New TherAemin - MIDI Functions - What do YOU want?

[b]I am on the scrounge for advice + opinions again![/b] - I have developed the ICs for the new TherAemin (these are custom PSoC chips which implement the LC HF oscillators and provide auto calibration, for both pitch, volume and reference oscillators - all entirely analogue) but I have a number of digital modules within these IC's which are, at present, unused.

So - I thought I will fill these available slots with something useful - like MIDI.

What I have is this.. There is one IC which implements the volume circuitry (connects to external components - inductor/s capacitors/s resistors/s etc which connects to the volume antenna) and outputs a voltage (0 to 5V) proportional to hand position. A single pin defines the polarity, and an input voltage (potentiometer) sets sensitivity.

The output voltage can be taken to a VCA (or for that matter VCA/VCF) to implement a stand-alone volume controlling circuit.

The Pitch circuit uses 2 PSoCs and external Rs,Cs and Ls - Note, PSoC Capsense is not being used - I am using the PSoCs as part of a LC circuit and the PSoCs are configured to be a circuit containing OpAmps, Filters, Switched capacitor circuits etc - and to hide the circuit so it cannot be copied.

My plan is to include MIDI In and Out on each IC - the In will be merged with the signals processed by the IC, so that pitch and volume control signals will come out of the 'final' IC ready for connection to MIDI equipment.

I can detect the high frequency seen by each PSoC by feeding the oscillator signal internally (from the Analogue section) to the (presently unused) digital section. I can (at this time) add MIDI with no extra hardware cost - The only cost is the time I spend on it.

[b]What I want to know is this: What MIDI signals would you like, and what makes sense for a Theremin?[/b]

1.)[b]Note on/off, Velocity (on/off?)[/b] Any ideas on this? In particular,
(a)what should trigger a "new note" message?
(b)Velocity - Should this be the value from the Volume antenna? - or should it be a constant (say 64) with
(c) Volume Antenna data being sent as a continuous controller message?
(d) Should trigger perhaps be a function determined by changes on the Volume antenna?

2.)[b]Continuous Controllers:[/b]
Would transmition of pitch and volume messages as continuous control signals only (no note-on / note-off messages) be useful? - This is the simplest way to implement MIDI.. and may be all I can manage to do quickly.. A 14 bit pitch and 12 bit volume message would be available, and these could be assigned simply to any continuous controller and any MIDI channel.

[b] I would be extremely grateful to hear your ideas - Particularly if you have ever used MIDI from a Theremin, or are MIDI proficient.. Also, I would be most interested in hearing from anyone who could write MCU level C code to implement MIDI functions .. I cannot afford to pay anyone at this time, but would be happy to discuss royalty payments based on sales.. I have too much to do and am running out of time and money - so am looking for help! [/b]

FredM,
Very interesting set of questions.
1)I have thought about this in reference to the ThereMax and the PAiA MIDI board. I can control pitch in the Theremax MIDI board, but, I have to find a suitable patch on my Roland keyboard that allows 'PITCHBEND'. My Roland XP-10 has a pitch bend range adjustment of 24 semi-tones. Also, because the PAiA MIDI board does not send MIDI 'NOTE ON' info, I, must physically weight a key,(of an appropriate starting musical pitch),down to intiate a 'NOTE ON' MIDI message. The 'VOLUME' MIDI messages seem to generally apply to most voice patches.
I have never owned a MIDI generating Theremin that was manufactured by MOOG. I don't think MIDI could be implemented better on a Theremin, than the extensive treatment given to it by MOOG.
2)I would like at least a 'pitch null' baseline 'note on' MIDI message being sent to the slave instrument to achieve a lowest pitch reference for theremin control/play. Also, the ability to have pitch bend as well as semi-tone stepping would be sweet. Did I make some sense? I am not totally up to speed on MIDI funtionality ---sorry!

Good Luck, Fred!

teslatheremin

Hi Tesla -
I cannot at this stage implement anything close to what Moog did with the EtherVox - but I am extremely interested in everything about MIDI implementation for the Theremin, as I intend to expand the MIDI facilities at a future date.

What I have is spare capacity with which I can create UARTs / Timers / Counters etc.. So the hardware required to implement MIDI is available.. BUT, at this stage, I do not want to implement a comprehensive user interface ..

