Let's Design and Build a (mostly) Digital Theremin!

Ah OK, it is under 1_OSC xmix < 0.

It is fun to play and can go from extrem to subtle. I think it is a keeper.

"Ah OK, it is under 1_OSC xmix < 0."  - Mr_Dham

Oops!  Yes, I left out that one little detail...

"It is fun to play and can go from extrem to subtle. I think it is a keeper."

Thanks for testing it and playing around with it!  The two 1_OSC:xmix < 0 oscillators are fixed at A440, but can be offset with 1_OSC:offs (and to some degree the hmul knobs).  

As usual, all of the oscillators can be offset via the 0_OSC:oct knob, but the two fixed oscillators are NOT offset by the D-LEV:bank knob, so you can do lower frequency beating by setting e.g. D-LEV:bank[2] and 0_OSC:oct[-2].  Though this pegs the upper pitch 2 octaves earlier due to design constraints, so for this reason it's best to do the initial +/-1 octave offset via the 1_OSC:offs knobs.  

The A (& D) beating can be offset to e.g. C (& F) by setting both 1_OSC:offs knobs to 51.  I think A is the most natural drone for the D-Lev because both ends of the pitch range peg at A8, and A minor is a bit more harmonious than C major.

Yes, I didn't catch initially that you can choose beating frequencies, that's rather nice.

A440 + offset makes it quite predictive, especially when playing with sine waves (harm = 0). This can really help the user understand what's going on.

I like the fact that if you know where you are in this distorted audio field, you still can play a simple melody and add a pinch of oddity.

I am just wondering if user wants to have rich harmonic content for all oscillators (fix and variable oscillators).
High harm value make the sound quite extreme.

"A440 + offset makes it quite predictive, especially when playing with sine waves (harm = 0). This can really help the user understand what's going on."  - Mr_Dham

Good point!  So I was playing around with that this morning and saw that zero beat was happening at A880 rather than A440 - oops!  Fixed it.

"I am just wondering if user wants to have rich harmonic content for all oscillators (fix and variable oscillators). High harm value make the sound quite extreme."

Another good point!  I played around a bit with that too.  Sines (harm[0]) for osc 1 & 2 would produce something very similar to the Moogerfooger MF-102* and offsetting them via the sprd knob would even simulate the LFO in the Moog unit.  The thing though is that harm[0] kills all PM/FM, so the hmul knobs then do nothing, and the fm knob turns into a sort of reverse harm knob which can only reduce harmonics.  One could switch in ring modulation with the fm knob being negative, rather than the xmix knob being negative, and label it "amfm" which would be cool.  But it all sounds kinda boring compared to the extreme (as you say - and I agree! - and made even more so by the addition of some extreme resonator) ringmod + FM presets.

If you want to give them a spin: https://d-lev.com/support/2023-09-28_ringmod_test_loads.zip.  The 517b3691 file has osc 1 & 2 harm[0] for xmix < 0.  The e88c8d50 file just has corrected A440.  For both of these you'll need to go to the P_FIELD and V_FIELD pages and double the values of Ofs+ and Ofs- (I'm considering doubling the resolution of these knobs to better accommodate another project).

*In other news, farewell Moogerfooger: https://www.moogmusic.com/news/farewell-moogerfooger.  Kinda ironic that they market the digital plug-ins of these as equivalent to the physical "genuine" analog pedals being phased out.  "They effortlessly conjure any sound you seek..." why would they discontinue this self-described holy grail line of pedals?

"But it all sounds kinda boring compared to the extreme" -Dewster

An irresolvable dilemma: https://youtu.be/3V-saYmzUyg?t=145

517b3691 has more chance to be in my music practice but that's just my own taste...

In 517b3691, when harm > 0 it looks like a constant pitch can be heard ((continuous component of osc 0) * (osc 1 + osc 2) ?).
Maybe it is there in e88c8d50 but it is hard to distinguish.

I played more with e88c8d50 today.

Reaching extreme sound with extreme setting is not a problem and could even be expected.
That's just that extreme arrives very early with low harm values (from 12 maybe) and then we have a very binary choice between very gentle or very brutal sounds.

How about : 
xmix < 0  then ring mod ON
harm < 0 then No harmonics for osc1 and osc2 (maybe also useful even when ring mod is off ? )
 
 

"In 517b3691, when harm > 0 it looks like a constant pitch can be heard ((continuous component of osc 0) * (osc 1 + osc 2) ?)."  - Mr_Dham

Hmm, could be the xmix code.  Leakage is always something to watch out for.

