Trackers: The Rise, Bloom and Later Developments of a Paradigm

trackers, music software, chipmusic, home computers, demoscene

Markku Reunanen
markku.reunanen [a] aalto.fi
Senior university lecturer, Aalto University
Docent, University of Turku

How to cite: Reunanen, Markku. 2024. ”Trackers: The Rise, Bloom and Later Developments of a Paradigm”. WiderScreen Ajankohtaista 5.8.2024. http://widerscreen.fi/numerot/ajankohtaista/trackers-the-rise-bloom-and-later-developments-of-a-paradigm/

Tulostettava PDF-versio

First published in Finnish in Musiikki 2–3/2019 as ”Trackerit: paradigman synty, kukoistus ja myöhemmät vaiheet”. https://musiikki.journal.fi/article/view/87867. Link to errata.

Trackers are music software whose history spans over three decades. They are tightly linked to the history of home computers, and they made previously unreachable digital composition tools available to enthusiasts in the 1980s. Over the years, trackers have been used both for commercial and non-commercial purposes, for example games and demos, and they have given rise to lively hobbyist communities. In this article, I go through the history of trackers, their characteristics and typical uses. The study is based on the analysis of 60 tracker programs and six interviews of their creators. The findings highlight the tight link between trackers and computer hardware, the reasons why authors wanted to create such software, and the gradual evolution of the paradigm.

Introduction

Tracker programs, or more commonly just trackers, are music composing software which were first published for the 1980s’ popular home computers, such as the Commodore 64, Amiga and Atari ST (see Saarikoski 2004). In the big picture, trackers can be placed in a larger historical continuum where expensive, professional studio technology gradually entered homes and became available for hobbyists. Now, in 2019, the paradigm is still well alive: new software versions are published regularly and there are various active tracker user communities.

Thomas Kuhn (1962) has defined paradigms as practices that define a certain field of study at a given time. Even though it is debatable whether trackers can be considered a field of study, the concept is still useful, as their use is in many respects a ”paradigm” with its distinctive tools, practices and terminology – even a holistic lens through which a composer understands music at large. In addition, Kuhn’s model of a paradigm shift is also applicable here, as trackers replaced earlier approaches to digital composition and were eventually challenged by other competitors representing a different paradigm.

Tracker software is tightly linked to several parallel developments, such as the history of computing, digital games, the demoscene and other hobbyist communities, as well as music production. Therefore, it is natural that existing research dealing with trackers also represents different fields of study. Technology, history and cultural studies have so far been the most common perspectives, while there is a clear lack of analysis of tracker music itself. Hanna Lönnblad’s journal article and her later master’s thesis on demoscene music are rare examples of such an approach (Lönnblad 1997; 1998).

Chipmusic – music that is either composed for or mimics the characteristic sound of vintage computer and game console sound chips – is not necessarily tracker music, but there exists a notable overlap between the two. Nils Dittbrenner’s Chip-Musik (2007) and Kenneth McAlpine’s Bits and Pieces (2019) are large overviews on the essence of chipmusic (also known as chiptunes), whereas the cultural side of things has been studied by Carlsson (2008), Driscoll and Diaz (2009), Karila (2013) and Polymeropoulou (2014). Sebastian Tomczak’s doctoral thesis, On the Development of an Interface Framework in Chipmusic (2011), is an in-depth look into the features of vintage sound chips and how they can be connected to modern computers.

Demoscene, the enthusiast community focusing on computer demos, has been both an active user and a creator of tracker software. Reunanen (2010, 66–70) discusses tracker music as a common format for demo music and Jimmy Maher (2012, 171–205) dedicates several pages to the iconic ProTracker in his platform studies book The Future Was Here: The Commodore Amiga (for more on platform studies, see Bogost & Montfort 2009). Lönnblad’s (1997; 1998) studies also took place in the context of the demoscene. Esa Hakkarainen’s (2011) An Overview of Retro Game Music mentions demos, too, but otherwise the focus is on retro themes and game music. So far, the use of trackers for commercial game music production has been explored little, even though there was a heavy connection between the two in the 1980s and 1990s, as discussed later in this article.