My plan is to implement basic MIDI functions, and an additional high speed serial output port.. This way, MIDI can be implemented with the basic unit/s quite quickly and simply.

The high speed data (which will not be of any use until the MIDI expander is developed) can be sent to a comprehensive MIDI expander (a unit which will process incoming data streams, intellegently package this, have a comprehensive user interface, so that whatever MIDI control functions are required can be provided) which I can develop when I have an income!

Originally I thought I would just provide the high-speed output, and not tell anyone about it - But realised that I could also implement a simple MIDI output using the data available.. Having MIDI available may mean that there will be less interest in the expander, but I am also hoping to sell my ICs to other manufacturers, and would provide data to allow them to do their own MIDI implementation from the output data.

So, for the basic ICs, There will be Midi In and Midi out, and pins to configure the channel and message type - options will certainly include continuous controller output (pitch, volume, and probably a few others, sending MSB and LSB) -

It is note data which I am most unsure of.. Firstly, because I am not sure I can process this accurately in the simple implementation, secondly because it will rarely be 'spot on' so pitch bend data will also need to be sent, and the scaling of pitch data can cause problems - But mainly, because note-on will cause triggering of envelope generators etc on the recieving instrument - And this doesnt seem correct to me - I would think that one would only want to trigger the envelopes when an action is initiated from the Volume antenna - and for all other situations, pitch bend and volume controller should be the only messages transmitted.

One approach to note-on, note-off is to consider how thereminists generate distinct notes, or the appearance thereof.

The first, and obvious way, is by use of the volume loop as an envelope shaper. The second way is by using sudden changes in pitch to give the impression of two distinct notes, one following immediately after the other.

The first way is straight forward - options would be at what volume to send a note-on as the volume rises, and at what volume to send a note-off as it falls.

As for the second way, the MIDI-Wave I tried out did not have this - it had the option to generate note-off, note-on when the pitch bend exceeded a specified amount - a semitone or a tone, and so on. A sort of air-harp mode. I think generating note-on/off on sudden pitch acceleration/deceleration would be more consistent with existing playing techniques. Probably very complicated though - you'd need to be able to disregard vibrato when detecting a note change - one for the expansion module. :-)

Air-harp mode is a bit rubbish, really. Great for swirly ice-cream bells (Peter Pringle demonstrates (http://www.youtube.com/watch?v=QPLNEw6KH4E)) but that's about it - the trouble is that you can't zip from one note to the next without triggering intervening notes.

I try a mathematical approach (my 5 cents):

Let V be the volume level, M the muting threshold,
P the pitch level (level means a proportional voltage) and G the maximum glissando speed (expressed as variation of the pitch level)

a)
V <= M ====> Note off
V > M ====> Note on

b)
Abs(dP/dt) > G ====> Note off
Abs(dP/dt) <= G ====> Note on

a) is an obvious muting function
b) means that if the speed of changing between two notes exceeds a limit, this will trigger Note off, while if you slow down, you'll get Note on.
Quick vibrato will not be a problem, as long as dP gets not too high. But as a music teacher told me that a "good organic musical" vibrato should never exceed a small third and should not be quicker as one could reproduce it with his phrenic, I see no problem.

[i]What I want to know is this: What MIDI signals would you like, and what makes sense for a Theremin?[/i]

...I'd like to be able

[i]minimum[/i]

- to produce note on/note off events with
-- a change in volume to <= 0
-- a change in note <= 0,5 (perhaps this must be combined with something like Thierry developed - speed-dependant...but not only...it should also trigger on slower changes...'read my mind' ;o)

- to have a dedicated 'gate' signal

- to send volume information

- to send pitch information

[i]next step[/i]

- to send on different MIDI-channels (some means of switching channels during perfomance would be great)
- to hold a note on one channel and switch to another channel, play there, switch back

[i]another step further up[/i]

- to receive MIDI-data and have the theremin
-- play it
-- mix it with the input via the antennae

...this is just a very quick, un-thought-through-brainstorm...but with some ideas I am carrying around with me...

...keep on trying, Fred! I hope you'll succeed!

Ye Gods! - Well, I asked for it ;)

I have been scanning all the data I can find re MIDI and Theremins - but the one thing I have not been able to locate is a user manual (or any techincal documentation) for the EtherVox Midi Theremin - This seems to be a 'gold standard' so if anyone can advise me where I might get a manual, I would be grateful.

