Soundtracker Origins, interlude: The coders behind the Cambrian explosion

Temps de lecture / Reading time : 48 minutes.

Remember Part 3¹Wow, that previous part had an estimated 21-minute reading time, and back then I thought that was really too long for an article about my little nostalgic trip down memory lane. Well now, GUESS WHAT? of this series of articles on Soundtracker?
Those were the days!

I fondly remember writing the following, near the end of that article:

The next parts will focus on each of the two parallel paths that I highlighted above:

• The Fairlight CMI path: finding out how they came up with the idea for their Page R sequencer.
• The « Commodore » path, going down the JMS rabbit hole, and their Multitrack Composer.

Will there be more parts? Who knows? I sure don’t!

Me, the hopeful pessimist.

Good times! Look at me, making sincere promises and everything:

• Sure, I could easily write that Fairlight CMI article, mostly just by copy-pasting the content of email answers dating back from 2019.
• And as for the Jellinghaus (JMS) article, along with a couple other ideas thrown in in order to tie it all together, well it’s a quick road to publication, methinks.

But where’s the fun in that?

I say we go the usual ADHD route: stop at 80% completion, and start something new altogether!²But let’s keep it in the same overall context, so as to not feel too much guilt. Structured procrastination: 60% of the time, it works every time.

Let me therefore add an interlude to this series of articles, in order to tell you a story.

A story of pioneers, of sharing, of standing upon the shoulders of giants, of youth, of friendly competition, and of bad blood.
A story of humans making History without even knowing it.

As with all good, heartwarming stories, it starts with…

A graph

Behold!

On November 4th, 2007, Claudio Matsuoka pushed the first commit of a very useful³To me and this silly project of mine, at least., techno-archeological project: the Tracker History Graphing Project.

That graph above presents a « family tree » of all known trackers, from Karsten Obarski’s original in 1987, to Olav « 8bitbubsy » Sørensen’s clones/ports of ProTracker v2.3D and Fasttracker II started in 2019.

Inspired by the UNIX History Graphing Project, the objective is to collect release dates and dependencies of music trackers and produce a graph of the lineage of music trackers in different operating systems.

From the original homepage at helllabs.org.

For 12 years, until his last commit in July 2019, Claudio explored the insides of all Soundtracker-like tools that he could find, reading through their documentation or even their source code when available, in order to discover who did what first, and who got inspired by whom.

Then he seemingly abandoned the project: There hasn’t been any further commit since July 2019, and the helllabs.org domain that hosted it has been down since 2023⁴I tried to contact him in October 2023 in order to at least have the Wikipedia Commons version updated with the file available on GitHub, but didn’t get an answer.. It could be because Claudio joined Canonical in January 2019⁵As per his LinkedIn page., or simply because, in the words of Mark Pilgrim, it was time for him to find a new hobby⁶Of note: He is also the author of LibXMP (for Extended Module Player Library), a project he launched in May 2009 and contributed to until January 2021 — so we can see he had a keen interest in having extensive knowledge on tracker formats. Moreover, the LibXMP project has a very interesting /docs folder….

Still, through this graph we can see who « took inspiration » (*coughs*) from whom, and how we got to modern trackers. Neat! Thank you Claudio!

Note that for this article, I used the latest available visual version, preserved by the essential archive.org project ❤️

Cambrian explosion?

Of particular interest to me⁷As a Soundtracker scholar, obviously. *pulls deeply on his imaginary pipe* is the top part of the graph, showing when and how it all started.

Let’s zoom in on the years 1987 to 1989, shall we?

It’s not possible to read it this way, so I advise you to open the SVG file in your browser and wander around from version to version, the way you would zoom in and out in OpenStreetMaps.

As you can see, while there were several parallel developments, it seems that the developers were (mostly) respectful of lineage: there is only one Soundtracker IV, it’s AFL’s; there is only one Soundtracker V, it’s D.O.C’s, etc. This seems to prove that new versions spread quite quickly — through a network of swappers, of course⁸Stamps back!.

Just for 1988, 15 new versions (or variations thereof) of Soundtracker were released — 7 of which by Unknown/DOC alone.
All from a single seed, The Ultimate Soundtracker, planted by one Karsten Obarski in 1987.

Hence, Cambrian explosion.

Cambrian explosion (plural Cambrian explosions)

1. (evolutionary biology) The relatively rapid appearance, during the Cambrian Period around 541 million years ago, of most major animal phyla, as demonstrated in the fossil record.

Synonym: biological big bang

Wiktionary

I like big words and I cannot lie.

How come there was so many versions? Why such a frantic pace of improvements, one building upon the other? What made it so special?

And most importantly, I was wondering who was the first to actually disassemble Karsten Obarski’s Ultimate Soundtracker — because I was pretty sure that most « clones » were based on that first available source code.

So what did I do?

I did what any sane Internet-person would do: I picked the names of all those who created new Soundtracker versions in the first two years following Obarski’s initial release, and promptly sent them a handful of questions.

And guess what?

They all answered.

Unusual suspects

This article is the result of email exchanges with some of the earliest developers of Soundtracker versions, as listed by Claudio’s graph above: Exterminator⁹Apparently the first one to « update » Obarski’s Ultimate Soundtracker, with TJC Soundtracker II., Unknown¹⁰One of the most prolific coders here, adding several features, not the least being extending the format to 31 instruments instead of the initial 15, the creation of the « module » format (MOD), and allowing to both save songs in that format. He released 7 versions in 1988., Marco Nelissen¹¹… who according to the graph was also an early disassembler of the original Soundtracker — but no one expanded on his version, curiously., TIP¹²… who went from upgrading Soundtracker, to creating his own, commercial Oktalyzer, which supported eight channels — an impressive feat in itself, long before OctaMED and TFMX., MnemoTroN¹³Another very prolific author (4 versions in a year), bringing his own share of key features, notably the ability to load a module, thus closing the loop., Mahoney¹⁴Of « Mahoney & Kaktus » fame; the first ones to depart from the « Soundtracker » name, creating NoiseTracker, one of the first « nice to use » trackers, and a key inspiration for the now-legendary Protracker. It was deemed nice to use because besides several bug fixes and cosmetic improvements, they added ways to use it with only a keyboard, instead of having to use the mouse for everything besides adding notes.… and, because why not?, I also contacted Lars Hamre¹⁵The man behind the hugely successful Protracker. It’s a later addition (1990), but one cannot talk about Soundtracker without mentioning Protracker, really..

As you can see by clicking each link, I didn’t have to go far to discover their real-life names, as Demozoo lists them. Unlike the warez scene¹⁶Y’know, pirates, « You wouldn’t steal a car », and all that shady stuff. Arrr!, from which it was born, the demoscene was never about being anonymous, but rather pseudonymous: while having a scene handle was necessary when doing software piracy, for those creating demos it was just a cooler way to present oneself 🙂
From their real names listed in Demozoo, contacting them was easy: some have their own websites, others their own company, and the last ones… were contacted through LinkedIn 😅

I need to point out here how incredibly useful are scene-related databases such as Demozoo, Pouët, the Exotica wiki and its Kestra Bitworld sub-site. I was able to not only cross-reference information from the graph, but also to dig deeper thanks to community comments, such as in this page.
As a bonus, they are happily cross-linking between themselves, showing the camaraderie¹⁷*pulls even deeper on his imaginary pipe, nearly chokes* of the demoscene community ❤️
All the screen captures on this article come from these websites.
The French Amiga-focused Obligement webzine was also a great source.

Who’s on first?

Let’s dive right in with the initial question: Who was the first to build upon Karsten Obarski’s Ultimate Soundtracker?

The graph here is helpful:

According to the graph’s datasource, the first version of Ultimate Soundtracker was released in August 1987, and Ultimate Soundtracker 1.21 was released in December 1987.
Now, there might be a mistake in the datasource since Demozoo shows a screencapture of a demo version (displayed here on the right side), signed « Obi« , with a November 1987 release date. But we’ll let this slip by for now…

The important indication from the graph datasource is that Soundtracker Pro I is listed as released by Marco Nelissen on January 20th, 1988. Just one month after the original’s release date: That’s some quick reworking!

