"Original" theremin sound?

The question of "sounding like the original instrument" seems to come up here (and on levnet) more often in the last weeks, especially when speaking about theremin construction.

I discovered more or less accidentally that there are some different basic opinions:

a) Imitate at the best the sound of the RCA theremins or Clara's custom built one.

b) Don't care about all, create a theremin with a rather simple sinewave sound and add some waveshaping, mostly by post-mixer distortion or variable coupling (Silicon Chip, Kees Enkelaar, Theremax, tVox tour, Henk).

c) Robert Moog, as he stated himself in an interview: "I see no need to imitate or reproduce whatever instrument's sound, I want to create absolutely new sounds".

I'd like to know what you think about that and which style you prefer. Thank you for taking the time.

I can't comment on playing them yet, but sound-wise I prefer the more "mellow" distortion resulting from not-quite-perfect RF oscillators heterodyning than the "pure sinewave" sound.

To draw an analogy about pure-sine/post distortion ... if you run an acoustic guitar through an effects processor you will get similar, but just not the same sound as running an electric guitar through the same effects. In other words, post mixing effects can't "fix" everything.

Don

I think everyone probably knows by now where I stand on this subject! ;-)

I want the lot!
well - not exactly "the lot" - I have no interest in copying the sound of acoustic instruments.. I would see no point in making a theremin with sampling technology that could 'emulate' a violin or trombone..

I feel that analogue emulation of classic Theremins would be great - these could be presets.

I think that for historic reasons (if nothing else) we should, somewhere, have a full collection of samples of the RCA (both through its original speaker and from DI) and other rare / historic Theremins (some of the Moog Theremins definately need to have their sounds 'preserved')

Then I would want full control of the parameters making up the sound, so that everything from unique instruments to emulations of classic electronic instruments could be achieved.

If I was limited to a few tone paletts, I would be inclined towards sine - but I would definately want control of harmonic content as a function of pitch -

I think that, whatever "sound" one prefers, the sound (harmonic content) needs to change as the pitch changes.. even if this change is almost un-noticable.. without this, any and every sound becomes boring (or at least I think it does).

So I am being awkward again.. neither a,b, or c fits completely! - and I see no reason why we should limit ourselves to those choices.

For cost / simplicity, there are probably 2 main 'topology' choices -

The Moog EW whos sound is primarily formed by a combination of the distortion on its HF oscillators, its diode mixer, and its deliberate distortion added at the VCA - Its tone varies as a function of pitch primarily due to the fixed coupling between its HF oscillators which 'skew' the waveform.

The S/C Jaycar EPE Enkelaar type topology, which produces more sine-like waveforms - primarliy due (I believe) to the HF oscillators being 'cleaner', the fact that these HF signals are buffered by FET's prior to being mixed, and the fact that the mixing is done with a balanced modulator IC which gives less distortion, and provides more mathematically correct multiplication of the signals. These Theremins have controllable levels of oscillator coupling, so 'skew' can be adjusted - therefore harmonics change as a function of pitch - but you can control the degree to which this happens.

I prefer the latter topology, so if I must choose, I would say "b".

I have not seen schematics for either the tVox tour or the Henk Theremins, so I cannot say anything about their topology - but I love the sounds I have heard from these instruments.. I am not sure it is fair to lump these with the S/C and other 'sine class' instruments.

I am not sure that the PAiA fits in this 'sine' class - From what I have heard, I think perhaps it sits somehere between the EW and the 'sine' instruments.. but I have never played one, and have almost ignored this Theremin in my explorations.

Hello everyone (sorry my english)

I think that the topic its very interesting and every each thereminist have their own opinion. I can say that the sound modified by an external effect makes the final sound "artificial" by a fan of theremin, in others ears would sound amazing. I think the best way to obtain good timbres its pre-etherodine, i know, it changes the pitch, but a Thereminist try to dont change the timbre meanwhile is playing.

Now i wonder how Thierry knows the tecnology of Tvox, i think that theremin has one of the best timbres and also i know that Theremin doesn't has coils to get lineal the responde of antennas, please Thierry could you explain more about that.

And the last thing, the "myth" says that only Clara Rockmore could make sound her Theremin with so beauty sound (or Timbre), the capacitance of the body really can change the timbre of a Theremin?, i guess yes in my experience but i think that the electronics has other opinion.

Thank you
airmusician1

I have only very few informations on the tvox tour.

But there seem to be small ferrite coils soldered between the circuit boards and the antennae for linearization.

I deduced my theory of its waveshaping only from hearing since I had the occasion to attend workshops with Lydia Kavina. In the leftmost setting of the waveform knob one hears an almost clean sine wave. Turning it clockwise seems to add some overtones by soft clipping which makes it sound rather like a slightly asymmetrical triangle wave.

