"there is some schematic of digital microcontroller based théremins around ? " - elther
Google >> theremin microprocessor << and one gets loads of hits ..
It depends what you are looking for as to whether any are of use to you - most are optical, and those which are capacitive tend to be extremely crude at their front-end, using RC oscillators..
Most (if not all) are, IMO, just a waste of time. Dewster here has, IMO, the most sophisticated ongoing development of this sort of theremin. This thread from page 1, download his files, this is as advanced as digital theremins get at present! >> http://www.thereminworld.com/Forums/T/28554/lets-design-and-build-a-mostly-digital-theremin?Page=0
The real problem is that designers often look at the theremin concept, and think "I can do this simpler / better using modern technology" - they are often stupified by the "crudeness" of the "old way" - feel "scornful" and push ahead designing their modern replacement... Yeah - that was ME! .. I could do the lot with a simple PSoC.. LOL.
But the truth is that the theremin is more than capacitive sensors controling pitch and volume - The "simplicity" of Lev's original design obscures its enormous complexity.
IMO, the digital theremin which can match the analogue is coming - probably soon.. But its not in production yet. At this time, I would advise anyone not willing to devote huge resources of time and learning and money in actually developing a digital / MCU theremin, to go the analogue route... IMO, there are no digital theremin circuits out there which are worth building.
A couple I found: http://thraxil.org/e3940/project.html Complete "digital theremin" driving VCO + VCA - Front-end apalling!
Neither are really worth reading IMO.. But a lot depends on what you want, and why..
Also - Dont forget that when theres a microprocessor, the code is as / more important than the schematic!
>> There are some good "mixed signal" designs about - these use analogue and logic level components - they are not truly "digital" in that there is no quantization in the time domain, but they are not truly analogue as amplitudes are constrained - as in, harmonics on the source oscillators are "thrown away".. The Moog E-Pro was a mixed-signal design.
"ugh !" - elther
You can do quite a lot with the PIC-32.. They do a low cost development board (think I paid about £50 with some add-on prototyping boards.. I used it in my Epsilon instrument - together with a handfull of PSoC's - it was nearly ok as a theremin..) - Its fast enough and has enough in the way of on-chip stuff to deal with non-critical theremin applications.
But it really comes down to what you are aiming for - In particular, I would only go the digital / MCU route IF there was absolutely no discernable quantization on pitch - as in, pitch change occurs at least every cent.. But I never quite got to one cent over the whole range, and could hear a distinct "zipper" effect even in the playing area where tracking was less than 1 cent - so have now changed my opinion and set "acceptable" at 1/10th Cent.. Probably need to wait a few years yet before an affordable MCU runs fast enough to achieve this.
To me, it is the natural distortion which occurs to a waveform as the pitch changes (as in, during the "sweep" or "portamento" between musical intervals) which is one of the things that gives both theremins and analogue synths their distinctive appeal.. I can hear it, so am sure others can - and to me, building an instrument without this is losing too much.
Volume is the least of the problems - Volume resolution does not need to be anywhere near as high resolution as pitch.. If you find an adequate pitch-only MCU based theremin design, you should be able to duplicate the pitch hardware and algorythms, adapt them slightly, and use this duplicate section to control the volume via a VCA - MCU based theremins have "numbers" which are, by whatever means, used to generate the pitch - just take the numbers generated by the duplicated section to control the volume.
But if you just want to buy a MCU based theremin, or copy some design from the web without doing any work on it, well - at this time I think you are not in luck - And certainly not in luck if you want a good playable theremin.
And even if you did find something - I will bet that an analogue theremin based on something like the EM, will run rings 'round your MCU theremin both in terms of quality and cost - If you got a MCU theremin to perform even as well as a cheap Silicon Chip kit, I would be surprised!
I think Dewster is on the right track though - of devices available the FPGA is perhaps the most ideally suited to this application - he may get there!
"The real problem is that designers often look at the theremin concept, and think "I can do this simpler / better using modern technology" - they are often stupified by the "crudeness" of the "old way" - feel "scornful" and push ahead designing their modern replacement... Yeah - that was ME!" - FredM
(raises hand) Guilty!