Congratulations for being the first one on the line, Marco!

Marco Nelissen: I’m pretty sure I wasn’t the first one (and the name, « Yet another soundtracker », certainly seems to be saying that there were many soundtrackers available already at that time).

I suspect that the release date of 1988-01-20 is wrong.
January 1988 is only 9 months after the release of the Amiga 500, and my first Amiga was a 500. I think I would have needed more than 9 months to buy one and get proficient enough with Amiga programming.

The release year of 1989 as listed on Pouët seems more realistic.

E-mail interview from January 2024¹⁸After contacting him through LinkedIn….

I’m not sure that the « Yet Another Soundtracker » argument proves anything, since Karsten Obarski did name the very first Soundtracker as being « The Ultimate Soundtracker » — but, you know, teenagers 🙂

In any case, I agree that 9 months is short to buy a computer, learn to code well enough, obtain original software, disassemble it and rework it into something notable.
So the new date of 1989 would put Marco’s version after many of the other trackers released in 1988, and therefore he cannot be the original disassembler of Obarski’s UST (Ultimate SoundTracker, for the rest of this article).

But the graph does connect his version directly to UST, so is that wrong too?

Marco Nelissen: I’m fairly certain I disassembled an existing tracker myself, and it was probably « the ultimate soundtracker » (though not sure which version), given that « yet another soundtracker » says it’s based on the original by Karsten Obarski.

I’m certain I was not the first though. There were many alternative trackers already available when I got involved, hence the « yet another » description in my version.

So the graph you linked to earlier is probably wrong. While it’s possible the branching itself is correct, the timeline is almost certainly incorrect. It seems much more likely that my « yet another soundtracker » came around somewhere in 1989, after many of the other variants.

Ibid.

Alright, so Marco is not The One. But still he did some original work, why didn’t it inspire others in the demoscene? Why did they all build from TJC Soundtracker II?

Marco Nelissen: TJC and other trackers I think came out of the crack/demo scene, so had wider reach because of that. I also didn’t release a standalone player routine (except maybe towards the very end, I’m not sure), so it was hard for people to use such mods in games or demos.

Ibid.

Thanks Marco!
Another possibility for Soundtracker Pro not being quite adopted by the scene, is the apparent fact that it extended the MOD file format (at least for ST Pro II), making its saved file incompatible with other trackers and playroutines — the subject of file format compatibility was a hot topic in the tracker community, and still is.

Therefore, still according to Claudio’s graph, that leaves us with:

The Exterminator/TJC

The first coder to release a « fixed » version of Soundtracker, The Exterminator, member of the group The Jungle Command, lighted the necessary spark and showed that it was possible to improve on Obarski’s game-changing music editor.

Exterminator was not a newcomer in the realm of sound editors, as he had already produced MegaTraxx for TJC, which was implemented into Soundtracker¹⁹« Then Exterminator (…) made for us a breakthru with soundtracker ii it was just megatraxx build into soundtracker. » in order to get TJC Soundtracker II²⁰I can’t for the life of me find any visual of MegaTraxx, only remains its source code here, and a list of productions that used one module contained in this early format. The best I can guess is that Exterminator implemented the MegaTraxx playroutine in Soundtracker — which might explain why the playroutine’s functions have mt_ prefixes… and we keep seeing the mt_init function up to Protracker’s playroutine, and probably later tracker replay routines..

Mark Langerak released only one version of TJC Soundtracker II, and soon moved on to other hobbies — namely, creating video games. Indeed, the demoscene has always been a source of hiring talents for game companies, from programmers to musicians and graphic artists.

He released his first game on the Amiga, The Plague, in 1990. He worked solo on the code and the graphics.
35 years later, he is Principal Software Engineer at Microsoft, having worked notably on Gunship 2000, The Sims 2, Kinect Star Wars, and many more.

I’m pretty sure he’s not too fond of his handle from his Amiga days 🙂

I used the most uncool way to contact him: I sent him a LinkedIn message 😅

So, was he The One who first disassembled UST?
Or did he receive the source code from someone else?

Mark Langerak: I think I was the first yes. I definitely did not get the source from anyone, I would have remembered that.

But I don’t actually remember how or if the original was actually disassembled (to source), I don’t think there were any decent disassembler tools available at the time that would’ve made that possible?
If I were to guess, the original Soundtracker would have remained in its original form as a binary blob, which would then have some strategic patch vectors poked in at the right places to extend it.

That is something I recall doing even to my own code when developing on C64 because there was not enough memory to have both source + binary in memory at the same time — I had no money then to buy a separate development machine 😉

LinkedIn interview from November 2023.

Oh, that’s clever! Even without the source code, it was possible to extend and rework a program by changing its representation in memory (then saving that representation to disk, I guess).

But that means… The Exterminator is not the one who initially spread the source code for UST.
Still, he is the one that showed that it was possible to improve upon a commercial software that was, apparently, not quite stable yet. And THAT is an important step for the Soundtracker saga.

Also, it was all meant to be fun!

Mark Langerak: One little factoid I do remember — the playback routine for TJC tracker has a hidden Jungle Command easter-egg.

The idea was that we wanted the fame of the playback routine getting used in other groups’ demos, and the way to show that the TJC tracker was used would be to invoke a magic mouse button + key combo when running a rival group’s demo.
That would then trigger a piece of code hidden in the (fake) arpeggio table in the TJC playback routine. This piece of code was designed to take over the machine and display a Jungle Command related message.

I completely forget how it was supposed to work, but it wasn’t very reliable — I only managed to ever find one demo from another group where I was able to trigger it and bask in the glory 😀

Ibid.

See? All in good fun! 🙂

The TJC Soundtracker lineage didn’t last long, though:

Mark Langerak: TJC Soundtracker II was done in isolation as I remember. It was also very much a one and done kind of thing, we did not contemplate or feel any incentive to continue with trackers in general. From what I remember we were much more focused on the next demos and making our own games.

Ibid.

Alright, it’s all fun and games²¹Hoho!, but I need to continue my quest:
Who’s The One behind the Soundtracker Cambrian explosion?

---

Added May 21st, 2024:
Reader Alex Limi kindly pointed me to this comment that he archived from a YouTube video, written by Daniel Ockeloen (aka MCS/TJC):

We are getting old but it’s nice to wonder in a old video and see my memory confirmed.

Soundtracker 2.0 was indeed the first split.

With a friend I started The Jungle Command, soon to be joined by Mark. We came from the C64 and needed audio routines; work was underway with MegaTraxx, but that had no GUI at all.

So Soundtracker came out and [with] some merging and fiddling, we had Soundtracker 2.0. A nice guru meditation hack was added and hidden in the source for much fun later :).

It was a small step but combined with someone who did some great tunes (our member Ron Klaren) and a few music demo disks (« Music invasion » was one series) it all got out of hand a little.

It was never really done as an angry act against the orginal maker, we where just kids. Jungle Command was never really much in the cracking part, more the making part.

Later we also had a second music routine for games like Battle Squadron which Ron also used to show his music skills and I still feel [his tunes] are among the best Amiga tracks of the time.

It was a great time, thanks for reminding me and others of the fun we had now 37+ years ago….

— daniel (mcs) from the jungle command.

edit: most groups knew who worked on it at first and communication was happening but indeed the MOD fileformat part by DOC was a massive step in making it a household name…

edit2: in my memory we had some communication with Karsten and he forgave our youthful sins. I hope I am not making that up and he also looks back with a smile on what his work did and how we can still enjoy some of these amazing tunes of the time.

From a comment on an awesome Ahoy video on Tracker history, edited for readability.

This comment explains why I couldn’t find any screencapture of Exterminator’s MegaTraxx: it only ever existed as an early set of audio routines!
Instead of spending time programming the whole interface for MegaTraxx, Exterminator probably chose the easy path and adapted his playroutine into The Ultimate Soundtracker, thus creating TJC Sountracker II.

And really, just comparing the source code for MegaTraxx and the TJC Soundtracker II playroutine confirms that: Apart from minor additions, removals and changes (and the whole Sample Info and Parts sections), they are pretty much the same.

