New SOMA REFLEX Synthesizer Promises Unparalleled Real-Time Control Over Sound (Sneak Preview)

SOMA Laboratory, creator of unique instruments like the Lyra-8 Organismic Synthesizer and the Pulsar-23 Drum Machine, shared this sneak preview of the REFLEX, an instrument that promises to let you play synthesizers in a completely new way.

For most synthesizers, you create a patch and then you play it. The core idea of the REFLEX is to give you an unparalleled level of real-time control over both timbre and sound, so that you can create new timbres as you play.

Here’s what developer Vlad Kreimer has to say about it:

REFLEX is the first complex synthesizer where the main goal is to give musicians direct access to the synthesis core, instantly playing dozens of synthesis parameters as artistically and intuitively as we play notes and velocity on, for example, a violin.

To make this possible a unique multidimensional sensor-controller was developed. The main part of the sensor-controller consists of 16 small hemispheres, arranged ergonomically in a way that provides comfortable and simultaneous access to all of them for the fingers of your hand.

Each hemisphere contains a tiny version of a high frequency transmitter-air-receiver chain, where the human finger functions as a kind of obstacle that affects the connection. It captures the position of your palm and fingers from 16 points of view. The sensor data collected by each hemisphere is connected in a smart way with one or more synthesis parameters. Combinations of several sensors can be processed in a separate way and attached to a unique set of synthesis parameters.

Sensor output is processed at a speed of 23 kHz, which is higher than the audio range, and reaches the synthesis core with latency of no more than 50 microseconds. Tiny movements of your fingers can drive the very synthesis core, letting you play with the process of synthesis directly and instantly.

To make this approach possible, special synthesis algorithms were developed that I call “pure math synthesis”. The main idea of this type of synthesis is the absence of any wavetables, samples and anything that is a kind of “frozen past”. Instead, it uses special dynamical recursive equations and high-resolution calculations that generate sound here and now, and which are very sensitive to even the tiniest signals from the sensor-controller. As a result, it gives you the unique experience of playing a very complex but also very responsive live real musical instrument that literally breathes under your fingers.

The hardware part of REFLEX is also unique. Starting the work on a digital synth I was very concerned with how to bring our beloved analog LYRA-like sound quality into the digital realm. Existing digital-to-analogue converters available in industrial volumes did not meet our demands. So I decided to develop my own conversion system. It works with several Pulse Width Modulators with direct raw outputs, scaled relatively to each other in order to gain the necessary resolution. At the end, there is a special analog part that finalizes the signal. This analog part is built with discrete through-hole components and discrete transistors (there are no OPAMPs in the circuit!) without significant long negative feedback loops (the famous Hi-End approach). The analog part uses unique germanium transistors and WIMA capacitors that give warmness and nobility to the sound.

REFLEX is also a multiprocessor system where each of the six voices is actually a separate mini computer (a special dedicated high-performance microcontroller) with its own clock crystal and analog part. REFLEX has no global FX but instead six separate FX units built in each voice, each of which can be driven according to the current state of its voice and the note it plays. Each FX unit also takes part in the synthesis of its voice including complex feedback.

The end result of these innovative solutions is a very high sound quality, actually rare for a digital instrument. The analog mixing of voices output makes it possible to hear and distinguish each note as brightly and as separately as in old analog synthesizers because each note actually is generated by its own complete monophonic synth.

From the user interface side, REFLEX strikes the perfect balance between amazing possibilities and simplicity. Despite the big underlying complexity of this synthesizer, it is surprisingly easy to use. There are just 17 analog-style knobs with fixed functions (no multifunctional knobs). There is no menu diving. Everything is as simple as possible, letting you focus on making and playing music!

The SOMA REFLEX is still in development – but it’s clearly going to be one of the most innovative approaches to synthesis in a long time.

Check out the SOMA REFLEX sneak preview and share your thoughts on it in the comments!