Trackers and related hobbyist cultures are too colorful to cover in depth in just one article. Therefore, I have limited the scope here to the following two, closely related research questions:

1. How has the tracker paradigm evolved over time?
2. For what purposes have trackers been created and used?

The research material consists of two main parts: tracker software and interviews with tracker creators. At first, I compiled a list of 60 tracker programs with their basic information (platform, release year, author, screenshot, and basic features), after which I chose the most influential or exceptional examples for further in-depth analysis, which could be labeled as close reading. Sofware studies is a relatively new approach in the field of digital humanities and often technical in nature, even though it is equally possible to study computer software as culture (Manovich 2009; 2013; Fuller 2008; see Haverinen and Suominen 2015). Even though the source code of the selected trackers was not part of the research material, my own previous programming experience with tracker music playback served as a useful foundation which provided extra understanding of the inner logic of tracker software. To explore the creators’ point of view, I conducted six semi-structured interviews, coupled with excerpts from three interviews conducted by others (Amiga Music Preservation n.d.; Sidmusic.org n.d.; Game Audio Network Guild 2011).

Defining Features of Trackers

Composing music using a tracker differs notably from other composition and notation paradigms. There is no stave or a piano roll view like in sequencers, such as Cubase. Next, I will go through the most relevant concepts in order to provide the reader with an understanding of what defines a tracker and how it is to compose music with one. It should be noted that there are considerable differences between trackers, so the discussion here focuses on the core features that constitute a tracker in the first place.

At the heart of tracker software is the track, a column where the composer enters notes, instrument numbers and various commands (see Figure 1). The tracks proceed vertically from top to bottom, and each typically represents an independent sound channel. Especially in the case of home computers, such as the Commodore 64 (C-64) or Amiga, the mapping is one to one with the channels offered by the sound chip. Along with increasing computational power this hard coupling eventually vanished, which will be discussed in more detail in the next section. When entering notes on the track the computer keyboard behaves like a double piano keyboard: letters and numbers correspond to notes which can be either typed manually or played in real time.

A full song consists of parts called patterns. The traditional length for a pattern is 64 rows, which gives a certain direction to music created using trackers. Each pattern comes with a number, which is placed on a list called song order. Part of the song order can be seen in the top left corner of Figure 1 (Pos, Pattern, Length). The possibility to repeat a pattern saves effort, as there is no need to re-enter the same notes. In addition, on vintage computers there were also technical reasons for it, as reusing the same content saved precious memory.

In addition to notes, tracks contain instruments, which refer to the sounds that are to be played back. An important difference between trackers is whether the sounds are synthesized either using the sound chip or computationally, or whether the sound is digitized from an external source. Roughly speaking, 1980s’ early 8-bit home computers (C-64, MSX, Amstrad CPC etc.) rely on their sound chips, whereas Amiga and later PC trackers are based on samples (cf. Dittbrenner 2007; Carlsson 2008; Tomczak 2011). This division is inevitably blurry, as even old machines might be capable of limited sample playback and trackers on new machines may feature also real-time synthesis. Typical features of an instrument are its volume, finetuning, looping, envelope and panning – differences between programs are particularly pronounced when it comes to instrument handling.

The third kind of content that goes on a track are the commands which can be controlled by using parameters. The parameters are traditionally given in the hexadecimal notation (as opposed to common ten-based decimal numbers, they are based on 0–F), so for example the number 255 is written as FF. Most likely, the original reasons for using the hex notation were easier implementation and saving some screen estate. As can be seen in Figure 2 with NinjaTracker, a tracker musician often works with plenty of numerical values. Table 1 contains the commands supported by ProTracker. Three main categories can be recognized among them: commands dealing with volume, pitch and song playback.

0 – arpeggio	E0 – filter on/off (Amiga hardware feature)
1 – portamento up	E1 – fine pitch slide up
2 – portamento down	E2 – fine pitch slide down
3 – tone portamento (to a note)	E3 – glissando control (sliding type)
4 – vibrato	E4 – vibrato control (modulation curve)
5 – tone portamento + volume slide	E5 – set fine tune
6 – vibrato + volume slide	E6 – pattern loop
7 – tremolo	E7 – tremolo control (modulation curve)
8 – not used	E8 – not used
9 – sample offset (jump in playback)	E9 – retrigger note
A – volume slide	EA – fine volume slide up
B – position jump (jump in song)	EB – fine volume slide down
C – set volume	EC – note cut
D – pattern break	ED – note delay
E – subcommands (see right column)	EE – pattern delay
F – set speed	EF – invert loop/funk repeat (non–standard)

