theremini

Let’s get one thing straight: Theremins aren’t digital. (Apologies for the headline for a moment, but stay with us.)

Theremins are based on perhaps the most important analog process there is in sound, heterodyning. And Leon Theremin can’t claim credit for this, nor Bob Moog or USA or Russia or Germany – it’s initially Canada, in the form of one Reginald Fessenden. (Go, Canadian engineering! Later on, Edwin Howard Armstrong develops the superheterodyne circuit that most resembles the Theremin, so that’s actually one gold medal each to USA, Russia, Canada.)

It’s a fancy word, but the concept is dead-simple: you mix frequencies to create new ones. The magic of the Theremin is, you use your own body and gesture to control the heterodyning process. Anyone who had trouble moving around while tuning their radio knows something of this effect, but the Theremin makes the process musical.

And so, the new Moog Theremini is a thoroughly analog control instrument. What Moog is adding to this control mechanism is a digital sound engine.

In simple marketing terms, yes, this means you can crank a knob and never hit a “wrong” note. But in practice, it means more than that. A new sound engine and digital musical applications of the sensitive analog control mechanism unlock new possibilities for sound and playability.

Leon Theremin himself applied his circuit to a vast range of imaginative applications. I spent yesterday pondering that with Andrey Smirnov, one of the foremost keepers of Theremin’s legacy, as we toured the exhibit here in Berlin at CTM Festival on the early Russian avant-garde, and let our hacklab participants play with their own Theremin circuit. Andrey emphasized just how broad Theremin’s applications of this deceptively-simple concept had been. And Theremin himself, coming from a background as a cellist, was richly musical in his imagination.

To learn more, we spoke again with Moog’s prolific engineer Cyril Lance about how the Theremini works and why.

Leon Theremin, and a modern recreation of his invention.

Leon Theremin was constantly reinterpreting the possibilities of what to do with the ingenious circuit in his invention.

First, he explains the all-important matter of pitch quantization. This has, of course, been an area of intense experimentation for people playing with Theremin inputs today, because digital processing does suggest ways of interpreting the raw input signal. Here’s how Cyril says the Theremini addresses this question:

The pitch frequency is measured extremely precisely from the heterodyned output of the pitch and fixed frequency oscillators. When in fully “continuous” mode (traditional theremin), there is no pitch quantization. When you are fully in quantized mode, the frequency is biased fully towards the root/scale note that it is closest to. If you are playing a C Dorian scale and the frequency is within +/-50 cents of D4, the output pitch will be quantized fully to D4 and there can be no variation from these defined frequencies.

The instrument can be biased to any amount between and this is where you can get very musical allowing for microtonal control with a bias towards the selected musical scale. This might be compared similarly to a fretted instrument where a musician approaches a note by bending or a saxophonist adjusts their pitch quantization through their breath-support and embouchure. In fully continuous mode, it would be like a violin where the violinist can choose any microtonal frequency – to be musical on an unfretted instrument such as the violin or traditional theremin requires refined and disciplined technique.

Since the theremin is particularly challenging in this regard, the ability for the musician to dynamically choose the amount of pitch quantization creates a more versatile and accessible instrument. When playing the Theremini, the musician will discover that even during the playing of a particular passage, or song, adjusting the amount of quantization allows them to create different ranges of expressivity.

What does the MIDI connectivity do? As it happens, while the engine always responds to the Theremin for pitch control, you can both use the Theremini as a controller and apply control to parameters from external MIDI sources:

There will be CC MIDI output for Pitch and Volume. The Pitch CC output will be scaled to the range of the Pitch Antenna and will also be “post-quantization” – this will allow for CC tracking that is either fully-contiuous, or fully-discreet and anywhere in between which will create some very interesting controller possibilities.

Many parameters are also available for modification via MIDI Input. These will allow users to modify the presets and the way the instrument sounds as well as several global setup parameters.. We are not currently announcing an “editor” for the instrument, but customers who use MIDI will be able to access these very easily. An example of this would be how fast the scan-rate is for the wavetable synthesis engine.

We’re keen to see this one released. Stay tuned.

More on the baguette from space:
http://www.moogmusic.com/node/92916

  • slabman

    Well, if you go down to a lower level, Theremins are based on measuring the capacitance of the human hand – and that’s eminently open to digital methods.

  • foosnark

    The MIDI implementation is kind of disappointing to me. It seems like they stopped a couple of steps short of making the Theremini into something much more flexible and powerful.

    I had visions of simply playing, for instance, Monark or Spectral through the theremin, via the combo of Note On and CC messages that guitar-to-MIDI converters use. It sounds more like the intent is just to use it as a kind of aerial Kaos Pad. It could add expression to lines played via keyboard or sequenced, which is cool, but so much less than just being able to play it directly.

    And the lack of full playback via MIDI in is also a disappointment. With that, I could edit a performance (because it wasn’t perfect, or I wanted to switch octaves, or I wanted to bounce it to audio and then transpose it to harmonize, or I wanted to arpeggiate it, or sidechain the volume to another track…). I could take a perfect performance and replay it with synth parameter automation drawn in after the fact. I could play the synth engine from a completely different source. Instead, I’m going to have to record this instrument live just like I record my fretless bass, as if it weren’t a digital instrument with MIDI.

    It really seems like they missed the mark there. (I might wind up getting one anyway, but I’m going to be watching for reviews and articles and videos a lot more first.)

  • petorr

    I think the midi engineering community needs to introduce a controller piggy-backing scheme to improve the resolution of midi control values. 1-128 is way too few values for the turn of a knob, and I can think of no better example than the digital control of a theremin. Such a concept was not mentioned in the article, rather they seem to get around it by smoothing or blending pitch between points. I think a good solution would be to multiplex one 32-bit control signal using four 8-bit midi controller assignments. That would be a simple way to implement near-continuous change without breaking compatibility. So, midi hardware would send a four-controller update representing one knob’s 32-bit precise value, and the synth hardware/software would be designed to read and use it accordingly. Has any hardware or software maker done this? Or, have any other useful ideas been floated?

    Also, as far as I know Edwin Howard Armstrong was the inventor of FM heterodyning for radio, so I think he deserved mention in that article.

    • http://pkirn.com/ Peter Kirn

      Well, okay –

      1. MIDI already *has*, has always had, a MSB/LSB scheme for achieving greater data resolution. Now, whether engineers actually use it is another matter. In some cases, they don’t simply because they really don’t need that level of precision. But pitch bend already uses greater resolution, so obviously it’s something you think about with pitch.

      2. If you read the article, Moog never says that they’re using MIDI input for pitch. I don’t know what resolution they are using, but everything you’re saying is conjecture. There isn’t MIDI pitch support, so this point is really moot.

      I think you’re right about Armstrong in that technically the Theremin is much closer to what Armstrong made. (Or, really, it *is* his superheterodyne circuit, in principle.) Adding a note there.