Theremin Development - (more) questions for thereminists.

Hi Amey -

Can tick the "and of course made to also accomodate lefties." box - the antennas will be pluggable into either RHS or LHS of the instrument.. This is something I really dont understand - why, for the price of a couple of extra sockets, this isnt done on all theremins!

Hmm.. "minimalistic design " - :-(

"looks great on stage" - Oh hell... This is the "arty" subjective stuff - I really cannot make any claims here, in MY head my ideas look great ;-) .. but..

"(easy to put together and dismantle)" - I am looking at flexible antennas - goose-neck "wire" for both the loop and stick - Because I actually want quite a long antenna, so the ability to bundle it up could be important..

Sadly though, in my head its not really a small simple theremin... "minimalistic" would only possibly apply for a version with a digital user interface. A build option I am looking at is having the instrument built into an adapted wooden pool cue case - with the lid removable to expose the controls and pitch antenna socket, and having two couplings for microphone stands (two microphone stands would be required.

Fred.

fred, 10 points for not implementing a TM.yes, after playing with that voice pedal,of course not to compare with the TM, i realised that it get's on ones nerves over time. no matter how bad i play, it just doesn't sound..honest maybe. but it's good fun too.  

the theremin community is maybe already splited into tubers vs chipers. since back longer. maybe since good old bobs days. i would really like to know what he would do these days, with all these new small Psocs etc... probably the same like you couple of ee's here are up to. frankly, i think what comes out of the speaker is important, because, that is what hits the ear, shakes the nerves and touches ones soul. it's different when you're just listening what you played or listen while playing. i don't know. i just regain my passion for that instrument and it's sonic potential.actually, i'm building my first theremax, ordered out of pure frustration over that theremini crux! and i like some painting by numbers from time to time.  o tempora o mores...

to complicate things a bit more: i like the 3.antenna idea. that would simplify the voice control rapidly... (i just say so because i control the filter of my synth with the volume. a bit dramatic,yes, but also a nice second articulated sound.)..but increase the engineers work too.

most probably: for earning the laurels of the whole true thereminists high sosolalasociety, you might need to put out a sound emitting true instrument anyway. and from reading some post here on tw, about that noble goal... it might get rough;-)

hopefully, some more people speak out their humble or not so oppinions.

dani

I'm not exactly sure what you're getting at. MIDI baud rate is 1MHz / 32 = 31.250 kb/s. Even getting raw 7 bit samples at this rate (+1 command bit, +1 start bit, +1 stop bit = 10 bits) would give 31.250 kb/s / 10 b/sample = 3.125 ksamples/s. Nyquist here is 1.5625kHz.
Your dialog box looks nice! What is it coded in and what is the windowing (widget) manager? -- dewster

Under "real time" I assumed that the plot is updated with a comfortable (reasonably fast) rate, reflecting the momentary changes  in circuit operating. In reality the data are collected in buffer and send when ready.

The dialog box has dual purpose:
1. Be sure on proper operating of volume resonance circuit + tuning of volume coil to reach the resonance when the hand is away of loop.
2. To calibrate the extreme hand positions "volume off" and "volume is max" for current configuration of used components.

Those are "factory" adjustments and I'm not sure the tab should be enabled in final "user" version.

   Hmm.. "minimalistic design " - :-( -- FredM  

I'm an apologist of minimalist design too. The theremin isnt a DJ's toy. The preparation should be done in studio/home with the friendly tools, the performance is the place for single button "go to next setup".

"I'm an apologist of minimalist design too. The theremin isnt a DJ's toy. The preparation should be done in studio/home with the friendly tools, the performance is the place for single button "go to next setup"." - Ilya

Yeah -

the ""go to next setup"" button requires a patch / preset memory - quite easily implemented, but presets require the ability to edit them at "studio/home" ... Lots of folks feel cheated it the instrument requires an external "appliance" to facilitate editing, be this a BCR-2000 or a software APP..

I am not hard on one side or the other - Most synths with digital user interfaces are never edited by the owner - they are left on their factory presets, but this doesn't happen with analogue synths, even those that do have presets - so, for me, digital interfaces appear to be mostly a frustrating waste of space.

If I was a performing thereminist, I would go for an instrument with minimal controls for sound, but have a BCR-2000 configured specifically to edit it at home... And I wouldn't bother with a DUI capable of editing the sound on the instrument.

But as I am not designing primarily for me, I will be looking at having sound editing available in the DUI if/when I implement it..  For now though, minimalism be damned! ;-)

I strongly feel that whatever I implement, the instrument must have directly accessible analogue controls for things like pitch and volume tuning and setup - that taking 'minimalism' to the extent of hiding these essential functions in some DUI, or automating them, is taking minimalism way too far.

