Reflections on the Loop summit

The last weekend of October 2015 found me at Loop, a "summit for music makers" organized by Ableton AG, the makers of Live, and taking place in Berlin. The summit gathered a wide range of notable people from many different intersections of music, technology and creativity, including musicians, producers, engineers, designers, researchers and more. Ableton did an amazing job with the organization and production of the summit, and when I arrived to register they gave me a document which included a description of what they wanted the meeting to "be about":

Basically it comes down to this: music making the days is largely a solitary pursuit. This of course has its upside; just think of the immense costs, mountains of equipment and teams of specialists it took to record an album a mere 20 years ago - something that any one of us can now do by ourselves on a laptop. And while we think these changes are for the better, this also means that collaboration, cooperations and exchange are now more important than ever for us to grow and learn as musicians.


The summit opened with a 30 minute performance by King Britt, Claudio and Gloria Jensen. KB ran sequences, pads and more from an Ableton Live instance controlled by their Push controller/surface, Claudio added voice improvisation with processing and a bit of live looping, and Jensen played a (possibly modified) violin that was also being treated with a relatively light level of effects. The music typified the kind of motion-free, circling-in-place work that I tend to hear a lot from people performing live with Live, which is to say that I neither liked nor disliked it. There's a lack of compelling form, but some interesting sounds and moments. Most interestingly, there seems to be very very little eye contact between any of the 3 performers, which seems odd to me by way of comparison with other improvisatory performances that I've seen. I talked to Gloria Jensen about this later, and she gave me a sense that for her at least, she was mostly just listening to the sound and responding to it, rather than actively looking at her fellow performers.

After this, Robert Henke took the stage for a keynote talk called "Success == Failure". Henke was one of the original founders of Ableton but left relatively early to pursue his own music and instrument designs. It was an interesting talk but the two highlights for me were (1) his remark that "Success points to the past, failure points to the future" and (2) his playing a long section of 10cc's "I'm not in Love", an iconic pop rock song from 1975 and talking about how the band made the record. I knew that Godley and Creme were quite interested in experimental stuff (Lol Creme basically invented the original version of what later appeared, as both the e-bow and Michaels Brooks' "infinite guitar" (played with such visibility by the Edge of U2). But it was amazing to hear the process they used to make this record, going deep into what were essentially music concrete methods to make a pop song. Henke used it as an example to encourage people to take more risks, be more daring, do more experiments.


First thing Saturday saw me on a panel about Open Source, along with Gianfranco Ceccolini, Marije Ballman and Soledad Penadés. Gianfranco is responsible for the Mod project (more about this below). I had met Marije some years ago at a couple of Linux Audio Conferences in the past, and she has been deeply involved with both Supercollider and a lot of custom code and hardware design as part of her work on instrument design, sound installations and more. Sole currently works for Mozilla as both an evangelist and advocate for web audio technology and solutions. Our panel was somewhat interesting, though I felt it was a bit too short after our 4 introductions. I played my typical role as the grumpy old man, trying to get people to understand that although I believe passionately in open source development, it doesn't generally solve any of the hard problems with software development. Sole was much more optimistic and positive than me, offering questioners useful suggestions on things to do. The questions were not particularly deep; one person wanted to know how to get more developers and users to try his new tablet control surface for Live.

I spent the rest of the day attending other talks and panels. This included listening to Matthew Herbert basically lament the current state of the world and apparently conclude that it is all so bad that we really probably shouldn't make any more music (he noted that 75% of the iTunes catalog has never been downloaded). A session after that saw Roger Linn (Linn Drum), Stefan Schmitt (Native Instruments), Gerhard Behles (Ableton) and Carla Scarletti (Kyma/Capybara) talking about instrument design (they all meant "electronic control surfaces that drive electronic sound generation"). I later saw Roger Linn demonstrating his new Linnstrument. It is a very exciting device on some levels, but on another is just another iteration of FSR technology. He seems particularly uninterested in the role of the sound synthesis/generation part of things, and to me this seems like a key issue if you want to build really expressive instruments (which Roger does want to do). Coupling sophisticated sensing technology to uninteresting sound generators isn't the way forward. He did congratulate me for Ardour - "saving the world from Protools" was how he put it. Stefan Schmitt seemed to me to have a very interesting take on this whole area, and I would have liked to talk to him at greater length, but I never found the time to track him down.