As I wrote in my article on Karsten Obarski (Part I of this series):

Exterminator had de-assembled the original program and released the resulting source code, along with its playroutine. This allowed anyone to not only create music in the tracker format (fine), but they could also freely (as in, illegally) use that format in their own productions (games and demos).

Source: myself.

Since the playroutine was a key part in selling The Ultimate Soundtracker rights (which is my guess on how usage worked at that time), well, once TJC Soundtracker II was released, Ultimate Soundtracker’s commercial fate was pretty much sealed.

Thank you Alex for this gem!

---

Added June 19th, 2024:
Well I had to contact Daniel Ockeloen, aka MCS, co-founder of The Jungle Command, just to see if there was anything more in his memories that he could contribute to the topic 🙂

…and boy did he have.

I’m 100% sure that TJC Tracker was the first, but as Mark writes, he might have later started from the 1.0 version — this would also explain the mt_ naming for a while in only one branch.

The hint that the source of the player was the same was well-known at the time, there is a reason why much of it starts with mt_ names.

From a Facebook Messenger conversation, June 2024.

The « let’s use the [Ultimate Soundtracker] GUI and semi bolt MegaTraxx in » was because of the lack of a GUI — the only thing around at the start was Sonic [Sonix by Aegis?] .

Mark should know best but in my memory, it was just the easier way to merge the two.

Ibid.

I can also confirm that Il Scuro/DefJam made the first upgrade [with Sountracker III], but the more important step was DOC. For the rest I think you are on track on what happened and why.

Ibid.

At the time, musicdisks where a thing, and many [sceners] sent us songs to put on one of our disks, kinda before megademos became a thing. So having a player routine we all could use and some GIU was handy.

I remember that we as a group where kind of proud of [musician] Ron Klaren‘s work so we had an incentive to push the player idea, even if Mark was already moving on [to focus on his game].

Ibid.

I also strongly disagree with the [reliability] of his hack: many demos and even games had the ‘Guru’ hack inside, even in later player releases. And John [van Dijk, aka Boil, co-founder of TJC] and I had fun testing new disks [to see] if they came in for it.

But like Mark, I can’t remember how to trigger it. I think it was space+mousebutton+joystick button. I tried to get it working on some things years ago, but failed — let’s call it a missing easter egg ready to be rediscovered.

I do clearly remember nobody found it until some English friends (Mic) found it and just changed the messages to say ‘hi back’ in the message.

Ibid.

Keep in mind, you probably need to merge like 10 people’s minds to get the truth — funny how memories work. Thanks for your great work on this!

Ibid.

Thank you for all these details, Daniel!

---

Let’s get back to our regular scheduling, and scroll down a bit on the graph:

The TJC seed has been planted, and we get to explore the sprouts.

Note on the left-side arrow of the graph:
I wasn’t able to contact Il Scuro/Dejfam (Ron Birk), but since, according to the graph, his Soundtracker III is a one-off version and it only inspired one tracker (Alpha Flight’s Soundtracker IV), I think we can safely skip it for our quest.

Thus, now begins the era of…

Unknown/DOC

In 1988 alone, Unknown of the demogroup Doctor Mabuse Orgasm Crackings²²Yeah, I know… released 7 versions of Soundtracker: Soundtracker III, then IV, VI²³Wait… Aaaaah, apparently someone released a v5.0 in 1988, which was just DOC’s ST IV rebranded, as per the scroller here (look for « fred »)., IX²⁴What? We skipped VII and VIII too??, 2.0²⁵Oh, back to more regular numbering…, 2.1, and 2.2.
And he stopped there, leaving others to go further, and having added an incredible amount of features to Soundtracker, not the least of which being the celebrated MOD format, which made it much easier to share music and include them in scene productions, be they demos or cracks.
And of course, as a proud marker of his contributions, he added « M.K. » to the headers of module files, for Michael Kleps, his name at the time. That marker²⁶Or « Magic number« . Which can be quite tricky to recognize. is still used today to recognize modules from back then, and to differentiate them from modules made by earlier or later Soundtracker-like tools.

Unknown/DOC seemingly found his passion with music tools, because Michael Kleps (nowadays named Michael Hartmann, possibly after wedding his partner) has since created the reFX company, which has been producing virtual synthesizers for several years now. See this YouTuber touring the reFX office and meeting Michael.

Of note: Even today Michael remains a fan of oldskool computer music, as in 2001 he created the reFX QuadraSID synth, which brought the sound of the C64’s SID chip to modern DAWs — a project for which he was interviewed in 2001 by the Remix64 portal — and I learned that he was working on a Spotify-like SID player.

True to his scene handle, it was very hard to get in direct contact with him, despite seeing he was active in his GitHub profile: I couldn’t find any working email address for him nor for reFX — you have to be a reFX customer in order to contact their support service.
BUT, after a lot of of poking around, I was able to contact him.

So, I wonder, as the most prolific author of Soundtracker versions, what inspired him to spend so much time on that tool?

Michael « Unknown » Hartmann: I was always interested in music, and once the source code got to me from one of Dr. Mabuse’s trading partners, I was super excited to play around with it and fix bugs or add features as I saw fit.

We made so many versions because we discovered bugs, and Dr. Mabuse couldn’t wait to add more ridiculous scroll texts to it.

Nowadays, with the internet and all, these would be considered minor releases nobody would even mention. But back then, every little fix, every little change was immediately a new version to be traded with others, etc.

E-mail interview from January 2024.

Indeed the scrolltexts were quite silly 🙂

Wait, « once the source code got to me from one of Dr. Mabuse’s trading partners », so he’s not The One who disassembled it?

Michael « Unknown » Hartmann: No, I didn’t disassemble it. I’ve got the source code from Dr. Mabuse, who got it from someone else. I got it pretty early on, but I can’t remember what exact version of what.

Since I was into music, I was super curious about how it all worked. I refined it a lot, as a lot of variables and labels had very generic names (param1, param2, param3, loop1, loop2, loop3, etc.) instead of proper names like « volume », « pitch », etc.

If somebody released a new tracker, I immediately tried to match the new features or even improve upon them. It was important for me not to fall behind and stay compatible.

Ibid.

Holy guacamole, Batman!

• Exterminator did not disassemble it, and
• Unknown obtained the source code from someone else!

Does it mean there is a missing link between TJC Soundtracker II and DOC Soundtracker III‽‽‽

Crossing the streams

Here I must take a quick step forward, and show you what (and who) comes next in the graph.

As those with good eyes (or with the SVG file open) can see, a strange thing happens:

1. Unknown releases DOC Soundtracker IX,
2. … which serves as the direct inspiration for Master Soundtracker 1.0, created by TIP/TNM…
3. …which, in turn, is the sole inspiration for DOC Soundtracker 2.0.
   I repeat: DOC Soundtracker IX is not the direct inspiration for DOC Soundtracker 2.0.

Who’s this « TIP » guy, swooshing by with tons of ideas and features, to the point of inspiring Unknown?
TIP, short for The Invisible Power, is in fact Armin Sander, who later created Oktalyzer.

But we’ll dive further in his work later; for now, I’d like to focus on a key part of his Obligement interview, by Guillaume Guittenit:

Guillaume Guittenit: Why did you make Oktalyzer? Tell us the story behind this software.

Armin « TIP » Sander: The story begins with me sitting in front of my Amiga and launching a new program called Ultimate Soundtracker for the first time and subconsciously realizing that this is the kind of software I wanted to develop.

The release of Soundtracker was a small revolution in the Amiga scene, and some of the best hackers tried to reengineer and extend the program. Unknown/DOC was one of the first coders able to put some additional features in it, and when I saw that this was possible, I was excited by the challenge to do it as well.

The first thing to do if you want to extend an existing program is to reassemble it, which is the process to create assembler source from the binary code of the program. I can’t remember exactly which tools I used, and if Resource was available to me, but shortly after that, I had created a fully functional source code of the original Soundtracker.

Later I had a telephone call with Unknown/DOC during which he explained to me that he never actually had had the source code, instead he extended Soundtracker by patching the binary code at specific locations. In that telephone call I offered to give him _only him_ the assembler source. And if Unknown/DOC broke his promise, this source could very well become the basis of a number of future Soundtrackers released in the Amiga scene.