The concept of tempo is distinctive to trackers and slightly differs from standard music terminology. On vintage devices, music playback is usually heavily tied to the screen refresh rate (50 Hz in Europe). It may have been the only technical means to get precise timing to begin with, in addition to which it matches well with game and demo programming which, too, are dependent on the same screen refresh cycle. For instance, on the Amiga the default BPM (beats per minute) is 125, which follows directly from the screen refresh rate: every sixth screen refresh the playback proceeds to the next row of the pattern, which correspond to 1/16 notes, so four rows constitute a beat. ProTracker’s command F lets the composer roughly alter the playback speed by changing the screen refresh counter, and later versions added the possibility to control the tempo directly using BPM values.

The last fundamental concept that goes with trackers is the module. The music composed by an artist must be saved in a format that can be distributed for further use by others, which has given rise to a number of file formats. The ”mod” file extension, used by the influential Soundtracker and ProTracker plus several other programs, has been so ubiquitous that tracker music has even been referred to as ”mods” (see McAlpine 2019, 142–4). Other common file formats are, for instance, Fasttracker II’s XM and Scream Tracker’s S3M. Modules are self-contained with all the necessary content for playing back the tune: the tracks, song order and instruments with their samples.

Many of the features seen in tracker software directly mirror the capabilities of the underlying hardware platform. For example, ProTracker’s E0 command switches the built-in lowpass filter on and off on Amiga computers. Yet, other features and terminology clearly originate from more general music theory and practices. On the one hand, programs have been used for certain purposes, which has shaped their feature set and, on the other hand, it is possible to recognize cultural reasons, as some functionality lives on mainly for historical reasons even if the original technical limits are long gone. In the next two sections, I address these factors in more detail: the history of trackers, their uses and creators.

Devices, Software and Pieces of a Family Tree

Even a quick glance at trackers from different decades and platforms reveals how features have been inherited from one program to another and how they have evolved over time. New trackers hardly ever radically differ from their predecessors, but they rather represent gradual evolutionary steps where technical or musical features are added, programming errors are fixed and usability is improved. These relationships form a family tree of sorts. It is, however, hard to explicitly lay out the tree, as features have been borrowed horizontally and not all the design decisions are made consciously.

As Petri Saarikoski (2004, 78) notes, computers were only one example of commodified electronics, such as calculators, electronic games and digital clocks, which become household items around the same time. Early 1980s’ ”microcomputer boom” machines were technically very different to today’s devices: their memory capacity and processor power were modest, and a typical sound chip could play back a few channels of simple synthesized audio. Graphics – if any – were of low resolution with few colors and the most common input devices were the keyboard and joystick (see Saarikoski 2004, 98–113; Forster 2005).

According to current knowledge, the first program that could be considered as a tracker is Soundmonitor programmed by Chris Hülsbeck for the Commodore 64 in 1986 (e.g. Game Audio Network Guild 2011; see Suominen & Sivula 2016). As can be seen in Figure 3, Soundmonitor already had several familiar features, such as the vertical tracks corresponding to audio channels, effects and plenty of numbers. The term ”monitor” and heavy use of hexadecimal numbers in the user interface bear resemblance to other tools of the time, namely machine language monitors, which were used for manipulating memory contents. Hülsbeck himself mentions analog synthesizers as his source of inspiration, as well as other C-64 musicians of the time and Steinberg’s early MIDI sequencer, which was the origin for the notation (interview with Hülsbeck 23.4.2019).

Even if tracks were already clearly present in the program, the term ”tracker” entered common use somewhat later along other software. Hobbyists soon started modifying Soundmonitor to suit their own needs: Rockmonitor II (1987) is an early example of tracker evolution. The Commodore 64 faded from the market toward the 1990s, but devoted enthusiasts still develop new software for it. The aforementioned NinjaTracker and Mats Andrén’s defMON (2013) are more recent trackers running on actual hardware (cf. Lindsay 2003).

The Commodore 64 was the birthplace of trackers, but their golden age was rather on Commodore’s next computer platform, the Amiga. The Amiga (nowadays often called Amiga 1000), introduced in 1985, was a highly multimedia-capable device compared to its contemporaries. The following cheaper model, the Amiga 500, achieved success also in the Finnish home market (Saarikoski 2004, 389). From a musical perspective, the main difference compared to earlier home computers was that the Amiga could play up to 28.8 kHz digitized 8-bit audio on four independent channels, which were grouped into two stereo pairs. Therefore, music on the Amiga is commonly based on sample playback at different pitches.