A final panel talked about genre in electronic music. This was a very interesting discussion. Dot, Electric Indigo and Peter Kirn all presented short talks about genre. Dot focused on how the sonic characteristics of a performance space affect our perception of the "genre" we're hearing, which was an interesting idea I had not come across before. Peter Kirn was very erudite and witty in his exploration of how an old english folk tune got transformed as it became a part of Appalachian folk traditions, and he closed with a remarkable video showing some sort of "traditional" folk dance from the 1950s or 1960s, overlaid with some massive house beats. It was simultaneously funny and profound, pointing to the things that link us together as humans, including the desire to dance and an appreciation of "simple" rhythms. Although this panel came to no real conclusions, it was a thoughtful exploration of some aspects of the question.

Saturday night saw most of us "presenters" at a dinner. It is always a delight to eat good food and drink good wine in the company of interesting, intelligent people, and this evening was no exception. Rebecca Fiebrink sat opposite, and told us a bit more about her work using machine learning techniques in a "meta-instrument building toolkit" called Wekinator (also open source). Gerhard Behles (CEO) of Ableton was on my left, his usual interesting self with some rich insights into the path of his company and technology, not to mention other aspects of life that he and I have both experienced and shared in the past (and present). By the end of the evening I had probably had just a little too much wine and was pontificating on the nature of evil, which I'm hopeful my neighbouring diners will rapidly forget.


At 10am on Sunday, after some Loop attendees had been out all night and even the conservative ones (like myself) had been up pretty late, 3 drumsets stood on the stage of the main hall, ready for a panel titled "On and off the grid: acoustic drummers and electronic music". Katharina Ernst, Kiran Gandhi and Zach Danziger took to their drummer stools and got the day started in fine style. Gandhi is a young drummer who has worked live with MIA, and has apparently boundless energy and enthusiam. Ernst seemed quite mysterious, and it was only later when she played solo that we got some sense of her quite unique style and skills. Danziger, who almost unbelievably had flown in overnight from a gig in Minneapolis, demonstrated his approach that he terms "playing in the cracks". After the brief performance, Dennis DeSantis proceeded to ask them all interesting questions about how their roles as acoustic drummers works in the context of electronic music. When asked about the somewhat common idea regarding playing "ahead" or "behind" th beat, Danziger made a very interesting point that he tends to feel that the question makes no sense to him, since merely by varying the sonics (e.g. bringing the snare or kick drum to the front), he can change the sense of "ahead" or "behind". There was a bit of a diversion (for me) into triggering technology. They closed with another trio-drumset improvisation which would have been great except that Gandhi seemed to lack the sense to really back off her own playing enough to really hear what Ernst or Danziger could do. That was disappointing because although she's a good drummer, her style is much more conventional than either of the other two. I talked to Danziger afterwards about a couple of questions I had regarding the whole area of "playing like a machine", but at the time I didn't know how jet lagged and tired he was. It would be interesting to talk with him under better circumstances, especially his joking about the "f" word, which in this case means "fusion".

I headed upstairs to catch the last part of another talk about instrument design, this time with people (including Rebecca Fiebrink from dinner) building much more "experimental" instruments than the quartet from Saturday. Joe Malloch tore into Roger Linn for his remarks about a lack of expressivity, making the excellent point that the piano is an instrument which lacks any of the pitch-related subtlety of any stringed instrument, and yet which caused us to explore a huge area of music that might never have been without the piano's almost "sequencer"-like qualities. I continue to feel that these "new" instruments focus too little on a rich sound generation model, and Malloch (along with Marije who has worked with him and now works at STEIM, itself a hotbed of "new instrument" design) explicitly said that he'd prefer a complex controller and a simple generator to the other way around. I don't agree with these guys, but I'm not building instruments. Talks like this make me wish that I did.