Most of what I have seen (U tube etc) has been absolutely horrible! - and those demos I have seen where music was being made with a MIDI Theremin, the Theremin character was utterly and completely masked by the "jingle bells" implementation of the MIDI.. Ok, if this is what people want, I may get 'round to doing it for them sometime in the future - but I am not inspired by what I have seen.

Funny - I feel I have gone full circle.. I LOVED the Mini Moog, and it was what inspired me to go into electronics and music.. I thought Polyphonic Synths + Midi were wonderful, but have realised that there is an intimacy with the old monophonic instruments which cannot be beaten.. And now I am looking at MIDI for the Theremin - I was all set to "improve" the "antiquated" Theremin concept - But have found that I have grown to love this instrument as an instrument in its own right, not merely as a controller for some other technology..

So I will have a play with the ideas, and try to get a simple MIDI scheme which does not sound like Santa on acid - but, for now, I am not going to spend too much time on it - It will be a free 'extra' on my chips - but will probably only implement Pitch bend and volume continuous control output at this incarnation..

Thank you all for your help -

<>

All I need is a good sounding playable instrument. Midi be nice, but I have nothing to chain it up with (I don't even know the terms) You do that, and I'm there

Andy

Are you planning on selling these chips with apps notes, or only complete theremins?

Also, what are your plans for long-term sustainability? After all, people are still using RCA theremins! Parts obsolescence haunts both the day job and hobbies for me, so lately I really try to use only things I am pretty sure will be around for the next 10 years or so.

Don

"Are you planning on selling these chips with apps notes, or only complete theremins?

Also, what are your plans for long-term sustainability? After all, people are still using RCA theremins! Parts obsolescence haunts both the day job and hobbies for me, so lately I really try to use only things I am pretty sure will be around for the next 10 years or so."

I am planning to sell IC's, kits and completed units.
The way I am doing things makes the problem of future component (un)availability almost inevitable - It is a sad fact of life that the pace of component redundancy(particularly LSI - but also some reasonably simple IC's such as Transconductance amplifiers etc) is increasing - I cannot gaurantee that, for example, PSoCs of the kind I am using will be available in 10 years.

I can, however, be reasonably sure that components capable of replacing the components I have selected are, or will be available in the future (For example, The PSoC was completely 'on its own' some years ago, but I could redesign the circuit using other mixed signal processors now, if i was forced to).

The actual Theremin PCBs have 'module' sections, and the module connection points are taken to break-out headers, allowing the standard on-board module to be bypassed and replaced with a small piggy-back board - This gives some future proofing.

The biggest risk that I see to the projects longevity is based on my longevity! While I am alive and dealing with problems, The Theremin products I design and supply will be maintained and repairable / upgradable - But I am keeping a few tiny details secret (What the Chips are doing) and without this information it will be quite difficult for anyone to reverse engineer my products..

So, what I will do is deposit full documentation, code etc - everything anyone, (even a reasonably skilled hobbyist) would require to build / repair / upgrade my Theremins with readily available technology - and specify that this is freely published on the WWW when I die.(I will also make all this data available if I stop my invovement with Theremins and no assignee takes on responsibility for its continuation)

A bit morbid - but I was reminded yesterday of the fragility of our schemes and lives, when my wife suddenly developed acute appendicitis and was rushed to hospital for emergency apendectomy.. We were married 13 days before.. She is making a good recovery, but events like this are shocking and make one realise our mortality

So yes - a good question. I would like my product to still be playing 30+ years from now.. I believe the best hope for electronics longevity is careful attention to detail in the design phase (never run ANYTHING at >40% of its rated parameters, and pay particular attention to high Z circuitry, and including protection components at every point where there is any chance of an ESD induced failure) - The sort of things I routinely do is to place a bridge rectifier on DC supply inputs (This caters for wrongly polarised wall-warts, and even for AC input), fit Transorbs, and gas discharge (neon) tubes to Hi Z inputs (like antennas)..

But, even with all the above - Competing with a valve RCA in terms of longevity is a challenge that I would certainly be pleased to achieve - but cannot confirm that I will!

The bigger problem is this.. With the way we are abusing our planet, will there be anyone to play my Theremins in 100, or even 30 years from now? Unless we take some absolutely radical steps NOW, I doubt it!