Linearity - players' point of view

Christopher,  PCH sounds interesting and I would like to see an application of it, but IMO the combination of proximity/capacitance is what makes a theremin a theremin.  The versatility that a player can develop with a theremin is a result of 'change in capacitance' being the main variable.  this means there are many ways to achieve the same change in frequency. Different fingering/hand/arm motions can be called upon in the performance of music, that offer the player a means to be efficient.

perhaps I've partly misunderstood your description. I think a 'proximity only' interaction can be useful in some situations, if someone were to want to move only the arm... but for myself, in order to be versatile and efficient on the theremin, the capacitance variable is a requirement.

dewster, I think I have to elaborate a little more about what's happening with this external portion of the equation that I've described. 

I conducted an experiment that ended up yielding some pretty interesting results.  I bought a spool of conductive thread (silver coated nylon thread...it was for another application, not originally having anything to do with this pitch linearity study)   I unwound approx 1.5 meters of the thread, did NOT cut it, left the spool on the ground, tied the loose end to the pitch arm on my etherwave pro about 2 inches from the point where the antenna extends...   my instrument was tuned the way it normally is, for my normal desired playing range/interval size.   I stayed in my normal position, but with my pitch hand/arm resting to the side of my body.    The resulting pitch was very high in frequency in the highest octave of the ePro range.    

the experiment proves one thing: that the total surface area of the conductive thread (the thread left on the spool sitting on the floor) had an effect on the resulting capacitance value...  I'm sure this sounds perfectly logical, but a spool of conductive thread is not a human hand/arm/body. when the proximity of the fingertips (the closest appendage to the antenna) changes, so does the proximity of the arm (which is also within range of the pitch field)... which creates a new equation at every place along the pitch range because a human arm does not offer a linear distribution of surface area as it is extended/retracted.  It can however, depending on the awareness of the player, be moved in such a fashion where there is a mostly linear change in capacitance... but that variable is up to each player.  I make it a point in my technique to always move my hand in a straight line from my body to the central point of the pitch antenna.

This external equation not only changes based on the shape of a players arm and how he/she moves it, but it is also affected by what location along the vertical antenna the player chooses to approach.  

I would bet that if anyone took measurements while approaching a the top of the antenna, the results would greatly differ from measurements take with the hand approaching the antenna base or center point.

The point I wish to make is that if the guy measuring changes in capacitance does not include his own body's non linear distribution of surface area (and how he chooses to move his arm) into the equation, then there are variables being left out, and a graph of plotted measurements can only be useful in the closed system from which the measurements were taken. each different person would create a different graph.

I hope I haven't lost anyone. I don't intend to undermine the intelligence of anyone here, but i get the impression that there are significant variables that are not being brought into question, not only by engineers, but theremin players also.  please prove me wrong if you believe I am mistaken.

my personal philosophy of improving on/with the theremin is very simple. it's awareness.  if I am aware of all the variables, then there is a greater possibility of working around the various shortcomings of the instrument. Not everything can be controlled by the player at all times, but a great majority of the variables can be tamed once they are understood.

what one person perceives to be freakish ability, another person perceives to be extensive methodical/practical study and immeasurable amounts of trial and error.

"... the experiment proves one thing: that the total surface area of the conductive thread (the thread left on the spool sitting on the floor) had an effect on the resulting capacitance value...  I'm sure this sounds perfectly logical, but a spool of conductive thread is not a human hand/arm/body. when the proximity of the fingertips (the closest appendage to the antenna) changes, so does the proximity of the arm (which is also within range of the pitch field)... which creates a new equation at every place along the pitch range because a human arm does not offer a linear distribution of surface area as it is extended/retracted.  It can however, depending on the awareness of the player, be moved in such a fashion where there is a mostly linear change in capacitance... but that variable is up to each player.  I make it a point in my technique to always move my hand in a straight line from my body to the central point of the pitch antenna." - Randy George.

Randy, I think you are on the right track, particularly with "when the proximity of the fingertips (the closest appendage to the antenna) changes, so does the proximity of the arm (which is also within range of the pitch field)... "

In some prior posting which I cannot find, I presented a similar hypothesis.. my thinking (and observation) is that linearity is improved when the antenna is thin and long as the horizontal hand-antenna coupling is less dramatic, and the angular coupling between the antenna and arm is more significant.. These are the pictures I posted in that thread.

These pictures only demonstrate the hand/arm/antenna couplings - as you have pointed out, many other factors come to play - movement of body bulk, background environment etc.. These are extremely crude pictures - it should be remembered that every point on the antenna has angular or horizontal coupling to every point on the hand and arm - too complex for me to draw in any meaningfull way! - Also, each point has coupling to the body bulk, to the floors / cielings / walls etc..