Fred.

"The theremin isnt a DJ's toy" - Funny you should mention this ;-) .. Two DJ's I know have opted for the theremini (rather than the less "minimalist" instruments I advised ) - they seem to be drawn to toys.. minimalist or otherwise!

I am also not certain that "the performance is the place for single button "go to next setup"" - will always achieve the best sound.. The environment affects the sound, and limiting to pre-stored configurations without any ability to 'fine tune' the sound to suit the environment wouldn't make me happy.

 Hmmm... Something to remember - Even if I have a BCR-2000 at home, simple access to the audio via the DUI is still needed...

Not having ever use a theremin for public performance, an editing DUI on the instrument seemed like a waste - But actually...

I know I've mentioned them before, but many guitar multi-effects pedals manage to function as dumb stomp boxes that also allow detailed editing.  A line of encoders could have handy and useful direct functionality in "playing" mode (null, sensitivity, offset, etc.) while having more "soft" and varied context functionality in a "editing" mode. 

A single spinner with arrow buttons and screen with soft buttons IMO doesn't constitute a practical, useful, UI - though that's often all you get even on many high-end instruments.  I'm on the fence re. touch screens: they seem useful and quite plastic, but I have an aversion to actually touching and smudging up a screen, they don't provide tactile feedback, etc.

Lashing your stuff to any outside computing device should be done with the full knowledge that (if well designed) your Theremin will almost certainly outlast any & every processor / OS / driver / electrical interface currently out there.

Making a UI in the instrument isn't that big of a job (<=big talk) if you are already planning on allowing manipulation of those parameters via external means (MIDI / USB / iCrap / etc.).

[EDIT] The three most disappointing HW synths I've owned relied heavily on a PC editor: Roland JV1010, Yamaha FS1R, and Yamaha MOTIF rack.  Roland dropped PC support almost instantly which left one with a box with 4 knobs + 3 digit LED display (=presets only).  FS1R was kind of half-baked in the first place, with the formant filters accessible only via external software, which itself was never well supported.  The MOTIF front panel UI was either designed by a super genius or a retarded monkey (hard to tell which) that pretty much forced one to to go to an external means of parameter manipulation.  Lesson learned, never again.

I get the feeling a poor panel UI means most in-house development bypasses it, which means it really doesn't get banged on / implemented properly.  The in-house guys spend their time polishing a PC interface that likely no one will ever see rather than the UI everyone sees when they walk up to the damn thing.  The JV1010 had a nasty habit of locking up if you spun the preset knob too quickly (encoder tied to interrupts?) which led me to believe no one at Roland ever actually used it.

[EDIT2] Please ignore me, I'm extra cranky lately (programming language hell).

Hello Dewster ;-)                                             (if not technical, ignore the blue  text )

"Please ignore me, I'm extra cranky lately (programming language hell)."

You have my sympathy! - I was in a bad state when the PSoC was causing trouble .. it can get to one!  The damn chip is touted as having freely assignable pins, which is almost true... But there are only two analogue busses shared by all the pins, and if you dont realize that the analogue parts have direct connection to specific pins so the busses arent used, the fitter just falls over if you lock pins to places they cant be routed because more than two analogue IOs are trying to share the same bus...

Its so damn annoying - one has to dig into the TRM to clarify details that should be in bold.. All sorts of functions like SPI and ADC etc have specific pins they must route to in order to get the most out of the part - sort of makes a lie of the claim that "if you make an error on your PCB layout, you can correct it in the PSoC" - Yeah, this is true for a blinking LED application - but may not be true for anything using analogue or comms ports..

Anyway - I have now added the things I couldnt fit before, after re-wiring the pins to their optimal functions (which arent my optimal layout, but its ok) - and the photos are here if anyone wants a look..

PSoC4 configuration photos

RE: UI.

I actually agree with you Dewster! - its possible to build a better UI than the crap ones produced (particularly by) Yamaha and Roland.. And I agree completely with your comments about the limitation/s of soft editors - been there, got the badge... My Yamaha TX16W had a god awful DUI, it only became somewhat useful when a new OS was written by an independent developer - and this (the "typhoon") completely transformed the UI and made it usable - but it was still horrible!

So I am now in a bit of a limbo.. Which audio board do I develop? - One using comms (midi) which I can drive from my BCR-2000 in the development phase, and develop (or have developed) A drop-in DUI before I sell / launch the instrument, or a (simpler) direct analogue board.

I dont know.. But I have a while (couple of weeks?) yet before I must commit - the core board is the top job, and CV build probably next - it will take me a week or two to lat these out, then I will decide the audio option.

