Modelling the theremin wave in software.

@Coalport - A reference to Stephen Hawking, is it my brilliance or my stubborn orneriness... no that would be you. (-'

My theremin research halts until the day I find that “local” talented musician. It has been an inspired week so that special person by chance might be reading this now.

Paul’s Box was lost during the clean up from the 1971 Sylmar earthquake. Coalport I think your fingers dazzled people in this area around that time? Hollywood called TheBox an Electro-Theremin so then he accepted the name.

The stock market wounded me on Wednesday, I heard a whisper and got out completely on Thursday, breaking even on the bounce back. Then came Friday and it was an "old shit moment". As an observer I now sit on the sidelines. This is called Victory!

Is it better to be brilliant or to be lucky?

Christopher

Edit: Where I go astray is when someone says "that theremin sound", can anyone share an example with me of what exactly they think this sound should be? We are all creative and have a different opinion of what a theremin is, just don't get hung up spending years redesigning the door handle so you do not have to climb inside of the Ferrari.

It is better to be lucky, because then you will be brilliant and remain undefeated by fate in spite of it!

Who the hell wants to end up like Tesla?!

Here's a static tone from my Etherwave with spectral plot and waveform. The spectral content does fluctuate so perhaps if you want to program from a csound perspective you could throw a random number generator at slightly altering the amplitudes of the harmonics. Sight changes in the spectral content I think will give you markedly different sounds but in the theremin soundspace. This is with the Etherwave Waveform and Brightness knobs at 12 O'clock. I like Thierry's mathematical description of the knobs that can be added to a simulation. Moog simply says: The WAVEFORM knob adjusts which harmonics are strong and which are weak, while the BRIGHTNESS knob adjusts the overall amount of harmonic content. Would be better to take video of this as you turn the knobs. Actually this is not too far off the Burns plot, but this has more grit in the sound.

Now let's take a look at a short extraction of Old Temecula's "Clara's Voice" sound file. This is not a static tone so you see more movement. Something I've noticed about theremin waveforms is that peaks on one side of the plot are often sine-like but the other side is flattened or sometimes more angular. 

Frankly, these plots don't really tell enough of the picture to be totally useful but if you want to model a theremin tone it appears you need to include the first 8 or so harmonics in a more or less decreasing amplitude and then play around with that till you get a sound you like. In Old Temecula's example above, the third harmonic was decidedly reduced as you look at the entire dynamic plot. Kind of going with what he says about enhancing even harmonics - at least for the first few.

I'm more or less convinced you can have a great variety of these semi-sinusoidal waveforms that can give you a theremin-like sound as long as you model the envelope and vibrato and portamento correctly - which is a heck of a lot harder than adding up sine waves. It's that playing technique that I feel is going to get you to the sound you want, not so much the waveform.

Frankly, I think sampling in different ranges is a better way to go.

Hello rkram53,

I was never the brightest bulb in the room so I developed my ability to cheat. When I first tried to figure out what caused the sound I was after there were many days just looking at waveforms on an oscilloscope. 

Master builder Art Harrison told me it was all about wave-shape, I just could not see it. As dewster says a pure sine wave does not have harmonics, but I say if you place a dink in the wave-shape correctly you will get a much better sound.

Looking at the harmonic representation in your graphic at the base of each one there is a bunched up squiggle on each side of the harmonic finger, what is that,  I see it on my computer software but have always wondered does this play into the sound, a character I do not recognize on an actual scope.

Looking at your software scope representation of my waveform I think it would be better if you chose a lower frequency (220hz) as I don’t think software is fast enough to display an accurate crisp wave shape. Not that it matters but I also think the waveform is inverted, in hardware-design this would be important to me.

In hardware I get my skew by slipping the audio signal through an audio transformer using a technique I developed after 10,000 attempts. FredM also mentioned something about this before. I am not an engineer but I would call it a slight phase shift in the wave form which creates the harmonics. Not all vacuum tubes will generate the big sound that creates the better vocal effect. After hand capacitance this fat tube effect is something important and unique to theremin design.

All my stuff is disassemble so I am unable to conduct further tests. My bedroom lab makes me look like a hoarder. If only I could throw something away. That's what happens when you grow up really poor and why today I happily give theremins away locally for free, always have my eye on Craig's List.

Christopher

Edit: While on the subject of sound, has anyone noticed the 15 watt Moog theremin amp sounds muddy? (extra dull)

Christopher,

Interesting. Not sure why that scope plugin inverted things, but if the goal is to use csound here to try and create a waveform it's more the spectral plot that's of interest. I chose a section in your file that was stable for a little bit to try and get the plot as you do a sweep up and down. 

But frankly, I'm convinced if the goal of the thread was to use csound to simulate some waveforms, adding up the first 8 harmonics with decreasing amplitudes as in the plots and then playing with emphasizing some harmonics is a valid starting place. I might actually download csound and give it a shot. But no static theremin tone is interesting in and of itself.

Rich

P.S. An I think I'll stick this link here for safe keeping.

http://files.csound-tutorial.net/floss_manual/Release01/Cs_FM_01_epub/ch023_a-additive-synthesis.xhtml

@rkram53, thanks for the plots, I did some of that myself with theremin recordings.

Actually Thierry's formulas give a pretty decent sound. (And yes the sound with the waveform parameter at 0 is pretty much those 8 harmonics) What I did to make the waveform fluctuate a little bit was stack a few slightly detuned versions of itself.

As I said I'm not specially interested in replicating a Theremin, It's just that I think that the Theremin timbre is nice, and that it would make a nice dynamic sound to play with the Linnstrument. Right now it sounds pretty nice, and being able to continously change waveform for each note like if I was turning the knob sounds good and it's satisfying.

What I'm missing right now, and that was the main point of the thread is the connection of frequency and amplitude to timbre that most acoustic instruments have and that the Theremin shares.
Thierry's model lacks that. I'm solving it somehow by changing brightness with amplitude. And by slightly changing the waveform with frequency. But it would still be nice to know how to model it to make it closer to the real thing.

@Rampoina,

Linnstrument? How do you like it?

I'd love to hear what you are cooking up there with this theremin-like sound. Can you set it for microtonal playing?

Rich