As all of these are variable, I cannot see that any scheme can "naturally" give constant linearity and note spacing regardless of the environment - Unless the theremin was sentient or had some means of "scanning" the environment and adjusting itself, an always linear, always consistant theremin is impossible.

In fact, technically, I do not think one needs to fully 'scan' the environment to obtain enough data - I have investigated a 2-point scanning scheme, where the player places their hand as close to the antenna as is reasonable, and auto-calibration occurs for this, and then moves the hand to the required 'null' position and auto-calibrates here.. Once this is done (for both pitch and volume) a reasonably linear sensing field can be calibrated between the two points, and a single user control for each would allow 'bending' of the curve (particularly useful, IMO, for volume / expression) to taste.

But it would not surprise me if expierienced players are able at present, even subconciously, to optimally orient their body position to correct some evvironmental distortions and give best playability - particularly if a long thin antenna is used which would perhaps tend to smooth discrepancies anyway.. And I actually doubt that the extra cost of implementing a automated tuning / linearization scheme would ever have payback - I think the cost could be greatly reduced and possibly feasible with a digital theremin, but its not economically viable in analogue.

Fred.

"This external equation not only changes based on the shape of a players arm and how he/she moves it, but it is also affected by what location along the vertical antenna the player chooses to approach.  

I would bet that if anyone took measurements while approaching a the top of the antenna, the results would greatly differ from measurements take with the hand approaching the antenna base or center point.

The point I wish to make is that if the guy measuring changes in capacitance does not include his own body's non linear distribution of surface area (and how he chooses to move his arm) into the equation, then there are variables being left out, and a graph of plotted measurements can only be useful in the closed system from which the measurements were taken. each different person would create a different graph." - randy george

I do totally get where you're coming from.  And I don't mean to imply there is a "one size fits all" solution to bringing linearity to the pitch field, but I'm hopeful that there exists a simple algorithm that makes the field feel much more linear to most players. 

The big challenge to the designer is that the algorithm pretty much has to be simple - with only one or two easy adjustments - or it will likely be impossible for the average person to understand and use in the field.

I could be wrong, but I would think that most Theremins (almost regardless of playing style) tend to cramp up near the antenna.  I would also think that most Theremins (that aren't highly coupled) similarly tend to cramp up near the null, and so the same algorithm may work for both.  I don't want to overplay a hand I haven't physically investigated yet, but a polynomial would likely make this a lot more linear.  It certainly seems promising in simulation (there is a page in my spreadsheet that investigates this method) but who knows until it is actually implemented and used extensively.

Thanks so much Fred!  your diagrams depict exactly what I had in mind with the arm and hand each contributing to the equation differently depending on proximity. The diagrams appear to depict a player approaching the base of the antenna. I guess you were intending to illustrate the point as simply as possible. approaching the center of the antenna seems like it would require a more complex diagram.

For the Etherwave Pro I notice (while I sit when playing) if my right leg is sitting further toward the direction of the antenna, that the bass range intervals expand... this happens because my hand is retracted more towards my body, and the leg begins to have some influence on the equation due to its proximity in the field.  For this reason, some years ago I changed my theremin chair to get me in more of a standing position with legs more or less straight, rather than sitting with my right leg bent and further into the pitch field.

a few months ago I had a day dream of a possible solution to offset body movements. similar to the goal of your two point scanning scheme to bend the curves...

Can you or anyone tell me if this has been thought of before or if it is even possible?  ::

Basically, a third capacitive circuit in the theremin (somehow engineered to not interfere with the existing pitch field, or volume field) that obtains a capacitance value of only the player's body [not arm]... so that the offsets created by a player's body movements could be subtracted from the significant changes [of the arm/hand capacitance values].... the goal being to eliminate one more variable from the "external" equation...  we could then be left with more consistent changes in pitch (more consistent interval spacings created from the learned muscle memory of the player) when moving the arm/hands/fingers only.

 The main requirement (which I imagine would be very difficult) would be for it to not detect any change in capacitance from the arm/hand movement. I was thinking this third field could be emanating from the center of the theremin housing, from a plate-like antenna, but lower than height which the pitch antenna extends so that the bulk of the player's body is closer in proximity than anything else... I imagine a vertical housing such as the E-Pro form factor or RCA style cabinet could facilitate the new addition the easiest.  Perhaps this needs to be a different type of circuit, like with an infrared distance sensor, but IR sensors are very one-directional (perhaps thats could be a good thing).