37 thoughts on “New SOMA REFLEX Synthesizer Promises Unparalleled Real-Time Control Over Sound (Sneak Preview)

  1. Holy bejeezus, this may be the coolest thing SOMA has done yet.

    I think it’s safe to say that this will be the most innovative new synth of 2021.

    1. It’s an early prototype, so it will probably be 2023 until you actually hold this in your hands. Pre-order waiting time for Pulsar-23 is several months right now.

  2. This is some mad scientist level stuff. When you see all the junk and remakes that are coming out, you realize just how special pieces like this are. SOMA is always doing amazing things, but this looks super exciting. Can’t wait to see more.

  3. This guy single-handedly demonstrates that there’s an alternative to Behringer’s knockoff factory and Roland’s endless variations on its hits from 30 years ago.

    The stuff he does with the Pulsar-23 is absolutely insane, too.

  4. I like the idea but what about MPE or a new keyboard that connects to the synthesis core or a overlay that could be mounted to keys. Simple the control surface could improve but nice idea and sounds

  5. I’m glad to see something that focuses on real playability & expressivity – even if I have to give up a hand to achieve that!

  6. Why is the violin the height of expressivity? ‘Unparalleled Real-Time Control’ – what exactly does that mean? There is always some lag between the brain, the finger and the sound output. Even if you implant an electrode into your brain there is going to be a lag. Can you control all parameters with a macro knob control? Well, it’s not ‘real-time’ control.

    1. With brain plasticity, there is quite a bit of behind-the-scenes calibration that makes it “feel” immediate. With synths, reducing latency is all about making it feel like our already-calibrated sense of “now”.

  7. Feels like a TouchOSC controller could possibly achieve a lot of this for any synth with some good sysex parameter edit messages configured.

    1. I agree. However, the above audio scan rates, and ultra-ultra-low latency, and other bespoke synthesis elements do raise this into a different realm.

  8. The idea of bubbling up controls is great. But there’s a reason these parameters are usually controlled by knobs, it’s easier to dial in precise values. While these hemisphere controls look tactile, they also look incredibly imprecise. Anyway, good luck making this work.

    1. One could produce the very same criticisms when looking at a guitar, a violin, or literally any other acoustic instrument. The idea that “imprecision” and “musicality” do not or cannot coexist has been proven as invalid time and time again; indeed, so much of the allure of vintage synth equipment can be narrowed down and attributed to the existence of “imprecision” within the equipment
      .

      1. A fingertip on a 1/2” hemisphere is going to offer degrees less of subtlety than the length of a violin string. The touch ribbon on a Hydrasynth might be comparable, this isn’t.

        1. The type of modulation that is occurring along the length of a violin string (pitch/note modulation) isn’t the same as the types of modulation that appear to be occurring on those 1/2″ hemispheres. If you wanted to accurately compare the type of modulation occurring along the length of a violin string you’d be comparing it to the length of the keyboard that is being used to control the pitch of the REFLEX.

          Meanwhile, the types of modulation that is occurring on the 1/2″ hemisphere are occurring on even smaller surface locations on the violin (where your find makes contact with the string on the fretboard, where your bow makes contact with the string), yet they offer such incredible degrees of subtlety.

          It’s almost as if the tactile contact point surface area alone isn’t the ultimate arbiter of the subtlety of modulation available on any given instrument, and that perhaps experience and skill with an instrument or a tactile control point may indeed be a factor in how much subtlety can in fact be wrung from that instrument. by the musician.

          By your measure the kick drum should offer more degrees of subtlety in regards to playability than any other instrument in the band. We both know that that isn’t the case at all.

          1. It is fun to compare new controllers that are the brainchild of one person, to expressive acoustic instruments that are the result of a long, refining process that represents the combined efforts of generations of makers and players.

            It is also fun to compare level of expression of a new controller being played by someone who as dinked with it for several months, with the degree of expression possible on an acoustic instrument played by a virtuoso with 30 or 40 years of dedicated practice.

            This instrument has already proven that it can make nice sounds and change it’s qualities nicely in real time. I think a person could work with it and it would probably be a worthwhile effort.

