver, since my senior already had one, I kept reading and referencing it, and I added specifications that were easy for me to use. Since I understood how the programming language worked, it was not very difficult at all. Of course, that is because the driver my senior had made was already there; if I had had to make it from scratch, I think it would have been more of a struggle. When I created a sound driver by myself, it was satisfying to be able to put all of my own ideas into practice straight away, such as being able to change the tone in detail.

Because The Tower of Druaga is set in ancient Mesopotamia, neither techno nor contemporary music seemed appropriate. I thought that it should be classical music, something with a slightly ‘Ancient Egyptian’ atmosphere. When the knight character comes out, I thought it would be good to create a brave and solemn atmosphere. The melody is the image of the knight gallantly marching forwards.Footnote ⁸ At that time, the sound integrated circuit could not output more than eight sounds including the sound effects, so I did not plan to write an orchestral melody. But I can’t help thinking that the sounds transformed into an orchestra inside the listener’s head …

The Status of Video Game Music in Japan in the Early 1980s

Regarding video game music at that time, I think that even the video game production companies themselves weren’t very aware of it. Of the companies that made video games, Namco was very aware of the importance of video game music – they recruited composers as company employees and composed original melodies, but it was also a time when the machines made by other video game companies were playing ‘ready-made’ songs without any repercussions. (There should have been copyright issues.) Mr Haruomi Hosono, of the Japanese techno-unit Yellow Magic Orchestra, produced Namco’s Xevious as a musical album, but that was the very first album of video game music anywhere in the world.Footnote ⁹ In terms of how the record was marketed, it was put with the film soundtracks and in the techno music corner, since game music didn’t yet exist as a genre. These days, first-class orchestras that usually play classical music often do eventually include video game music in their concert repertoires. It seems more people come to the video game music concerts than to the classical performances, and the orchestras say they want to continue doing these kinds of concerts in the future. The orchestras are happy because adults come along with kids for the video game music programmes.

About the Albums Super Xevious (1984) and The Return of Video Game Music (1985)

For both of those albums, the producer (Mr Haruomi Hosono) selected and arranged his favourite songs. Once everyone had decided the order of the songs, the engineer created the record by cutting and pasting songs from the game’s circuit board. For Super Xevious, there is a special BGM (background music) for the player name and high score in Gaplus and The Tower of Druaga. As for The Return of Video Game Music, my essay is written inside the album notes.Footnote ¹⁰ At the time, we discussed it with the director: we decided that the game music we had composed would be the first A-side. For the B-side, we decided to put in the music that we had made but had not ended up using in a game and that we personally liked. Later, the B-side compositions by Ohnogi-sanFootnote ¹¹ were used in a game after all.

To record the music I brought the whole game machine to the recording studio, attached a clip cord to the foot of the ICFootnote ¹² on the circuit board, and recorded it with as little noise as possible before the sound was amplified. During the break, I watched the producer, sound engineer and other people fight over who got to play the game. I think these kinds of records were starting to be produced because video game music had become trendy worldwide, and because there was increasing focus on its originality.

The Current State of Video Game Music Education

I do not think there needs to be a specialist BA or other university degree, but I do think technology and research are necessary for attaching sound to image in the same way for films and television – so a class on coordinating image and music in video games could also be good.

If video games as a medium are an important part of popular culture, then I think video game music is certainly also important. In the future, I think games will not just be a part of ‘culture’, but will also play a role in physical convalescence and in medical treatment, and that they will become more interlinked with human life in general.