Fred.

"...if you dont realize that the analogue parts have direct connection to specific pins so the busses arent used, the fitter just falls over if you lock pins to places they cant be routed because more than two analogue IOs are trying to share the same bus." - FredM

Sounds like the bad old days that are mostly over for FPGAs but used to get me all the time with CPLDs.  Move a pin and it wouldn't fit and / or meet timing.

Fred, those PSoCs are so inexpensive, isn't there some way to lash 2 or 3 together for a Theremin?  Maybe use one for the pitch side and one for the volume side?  Maybe have the two processors talk to each other via UART?

(if not technical, ignore the blue  text )

"Fred, those PSoCs are so inexpensive, isn't there some way to lash 2 or 3 together for a Theremin?  Maybe use one for the pitch side and one for the volume side?  Maybe have the two processors talk to each other via UART? " - Dewster

Hi Dewster, this is absolutely correct, and easy - its something I have done many times with the PSoC1 (more because the MCU was so slow that I needed to spread tasks)... And I actually planned to do this when I add another PSoC for the DUI.. The idea was to put as much as possible into the 'core' PSoC (enough for a stand-alone) and then in addition to the DUI, add other functions to the 2nd PSoC as well.*

So why am I fighting to cram everything onto one PSoC4? Pure bloody stubbornness! - Its insane, compulsive obsessive behavior - I justify it on the grounds that a £4 saving and saving of the required board area, will equate to perhaps £15 reduction in price to customer - but if I account for extra components / time / hassle, this probably reduces any reduction to £10 or less.. Its on the edge where I have enough reasons to stay with one, but only just..

If I was using the chips directly, then at £1 with perhaps £1 for fitting, and reduced PCB area, I would go this route - but the board is a large £4 'component' requiring pins and socket bringing the 'module' price up to about £6 - Its over the line where I can just add an extra one without carefully weighing up cost/benefit.

There is also a "satisfaction" factor - Kind of like banging ones head against a wall, because it feels so great when you stop... ;-) .. Getting a chip packed full and exploiting it fully feels a lot more satisfying than using two and "wasting" their potential.. Because its all hardware configuration, once you have managed to squeeze what you need into it, there's no problems like one has with over-stretched firmware.

Fred.

*its also a lot about 'partitioning' the design - I want it modular as much as possible - and tthere has been a real battle to do this.. One idea I am looking at is making the 'core' RF only (to take care of all the front-end stuff) and then doing the audio board with 2 PSoCs and including the DUI on this.. Here its the "commitment factor" which is blocking me - because doing it this way I would have nothing until I had everything .. I would need the large and comparatively complex audio board and DUI "set in stone" and couldnt get a sound out of the instrument until everything was working.. And the 'core' PSoC would be under-exploited unless I used it for functions like tuning / linearization / span under control of the DUI, Functions which I really dont want to hand over to any DUI.

With Dani's talk about CV, I was thinking perhaps of doing a "RF only" core, and replacing the audio functions with dedicated CV functions.. then having an audio board with separate PSoC and spare pins on which I could implement mixed analogue / digital control allowing storage of a few presets.. This would allow the front-end board to be a stand-alone CV theremin, or a simplified no-CV conventional theremin with the addition of the audio / control board..

Just thinking about it - I may even be able to compute the difference frequency numerically without needing to go to PLL multipliers - just by timing a few cycles of the pitch oscillator (and knowing the reference oscillator period) .. Then I could generate a CV which would actually track the theremin pitch if an audio board was fitted.

Oh hell - its back to square 10 ;-) ... This is just too important an idea to ignore, because I could get a V/Hz voltage out and generate a V/Octave CV out.. The V/Hz would come in real handy on the theremin audio board...

Ignore that - it was a brain glitch! - I was thinking about post-heterodyning times on unheterodyned frequencies, which, as I say, is a brain glitch ;-)

----------------------------------------------------------

Revision of ideas / Update;

(this is being updated as I think of other things) ..

----------------------------------------------------------

Been re-thinking the whole theremin construction and options in the light of feedback received.

Major issues I have looked at:

1.) CV Output.

2.) The balance between "functions" and "minimalism"

3.) Getting to market - Time / Cost.

4.) The user interface (which directly related to 2 + 3)

5.) "Partitioning" to enable modular construction and simplify development.

-------------------------------

These are my thoughts at this time:

----------------------------------------------

A.) MAIN BOARD: A primary "core" board which implements all the RF functions and features such as register switching, tuning, linearity and span for pitch and volume. This board has no stand-alone function and must be connected to one or more boards to become anything useful. All controls on this board are manual, there is no digital interface to it. (This board is probably price/ complexity level 6).

