--- orphan: true --- # Robert Henke ```{raw} html
English Deutsch
``` Robert Henke is a Berlin-based composer, visual artist, and software developer. With a background in engineering, he explores the intersection of sound, technology, and audiovisual art. Henke is a co-founder of Ableton Live, which has profoundly shaped electronic music production. His long-standing project Monolake, launched in 1995 together with Gerhard Behles, became an icon of Berlin's electronic club culture. His work spans ambient music, contemporary compositions, and immersive installations, often designing custom software and machines to generate sound and visuals, which he then performs live. [roberthenke.com](https://roberthenke.com) --- ## Live Setup and Hand Luggage **Benjamin Weiss:** What do you usually take with you for your live set? **Robert Henke:** There are two scenarios. One is the small live setup, where it's really important that I can travel with just hand luggage. That means a laptop, an audio interface, and some MIDI controllers that are easy to replace if they break or get lost—something like a [Novation Launchpad.](https://novationmusic.com/launchpad) The other scenario is the big projects, where I have to take flight cases anyway, like when I travel with lasers. For sound, it's usually just a laptop and a MIDI controller. That's basically it at the moment. **Benjamin Weiss:** Don't you have any special preferences? Is something standard like a Launchpad enough for you? **Robert Henke:** Since I control everything via [Max](https://cycling74.com/products/max), I often just need something very simple—something with buttons and LEDs. I handle the rest. The haptics are important to me, though. For example, in my project with old computers, I deliberately built more hardware to get exactly the right feel. But my general approach is: Either it's exactly what I want, or it's generic. Anything in between only satisfies me halfway. If it's generic, I accept that it's generic. I don't expect a Launch Control to be anything other than what it is. I accept the fact that I'm limited by the interface. When I think about how I use it, I already have the interface in mind. The interface is there before I start building my live set. Still, it has an impact: I know I have these two buttons down there, the fader, and the three knobs. Certain things the hardware dictates lead to decisions about how I build my software around it. If I had different hardware that I liked, I might use a different way of interacting. I align my software with what the connected MIDI controllers will be because I think early on: What do I want to take, and what do I know works? I think about it from the start when I'm designing something: How do I want to control it? The question is there from the beginning, not just at the end when I think, "Oh, now I have to perform this somehow." The starting point is always: What do I actually want to play, and how can I do it? ## Performative Mixing and Control **Benjamin Weiss:** Does this depend on the project, or is there a general setup? **Robert Henke:** It depends on the project. With the old computers project, it was clear that mixing would be an extremely important aspect. The sound sources are very raw and absolutely need filtering and effects. Filtering and mixing in effects is an integral part of what makes the whole thing come alive. That's why I decided to take a small analog mixer on stage—always the same one, so I get used to it and think carefully about which fader is assigned where and how the effects are routed. So the mixing process can be performative. For other things where the mix is more complex—if I'm working with a lot of tracks—it becomes pointless to mix live. It just needs to be well-mixed beforehand. Then I might have a graphic EQ that I absolutely need to control to have meta-control over the sound at a level I can manage in a live situation. Controllability in unexpected situations is important to me. And also that there are as few controllers as possible. I'm always amazed at how little you can do simultaneously on stage. In a studio or jam session, where it doesn't matter if you spend ten minutes adjusting, you can enjoy having access to everything. But when you really need to deliver a focused hour or two on stage, you can't afford to spend five minutes searching. Every decision has to be spot on. My feeling is, for example: You can't control more than eight parameters at once. ## The Cerebellum and Graphic Control **Benjamin Weiss:** How did you arrive at eight? **Robert Henke:** No idea, it's just a rule of thumb. Of course, I can operate a graphic EQ, no matter how many parameters it has—that's a gesture for me. My cerebellum handles it. I have a graphic EQ in every project, and it's always on my old [Doepfer-Faderbox](https://doepfer.de/pf.htm). I've had the same physical hardware for 20 years. I can use it at midnight, fresh out of bed; it just works. Meanwhile, I can do 30 other things because my cerebellum takes care of it. But for anything that requires a creative decision, I only need a few knobs. It has to be simple enough. I always look for efficient meta-parameters or macros