Basically you are trying to build something that rejects interference and is stable (at least in the short term) yet is exquisitely sensitive to the environment. Good luck with that!
And it needs to react non-linearly in several complex ways.
And the output waveform has to be "pleasant" to the player's and listener's ears.
Finally, if you do it digitally it may not be as readily accepted by the general analog playing community (not calling anyone out, just a general observation about how digital guitar effects have been perceived, rightly or wrongly) so make sure you hide any digital issues, or make them somehow worth it.
I've read that the human ear is sensitive to around 3 cents pitch change. A half step is the 12th root of 2, and a cent is 1/100th of this or the 1200th root of 2. Three cents is the 400th root of 2, or 1.001734..., or 0.17 percent, or one part in 577. This is probably applies more to the center and upper end of our pitch perception, lower frequencies don't need as much resolution (which can be convenient).
In terms of volume, I believe a 1 dB change is the accepted threshold of audibility, which is 10^(1/20) = 1.122... or around a 12% change. But more resolution here certainly wouldn't hurt.
"I've read that the human ear is sensitive to around 3 cents pitch change" - Dewster
I read the same sort of figure before I started Epsilon.. I now disagree with it. Perhaps someone with "perfect pitch" can detect that a note is not "perfect" if it is out by 3 cent.. But I think that most people can hear a dynamic change in pitch far less than 3 cents, even if they couldnt tell that a note was 20 cents "off key".
I have a lot of questions regarding pitch change.. All I know for "sure" is that I can Sometimes hear change less than 1 cent - and I believe that what I can hear is a function of the frequency and the rate of change.. So I am not actually sure if I am hearing the "pitch change" or if I am hearing other artifacts such as change in the resulting waveshape or phase relationships or whatever..
Its damn difficult to test and quantify - But having designed and built audiometry equipment for the Royal Free Hospital back in the 80's, I know how comparatively crude the tests performed can be, and how unreliable the data on human hearing (much of which comes from earlier tests using equipment that was inferior to what I was building in the '80's) is likely to be.
Certainly the testing is unlikely to cover the huge range of possible ways pitch can vary from an electronic musical instrument, and all the tests I have observed used pure sine signals - I think that the way we process sound cannot be evaluated correctly unless we experiment with complex waveforms.
Very interesting that you were so involved with audiometry Fred.
I just wanted to add that a linear numerical approach to pitch is rather convenient as to how our perception of pitch seems to work. If you have a 1 in 1000 resolution at 1kHz, then 1:500 is probably OK for the octave below that, and so on down to 1 in 20 at 20 Hz (though I haven't read much about pitch perception in the various ranges). This is one of the reasons why I decided not to transform between the linear and logarithmic domains in my prototype (to deal with perceived pitch linearity & hand movements).
There is also some dependence on how the interpolation or noise coloration between the notes works. Hard stepping can be audible, some white dither (as PM or FM) doesn't seem to present as much of a problem. Thermal or environmental noise may be statistically white enough to do this for you with the lowest notes.
"Very interesting that you were so involved with audiometry" - Dewster
Lol ;-) Apart from "defence" I have probably had some involvement with most areas of electronics - but my particular areas have been sound / musical instruments / synthesis, medical electronics, and safety critical design.
I worked as chief medical physicist (electronics) for several years in the Royal Free Hospital School of Medicine, where I designed and built equipment which was needed for research but could not be bought because the required instrument was not available, had not been invented, or was not good enough to meet the requirements.
The range of projects was wonderfully huge - from simple, like the audiometry stuff, to mind numbingly difficult - like counting the number of molicules being breathed in via a nebulizer, to bombarding a protien sample with a non-destructive pulse of lazer and determine the resulting rotations of the molicules, using photo-multiplier tubes..
The funding for all this exotic research was cut when Maggie and her henchmen came to power in the 80's - They needed the money for more worthy causes - like destroying the unions and descimating Britains remaining manufacturing industries.. I, like many others, was forced to seek employment in the shrinking private sector - As an engineer I was expected to get employment in the growing defence industry (which was desperate for engineers because Maggie's mob were a bunch of warmongering extortionists).
I ended up working in various industries thereafter - settling in the "safety critical" area for several years, developing monitoring systems for explosive atmospheres on oil rigs.