Remember 1995? Computers onstage were still a comparatively risky proposition – often relegated to MIDI, more prone than today to instabilities, and absent today’s DJ and live performance apps. Monolake, which is now just Robert Henke, was both Robert Henke and Gerhard Behles. (Gerhard is now plenty busy being CEO of Ableton.)

And then there was Monolake’s PX18 sequencer, a step sequencer – cum – timeline with loads of interesting tracker-style and mathematical-musical features.

We assume that software is ephemeral, likely to be lost to history. But in this case, software is leaving a record that’s actually far clearer than the one left by musicians using hardware.

And the PX18 is an amazing piece of history – one that exists in software form. You can theoretically even download and use it, though you probably wouldn’t want to – it’s somewhat idiosyncratic software and in need of some updates. That does, however, point to the possibility of documenting musical practice in software form as archival activity. (Having just been digging around Soviet-era light organs, software offers some opportunities for media archaeology that hardware may not.)


The PX18 is important historically for two reasons. One, its a glimpse inside the imagination of Monolake. This is, effectively, the template for a whole heck of a lot of what would come to be called IDM. It’s about grooves, but with elaborate compositional permutations, modified in sparing, thoughtful ways. Embedded in the design of this software, you can see that conceit – the ability to tickle brain and dancing body all at once. In fact, I think even “IDM” is really unfair, in a way: this is at the heart of a whole lot of what now defines the Berlin sound in dance music.

Two, connected to that, the PX18 is widely accepted as the spiritual and technical predecessor to Ableton Live. The ideas Monolake used as musicians had the PX18 as a testbed, and gave Gerhard and Robert a clear sense of what they wanted when they founded Ableton at the end of the same decade. One of the amazing things about Live as software is that it evidently sprung into existence more or less fully-formed — you can see screenshots of early builds of Live and they’re fairly identical to what shipped as 1.0. (I should know; I was a 1.0 user.) A commonly repeated misconception, in fact, is that Live was prototyped in Max. While that was later true of specific devices like Operator, Live was written in C. The prototype, then is really the PX18.

But look closer at the PX18 and you see things that actually aren’t in Live. Rotational compositional shifts here are the focus in a way they aren’t in Session View. If Live is about triggering, the PX18 is about variation. I don’t think it would have caught on as commercial product, but that isn’t the point. The PX18 is really its own musical animal, and I think you can hear some connection between how it was built then and how Monolake’s – and Robert’s – music has evolved to now.

It’s also significant in the evolution of this form of music that all those compositional parameters are so immediate and accessible. Part of what may stop this kind of music from being overly cerebral is making compositional ideas something you can play with – something open to improvisation.

That of course has made Ableton work (on its good days, anyway).

But I also admire the extent to which this is an idiosyncratic creation of artists made for themselves. Robert Henke today prides himself on coding his own laser visualizations, and that strictness of form is evident when you hear (and see) his performances. The PX18 is a reminder of why making your own tools can be meaningful. It’s not a matter of authenticity of performance. It’s a question of being able to shape tools to realize your own musical concepts, to embody your ideas in your work.

It’s impressive to watch this ’95 creation still chugging along – even if for no other reason than historical curiosity.

PX18 Step Sequencer Still Going Strong []

But looking at the ideas contained within, I hope it’s a reminder to others to create new tools and ideas. With Max for Live, Reaktor, SuperCollider, Pd, and others, not to mention coding tools that are friendlier than ever, it’s a great time to do so.

  • estevan

    This is a great bit of history. CDM should document all music software changes through the years. For example, Emagic’s Logic had something interesting going on in the day with their matrix view. Newer and newer producers aren’t able to contextualize their software within it’s recent histories.

  • djhokey

    I see that and all I could think of was toying the Jeskola Buzz in the late 90s. Looks like my weekend will be strolling down memory lane with that and playing around with this.

    • Tekknovator

      Buzz <3 The man is at it again. Just installed it two weeks ago and some of the features are still way ahead the commercially feasable, again.

  • Could you make an article about coding resources for dsp/audio stuff ?.I think it’s still difficult to find a proper tutorial or guide to get into vst/dsp development. On the other hand I have started with web audio and there’s a lot of resources and it’s lots of fun!

    • djhokey

      KVR Forums are pretty good: Also Martin Finke’s tutorial has a nice step by step

    • Random Change

      You might be interested in the books by Will Pirkle:
      Another great resource once you have learned some of the basics are the DAFX books. You might also benefit from having a book like the Computer Music Tutorial by Curtis Roads on your desk.

      And as a rather obvious note of warning: Writing plugins is a lot less interesting than toying around with DSP algorithms. There is so much menial work to do, so much overhead that you incur by having to use one of the frameworks which are out there. JUCE makes is somewhat easier to roll your own, and on top of that, it’s even agnostic to the backend technology (like VST, Audio Unit, AAX, etc.). What I mean to say is that you do not need to start with writing plugins and that it might even turn you off DSP becasue most of the stuff you’ll have to do even for the most simple effects (like a static delay) is boring as heck. You could play around with DSP algorithms and coding techniques in a more friendly setting, say, Python with some suitable library (pyo perhaps).

    • Rory Walsh

      The Audio Programming book is quite a resource. It covers a huge amount of topics ( You may also want to check out the following course on ( The two instructors are serious heavy weights in the world of DSP.

  • redgreenblue

    I would go even further back and give the TIGER part of Dr T’s KCS as a precursor to Ableton. You could trigger groups of midi regions from the computer keyboard in real time or in sync and even construct a song by having the program remember the sequence of things triggered. I loved it. Yet somehow I’ve never warmed to Ableton.

    • Huston Singletary

      Opcode Vision as well. Especially the earlier incarnations. Stepping and constructing together entire MIDI pieces into song form via the characters on your computer keyboard and having the ability to fold and create a single sequence out of them in 1993 was fantastic.

  • wndfrm

    nice article, thanks for sharing this!

    slight aside – “That of course has made Ableton work (on its good days, anyway)” .. is this a thing? i think ableton is pretty stable (as these things go.. there are always exceptions), at least i’ve never had any issues that i consider endemic or systemic to the host.. maybe it’s just due to my resistance to ‘early adoption’.

    i would be very interested in seeing the history of logic, i was an emagic user for many years, but found the apple versions to be bloated.

  • heinrich zwahlen

    Absolutely, i was using Dr T’s from 86-89 on an Atari and when Live came out i finally got my old workflow back in many ways after having been limited to work with mostly linear compositional tools like Pro Tools and Logic for a long time. I totally feel that Dr T’s offered the blueprint for many of the Ableton concepts.

  • nesnduma

    “A commonly repeated misconception, in fact, is that Live was prototyped in Max. While that was later true of specific devices like Operator, Live was written in C. The prototype, then is really the PX18.”
    Well, anyone can see that PX was programmed with Max.

  • Will

    Love this stuff. The design reminds me a bit of Slifty—a great beat mangling app from one of the SMEX developers (who went on to FXPansion and never finished Slifty). We’ve definitely lost some good software along the way. I recently rebuilt a DOS machine to run Voyetra Sequencer Plus and I’m totally loving it. Undo would be cool though!