that support a gesture. For example, I love playing with the decays of sounds so I can make everything staccato. Those are actually macros. I think: For a certain type of sound in a certain piece, it's interesting if some sounds can be shorter or longer. If I notice that certain things need to be separated because it makes sense to play them separately, then I do that separately. I have a reverb on a snare and a bass drum that decays very long—like an 808 through a compressor. And I have a hi-hat with decay. I might assign one knob to control the decay of the reverb and the decay of the kick, so that the snare and the decay are either really wide in the room or very short and accented. I might put the decay of the hi-hat on a second knob so I have the option: Everything stays short and snappy, then the hi-hat runs over the top, I pull the hi-hat short, and make the rest full again. I think in gestures. In another piece of the same performance, I'd tend to use the same two knobs for something similar to achieve a similar effect. Simply so I can learn it better, but also so it has a consistency. I can make everything extremely short or create an overwhelming moment where everything just opens up. ## Prototyping and Practice **Benjamin Weiss:** Do you try this out constantly while preparing, or do you just think, "This fits together"? **Robert Henke:** I make prototypes. I do a lot of proof-of-concept tests beforehand. Even just with placeholder sounds or in a very simple way—maybe I quickly sketch something with the rack, with macros, or with automations before I even build it. If there are things that are technically difficult to realize, where I'd have to build my own Max for Live device or the mapping would be complex, I prototype from the result. I cut things in Live and draw automations to see: Is this even musically meaningful as a gesture? If it doesn't sound good, I don't bother testing it. I also practice properly for live sets. I prepare, have my laptop open, and my MIDI controllers connected, also to do a technical check. Does everything run? Are there CPU spikes? Does it still run if the computer has had the set open for two hours? For big projects, we rehearse until we drop because I can't afford anything else. For Monolake live sets, I pack my controllers and do a quick check to see if everything works. That's it—no more rehearsing. It's the latest incarnation with material from the last two years. It doesn't matter if a sound isn't exactly how I want it. It's a techno club set in the broader sense. ## Optical and Haptic Feedback **Benjamin Weiss:** When you have devices in front of you, what elements or their execution are important to you for live performance? **Robert Henke:** Large devices have the advantage that you can dance around and still hit the buttons. The disadvantage is that you can't fit them in hand luggage. Clear optical feedback is important to me. It can be cryptic as long as I know what it means. I prefer a red LED that I know what it does over a display that tells me in 20 lines of small text that the value is now 1.371418. Abstract clarity is very important to me. It can be something I potentially have to learn, but once I know it, it must be graspable at a glance. The same goes for buttons. I want to feel: I press it, and something happens. A certain haptics. I love the encoders in the [Cirklon](https://www.sequentix.com/?srsltid=AfmBOoqKoMn8CHAWmIPzY1ixG7fqzYAt6uCGPbadfxsbBHTPvKJ8-CpC) because they're so nicely stepped. If you want to transpose something three semitones up, you can get it into muscle memory. You can do those three clicks without counting or looking at the display because you know how it feels. I actually practiced this for a while: Not looking and just being fast enough. Can I transpose a fifth up and down? I feel those five clicks in my cerebellum and don't count anymore. Just like on the piano, when I know I hit here and I'm a fourth down or play an octave. My idea was to get so good with the Cirklon that I could play the pitch knob and just play a melody in my head because I internalized the clicks. I got surprisingly good at it, but I didn't take the time to excessively practice the C minor prelude on the rotary knob. ## Haptic Encoders and Utility Use Cases **Benjamin Weiss:** Would you be interested in defining haptic feedback for parameters yourself? **Robert Henke:** The idea that an endless encoder could tell me through haptic feedback that it has steps, a center position, or a stop is cool in principle. My problem with encoders is that I generally prefer classic pots with a stop. I know: The send is now closed, or it's half open. But I've never played with a Behringer controller with an LED ring or on a big console like an Avid Venue, so I can't judge how well an encoder with an LED ring fulfills that. For a studio situation, it's great, but in performance, you ultimately have to look. **Benjamin Weiss:** Is there something you miss for live sets? **Robert Henke:** Right now, I'd like something freely configurable that's as small as possible. With a 32-character LED display—not LCD with backlighting, but LEDs. So it's dark enough. The darker, the better