Obligement interview, in April 2014.

SAY WHAT???
Wow, there’s a lot to unpack in that last paragraph.

It seems that the four « pre-TIP » versions of DOC Soundtracker (III, IV, VI and IX, for those keeping score) were not in fact created by improving some source code that was available, but by patching the software while it’s in memory — the same way The Exterminator first improved UST to obtain his own TJC Soundtracker II.
That would mean that when Unknown told me that « the source code got to [him] from one of Dr. Mabuse’s trading partners », either he was mistaken (which we can forgive, after 35 years), or he was talking about DOC Soundtracker versions starting from 2.0 onward, and therefore TIP was actually that trading partner.

Armin further confirmed that he was The One by email:

Armin « TIP » Sander: Actually I did a lot of reassembling [back then], I guess because I wasn’t good or smart enough at cracking software or modifying binaries, even considered this as dreadful boring and hard work.

As stated in the Obligement article, I can confirm that the SoundTrackers I’ve released were built on fully reassembled version.

E-mail interview in January 2024.

Seeing the drastic changes in version number (from Roman numerals to « 2.x ») and layout/design between DOC Soundtracker IX and DOC Soundtracker 2.0, I’m inclined to believe that Armin was indeed the source.

Yep, DOC Soundtracker 2.0 does look like a rework of Master Soundtracker 1.0.

Visually, Unknown used a cleaner font and added back the scrolltext for Dr. Mabuse’s deliriums (seen here on the right screenshot, above the « Disk Status: All right » section), moved/renamed some buttons, and added himself as 5th coder in the credits for this version.
Backstage, Unknown made several changes of his own, notably the addition of the « pattern break » and « position jump » commands.

The fact remains: Armin is The One who first fully disassembled Ultimate Soundtracker²⁷Which version? Possibly Obarski’s UST 1.8. Yup, I could have saved a LOT of research time by finding this Pouët comment….

---

But that certainly does not lessen the improvements brought by Unknown to Soundtracker — most notably one that came in DOC Soundtracker IX and has hugely boosted the popularity of the tool: the MOD format.

The Module format

Yes, I am adding an aside to this whole interlude…

See, initially Soundtracker and its clones saved music in a basic song-file format (.sng), which was roughly the data table, and that’s it. The song-file then pointed to sample files (raw audio data or IFF 8SVX, as .8svx²⁸Or not, since the Amiga didn’t enforce file extensions.) with their absolute path, including the name of the sample disk (ST-01:, ST-02:, etc.). This allowed Soundtracker to load samples once the song file was fully loaded.
This means that when sharing a song with, say, a friend, you had to provide both the song file and all the sample files that were used — possibly even with the correctly-named sample disk! What if you didn’t have the right sample disk, or if the sample file was misnamed, corrupted, impossible to read, or… ?
That early format was, to say the least, kinda cumbersome to share and use. It was a bit like providing the MIDI file of a song, along with all the attached samples, hoping for the best.

Unknown/DOC is apparently the one who came up with the « module » format, which combined music data with sample data, making it possible to share it all in a single file. And therefore, something much easier to include in a demoscene production.

This addition was so good that even Karsten Obarski included a « Save as Module » feature in his final UST 2.0.

So, being in direct contact with Unknown, I had to ask: How did he think of working that out?
…and, why did he choose to prefix module files with mod., rather than suffix them with .mod, as is the usage? 🙂

Michael « Unknown » Hartmann: It’s been such a long time that I am trying to remember how it came about. I must admit I’m not 100% sure about that.

E-mail interview from January 2024.

Damn, so close. 35 years is a long time ago, sure, but I was hopeful to have a hint at the creative process behind it.

I have to rely on secondary sources, then. There are several pages documenting the format itself, but none which seem to explain how it came about — and most attribute it to Karsten Obarski himself. They essentially take it for granted that module files have always existed, forgetting about the original songdata+samples mayhem of yore.

But one like-minded individual³⁰Y’know, a sane internet person, like me! did dive into that part of Amiga history, and wrote a Twitter thread about it (which I preserved in that blogpost). The author, retro-computing enthusiast Thomas Cherryhomes, set things straight about the whole affair. Granted, there’s little sourcing on this and it might be hearsay, but that’s the best I have, so I’ll take it.

I’ll summarize his thread and expand upon it³¹Because of course! here:

• Karsten Obarski’s Ultimate Soundtracker initially only saved songs (.sng files), with separate sample files (the first one coming from the ST-01 sample disk, and following). This is because UST was at first an in-house tool for the reLINE game studio, Karsten’s employer, until editor EAS Computertechnik saw it as commercially viable.
• Obarski expanded on the idea with the Preset-Ed tool, making it easier to manage the list of samples. It was promptly « improved » by others too (*coughs*).
• The idea of a self-contained file format was brought to life by Unknown/DOC with his « module » addition, introduced in DOC Soundtracker IX. But Unknown only coded the ability to save a module, so it was really a « last step before release » feature, like an « Export as PDF » option that we see today.
• The « Load module » feature was introduced by either TIP/TNM (according to Thomas’ Twitter thread) or by MnemoTroN/SPT (according to Claudio’s graph), finally making the format truly sustainable.
  Now musicians could freely load and save modules, exchange them with other musicians without having to worry about a missing or corrupted sample file, include the module in a scene production, etc. The initial song+samples (.sng) format was all but forgotten.
• Bringing it all full circle, Karsten Obarksi included the ability to save as a module in Ultimate Soundtracker 2.0.

Phew! What a ride.

  

Postscript to the aside:
While finishing this article, I stumbled upon this demo of the Weasel Audio Library that uses several early Soundtracker files³²Aptly named « Karsten Obarskis’ Ultimate Soundtracker sound module playing jukebox! » 🙂, among which those of Karsten Obarski.
And while some tracks do use .mod or mod., most of the early songs use the stk. file prefix.

So, could it be that it was Karsten Obarski himself who chose to prefix Soundtracker files with stk. rather than suffix them with .stk?
And therefore, when the time came to name the Module file format, Michael Hartmann just used mod. as the logical continuation of Obarski’s naming scheme?

I’ll update this section once I find more clues…

---

Sharing is caring

Now, how did the source code spread from Armin and Michael?

Indeed, starting with Master Soundtracker 1.0 and DOC Soundtracker 2.0, we had two coders working from the initial same disassembled codebase: TIP and Unknown.

And neither were supposed to share that source code.

It seems that in the graph, each version’s arrows should have gone on and on, and in parallel, but actually… not.

Three things happen in this part of the graph:

• Armin « TIP » Sander released a second and final version of his Master Soundtracker³³v3, possibly because of DOC Soundtracker v2… in September 1988, then moved to a new project, the 8-channel Oktalyzer (seen here in the bottom right of the graph).
• Michael « Unknown » Hartmann released two/three³⁴Depending on the source. final versions of DOC Soundtracker, the last one in December 1988, and then no more.
• The next 2.x version after Unknown’s 2.2 comes from a newcomer, MnemoTroN, in April 1989.

Also, we see the very last version produced by Karsten Obarksi, The Ultimate Soundtracker 2.0, with no lineage following it. As the graph shows, UST 2.0 took inspiration from TIP’s own Master Soundtracker 1.0 on some part³⁵Most importantly, it made official the module format introduced by Unknown in DOC Soundtracker IX., and was released roughly at the same time as Master Soundtracker 2.0.

We can understand that, despite having created a new kind of musical software, Karsten wasn’t able to keep up with the changes brought by other coders (and which were now expected as standards by the community), and decided to drop his commercially unsuccessful creation³⁶You can read more about Karsten’s path in this previous article of this series..

So long, Master Obarski, and thank you!

MnemoTroN/SPT

A member of the Spreadpoint group, MnemoTroN (a.k.a. Thomas³⁷No last name, because « I don’t really want to share my real name, since after all I was not only doing Soundtrackers, but also did quite some cracking back then. 😉 » ) was also quite prolific, with 4 Soundtracker releases to his name.

What inspired him to spend so much time on that tool?