To walk through all the Amiga trackers would require an article of its own, so I focus here on some typical examples only. Karsten Obarski’s (The) Ultimate Soundtracker, also known simply as Soundtracker, from 1987 is the oldest of them. Obarski mentions that he knew of Soundmonitor, but also used C-Lab’s ScoreTrack, which may have affected the initial design decisions (Amiga Music Preservation n.d.). As can be seen in Figure 4, Soundtracker notably differs from its predecessor: instead of a textual interface you can use a mouse and click buttons. Overall, the program already looks like a typical tracker, and its features can still be recognized in software that came out thirty years later. As a disappeared curiosity the tracks have fixed names to indicate their expected – even if unnecessary – use (Melody, Accompany, Bass, Percussions).

Pex ”Mahoney” Tufvesson developed his NoiseTracker in 1989 using Soundtracker as the starting point. His motivation was to fix errors that were left in Soundtracker and, while doing that, he also added new features, such as improved instrument handling (interview with Tufvesson 24.4.2019). The mod file format, which became a standard of sorts on the Amiga, is based on Tufvesson’s additions. These roots are still visible in the ”M.K.” (Mahoney & Kaktus) tag which can be found in such files.

Shortly thereafter followed the ProTracker family of programs, which quickly took over the Amiga tracker domain not only among hobbyists but also game audio professionals. The first version from 1990 was developed by a group called Amiga Freelancers, after which others started modifying the software to their liking. The ProTracker 2 series was developed by Noxious and version 3 by Cryptoburners. When looking at ProTracker’s user interface (Figure 5), it is evident how after Obarski the changes were mainly cosmetic. The information density has increased, partly to improve usability and partly to accommodate for new features (see also Maher 2012, 171–205; McAlpine 2019, 137–42).

Commodore’s dominance at homes started fading at the beginning of the 1990s and, eventually, the company went bankrupt in 1994 (see Saarikoski 2004, 389–95; Bagnall 2005). Home computers were mostly replaced by IBM PC compatibles over the decade, and notable trackers started appearing on them as well. The transition did not, however, happen overnight and the Amiga remained an important enthusiast platform in spite of its commercial hardships, because of its established trackers and sound quality, which was not surpassed before 1992 with the release of the Sound Blaster 16 and Gravis UltraSound expansion cards for the PC, both of which were capable of 16-bit sound.

From a technical standpoint, the Gravis UltraSound (GUS) is like an advanced Amiga: it is possible to play up to 32 individual digitized sounds simultaneously at different pitches. Therefore, extending the existing four-channel model to a greater number of tracks was straightforward. The Sound Blaster 16 was not capable of similar hardware mixing, but had to rely on software implementations instead, which was not feasible before the processor speeds eventually started facilitating it. Similar attempts had been made on the Amiga even earlier, as Oktalyzer (1989) had low-fidelity support for eight channels. It is evident how the need for more than four channels was recognized early on, even if it took a few years more for the hardware to catch up. Finnish OctaMED acquired a similar feature in 1991 and proper multichannel mixing in 1996, when more powerful next generation Amigas became capable of it (Kinnunen 1999).

The first PC-based trackers, such as Scream Tracker 2.2 (1990), ModEdit (1991) and Whacker Tracker (1992) were clearly behind their Amiga counterparts in usability, features and sound quality. The situation started changing in 1994 when Scream Tracker 3 (ST3) and Fasttracker II (FT2) were released within the demoscene. Out of these two Fasttracker II is clearly a closer relative to Amiga trackers: the tracks, song order and commands resemble ProTracker to a high extent (Figure 6). New soundcards, increased memory and improved processors, however, facilitated new features, such as up to 32 channels, 16-bit samples, free panning and volume envelopes for the instruments. FT2, in turn, inspired later software: for example, the PC-based Skale Tracker (2005) and the multiplatform MilkyTracker (2005).