"That's just that extreme arrives very early with low harm values (from 12 maybe) and then we have a very binary choice between very gentle or very brutal sounds."

Probably also the weighting of xmix, which should be linear for this I think, rather than 1/3:2/3 at the center.

"How about : xmix < 0  then ring mod ON; harm < 0 then No harmonics for osc1 and osc2 (maybe also useful even when ring mod is off ? )"

Ooh, interesting idea, thanks!  I'm just about out of real-time on that thread but I'll certainly try to give that a go.  Half of this work is coming up with ideas, which I'm not always the best at, so I definitely appreciate the assistance!

"The issues with Claravox continue to amaze me. Many of these problems seem incredibly bush-league and cold have been avoided through some basic acceptance testing and/or very low cost changes to some of the components used."  - Bendra

My (limited) experience is that musical product development requires a very tight and multi-disciplined feedback loop.  The tightest loop is the developer working alone, which means, in addition to a varied technical background, they must have at least some musical chops re. the target.  Focused suggestions from highly interested and motivated musical technicians / technical musicians can be extremely helpful and needn't complicate the loop, but these multi-disciplined folks are necessarily few and far between, and any project is super lucky to have any at all.  It helps a lot I think if the development is not shrouded in typical corporate secrecy and arbitrary hard deadlines.

I don't know what the Cvox case was, but hardware types with little ergonomic background, coders with little musical background, and musicians with little technical background all working together is likely a recipe for disappointment - there's not enough common ground between them to facilitate the specific detailed communications necessary to drive effective development to a satisfactory conclusion.

My (limited) experience is that musical product development

Oh, pshaw! You've been waist-deep in Theremin development for the last decade

requires a very tight and multi-disciplined feedback loop. The tightest loop is the developer working alone, which means, in addition to a varied technical background, they must have at least some musical chops re. the target. Focused suggestions from highly interested and motivated musical technicians / technical musicians can be extremely helpful and needn't complicate the loop, but these multi-disciplined folks are necessarily few and far between, and any project is super lucky to have any at all.

Seems about right to me. Bob Moog himself was a really unique asset, irreplaceable really.

It helps a lot I think if the development is not shrouded in typical corporate secrecy and arbitrary hard deadlines.

The deadline pressure must have been crazy. For the company to announce that it would ship before the end of 2020 was so irresponsible and unfair to the development team. If my employer did that to me I'd seriously think about quitting.

I don't know what the Cvox case was, but hardware types with little ergonomic background, coders with little musical background, and musicians with little technical background all working together is likely a recipe for disappointment - there's not enough common ground between them to facilitate the specific detailed communications necessary to drive effective development to a satisfactory conclusion.

This is all true and fair, and there are shortcomings in the Claravox that can surely be traced to these factors. I'm actually talking about really basic stuff than that, like:

* Varnish coating preventing antennae from making metal-on-metal contact with the jack it plugs into
* Loose connection for volume antenna causing the loop to wiggle and droop
* Stand legs not locking in place/falling out

I mean, really?

"Oh, pshaw! You've been waist-deep in Theremin development for the last decade" - Bendra

You're too kind!  But this is my first time to the rodeo, so there have been a ton things that I needed to figure out how to do.

"Bob Moog himself was a really unique asset, irreplaceable really."

He was!  An all-around innovator / developer / manufacturer.  His vision led him well, and he had the authority within his company to carry it through.  I think a lot of the failing of the corporate approach to creative development is the rigid hierarchy - those who know what needs to be done often aren't allowed sufficient control to do so.  Those with the power are usually too distant from the project and too non-technical to do any good.  That, and money ruins everything.

"This is all true and fair, and there are shortcomings in the Claravox that can surely be traced to these factors. I'm actually talking about really basic stuff..."

It's the physical design stuff that's the hardest to control.  Wood shrinks, no one really checks anything as long as it looks OK and works the firs time, etc.  Pushing an entirely new product out the door must be really hard for a smaller company to get everything right.  Larger companies can afford to do more QC of the incoming components and outgoing assemblies.  That said, it's a little scary seeing how unevenly populated the Cvox PCB is, like they just stopped at some point and didn't do a real final spin.