Thomas « MnemoTroN »: The story behind my initial SoundTracker releases is as follows: At a CeBIT show in Hanover (must’ve been in March 1989, because we released SoundTracker 2.3 in April 1989 according to Demozoo) I met Unknown. I can’t remember if we were introduced or if it was a chance encounter.

He gave me a disk with his SoundTracker source code, probably because I was a bit known in the scene by then due to some small intros and game cracks.

The source was based on previous versions done by Exterminator and TIP, so obviously they exchanged sources before. I kept all that in the credits and the start screen, including Karsten’s name. 😉

E-mail interview from February 2024.

Wow, Unknown literally handed the reins to the project to MnemoTroN, live in person! Well, of course not quite « handed the reins », but it’s funny how things turn out.

Also funny is how I mentioned the CeBIT’s importance in my previous article in this series, talking about how Chris Hülsbeck could have been influenced by Steinberg 🙂

But still, it’s weird that Unknown would just give the code away like that, since Armin « TIP » Sander specifically asked him not to — remember the Obligement interview?

« I offered to give him _only him_ the assembler source. And if Unknown/DOC broke his promise, this source could very well become the basis of a number of future Soundtrackers released in the Amiga scene. » — TIP.

MnemoTroN remembers:

Thomas « MnemoTroN »: When I met Unknown again (probably at CeBIT 1990) he said that he gave me the source with the instruction not to release it further, but I really, really cannot remember him telling me that. Maybe it was too loud at the Commodore booth or I was a bit excited of all the dozens of people there (you cannot imagine how many people were there, it was insane).

It was a bit awkward to hear that because I’m very honest and wouldn’t have released anything based on his source if I actually had heard or remembered that.

Anyway, the deed was done already, the SoundTrackers were released and it was history. So, apologies to him again.

Ibid.

So TIP gave his source code to Unknown with the promise of not sharing it with anyone else; and Unknown, maybe out of pride, gave it to MnemoTroN with the same promise in mind — but maybe not in speech, or not loud enough at least.

This might be a misunderstanding that dates back 35 years, for both TIP towards Unknown, and Unknown towards MnemoTroN. At a time when introspective people (well, geeks) couldn’t rely on the Internet or even SMS to patch up misunderstandings without meeting face to face or, Zeus forbid, through a phone call, it was quite a blunder.

UPDATE
Right after publishing the article and sending it to the various interviewees, Unknown got back to me with these details:

Michael « Unknown » Hartmann: I’ve just finished reading your article. (…) It all brought back a lot of memories, as well as the Cebit stuff. That was after I moved out and lost contact with Dr. Mabuse³⁸See section « Let it go » below., so I figured somebody else could benefit from it and handed it to Mnemotron.

E-mail message from March 2024.

More details about the source code itself:

Thomas « MnemoTroN »: The source code that Unknown gave to me in 1989 was the first one that changed the number of instruments in a SoundTracker song from 15 to 31. That version was unreleased and it seemed they wanted to keep it that way to keep ahead of the competition or something. That is also why he introduced the « M.K. » signature with his real name initials in the song data to distinguish the 15 and 31 instrument song data.

E-mail interview from February 2024.

Ah! Browsing Claudio’s graph, I always found it weird that it would indicate MnemoTroN’s version as being the one introducing Unknown’s famous « M.K. » marker.

Now that explains it: In working with Unknown’s extension of TIP’s source code³⁹Are you still following? 🤪, MnemoTroN was key in spreading that famous marker.
Without him, the marker wouldn’t have spread, since it seems Unknown stopped working on Soundtracker before he got a chance to release it himself.

And, why did Unknown add that marker in the first place?

Michael « Unknown » Hartmann: One day, our musician Frog (Timm Engels) came to me and asked me for more than 15 instruments because he was making music for the game « Datastorm, » and he had already used all 15.

I used a free bit (there were four bits free, but I was sure nobody would ever need 255 instruments), so he instantly had an additional 16 instruments. This change took only a few hours to make, but I had to detect during loading if this was 15 or 31 instruments, so I used the old type marker (can’t remember what it said before) and replaced it with « M.K. » (My initials at the time).

Once the new version was officially released and spread, I got upset when people attributed this 31-instrument format to « Mahoney and Kaktus ». This was the first I’d heard of those guys. But again, this was before the internet, so I couldn’t « correct » it or even really talk to people.

E-mail interview from January 2024.

In a world were scene handles are more well-known than real names, we can’t really blame sceners for thinking « M.K. » was tied to the already-famous Mahoney & Kaktus. But anyway, so the story goes…

Of note, his « I was sure nobody would ever need 255 instruments » rings a bit like Bill Gates’ apocryphal « 640K ought to be enough for anybody », since users of Renoise, OpenMPT, or Polyend, the modern man’s trackers, have hit this limit already. Ah, to be young again 🙂

UPDATE
Right after publishing the article and sending it to the various interviewees, MnemoTroN got back to me with these details:

Thomas « MnemoTroN »: Regarding Unknown’s quote that he changed the old signature to « M.K. »:
There was NO signature in the original 15-instrument Soundtracker songs. The whole format is quite a mess from an architectural view and also from a coder’s view concerning the CPU time it needed on a stock Amiga with the original player code.

When the format was changed to 31 instruments it would have been a really good time to streamline it, and maybe Unknown even thought about it, but [he] just wanted to get the extension to 31 instruments done and added his signature. The rest is history.

E-mail message from March 2024.

---

Let it go

Both Unknown and TIP stopped releasing new Soundtracker versions around the same time: We don’t see any release from them in 1989 nor later.
As far as I can tell, it’s not because Michael mistakenly spread Armin’s source code through Thomas.

For Armin, it’s a story of focusing on more interesting technical issues:

Armin « TIP » Sander: (…) while I was working at extending the tracker, my motivation soon turned to solving a big problem the composers I knew had: the limitation of the four hardware audio channels.

(…) The development of the playback engine took a few weeks, and the development of Oktalyzer itself took about six months.

Obligement interview.

Armin has since moved on with several other projects, commercial or not, for the past 30 years. Contrary to this little writing project of mine, Armin is not the nostalgic kind, as his last answer to the Obligement interview will show you. I do invite you to read it; I’ll only copy-paste a small, telling portion:

Armin « TIP » Sander: I realize that [my old Amiga 3000] strongly connects me to a time when I had a lot of fun making things and met a lot of interesting people.

But even if my feelings make me wish to hold on and remind me of the past, I must also accept that this time is over and will never come back, and that the only thing that exists, is this very moment, which gives me a tiny opportunity to change the future for me and – if I am very lucky – also for others.

Ibid.

(a pause)

For Michael, it’s a story of growing up:

Michael « Unknown » Hartmann: Dr. Mabuse got a girlfriend, so he rarely met up anymore [with the rest of the DOC crew], and he didn’t trade as much anymore.

A month later, I turned 18 and moved out of my parents’ house. So, instead of having no responsibilities (hanging out with friends all day and coding/gaming for fun), I now had a « real » life to take care of.

Even if I had continued to refine our Soundtracker, I would have had no way of « getting it out there » since Dr. Mabuse had disbanded D.O.C by stopping to show up.

Since my move in January 1989, I have only met him once more, when he was driving a taxi and dropped off some passengers in front of my apartment. He rang the doorbell, we talked for about 10 minutes, and then he had to go. That was the last I ever saw of him.

We basically both aged out of the scene as our adult lives began.

E-mail interview from January 2024.

Bittersweet ending for two guys who had quite an impressive run in 1988.

---

Back to MnemoTroN.

Thomas « MnemoTroN »: It’s a bit difficult to remember the details from about 35 years ago, but I know that I have spent some of the time coding at a friend’s house where we were gathering (mostly on the weekends) and just looked at new Amiga demos and games our organizer received from his contacts.

Working on SoundTracker itself was just fun. I am no musician myself, but friends dabbled in creating their own songs and came up with some ideas how they would like to improve the tool. Since I was the front coder of the group, I tried to accommodate their ideas.

E-mail interview from February 2024.