[i]"And the last thing, the "myth" says that only Clara Rockmore could make sound her Theremin with so beauty sound (or Timbre), the capacitance of the body really can change the timbre of a Theremin?"- airmusician1 [/i]

I dont think so - I think Clara's sound comes more from her superb vibrato combined with extremely subtle volume 'modulation' - but mainly vibrato.. I think that getting the vibrato to modulate the frequency at [b]exactly[/b] the right "point" (ie where the centre frequency of the modulation is with respect to the actual frequency of the note being played) with [b]exactly[/b] the right depth and [b]exactly[/b] the right speed, and perhaps even the 'waveshape' of the vibrato - perfectly timed and suited to the music being played.. that, I believe, is Clara's "secret" - And no electronics will be able to give anyone "Clara's sound".

Tvox pictures (http://www.mb-labo.com/Theremin/t-Vox.htm)


[i]"But there seem to be small ferrite coils soldered between the circuit boards and the antennae for linearization" - Thierry [/i]

Yes, there is some axial component between the antennas and circuit boards - and I assume these are inductors (capacitors of that size would be pointless! ;) .. but in some ways this only increases the mystery - This Theremin has good linearity - Ferrite inductors like this have gross inter-winding capacitance.. All the trouble designers like Bob Moog went to with specially selected or produced linearization coils..

I think an additional linearization method is being used.. Perhaps something along the lines of voltage control applied to a HF VCO.. This does work - but to simplify this method down to the few components in the Tvox - That is quite an achievement!

I don't think that "What is the player's favourite sound?" is a relevant question. A better question is "What does the music dictate?"

If, for instance, the theremin is substituting for an instrument in a string quartet, then one wants a waveform with similar harmonics to a stringed instrument.

I imagine that, in a classical setting, four presets would cover the majority of cases - namely Woodwind, String, Brass and Vox.

In the world of electronic music where the distinction between composer and player is more blurred we should consider the expectations of electronic musicians, which are largely informed by experience with synthesizers. Here I would look at ways of combining and manipulating primitive waveforms - sine, pwm, triangle, sawtooth and so on as a starting point for a broad palette of sounds.

"I think that, whatever "sound" one prefers, the sound (harmonic content) needs to change as the pitch changes.. even if this change is almost un-noticable.. without this, any and every sound becomes boring (or at least I think it does)."

I couldn't agree more here. In fact I believe that the harmonic content needs to change not only with pitch but also with volume as in any acoustical instrument. I'm always using the envelope of a low-pass filter to get close to this behavior but I'd really prefer a more subtle and natural result than the rather artificial sound of the filter...

[i]"In fact I believe that the harmonic content needs to change not only with pitch but also with volume as in any acoustical instrument. " - Marble field[/i]

Absolute agreement on this, particularly if modeling acoustic instruments - Fortunately this is extremely easy to implement in a Theremin, as a CV following volume is usually available - In order to 'craft' the pitch->harmonic relationship in a controlled manner, one needs a control signal of some sort (CV) which accurately tracks the pitch rapidly over the whole audio spectrum.. this is not so easy to obtain - particularly for frequencies below 200Hz.. I have only managed this with complex techniques using phase-locked-loop frequency multiplication.

I think this complexity is the reason that variation of harmonic content with pitch has only been crudely implemented in theremins (by means of oscillator skewing and fixed filters).

[i]"I'd really prefer a more subtle and natural result than the rather artificial sound of the filter" - Marble field[/i]

I think the best way to get this is probably by splitting the audio into two paths - feed each path to seperate waveshaping / filtering circuits, each being independently adjustable.. then feed each output to its own VCA, and combine (mix) these and take the output of this mixer as your audio out..

Now, use the CVs (mixed pitch and/or volume)and put one through an inverting amplifier - feed the non-inverted CV to one VCA, and the inverted CV to the other VCA.. When one VCA is fully 'closed' the other should be fully 'open'..

As the CV changes, the mixture changes - and the signal will 'morph' from the one sound to the other... I have found that this is more 'natural' and less 'electronic' than sweeping a filter.

I think it would be good if theremin manufacturers published linearity curves for their instruments and, when CV is used internally or made available to the player, the range within which it is reliable.


My personal "ideal theremin" would (I think) have one parameter of the timbre controllable via left-right movements of the volume hand. I already use up-down and forwards-backwards for envelope shaping and expression, and my experience with a wah-pedal suggests that I would generally be changing the timbre quite slowly compared to the volume, so it should work out fine. I suspect the loop would need to lose its bulge.