Scream Tracker 3 (Figure 7) can, likewise, be recognized as a tracker but it stands out more from the mainstream of the time. The interface is not operated with a mouse and there are separate views for the tracks, song order and other features, which resembles the earlier version of Scream Tracker and even Soundmonitor, rather than Amiga trackers. The commands are largely in line with their ProTracker counterparts, albeit with a different numbering. As with FT2, there are as many as 32 channels available, even though their precision is limited to 8 bits. The legacy of ST3 can be observed in later software too, in particular Impulse Tracker (1995) and Schism Tracker (2003), both of which highly resemble their predecessor (cf. Lönnblad 1998).

Renoise (2002) is, in many respects, a prime example of how far trackers can be stretched. It is a commercial digital audio workstation (DAW), which supports, for example, VST plugins, several sound file formats, scripting and multichannel audio devices. Many of these features are comparable to and familiar from other competing professional audio software, but where Renoise stands out is its adherence to the tracker paradigm.

Some of the trackers that constitute the research material do not conveniently fit into any historical continuum or family tree. Little Sound DJ (LSDJ, 2000) runs on the tiny screen of a Nintendo Game Boy handheld console and is operated using its directional pad and buttons. On the one hand, it is easily recognizable as a tracker, but on the other hand, its features are so closely tied to the particularities of the device that it highly differs from more common computer-based software (McAlpine 2019, 186–8). An emerging trend of the 2000s is cross-development, meaning that music is composed on one platform but used on another, whereas traditionally the two were the same. In my research material examples of such are GoatTracker (2001), which is used for composing for the C-64, and Arkos Tracker (2010), which targets early 1980s’ MSX, Spectrum and Amstrad CPC computers. Both of them run on modern computers, but the outcome works on vintage hardware. Similar traits can be observed in software development, too:

defMON is coded using TextMate on Mac, the ACME cross assembler, and makefiles. I’ve also used some different solutions for executing the assembled file directly on the C64, such as Graham’s CODENET and more recently a similar solution that uses the 1541U2 hardware. The very first versions of the player were coded in TASM on the C64 though, but it didn’t take long until I migrated to a cross assembly environment. (interview with Andrén 24.4.2019.)

Even though the final product, defMON, itself runs on an actual C-64, Andrén prefers the comfort of a modern computer for the development. Themes such as authenticity, skill and authorship are frequently discussed in the context of retro computing, for example chipmusic (Polymeropoulou 2014), hardware hacking (Lindsay 2003) and the demoscene (Reunanen 2010, 34–6).

In this study, I will omit so-called players, although there is at least a similar amount of such software – for just listening to tunes there is seldom a need to use the original tracker, which would be clumsy for that purpose to begin with. In addition to sharing and listening, songs are also used as content in multimedia productions, such as games and demos. To get a tune playing outside of trackers or players requires even more software, namely programming examples and libraries, which improve – or limit – the utility of music composed with a certain tool. It is important to realize how trackers do not exist in isolation but are part of the larger sound ecosystem of a platform, which consists of much more than musical composition.

Creators and Purposes

To answer the second research question, let us give the voice to tracker software creators: who programmed them, why and for what purposes? Based on credits found on trackers and their documentation, it is evident that most trackers have been created by either a single person or a small group of people. All the interviewees of this study had designed and implemented their programs chiefly alone, although they had received help, inspiration, comments, source code and testing support from others. Some of the trackers were credited to a demo group, even if there was effectively just one author.

All the interviewees had some prior experience with music before starting their tracker project: in particular, piano playing was a recurring theme, and most had experience of other instruments, too. As to their computer hobby, playing games, watching demos and programming were the most common ones and, thus, we could say creating a tracker brought together several of their already existing interests. Lasse Öörni crystallizes plenty in his short answer: ”Gaming, drawing, coding. First as a hobby and finally (game) programming also as work.” (interview with Öörni 23.4.2019)

When asked about the motivation to start creating a tracker – a challenging and time-consuming project – programmers often mentioned a personal need that was not satisfied by any existing tool. Chris Hülsbeck and Lasse Öörni needed game music for the Commodore 64 (interviews with Hülsbeck and Öörni 23.4.2019), and, in the same lines, the first tune composed by Karsten Obarski with his Soundtracker ended up in a 1987 ball game called Amegas (Amiga Music Preservation n.d.). An equally common reason was that existing tools were found lacking or of low quality. Quoting Pex Tufvesson: ”Since the previous versions of Soundtracker contained a lot of bugs, I felt the urge to fix them. And when they were fixed, I took the opportunity to add a couple of features I thought were missing!” (interview with Tufvesson 24.4.2019) General interest in the topic provided extra motivation, and for Julien Névo creating Arkos Tracker 2 was a conscious learning project to improve his C++ language skills (interview with Névo 24.4.2019).