One funny thing is that fixing Soundtracker wasn’t just about fixing Karsten Obarski’s code, but also fixing the code « improvements » from previous coders:

Thomas « MnemoTroN »: I think one of the first changes was to replace the irritating animated mouse pointer from D.O.C. with the blocky design that survived so many updates and changes to the tool, even in its NoiseTracker and ProTracker variants.

Ibid.

Indeed, some coders started writing their own Soundtracker in reaction to the changes brought by DOC Soundtracker — mostly cosmetic ones 🙂

Thomas « MnemoTroN »:The new mouse pointer was drawn by my friend Tuca. Tuca and me also created the needed updates to the SoundTracker graphics because the text in them was not printed by the program, but was drawn by hand in Deluxe Paint and then converted to the needed bitplane format of the Amiga.

Ibid.

… but they also added some new ideas that I didn’t know about:

Reed/FLT: One of the more interesting features of Soundtracker 2.6 was the sub-patterns, an idea which didn’t really catch fire in other trackers… quote from the documentation of Icetracker v1.1:

« Basically now a pattern, instead of being made up of 4 joined tracks (i.e.: Channels 1,2,3 and 4 were previously classed as one pattern) are now individual sub-patterns in themselves, that is that now it is possible to independently call up a single sub-pattern to repeat on a certain track (i.e.: the main beat, for example) without constantly copying it into every 4-track pattern as was previously required. »

MnemoTroN/SPT: The sub-pattern stuff maybe was a bit too « advanced » and since there was no easily viewable step/pattern list it didn’t catch on. People would just copy full patterns to the next and modify them instead.

Selection of comments about Soundtracker 2.6 — containing an interesting exchange on Soundtracker history, and this comment from hitchhikr, which I took for myself: « for a couple of minutes i thought you were deeply interested by the exhumation of the early history of soundtracker » 👀

Nifty!

---

So now we know.

From then on, I suppose the source code for each version was shared as openly and widely as the version itself, and anyone could pick it up in good faith and give it a twist.

Our question is answered.
We have found The One — and The Ones After Him too.

Our quest has ended. We are free, finally.

*pauses while looking at the sea*

*takes deep breaths*

---

But let’s not allow that to stop us from digging further!

Open the graph again, zoom in to Soundtracker 2.3, and scroll down a bit lower. There! Can you see it?

While MnemoTroN was enjoying himself producing a couple of Soundtracker releases, in comes the dynamic duo of Mahoney & Kaktus, with their own Noisetracker — those same Mahoney & Kaktus that Unknown complained about for inadvertently stealing his « M.K. » meaning!

And it seems that their clone of Soundtracker had fans: « The most important tracker developed since the ultimate soundtracker (until protracker came along). » — said the aforementioned Reed/FLT, in a 2004 comment on Pouët.

Mahoney & NoiseTracker

In the first article of this series, about Karsten Obarski, I already mentioned Mahoney, through an interview of him that I found online:

« The basic idea of ​​Soundtracker was delightful, but unfortunately there were some serious bugs and other shortcomings. I tried to contact Karsten Obarski with bug reports, but was told that he was not going to do anything about it and that Soundtracker was not a commercial success. Today it would be classed as abandonware, but at the time it was just frustrating. »

Pex Tufvesson took matters into his own hands and disassembled the entire program. After much effort, he had the entire assembly code for Soundtracker on his desk. He fixed the bugs, increased the number of samples from 15 to 31, and released Noisetracker 1.0 on August 1, 1989. However, this first version also contained some bugs, so version 1.1 was released just a week later.

Techworld interview (automatic English translation)

Half of Mahoney & Kaktus, a duo quite legendary through their quality/fun contributions to the demoscene, with not only NoiseTracker but also milestone releases such as the double-disk, 101-songs, 4-hours long His Master’s Noise musicdisk, Pex « Mahoney » Tufvesson has remained quite involved in the demoscene to this day, documenting all of his contributions along the way!

I could explore how and why he (along with Kaktus) created NoiseTracker), but he was kind enough to do a talk just about that in 2017, at Datastorm⁴⁰Yes indeed, that same name as the game for which Unknown/DOC expanded Soundtracker from 15 to 31 instruments., an oldskool Amiga/C64 computer party in Gothenburg, Sweden.

The talk is fun and shows the character of Pex and his willingness to share, but if you’re in a hurry, jump at the 20-minutes mark, where he starts talking about NoiseTracker itself.

I’ll pick a few citations:

Pex « Mahoney » Tufvesson: [The Ultimate Soundtracker] did things pretty well, but it didn’t show much. (…) When you played a song, it didn’t scroll the patterns, so it was quite frustrating.

From the above video.

Pex « Mahoney » Tufvesson: [The Ultimate Soundtracker] was not the only tracker; I’m going to show you another one, which I was so disgusted of that I had to change it: The DOC Soundtracker.

There are two things that I really didn’t like about it, and this (points to the scrolling text) is the really really bad thing: when I’m in the creative zone of making music, I don’t wanna read a fucking scrolltext, that’s a no-no.

My solution at first was to put some tape on my monitor to hide the scroller. But I found a better way and decided to rewrite it.

Also, you couldn’t edit your own samples in that program. (…)

So I took this program, disassembled it into some 1,400 lines of code, 200 kb of sourcecode, and made something like this (loads NoiseTracker).

Ibid.

Pex « Mahoney » Tufvesson: There’s quite a lot of sceners and people that have grown up with NoiseTracker and used in as an experience platform to get to know music, and to get to know how to make music, [including Axwell of the Swedish House Mafia].

Ibid.

Pex « Mahoney » Tufvesson: NoiseTracker has sold 0 copies, it has earned 0 billions dollars, but what I got from NoiseTracker was a bunch of letters from people being happy with our music program.

Ibid.

Pex « Mahoney » Tufvesson: I still receive letters and messages from people who were introduced to NoiseTracker as young people, including a few weeks ago when someone wrote: « You and Karsten Obarski destroyed my youth but shaped my whole life, so that’s ok mate! » That still puts a small golden edge on existence.

Techworld interview (English translation).

Coming back to the Techworld interview above, it has this to say about Pex: « After much effort, he had the entire assembly code for Soundtracker on his desk. He fixed the bugs, increased the number of samples from 15 to 31, and released NoiseTracker 1.0« .

That seems to contradict all that I’ve written so far. Gasp!

What’s his take?

Pex « Mahoney » Tufvesson: I did reverse engineer the whole program myself.

Email interview from January 2024.

Double-gasp!

… and which tracker did he work from? Was it DOC Soundtracker, according to his talk, or MnemoTroN’s Soundtracker 2.3⁴²Which, as we saw, is based on the unreleased DOC Soundtracker codebase, so a mix up is of course possible., as the graph says?

Pex « Mahoney » Tufvesson: My final word is that so much water has passed under my bridges since then that these kind of details are long forgotten.

However, I did not extend the number of samples from 15 to 31, that was already done in the version I reverse engineered.
And, the magic word « M.K. » in the module file format « .mod » does not come from « Mahoney & Kaktus ». It’s the initials of the guy who actually did the 15-to-31 instrument change.

Email interview from January 2024.

Coming from Mahoney himself, that remark about « M.K. » will certainly please Unknown/DOC 🙂

So, the 31-instrument extension thing is telling: NoiseTracker is indeed based on MnemoTroN’s Soundtracker 2.3 at the earliest, which is the version that introduced the « M.K. » marker to the World.

Mahoney & Kaktus brought heaps of improvements and new features of their own, turning the tracker into a more streamlined music production tool.

The really big improvements came with Noisetracker version 2.0, which was released in April 1990. In it, the user can sample his own instruments and manage his instrument collection better, including through direct editing.

[Pex] wrote a real-time zooming tool so that the musician can more easily control the amplitude of the samples.

Pex « Mahoney » Tufvesson: « I also added simple support for MIDI. »

« That version of Noisetracker was roughly equivalent to the music program I myself wanted. »

Techworld interview (English translation).

… and it sure was what many people wanted!