The solution does require that a player not use the full 'body swimming in the field' technique and I realize that it's not going to work for everyone, but ... for those players whose body movements are minimal, I think it could work wonders.   I normally keep my body as motionless as possible, but damn, I sure wish I could relax even more... because keeping the body motionless does require an added degree of concentration.   Perhaps all this is a pipe dream, but in my dream quest to reduce variables, I have to bring the question to the table.

Thanks dewster, i think I'm pretty much on the same page as you.

RE cramped up in the highest octave: My Etherwave Pro has a very linear pitch response right up to the antenna, and also I've tweaked the bass interval response to get linear response in the bass range too.  but the catch is, that this linearity works only for me because of my body size, and playing method. even the amount of 'meat' on a players hand even has an effect on the amount of capacitance changed.  I have very slender fingers. 

 I've had a friend step up to my theremin recently and do some tests, he is quite a bit more bulkier than me.   none of the linearity was there when my friend played my instrument.   He was not a theremin player, which doesn't really help but  I paid close attention to his movements.

The next time I have the opportunity, I'm going to visit Thomas Grillo again, and we'll document a series of linearity related tests.  I'm sure the results will be interesting...  I'm convinced that no Etherwave Pro plays exactly like mine either, and I'm very interested in doing some real-world A-B tests between multiple EPros. 

"The diagrams appear to depict a player approaching the base of the antenna. I guess you were intending to illustrate the point as simply as possible. approaching the center of the antenna seems like it would require a more complex diagram." - Randy

The diagrams are extremely crude, just to illustrate the concept.. I think that approaching the antenna at some mid point will increase the angular effect.. but the maths is way beyond me - the thing is that the further from the hand <-> horizontal <-> antenna plane, the less effect any extended length of the antenna will have (I think) - there is probably some optimum length, and I would not be surprised if this was the arms length - it would not surprise me if the optimum antenna extended about say 120cm, with the optimum 'centre' at 60cm with the hand moving horizontal to this plane..

"Basically, a third capacitive circuit in the theremin (somehow engineered to not interfere with the existing pitch field, or volume field) that obtains a capacitance value of only the player's body [not arm]... so that the offsets created by a player's body movements could be subtracted from the significant changes [of the arm/hand capacitance values].... the goal being to eliminate one more variable from the "external" equation..." - Randy

The above would be difficult, perhaps impossible with conventional theremin topology - perhaps slightly more plausable with some of the stuff I have been working on (direct to voltage theremin front ends, which drive analogue computation circuits and then voltage controlled heterodyning theremin modules)..

But I think there is an easier way - this is directional antennas.. These antennas eliminate all capacitive influence outside of their focus.

Alas, again though, we hit the barrier of cost - Unless produced in quantities of >100 units per run, these instruments would be unaffordable for all but a few thereminists.. If less than 10 were built, I would need to charge perhaps £6k  each just to break even on the time, let alone components, if 100 were built and sold, then they could still clock in at about £1.5k each.. Ok, I havent really costed your idea accurately, or even confirmed its viability - but I usually underestimate costs!

I dont really think the extra extremes, with the extra costs, is financially viable - I think that reasonable linearity, with perhaps a user control to 'bend' the equalization to taste, and an adjustable (say 3 to 5 octaves) playing field user control allowing the player to set the distance between notes, combined with an E-Pro like register switch, is the best that I might one day bring to market (or at least publish).. I cannot do this with conventional topology - the instrument would generate control voltages which could feed 1V/Octave synthesiser modules, or drive my 1V/Octave heterodyning theremin modules.. It may be possible to produce a fully analogue heterodyning theremin using the above topology for about £800 if 10 or more were built.

I would love it if someone with money sponsored this venture (and took care of the busness side), or if someone with money and skill produced these theremins in exchange for some crumbs for my guidance.. I simply do not have the business accumen to stay solvent!

Fred.

Thanks Fred.  Its interesting to me to become aware of all the details that must fall into place to actually carry a design all the way to production. from a musician's POV it's easy to forget that the dollars/pounds/euro/moneys provide another limitation.

The engineer in Seattle, whom I'm speaking with claims that he has a nearly finished design that will allow a player to customize the physical width of every octave. i'm not sure how exactly he achieves this, but he has asked if I would provide him feedback and will send me a prototype unit to evaluate... it is a digital design and I think it must be a cost effective endeavor, but I can't say anything more, he may perhaps wish to keep his ideas a secret if he intends to mass produce his theremin design. I'll ask if it's okay to share feedback about the instrument.