A little later, Thor Magnusson gave a great talk on his experiments with live coding. Most of the live coding stuff I've seen has been frankly uninteresting: a mess of Supercollider or Chuck or some other code all over the screen, accessible only to initiates and just some kind of mystical totems for everyone else. Magnussion's Xiilang is quite different, and utterly amazing as a conceptual and practical piece of work. It is open source (check github) and from this brief exposure seems full of potential. It is, of course, built on top of Supercollider, the giant that hides behind so many of the projects in this sort of area. Marije mentioned to me that SC is now more at home on Linux than OS X, which is deeply ironic given that Apple originally hired SC's developer James McCartney. Several people also mentioned that SC "needs" JACK, which I also found puzzling.

Sitting around in the same room, I then continued to listen to Peter Kirn and Matt Black (Coldcut and Ninjatunes Records) discuss "Why electronic music needs to be a lot wilder". This somewhat wandering discussion left me a bit confused. I was very sympathetic to Black's basic point that "four-to-the-floor" house music has become homogenized, bland and much too dominant in the music scene of today. But the more I listened to (and later reflected on) the conversation, the more convinced I felt that (somewhat like me) Black is just growing older and feels marginalized by imagined, unknown but possibly real trends that he is not connected to. I did ask a question that drew comparisons between what he described (how early house/techno was a lot more "wild" than most of today's equivalents) and the way that Miles Davis' wild invention of electric jazz and fusion gave way to the bland mess of highly skilled but ultimately musically empty fusion by the 1980s. I really don't think that what Black is disturbed by is unique to electronic music, but is common to all music, across all traditions. I don't have any idea when the next "wild" era will arrive.

A final session for the day was called "Again & again & again: loops and repetition in electronic dance music", and saw Henrik Schwarz, Hillegonda Rietveld, King Britt and Recloose in conversation with Tony Nwachukwu. Each panelist got to play (part of) a track and then talk about it from the looping/repetition perspective. Rietveld played a track that frankly sounded amazing in the room we were in, but hasn't held up when I listen to at home. She had a copy of a book called "Repeating ourselves" by Robert Fink which I read last year, but didn't really draw from it too much in the limited time she had. Henrik Schwarz played an interesting piece of his own that uses white noise sequenced using a rhythmic pattern and practice derived from an african tradition. Despite the white noise as the only tone, it was very propulsive, and his demonstration of how he has used this pattern along with phase shifting was quite exciting as we saw form emerge naturally from something quite simple. King Britt got to play a short section of a new piece from him, that drew from both the opening to Baba O'Reilly by The Who and various Phillip Glass pieces. He seemed a little frustrated that we got to hear only a few minutes of a 13 minute piece ("That wasn't even the intro!"), so it is hard to know how this piece evolved in the long term. Recloose played a piece of a 1960s soul track and then showed how he had looped parts of it and used the loops in a work in progress. Finally, the moderator played the opening of E2-E4, the track by Manuel Gottsching from 1975 which many people (including King Britt) retroactively credit with being the birth of what we now call techno or some subset of EDM. I talked to Nwachukwu afterwards and pointed out that I first heard E2-E4 in the late 1970's within its original context, just another record from the "berlin electronic" school, and that at the time, most people didn't really think that it was very good and would be forgotten. It was only when the 80's dawned and some DJ's (Dave Mancuso, Larry Levan) used E2-E4 in clubs that people like King Britt would hear it in an utterly different context, and for most people it wasn't until it was remade as "Sueno Latino" that it became part of that scene at all. It is curious to me how a work like this can cross over between genres, where it is largely ignorable in one and a "classic" in the other.

And then ... we all descended back into the main hall for what turned into an Ableton product announcement. Dennis DeSantis closed out Loop by (correctly) pointing out how they had worked hard to make the summit about people and things other than Ableton. Then he said "But...", and the hall laughed. Gerhard Behles took to the stage to start an Apple-esque launch of the new Push 2, version 9.5 of Live and to reveal a new technology that Ableton calls "Link" which provides tempo and timeline synchronization between both Live instances and iOS applications. Apparently the Link SDK will be freely available so that other developers can use the same technology.

In general, I have to say that the whole presentation made me feel a little sick and rather angry. Although I understand that what was presented was useful technology for many people in the hall, their standing ovations made me uncomfortable. I also felt that as funny and personable as the 3 Ableton staffers who presented these things were (and Jesse Terry was pretty funny), they also used misleading language as a "sleight of hand" to talk about the new developments. This didn't bother the majority of the audience, who audibly gasped at seeing someone sampling vinyl "directly" from the Push 2 and were clearly taken with the almost complete absence of a computer from the promotional video about it. It didn't bother them that Michaela Bürgle started off talking about "playing in time" and then managed to switch the meaning of this to "starting in time" as she revealed the Link technology that still doesn't do much other than ensure that Live's beat/bar quantitization can cross machine boundaries.