            All instruments have their different qualities and features. They don’t all have to be “impressive” in the same ways.

            As for precision, after-touch is a pretty good example of a control source that’s difficult to control with precision.

          2. My measure is: how reproducible and controllable any desired modulation? You take a fat finger on a small curved metal piece, you’re not going to get subtle results. But ignore logic all you want. There’s already many synths out there that have every parameter mapped to a controller, just with more accurate results. But sure, someone might make it work for them.

            1. “You take a fat finger on a small curved metal piece, you’re not going to get subtle results.”

              So you’re saying that nobody is getting subtle results on any stringed instrument that has an even smaller curved metal piece as the sole point of contact than what the REFLEX offers?

              1. I think this is an open question. With all of these types of controls that depend on subtle, tiny movements, it is down to several factors: quality & consistency of the controls, range & resolution of parameter, years of consistent practice required to gain mastery, etc.

                I agree with the Snazzy Wonton that out of the box, those controls are probably going to be easy to use, but difficult to control with precision. After some time, the user will gain greater control.

                As an aside, I love using a breath control because it lets me use a modulation source that I’ve already mastered from another discipline (I’m a brass player). There’s a breath controller in development called Photon which uses breath pressure plus a number of other sources (lip pressure, tongue position, etc.) even though these are going to be very subtle movements, I have already developed them from my years of training in a related musical use.

  9. This has some mold-breaking aspects: ultra-fast/responsive, high-fidelity, deep DSP design, experimental possibilities– all fantastic!

    If the ergonomics are engineered right, it could work really well. Is it always RH on keys, LH on controls?

    Snazzy Wonton makes a good point above about it possibly being difficult to control with precision. But perhaps that is not the point. In the demo, it reveals a kind of soundscape where because pitch is handled by the MIDI note #, then many other aspects are modulated (impressively) in realtime.

    The meta concept of putting the entire synthesis palette under your fingers is compelling and weird (in a good way).

    1. Theremin only gives you control over pitch and volume.

      This is designed to give you the same level of control over all the synthesis timbre parameters in realtime.

  10. So it basically creates some kind of random chaos by touching little nipples. Meh.
    I prefer morph groups myself.
    Fx per voice on a polyphonic sound isn’t exactly the bees knee either.

  11. Not sure what “unparalleled level of real-time control over both timbre and sound” is supposed to mean. It’s just marketing blah blah.
    Anything that isn’t pitch is timbre by definition.
    MPE is way more though through (controlling several parameters of timbre and pitch with the same finger on the keyboard).

    1. It’s not necessarily “marketing blah blah”. In the synth world we generally understand “timbre” to refer to both oscillator waveshape (or other various oscillator focused functions) and filter. “Sound” is probably purposely vague– but could refer to things like delay/reverb settings, chorus depth/rate, various envelope and LFO settings, pan, stutter, etc. — none of which are not “timbre by definition”.

      MPE is th(or)ough when it comes to hands on keys– but it’s not necessarily better. Having one-handed playing keys while another has bunches of fixed function knobs and touch-based sensors seems pretty thorough to me.

      I’m not saying this is better or worse than MPE (I don’t think you are either). It’s certainly valid to compare them. Would a person have a better experience with one or the other— it would depend on what features they are after.

      1. Yeah, apart from the precision issue, there’s the “unparalleled level of control” claim. Many synths already have knobs for every (or most) controllable setting. This has simply traded in knobs/sliders for some movie future UI interface shaped to fit a hand. As you’ve said, maybe it will be amazing for someone, but it certainly can’t claim to be “unparalleled” control, reimagined control would be more accurate.

        1. With super-audio scan rates, that aspect is unparalleled. I expect these touch sensors will do things you can’t do with knobs. However, the always-start-and-end-on-zero is a limitation/feature.

  12. a comparison with leapmotion + geko midi is possible, even if they implement two different gesture methods …

Leave a Reply