The history of trackers is heavily intertwined with the demoscene, as several important trackers have been created in that context and for its needs (e.g. Reunanen 2010, 66–70). For instance, NoiseTracker, ProTracker, Fasttracker, Scream Tracker and Arkos Tracker could be categorized as such programs. Most of my interviewees had participated in the demoscene one way or another; Tufvesson as a still active author (interview with Tufvesson 24.4.2019). Demo background music is an evident use for tracker music, but at least equally important are various music competitions organized at demo events, parties, and nowadays also on online forums (see Ratliff 2007; Reunanen 2010, 37–9).

Games and tracker music, likewise, have a long shared history, as we have seen above. Their relationship is not quite as straightforward, though, as even in the presence of Soundmonitor and its derivatives game music was still commonly composed by directly programming it in symbolic machine language, assembly (Sidmusic.org n.d.; interview with Hülsbeck 23.4.2019). Although from a musician’s perspective there are notable advantages to using a tracker, such as interactivity, the paradigm shift did not happen immediately (cf. Kuhn 1962). Another factor is that Soundmonitor was released only in 1986, when the C-64 had already been in the market for more than three years. The peak of tracker music was rather on the Commodore Amiga in the late 1980s and early 1990s, as confirmed by the World of Game MODs collection, which features about 10 000 tunes – most of them extracted (”ripped”) directly from games (World of Game MODs 1999). Increased storage capacity, such as the CD-ROM format, and generally improved multimedia capabilities eventually made it possible to use CD-quality audio in games, after which game music largely diverged from the tracker world (Game Audio Network Guild 2011; cf. Ratliff 2007).

Mentions of trackers in other professional uses other than games were rare in the answers. Pex Tufvesson said that he had received thanks from people working in the music business who had kickstarted their career on NoiseTracker (interview with Tufvesson 24.4.2019). Teijo Kinnunen mentioned that Calvin Harris created his first published album (I Created Disco, 2007) using OctaMED (interview with Kinnunen 28.4.2019). In both cases, there is a similar narrative where musicians move ”forward” from trackers to other tools. The Renoise (n.d.) website features a lengthy list of its users ranging from producers to DJs and composers. In this particular case, we are dealing with marketing material, which needs to be considered when evaluating the claims.

The third link between trackers and business is the fact that some of them were created as commercial products. Renoise is an evident example of such, but also Soundtracker and Oktalyzer were sold. Teijo Kinnunen’s MED was initially distributed for free, but after the British company RBF Software made him an offer, the following versions titled OctaMED became commercial (interview with Kinnunen 28.4.2019). Most trackers, however, have been published for free use by the community – often there is even no explicit mention of it, as it has been such standard practice. This kind of liberty is different from the free software movement, which has also gained traction among tracker developers: GoatTracker, MilkyTracker and Fasttracker II Clone (2017) represent such recent ethics (see Vuorinen 2007).

When the project that started out of curiosity and personal need becomes public, the role of the creator changes. All my interviewees had received user feedback, which had been mostly positive. Typically, the thanks were coupled with development ideas and feature requests: ”At times there have been a lot of feature requests, and sometimes I’ve just had to answer that the source code is available, you can do it yourself.” (interview with Öörni 23.4.2019) Teijo Kinnunen related very similar experiences:

A while ago I tried OctaMED on an emulator and found it amusing how I had added this and that feature based on user requests. Some of them were probably never used… But at the time I didn’t yet have enough experience on software design so that I could have just said ”no” to feature requests, if there didn’t seem to be a real need for them. (interview with Kinnunen 28.4.2019.)

Whether you fulfill the requests or not, the increasing popularity of a tracker brings the author closer to the community when private development becomes public and when the tracker starts to become an important tool for others as well. As we can observe in the interviews above, the planned and actual uses for a tracker are not always in line: programs originally developed for game sound have been used for demos and vice versa. The final phase in the life cycle of a tracker is its end, when the author loses interest or the users migrate elsewhere.