I had not previously heard of him and he hasn't ever posted anything online but he told me he has read some threads on TW before...  that may suggest that there are other people who read these forums, who could potentially publish/manufacture a next great theremin design. who knows?

When I visited Spain in 2008, at the University where I made a performance there were some teachers there that also suggested the idea of directional antennas... That could be a great solution, but It seems to me that the traditional appearance of the theremin would likely be affected... people might even not wish to recognize it as a theremin... 

I would like to hear more feedback from players...   I'm sure there are people that do not care much about a linear pitch response, perhaps players using the theremin for experimental/control/electronic effect. This discussion would probably not matter much for that crowd.  But, are there others in this community that do care and have something they would like to contribute? please share.

"I had not previously heard of him and he hasn't ever posted anything online but he told me he has read some threads on TW before...  that may suggest that there are other people who read these forums, who could potentially publish/manufacture a next great theremin design. who knows?" - Randy

Yes, a lot of people read these posts - and quite a few attempt to take ideas posted here and quietly combine them to put into production.. The ethos of playing "open hand" is not one generally adopted by those who are looking to make money, or those with "good" business sense, LOL ;-)

From what you are saying, it sounds like post-antenna processing of some kind (almost certainly digital) is being applied.

Only time will tell if a 'great' theremin will evolve from someone quietly absorbing and implementing the ideas publicly presented.. But I am reasonably sure that such a theremin will not come from someone who has not "saturated" themselves in the musical aspects - as in, I doubt it will come from someone who has never (or rarely) witnessed theremins being played live by good thereminists, or has not had their own (even if limited) expierience playing the theremin under competent guidance.

I have made several prototype theremins with directional antennas - these look like conventional antennas (slightly fatter than average, because the focussing 'shield' must be seperated by a reasonable distance from the sensing antenna / electrode) - but when using a long thin antenna electrode, the proportions become more normal.

But all developers of anything more than a 'standard' theremin will, I think, face real problems with cost and sales.. There may be a "big" market for mid / low end theremins up to about say £500 - but this market is choaked full of "acceptable" theremins, including the EW.. Above this price one moves, I think, into more 'sparce' territory - but the number of those willing to buy at this price drops hugely, so one cannot risk larger production runs, and economies of scale kick in making the instruments even more expensive.

Even for custom built instruments, I have found a barrier at about £600 - Pro (or people who think they are "Pro") thereminists who have contacted me for special theremins doing particularly unusual things, are simply not willing to pay for the time required to develop their one-off.

There is NO money at the "High End" unless you can make a High end theremin at a low end price.. Or unless you are a salesman / conman who can sell a Mini for the price of a RR.. Alas, with my "sales" skills, I would manage to persuade people not to buy my RR at the price of a Mini, by pointing out fuel consumption and other "down sides" of the RR !

Anyway - this is probably my last post on this subject for a while.. Will be interesting to see where this all goes and what this new theremin will really be like.

Fred.

"My Etherwave Pro has a very linear pitch response right up to the antenna, and also I've tweaked the bass interval response to get linear response in the bass range too.  but the catch is, that this linearity works only for me because of my body size, and playing method."  - randy george

I think this is an argument against the inclusion of fixed linearizing compensation circuitry in the EQ network, and perhaps also against the aggressive application of any linearization method.

"The engineer in Seattle, whom I'm speaking with claims that he has a nearly finished design that will allow a player to customize the physical width of every octave."  - randy george

Off hand, I can't imagine a use for this, particularly if the octave widths change abruptly at the transition points.  It sounds like it might be a nightmare for the average player to adjust, and if so, the added benefit would have to be huge.

"But I think there is an easier way - this is directional antennas.. These antennas eliminate all capacitive influence outside of their focus"  - FredM

And if used to reduce the bulk antenna capacitance, a directional antenna could dramatically increase raw (%F/pF) sensitivity.  You might be able to make a Theremin you could play with your hand a meter or more distant.

"There may be a "big" market for mid / low end theremins up to about say £500 - but this market is choaked full of "acceptable" theremins, including the EW."  - FredM

Every time I open up my EWS (or have a look at the schematic) I'm a bit shocked at how little there is inside it.  The wood case probably costs as much to manufacture as the guts.  I know it's a low volume item, but I can't imagine there's more than maybe $50 worth of stuff total.  But I guess Moog Inc. must at least double that (=$100) and Sweetwater or whoever you ultimately buy it from must double that (=$200).  Profit margins for this market seem kind of high for what you get, but then again that's probably due to the low volumes.  I would bet no designer ever gets fully compensated (at least monetarily) for their R&D time, but I reckon none of us here (players included) got into Theremins for the big bucks, international name recognition, or cute groupies.