However, it did force me to reflect on a fundamental aspect of product development. The JACK community developed technology back in 2005 or earlier that could do what Link does (and a whole lot more besides). But the developers failed to package it, failed to make it usable for anyone other than a small group of tech-oriented tinkerers, failed to give it an inviting face. In addition, the developers failed to understand how significant platform is when trying to create products for people. Designing the most awesome thing in the world is no help if it doesn't run on the platforms that people want to use (for whatever reason). Ableton not only chose OS platforms with huge numbers of users, but has proceeded to develop its own product-centric platform that provides compelling reasons for users to want to remain within its walls. The open source community shouldn't be trying to follow or copy what a company like Ableton does, but there is a strong lesson here: our technology is not as important as we like to think.


I took the time on Monday to visit with the people behind the nearly-ready mod technology. I had heard vague rumblings about this from various people but had not bothered to read up on what it really is about. I am almost glad I did not, because the demonstration that Gianfranco gave me was a lot more impressive than anything I would have picked up online.

The mod is basically an FX pedal. However ... it runs Linux (and JACK!), and has incredible web/cloud integration allowing the user to download new plugins, new settings, new combinations and routings of plugins and more directly from the web. The whole device is managed via a web interface (so it can be done from any web-enabled device), and it even exists as a completely standalone application that can run on your computer - you miss out on the foot pedal hardware if you do that, of course. For plugin developers, the integration is almost as seamless: upload code to mod servers, where it is built and automatically merged into the web-centric system, available for all platforms. The plugins are not limited to FX: you can run instrument plugins and send them MIDI.

Gianfranco, Felipe Coelho and other developers have done some amazing work here to take the underlying aspects of the Linux audio ecosystem and package them into a form that makes a compelling product for actual users. I have no idea how the outlook for FX pedals is in the long run - I can imagine a downward curve over time, but I've been wrong before. Meanwhile, for anyone interested in an insanely powerful and flexible pedal backed by a similarly powerful web-centric support infrastructure, the mod is something you need to investigate.

One Other Thing

I'll be 52 soon, and I grew up in a world in which roles for women were still very limited. It was incredibly encouraging to see a conference at which, despite the male dominated audience, panels and presenters and performers were so much closer to 50/50 men and women. It was really noticeable to an older guy like me that I was actually being given the change to listen to women (like Dot, like Electric Indigo, like Kiran Gandhi) talk authoritatively, insightfully, imaginatively about what they do in a context that wasn't actually about women at all. I would like to congratulate Ableton for having gone the extra mile here, at least compared to habits and practices I've seen as I've been alive, in finding and bringing smart, intelligent, relevant women into a context where they could just be smart, intelligent and relevant people. So refreshing, so excellent. When I think about this kind of world as the one my daughters will take over, it is exciting and right.

Thanks for the great article!
You’re right that expressive controllers need complex sound generators (take the Madrona Labs Soundplane controlling Kaivo or Aalto, for example).
Do you think that SuperCollider should become less dependent on JACK? The devs don’t seem to be moving in that direction at the moment. What did you find odd about SuperCollider “needing” JACK?


I would have expected JACK to be merely be one option in how to move audio+MIDI in and out of Supercollider. People made it sound as if it was the only option.

Great article, thanks Paul. I’ve heard the argument about dance music not being wild before, and you are absolutely right. It has nothing to do with dance music. People are doing incredible things in dance music all the time and I’m regularly hearing tracks that bring something new to the table. But that doesn’t mean that that music is the music that gets heard by the masses.

And for your last point, about women in music, out of the four lecturers I most respected at university two were women. And one of them encouraged me to use linux and Ardour in my music. Needless to say I’m very grateful!

Some great thoughts here. Thanks for the post.

I looked it up and the album E2-E4 was recorded live in a studio in 1981 and released in 1984, so you could not have heard it in the late 1970s :slight_smile:

Manuel Göttsching’s first album was Inventions For Electric Guitar, released in 1975.