Conclusion

Studying trackers revealed how they are, first and foremost, cultural artifacts that reflect their time, such as the computer market, software development, videogames, popular music and hobbyist cultures. The first trackers were among the first publicly available interactive software that let musicians compose on home computers without extensive programming skills. Even if trackers do not currently represent mainstream music production, they still hold a strong position among enthusiast communities, such as the demoscene and chipmusic circles.

We can recognize a few different periods and turning points in the history of trackers: the birth of the paradigm on the Commodore 64, the popularization and formation on the Commodore Amiga, and, finally, the improvements in sound quality and channel count brought about by later hardware development. Few programs have been completely new and revolutionary, but rather evolutionary steps over their predecessors – even Obarski’s Soundtracker can still be recognized in software created thirty years later. The existing culture of use has, on its behalf, shaped the developments with factors such as familiarity, tradition and audience expectations in play.

Programmers have started developing trackers, first and foremost, out of their personal interest and needs: for example, to compose music for their games or to fix the shortcomings of existing tools. The popularity of a tracker forms a user community around it, and the role of the programmer changes when there is a userbase with their requests to consider. Games and demos have been traditional homebases for tracker music, coupled with other enthusiast communities and, at least to a certain degree, professional musicians.

In this article, I have studied the history of trackers, their creators and uses, while several other perspectives would still warrant further attention. As mentioned in the introduction, there is already some research on the social aspects of tracking, but less so on their actual use: how is it to compose music using a tracker? From a software studies perspective there would be room for further depth, as I have here remained on the surface consisting of visual appearance and features, instead of considering their technical implementation or relationship to music technology at large. Furthermore, the analysis of actual tracker music is still an effectively untouched subject.

To conclude, let us go back in time to the end of the 1980s, forget about our current digital tools, and sit down for the first time in front of a tracker to marvel how the home computer is opening a new world for us:

Making music, of course. But more importantly, which you tend to forget these days, the Commodore Amiga and Noisetracker was actually a very cheap way of being able to experiment with sample based music, and learning the tricks of the trade. (interview with Tufvesson 24.4.2019.)

Acknowledgements

Thanks to Petri Saarikoski and Yrjö Fager for their comments, and the Academy of Finland for funding the Centre of Excellence in Game Culture Studies (CoE-GameCult, decision 312395).

References

Interviews

Interviews conducted by Markku Reunanen.

Chris Hülsbeck, 23.4.2019.

Julien Névo, 24.4.2019.

Lasse Öörni, 23.4.2019.

Mats Andrén, 24.4.2019.

Pex Tufvesson, 24.4.2019.

Teijo Kinnunen, 28.4.2019.

Websites

All websites accessed 1.11.2019.

Amiga Music Preservation. n.d. ”Interview: Karsten Obarski”. http://amp.dascene.net/detail.php?view=3982&detail=interview

Game Audio Network Guild. 2011. ”Interview with Chris Huelsbeck”. 29.4.2011. http://www.audiogang.org/interview-with-chris-huelsbeck/

Kinnunen, Teijo. 1999. ”A Brief History of OctaMED”. Accessed through the Internet Archive. http://stekt.oulu.fi/~kinnunen/omhist.html

Renoise. n.d. ”Artists | Renoise”. https://www.renoise.com/artists

Sidmusic.org. n.d. ”Interviews with Rob Hubbard”. http://www.sidmusic.org/sid/rhubbard.html

Literature

Bagnall, Brian. 2005. On the Edge: The Spectacular Rise and Fall of Commodore. Winnipeg: Variant Press.

Bogost, Ian & Nick Montfort. 2009. ”Platform Studies: Frequently Questioned Answers”. In Proceedings of the Digital Arts and Culture Conference 2009. Irvine, CA: University of California. https://escholarship.org/uc/item/01r0k9br

Carlsson, Anders. 2008. ”Chip Music: Low-tech Data Music Sharing”. In From Pac-Man to Pop Music, edited by Karen Collins, 153–62. Farnham: Ashgate. https://doi.org/10.4324/9781351217743

Dittbrenner, Nils. 2007. Chip-Musik: Computer- und Videospielmusik von 1977–1994. Osnabrück: Osnabrück University.

Driscoll, Kevin & Joshua Diaz. 2009. ”Endless Loop: A Brief History of Chiptunes”. Transformative Works and Cultures 2/2009. https://doi.org/10.3983/twc.2009.096

Forster, Winnie. 2005. The Encyclopedia of Game Machines: Consoles, Handhelds & Home Computers 1972–2005. Utting: Gameplan.