in the room. I can't tell people not to use their phones and then sit in the middle of the audience with Christmas tree lighting. A universal device with a display, a few buttons with LEDs, and an encoder. I can think of 100,000 things you could use that for. For my new show, I need to navigate through scenes. I decided to work with a session view because I want to control the general flow of time. At the same time, I want to completely avoid looking at the screen. I have a track in the live set that controls the lasers. There's meta-information encoded in MIDI notes—as cues for the lasers, but also about clip color and clip name. When I load the set, I iterate through all the clips, and then my Max patch knows the structure of my pieces. The first piece has five clips, the second has 17 clips, and so on. During the concert, I'd just say: next, next, next. It would be nice to see what's currently running and what's next. The display shows the scene, ideally with a progress bar so I roughly know where I am. If a complex drone unfolds over three minutes, I can relax, watch, and play with the lasers. Then a button for "Next," a button for "Again," and possibly one for "Back." I'd put those on the lower buttons. I'd turn on the LEDs beforehand so I know: These three buttons are assigned. When I press them, maybe red as confirmation. Another use case is a modular helper. For example, as a tuner or scope. A utility part for a modular system: Take the first jack, plug in a cable, and I have a scope, an FFT analysis, or a tuner to help me tune the oscillator. Or a reference signal, a 440 Hz reference or 1 kHz reference. A noise generator. Or something that quickly outputs a trigger signal. Very basic, not a step sequencer, but five knobs: A press is a gate or trigger. Often when testing, I want a quick trigger signal before I continue patching. If I have this thing on the table: Plug in, plug in, bang. Such utility use cases often fail if they're too complicated. It has to be very simple. ## Programming and Community **Benjamin Weiss:** How should such a device be programmed? **Robert Henke:** Ultimately, there are two types of users: Those who are happy if there's something at all and never think to press an edit button. And then there are those who build things. Ideally, you'd have a website that you open in Chrome. You plug the device into your notebook and push the firmware from the browser without any other app. That's how it works with the [Shuttle Control from Endorphin.es](https://www.endorphin.es/modules/p/shuttle-control). You type in the IP or go to the website, click return, and you have your configuration page. I wouldn't market the device as an "open canvas." The feeling of "can do anything if you program it yourself" is off-putting for many professionals. I'd market it more like: "These five things are included, and they're great." Then an update comes with six things, and somewhere on the third page it says, "Oh, by the way, you can do all this yourself too." That way, you've covered it, but you avoid the feeling that it's something that gathers dust because I never want to deal with it. You unpack it and use it. That's the aspect: That you can use it. The fun has to be in there. ## The Hardware Demonstration **Benjamin Weiss:** What game would you like to have? **Robert Henke:** Pong, of course. Pong and Doom are always the minimum requirements. **Benjamin Weiss:** I'll show you the prototype now. [Benjamin shows a device]. It has 16 buttons with RGB LEDs. You have to go through scenes. Each piece has scenes. **Robert Henke:** The display shows the current scene. I like the buttons, by the way. Are the step buttons different? The haptic feedback is clearer. I actually prefer the top buttons. The difference is small, but the haptic feedback is clearer. I'd tend to prefer the top ones. **Benjamin Weiss:** [Explains connectivity: audio, MIDI, CV/Gate]. **Robert Henke:** I thought I could hold down a step and adjust something. My way of thinking would be: Hold it down, here are the four parameters for the step. As long as I hold at least one down, what I do applies to all. I could really see this as a utility part for a modular system. An automation looper: CV in, record, loop, overdub. Or parameter locks per step with four knobs that output four CVs. One input configured for the clock from your modular system, and the other four are outputs. Clock, reset, and four outputs. That would be cool. If the clock isn't connected, the internal clock runs. If the plug is in, it automatically uses the external clock. I wonder if it needs two MIDI ins and two MIDI outs. I like MIDI, but for modular, you already have other connectivity. As an effects device, it would also be interesting to collaborate with someone like Sean Costello. Write an effect that does amazing things with four knobs. Something special from someone who's state of the art. Something that sounds really good and is quick to use. No one really needs another chorus or flanger, but if it sounds amazing, it would be great. --- *Interview: [Benjamin Weiss (MID)](https://instrument-design.com/)* *Funded by [NLnet](https://nlnet.nl/project/TBD-DSP-Toolkit/)*