---

The last version of Obarski’s Ultimate Soundtracker was released in October 1988, and NoiseTracker 1.1 was released in August 1989 — some 10 months later.
By this time, I suppose Karsten Obarski had given up on the project, because…

(drumroll)

… because EAS, the publisher of Ultimate Soundtracker, actually contacted Mahoney & Kaktus in order to release NoiseTracker officially. You can even see a « (c) EAS » on the screencapture of NoiseTracker 2.0, released in June 1990! I suppose that makes NoiseTracker 2.0 the most legal clone of Ultimate Soundtracker.

We’ll talk more about the legal aspect later, because here comes another contender!

---

Towards the end of 1990, more trackers began to appear, for example Protracker which took over when Tufvesson stopped further developing his program. Noisetracker slowly fell into oblivion as [Pex] took on new challenges himself.

Techworld interview (English translation).

Lars Hamre & Protracker

We’ll scroll one last time down the graph.

Protracker might be the most well-known « old » tracker today — heck, even the MOD format is sometimes referred to as « the Protracker module format ».

Lars Hamre is the one behind Protracker, and just like Michael « Unknown/DOC » Hartmann, it seems that he kept on being on the tooling side of music: after being part of the Ultrafunk collective, he’s currently the lead developer in the R&D team at NewTone, a creative workspace in Oslo, which provides a recording studio, a repair shop, and of course, a workshop where they build all sorts of prototypes for clients, either software or hardware.

Back in 1990, Lars was 21, and somehow he came out of nowhere to swoop the tracker crown: even after he left the scene, many would try to continue improving trackers through the Protracker « brand » — so much that Protracker had its own little Cambrian explosion, with several developers wanting to build on it.
The Soundtracker « brand » was all but retired by the time Protracker came around.

So, why did Lars start Protracker? Let’s ask him…

« Hello Mr. Hamre! »

Lars « ZAP » Hamre: Hello Xavier, and yes, you can call me that 🙂

Email interview in January 2024.

« … or shall I say, Zap/Amiga Freelancers ? 🙂 « 

Lars _« ZAP » Hamre: Well, no thanks. It was a big mistake to even write « Zap » in the Protracker graphics. I didn’t like it then and I don’t like it now.

It was never a nickname that anyone used to address me.

Ibid.

Alright, sure!
With Noisetracker taking the tracker crown at the times, what pushed him to create his versions of Soundtracker? 
Did he find Noisetracker or Soundtracker too buggy, or did he have bigger plans?

Lars Hamre: My brother pushed me. He was the one who wanted trackers to do slightly different things, so he acted as the managing part while I wrote the actual code to implement the changes, of course adding ideas of my own.

It was never about being bothered by or wanting to fix other people’s bugs. For the most part it was only about adding functionality and features. Getting to add whatever we felt was useful and desirable at the time.

Ibid.

Was he in contact with other tracker authors? Did he feel there was a competition to make something better than the others?

Lars Hamre: No, I don’t think that I was ever in contact with any other tracker authors. I was not involved in any swapping of ideas or code with them or anything like that.

There was no real competition going on either. The actual reality was more that we were a very small group of people who mainly did it for ourselves.

Sure, we got useful input from others, and probably included some of that – not that I can remember specifics. People sent bug reports, help requests and so on. General « fan mail ».

Ibid.

It seems he stopped coding Protracker in 1992, and that the name was officially taken over by Peter « Crayon » Hanning. What made him stop?

Lars Hamre: After having enjoyed playing with more proper music hardware than what the Amiga had to offer, the limitations of the Amiga were slowing things down.

Other developers found ways to push it a little further, but it was too little and too late. Commodore itself gave up only a few years later.

The PC was barely there as a replacement in ’92 but things would accelerate quickly from there on.

There was also the realization that, « Damn, I helped create a tool that gave people who had no musical talent the ability to make some form of music anyway. »

The source was uploaded somewhere, and after that I didn’t really check in any detail who chose to develop it any further.

Ibid.

Indeed, just like Armin Sander, Lars Hamre is not the nostalgic kind, as his AMP interview shows:

Lars Hamre: After that I moved to PC and never looked back.
No more Amigas ever, and I’m happy with that. I think it’s utterly stupid to watch all the people who still try to keep the machine alive, making new hardware and os.. etc.. just got to laugh 🙂

(…)

It was fun while it lasted. I have lots of good memories from those years. But it’s over now.. come on, say after me: it is over!!! no more, gone 🙂

From his AMP interview.

Well, that’s it for Lars Hamre then. After Protracker, he produced more musical/creative tools with Ultrafunk, which was absorbed by Newtone as its R&D unit. Cool moves!

After Lars, there were more Protracker versions by several people, notably Cryptoburners which tried to push the enveloppe even further with the 3.xx branch, but that’s not a story for this series of articles.

---

Still, that’s a lot of very similar trackers by wildly different developers, all continuing on the tracks⁴³Pun intended 😎 of a single initial tool, isn’t it?

About forking software

Underlying the whole expansion of the tracker concept⁴⁴After clicking that link, scroll down to « Evolution tree ». Yup, there are a lot more clones/authors not mentionned in this small-ish article., from the original Ultimate Soundtracker by Karsten Obarski and the dozens of forks⁴⁵« fork (plural forks) — 6.2 (software) The launch of one or more separate software development efforts based upon a modified copy of an existing project, especially in free and open-source software. » (« Cambrian explosion » and all that), is the question of whether all of this was legal.

We know that Ultimate Soundtracker wasn’t a commercial success and was dropped pretty quickly from EAS’ catalogue. We also know that Karsten was less than happy with the competing trackers that took his work and got cheered on — hence why he didn’t wish to work any more on it.

It’s difficult to say what would have happened to the Soundtracker concept without the huge contribution from other coders.
Many things are intertwined:

• Would Soundtracker have been commercially successful quickly enough despite its initial issues and shortcomings?
• Wouldn’t the demoscene, and its host of amateur musicians⁴⁶And therefore, a shareware model., be a better target market than « real » musicians, who for their part were more keen on solfege and staff notation⁴⁷You know, 🎼. rather than hexadecimal numbers?
• Would Karsten have continued working on Soundtracker solo if it had been successful? For how long? Would he have had some help?
• Would we still remember Soundtracker & Kasten Obarski today if not for the demoscene, while other musical tools of the times have long been forgotten? Would I even be writing this series of articles?⁴⁸Which is the key factor to me, of course.
• Is success tied to money, or rather to recognition? As an aside, will it be possible, one day, to pay rent & food with recognition?

What was the feeling of each coder about taking code and running with it?

Of course most of them were kids in 1988/89, or at best young adults, so pirating games and software was pretty much the norm; but still, I wanted to address the elephant in the room with the adults that they all became since then.

Mark Langerak: Well… I am not necessarily proud to admit that things like copyright and ownership were not at all a consideration for me at that time :-O

Pretty much anything anywhere was pirated warez, be it games, assemblers, tools… So in that light I gave it not a second thought to take Karsten’s work and do whatever with it, copyright legal or not.

Of course, I was in some sense aware that Soundtracker was in fact not abandonware, but someone’s (Karsten’s) effort for generating income. However /everything/ was pirated in those days and so it went.

LinkedIn interview from November 2023.

Michael « Unknown » Hartmann: Back then, I thought the genie was out of the bottle (other trackers based on the disassembled source were already out), so I figured that having one more wouldn’t do any additional harm.

Nowadays, being older, having the benefit of hindsight, and running a business myself, I see this very differently, but back then, it was just « normal. »

E-mail interview from January 2024.

Armin « TIP » Sander: I did not think about that and – as I can remember – the topic never came up. The excitement and novelty was the salient north pole, and I guess it wasn’t so different for the other authors.

E-mail interview in January 2024.

Thomas « MnemoTroN »: I don’t think I thought much about copyrights back then. We all had our dozens copies of games and applications and since we were mostly in school at the time, no money to spend anyway. I received the source from Unknown and just started tinkering with it.

E-mail interview from February 2024.

Pex « Mahoney » Tufvesson: As the Ultimate Soundtracker was kind of « impossible to buy » where I lived in the South of Sweden, there was no other way to get hold of it other than getting it from a friend.

This kind of music making tool was breaking new ground, and me trying to improve this rough diamond was more in the spirit of « helping myself and my friends to having a good time ».