Fuller, Matthew (ed.). 2008. Software Studies: A Lexicon. Cambridge, MA: MIT Press.

Hakkarainen, Esa. 2011. ”Katsaus retropelikonemusiikin maailmaan” [An Overview of Retro Game Music]. WiderScreen 14 (1–2). http://widerscreen.fi/2011-1-2/katsaus-retropelikone-musiikin-maailmaan/

Haverinen, Anna & Jaakko Suominen. 2015. ”Koodaamisen ja kirjoittamisen vuoropuhelu? – Mitä on digitaalinen humanistinen tutkimus?” [Dialogue Between Coding and Writing? – What Are Digital Humanities?]. Ennen ja nyt 15 (2). https://journal.fi/ennenjanyt/article/view/108634

Karila, Walter. 2013. Pulppuavan saundin kovassa ytimessä: Commodore 64 -kulttuurin symbolinen rakentuminen Lemon64-internetsivuston keskusteluforumilla [In the Hard Core of Bubbling Sound: Symbolic Construction of the Commodore 64 Culture on the Lemon64 Online Forum]. MA thesis. University of Turku.

Kuhn, Thomas. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press.

Lindsay, Christina. 2003. ”From the Shadows: Users as Designers, Producers, Marketers, Distributors, and Technical Support”. In How Users Matter: The Co-Construction of Users and Technology, edited by Nelly Oudshoorn & Trevor J. Pinch, 29–50. Cambridge, MA: MIT Press. https://doi.org/10.7551/mitpress/3592.003.0004

Lönnblad, Hanna. 1997. ”Kahden tietokonedemon vertaileva analyysi” [A Comparative Analysis of Two Computer Demos]. Musiikin Suunta 19 (2): 28–34.

Lönnblad, Hanna. 1998. Tietokonedemot kulttuurina ja musiikin harrastuksen muotona [Computer Demos as Culture and Form of Music Hobby]. MA thesis. University of Helsinki.

Maher, Jimmy. 2012. The Future Was Here: The Commodore Amiga. Cambridge, MA: MIT Press.

Manovich, Lev. 2009. ”How to Follow Global Digital Cultures, or Cultural Analytics for Beginners”. In Deep Search: The Politics of Search Beyond Google, edited by Konrad Becker & Felix Stadler, 198–212. Edison, NJ: Transaction.

Manovich, Lev. 2013. Software Takes Command. London: Bloomsbury.

McAlpine, Kenneth B. 2019. Bits and Pieces: A History of Chiptunes. Oxford: Oxford University Press.

Polymeropoulou, Marilou. 2014. ”Chipmusic, Fakebit and the Discourse of Authenticity in the Chipscene”. WiderScreen 17 (1–2). http://widerscreen.fi/numerot/2014-1-2/chipmusic-fakebit-discourse-authenticity-chipscene/

Ratliff, Brendan. 2007. Why Did Freely Shared, Tracked Music in the 1990’s Computer Demoscene Survive the Arrival of the MP3 Age? MA thesis. University of Newcastle upon Tyne.

Reunanen, Markku. 2010. Computer Demos – What Makes Them Tick? Licentiate thesis. Aalto University.

Saarikoski, Petri. 2004. Koneen lumo: mikrotietokoneharrastus Suomessa 1970-luvulta 1990-luvun puoliväliin [The Lure of the Machine: The Personal Computer Interest in Finland from the 1970s to the Mid-1990s]. University of Jyväskylä.

Suominen, Jaakko & Anna Sivula. 2016. ”Participatory Historians in Digital Cultural Heritage Process – Monumentalization of the First Finnish Commercial Computer Game”. Refractory – Australian Journal of Entertainment Media 27. http://refractory.unimelb.edu.au/2016/09/02/suominen-sivula/

Tomczak, Sebastian. 2011. On the Development of an Interface Framework in Chipmusic: Theoretical Context, Case Studies and Creative Outcomes. PhD thesis. University of Adelaide.

Vuorinen, Jukka. 2007. ”Ethical Codes in the Digital World: Comparisons of the Proprietary, the Open/Free and the Cracker System”. Ethics and Information Technology 9 (1), 27–38. https://doi.org/10.1007/s10676-006-9130-2