@beluga: that’s a good point :slight_smile: I think my problem is that “late 1970’s” for me blends into 1981 (the year I finished high school), and I tend to fail to make a distinction. I’m now wondering/pondering if the altered chronology makes any difference to the way I should think about this. I distinctly remember hearing the album and reading reviews and the general verdict from people who were fans of the general krautrock/berlin-electronic school being that it wasn’t a very strong album, especially by comparison with, say, Blackouts (1978).

Thank you for the correction - I may edit the post at some point to reflect reality.

Hi Paul,

The thing that started to annoy me as I read through this aside from the Coldcut/Ninja Tunes guys, where was the rest of the early techno underground ,mostly referring to the Detroit scene guys from the late 80’s early 90’s - Juan Atkins, Jeff Mills, Stacey Pullen, e.t.c (could also to the UK, or Berlin with the Tresor artists). There seems to be mostly academic representation. Those types always seem to be a step back and are thinking alot some doing some very interesting performance work, but not pushing the industry along (from underneath)… The guys I have mentioned are like the Miles equivalent in the relevant scene, they aren’t obsessed with dance floor (although very welcomed on the dance floor) but more about pushing the boundaries of the electronic form, flipping what they do when they feel they have reached the end of the current experiment (especially Jeff Mills - recently doing arrangements with symphony arrangers and performing with symphony orchestras and drum machines)…

Some very interesting points you have raised.

I have noticed that Roger Linn does seem principally focussed on being an interface designer… But he is very good at that… The Tempest drum Machine that he developed with Dave Smith is amazing…
I see your point about Jack too… On it’s release I was thinking … this could and should change everything… but then the marketing was never there , and I don’t think there were enough discussions between musicians and developers… When android tablets came out … I also thought… here comes Jack’s day… and again, packaging and marketing weren’t there (and latency on Android was particularly bad)… I really think though the missing link is conversations between musicians (particularly electronic ones) weren’t there. I think Nick Copeland had the idea, but without other software devices (drum machines, samplers e.t.c.) … the ecosystem which could have been like the soft open “system 500” modular system… and if implemented properly could have been extended to use multiple processing devices, (e.g. P.C, tablets, phones e.t.c…)

oh well… maybe “Live Link” will kick everyone in the open sources butt to realise that we already have that functionality and to start utilising it…

But I think it will end up being like touch screen devices, it seems the corporate developers can sell the things better (e.g. bitwig 1.3)…

Paul, I wondered what’s your conclusion out of this, how will it affect the development/packaging/distribution etc. of Ardour and JACK if at all?

No conclusion. More a case of understanding some missed opportunities, and a possibly more aggressive pursuit of instant-usability over tweakability if anything clever ever comes along in the future.

If you have this officially yet unreleased Windows build of Ardour, not offering it for download to anyone interested (no limitations, no obligations) is such a missed opportunity, maybe.

Offering unsupported software that is known to have some issues to people expecting things to just work represents the very OPPOSITE of the conclusion I reached during the Ableton product announcement.

Yeah, better give that same software only to people who pay. That’s quite an interesting conclusion you draw.

offering it to people who subscribe to supporting the project is quite different than selling it. Nobody pays for it in the sense you describe.

I would be interested to hear you pontificate on the nature of evil.

I enjoyed reading this. Thank you for sharing your thoughts and experiences at Loop. :slight_smile:

Re. FX pedals: Maybe they’ll decline, but I still think there’s something unbeatable about the physical presence of a mechanically crude and abuse-resistant device that can alter/augment your sound. Handling it is more subconscious than having to worry “did that touch register right?!” on a touchscreen, or having to take your hands off your instrument to fiddle with knobs on, say, a MIDI controller.

I actually find myself going somewhat back to more fixed-function devices, simply because computers can be the Hell beneath Hell: Getting them configured to work properly is Hell; keeping them working across OS-and-what-have-you upgrades is yet another, deeper Hell! :slight_smile: I may not be 52, but old and grumpy – sure thing!

Thanks for the writeup, Paul, and thanks for Ardour. Technology may not be as important as we think, but it’s darn important nonetheless.