At that time, I did not have a feeling that this in any way could hurt anyone financially or breach any rights.
Being a 15-year-old boy, « rights and permissions » isn’t a big part of your life.

E-mail interview from January 2024.

Lars Hamre: To be honest, we never really cared much about rights and permissions and fair use or anything like that. Standard procedure was that everyone we knew, including ourselves, freely made copies of anything that was available.

It was never a commercial venture done to make money.

E-mail interview from January 2024.

Ah, those damned kids 🙂

Did they try to contact Karsten about this?

Mark Langerak: I never was in contact with Karsten myself, but I do have a vague memory that John van Dijk (Boil) heard through the grapevine that Karsten was (unsurprisingly) not happy that his Soundtracker was pirated and exploited.

I remember thinking at the time « well, we made it better, so…? ». Which of course is a terrible take on what is, in truth, stealing someone else’s work, but I was still young and naive then.

LinkedIn interview from November 2023.

Michael « Unknown » Hartmann: There was no way for me to reach Karsten Obarski.

E-mail interview from January 2024.

Armin « TIP » Sander: Yes, once, I dialed his telephone number and he picked up. I remember vaguely that my intention was to find out what he thinks about that the Amiga Scene improved on his works. I wasn’t thinking about that we did something illegal, I only thought in terms of creative advancement and expected, quite naively, that he was proud and excited.

I can’t remember what he said, but the sentiment was that he did not care at all. It was an overall very detached, cold, and short conversation, as if SoundTracker was unimportant to him at that time. Perhaps I lost a hero back then.

E-mail interview in January 2024.

Thomas « MnemoTroN »: No, I never tried to contact Karsten. Although one or two of my friends knew people at reLINE in Hanover, which was also for whom Karsten initially started to create his music.

I don’t even know if I knew or realized that SoundTracker was a commercial piece of software back then. I think the original version of it « passed by », but since I’m not into making music I only glanced at it and carried on. The potential wasn’t clear at first.

E-mail interview from February 2024.

Pex « Mahoney » Tufvesson: Me and Anders « Kaktus » Berkeman got contacted by E.A.S., the software company selling the Ultimate Soundtracker, as they wanted to sell our version. However, they filed bankrupt just before Noisetracker 2.0 was released, and the story ends there.

We never got hold of Karsten himself.

E-mail interview from January 2024.

Lars Hamre: Karsten Obarski was a name on the soundtracker screen. Not much further knowledge was available.

E-mail interview from January 2024.

---

Yeah I know, that previous section sounded a bit like a collective « Sorry Karsten, we were just kids having fun! », all contrived and guilty.

Meh.

But I’m keeping it, if Master Obarski ever browses by this page someday. Hi Karsten!

---

Now let’s bring back the fun!

I’ll try to finish this overly long article⁴⁹Some stats as I write this: 51,998 characters, 9,098 words, 34 minutes reading time, equivalent to 43 standard pages in LibreOffice. And I haven’t even started writing down the part about Lars Hamre yet. by not only bringing it all full circle, but also by closing the Möbius strip altogether.

Crossing the streams, again

At the same Datastorm 2017 event as the video above, Pex had the opportunity to interview Chris Hülsbeck — yes, the guy that you read about in Part 2 of this series!

Worlds colliding!

Being an oldskool demoscene event, it’s a deep dive into the early career of Chris Hülsbeck, with plenty of reference to non-demoscene productions and contemporary stuffs.

As an illustration during that discussion, Pex made an animation built on Claudio’s graph⁵⁰🤯, and added this box at the very top⁵¹In his defense, Claudio did add Soundmonitor to his graph’s datasource in 2012, and linked Ultimate Soundtracker to it as an « influence », but this somehow never made it to the visual representations available online.:

Pex: I want to rewrite history and say that Soundmonitor was the first tracker.

Chris: (laughs) Oh wow, thank you.

Pex: Do you agree on this version?

Chris: Well, I think it could be called « the grandfather of trackers ». Because it’s not exactly a tracker, but it has a lot of the concepts of a tracker already.

From the above video.

I’ll add another small « world colliding » anecdote here: Mahoney was a fan of the band Art of Noise⁵²As I read in this interview from 1991., which I understand is the reason behind the name NoiseTracker.
And for their part, Art of Noise were among the early adopters of the Fairlight CMI sampling station 🙂
I really need to get to that article about the Fairlight CMI, and how it ties to the story of Soundtracker!

---

Part(y)ing words

How fun was it to be part of the Soundtracker saga in 1988-90, and of the demoscene in general in those early days?

Mark Langerak: Being part of a demo scene was great fun 🙂 Lots of friends pushing each other in friendly competition to try and outdo each other to make the computer do things no one had done before.

I hadn’t thought much of those days or of Soundtracker until you contacted me. I wasn’t really aware there was a whole subscene of Soundtracker development that came after Soundtracker/TJC Soundtracker II.

I don’t want to claim I started anything (someone else would’ve if I hadn’t) but it feels good to be at the top of that graph 😀
It is like having the highscore at the local arcade 😀

LinkedIn interview from February 2024.

Michael « Unknown » Hartmann: We were all 16-17 at the time. We copied each other’s ideas only after somebody released something new. Every [tracker] release was a competition with others. You created demos and trackers to build a reputation, to have something unique to trade, etc.

(…)

Since we didn’t know we were in the « golden days, » it felt like any other day. It was just our life. Nothing special.

E-mail interview from January 2024.

Armin « TIP » Sander: May be there was some competition [with other tracker developers], and of course being at the front lines of creative development is a great feeling.

But for me – in retrospect – competition wasn’t a primary motivator. I liked the impact it had on the musicians around me and the challenges that came with putting new cool features into code.

E-mail interview in January 2024.

Thomas « MnemoTroN »: It sure was fun. I never had the contacts to others in the scene, though, that was handled by the organizers. So I never really had feedback if others received or saw my productions.

When the 2000’s came and sites like Pouet started tracking all the productions from the scene I realized « Oh, wow, people actually saw the stuff and remember my handle ». That made me smile.

E-mail interview from February 2024.

Lars Hamre: It was a fun and youthful time. 

E-mail interview from January 2024.

(grmbl, I forgot to ask that question to Marco and Pex, oh well…)

---

And that’s it!

This closes what I thought would be a tiny interlude between two bigger articles in this series. But it turns out, when you send a handful of questions to a handful of people, you receive a LOT of content to use in a single article 😅

Thanks to all these guys for still being there to tell the tale, 35 years later!

Greetings go out to Mark Langerak, Marco Nelissen, Michael Hartmann, Armin Sander, Thomas X, Pex Tufvesson, and Lars Hamre.
Thank you all!

And thank you, dear reader, if you are one of the happy few who read the entirety of this overly long article — including the many notes 🙂

Now, let’s get some sun on that skin…

  

   

  

… ah well, first let’s listen to one more module.

---

One last thing

This article was made possible by the awesomeness of scene.org, who hosts Pouët and Demozoo; and exotica.org.uk, who hosts Kestra Bitworld. Those were my three main sources for tracking down the what, the when and the who of this saga.
Huge thanks to all staffers for these awesome online projects: you’re the true l33t!

Thanks also to Obligement for the Armin Sander interview, and to other projects with great documentation, which stirred me in the right direction: modland, Weasel Audio Lib, and Textfiles.com.

---

Other articles in this series:

1. Soundtracking sur Amiga : passion, explications et exemples — The Twitter thread that started it all (in French).
2. Soundtracker origins, part 1: Where in the World is Karsten Obarski? — About Karsten Obarski, author of The Ultimate Soundtracker.
3. The origin of Soundtracker’s MOD format — When you see a Twitter thread with key information, it is your duty to preserve it.
4. Soundtracker origins, part 2: Welcome to Turrican, aah hahahaha — About Chris Hülsbeck, author of Soundmonitor.
5. Soundtracker origins, part 3: Facing a stone mountain — About Karl Steinberg, author of MIDI Multitrack Sequencer.
6. Soundtracker Origins, interlude: The coders behind the Cambrian explosion — Where I get to interview a few key people in the Soundtracker saga.