B.) "Classic" AUDIO BOARD: (Optional) This will be a manually controlled analogue board soley and entirely dedicated producing classic theremin sounds - Two knobs set the "excitation" wave shape, and this is followed by a complex resonator / formant circuit with 8 knobs. The intention is that resonances / formants typical of instruments and the voice can be adjusted to shape the tone. (This board is probably price/ complexity level 4)This board can be used alone or with boards E and/or F.

C.) CV BOARD: (optional) This board is low cost (board is probably price/ complexity level 1) CV interface outputting a linear voltage compatible with V/Octave for synths and also (perhaps) a separate V/HZ CV, it also outputs volume CV and gate signals. This board does NOT track any audio - it will output CV with approximately the same span and linearity one would get if an audio board was connected, and switch registers, but could be semitones out and irregular if audio was being simultaneously output from an audio board. For this reason, if both audio and the CV board are fitted, a selector switch will give the thereminist choice of "theremin" or "controller" modes. In Controller mode, there will be no audio output - none - not even preview.. In Theremin mode there will be no pitch CV output. The instrument owner will have the option to bypass this lockout - but this will be an explicit action they must take, as the audio (pitch) output would not be musically related to the CV.

D.) AUDIO OUTPUT BOARD: (Essential if any audio boards fitted) This board will take signal(s) from the audio board(s) and mix them. This board will also take an external stereo audio signal and allow this to be mixed with preview and the main output signals so that the thereminist can monitor and practice using headphones / ear-buds. This board will also contain the VCA  (This board is probably price/ complexity level 1.5). One of the things which astounds me about even top theremins like the EWP, is that monitoring by the thereminist is almost completely ignored - One needs a dedicated monitoring path with good headphone amplifier, so you can hear your theremin, hear the preview signal, and hear what you are supposed to be playing along to.. I dedicated a small Mackie purely for this function when I thought I might become a thereminist! ;-) .. I want that same functionality built into any theremin I design, and its not as if this costs much extra.

I will be needing the services of an artist to do my panel designs before I put anything on the market! ;-) .. but it will need to be an "old school" artist - I detest the gaudy look of some modern panels, particularly on some modern Japanese instruments and American stomp-boxes! (oh, I call "preview" "Cue")

-------------------------------------------------------------

At present, I think the above will be my first target.

-------------------------------------------------------------

E.) Manual Audio Board 2: (optional) This will produce a range of mixable waveforms using the mixed signal heterodyning engine and classic subtractive synthesis. This board can be used alone or with boards B and/or F. (This board is probably price/ complexity level 3)

F.) Manual Analogue Heterodyning board: (optional) This will consist of two classic analogue mixers - the EM (diode) and a 4QM for more sine-like waveforms. This board can be used alone or with boards B and/or E. (This board is probably price/ complexity level 3)

G.) CV Synth board: A  CV synthesiser for use when board C is fitted .. Will NOT work with any other audio board (except board D which it will need). If an instrument is fitted with both theremin audio boards and this board, this board will only play when the instrument is in "controller" mode, and the theremin audio boards will only play when the instrument is in "theremin" mode.. All audio will go through board D, so it would function as a switchable voice theremin.(This board is probably price/ complexity level 3-6 depending on the complexity of the synthesiser).. The CV output from the instrument WILL track the pitch from this board.

-----------------------

DIGITAL CONTROL:

-----------------------

All the above are controlled entirely by mechanical analogue means, this results in a lot of knobs for a full system, and even the minimum audio system has its fair share! - Also, there is no way to store any setup.

Digital control would replace these mechanical controls with electronic ones (digitally controlled potentiometers for example) and the controller (DUI) would be able to store these settings in presets - The controller would also support MIDI communication OF CONTROL SIGNALS ONLY so that external hardware and software appliances could be used to alter the controls, and presets could be uploaded and downloaded.

H.) Digital User Interface (DUI) -

I.) Digitally controlled Audio board. (all functions of boards B E and F, and perhaps more)

I will think about these after I have A,B,C and D running! ;-)

------------------------------------------------------------------------------------------------

* Price / Complexity level.. I am thinking purely speculatively here, may be miles out either way.. bit each complexity level will probably equate to about £40 on the the price.. So a CV controller (6+1) would be about £180, and a "classic" (A+B+D) = 6+4+1.5 =11.5 = 10*£40+60 = £460 .. This is board prices, cost of cabinet antennas etc will probably add another £40 ? - Depending on whether wood or plastic... I always tend to underestimate costs, so the above are probably on the optimistic side..

------------------------------------------------------------------------------------------------

Comments / ideas are REALLY welcome at this time .. Nothing is yet set in stone!

Fred.