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

Posted: 12/23/2017 6:03:29 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"maybe you created the first theremin that got x-rated."  - xtheremin8

Ha ha!  I must give credit to a Thereminist here in NJ for whom I've done some surgery on his instruments (ESPE01).  When he came over the first time he did a very identifiable orgasm-y thing on his Theremin.

"...that wobbly pitch on the second video at 0:05 and the in the third at 0:33..?"

It's nothing really, though I'm glad you brought it up because it could be misinterpreted as a flaw.  The prototype is on a carpet with pad underneath, so it shakes around some when I'm fiddling with the knobs (sorry, not shouting at you, just wanted to emphasize that).  There's a fluorescent desk lamp a foot or so from the pitch plate (my cramped work space) which makes jiggling things worse (in a capacitive sense, not a mains hum sense - the hum filter crushes all hum).  If I shake it on purpose with the formant filter on it sounds a little like giggling.

This is why the Kowalski Theremin - where the left hand is manipulating performance controls - is a large-ish box with four firmly supporting legs.

"quantisation is something i still can't get around with. i think it's the spacing between notes. the start/end point..thing. but i see benefits like to learn phrygian or other more exotic scales, that are not obvious to our western-chromatic tuned ear. or different tunings like the lucy scale for example. colundi everyone?  which scales have you in peto on your machine?"

Pitch quantization is isolating the note value, then applying a non-linear function to it so that the value is smoothly biased towards the end points.  Stronger bias gives stronger qantization.

The Theremini has pitch quantization, as well as scales and base key settings, so I pretty much knew the value of it, which is not a whole lot to me.  The prototype only has chromatic quantization so far, though I might make it more complex later (it might only take a day or so of work to make it like the Theremini).  I turn it on a bit (setting 1 or 2) sometimes when practicing, it helps make things sound better, though it interferes with vibrato.  It's an interesting toyish feature, which is probably why it was included in the Theremini.

You bring up a good point though, that quantization with scales could help train the ears of those who are new to music theory -  or it might just make them lazier!

"that leads me to another question i had and was always to afraid to ask: that led tuner. now in action it makes a bit more sense...i worked some years ago some days with a local synesthetic artist and she told me much about how that is different and such, to see tones in colours. so that strikes me every time i see rgb leds and the possibility to create almost every colour with. but i don't know if such a thing would be possible, to have a big led changing colour according to the played note."

FredM waved me off of color changing due to some people being color blind.  I thought emphasizing certain notes on the tuner with a different color (very much like the black keys on a piano) would help, but it really interfered with readability.  

Having a single LED changing color might be interesting to watch, but I don't think it would be discrete enough to really be useful. With the prototype tuner certain scales have patterns that are pretty obvious (major, minor) - in different keys they are rotated.

In the end, color is another axis of information, though for some reason my brain can't process it quickly enough.

"a nice book: "push turn move" it's all about interface design in electronic music, if you like fat heavy prints,1.7KG, it's published by bjbooks in denmark."

Thanks for the pointer!  I'll look into it.  I've read a certain amount about UI design on the web and a lot of it is fairly practical (group common things, don't have more than 5 or so things in a group, place the most important things at the edges of the group, etc.).  Synthesizers have a lot of adjustments, and tend to have poor UIs.  Yamaha has made some notoriously bad UIs (FS1R, Motif Rack, AN1x - I owned 'em and now I don't, and the UI's were largely to blame).

[EDIT] Ooh, that's a pretty looking book!

Posted: 12/23/2017 11:30:55 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

On the twelfth day of xmas, your dewster gave to you: 12 (previously posted) Theremin prototype videos:

https://www.youtube.com/watch?v=crjbTuCnlvI

https://www.youtube.com/watch?v=lm7DPvUaquw

https://www.youtube.com/watch?v=m6nnEG6A0C0

https://www.youtube.com/watch?v=_tVfaZDg7wE

https://www.youtube.com/watch?v=onTCRCq8C6o

https://www.youtube.com/watch?v=vn2uuYh9q-I

https://www.youtube.com/watch?v=jcIq-OnQqC4

https://www.youtube.com/watch?v=ExEyotYfeVE

https://www.youtube.com/watch?v=nzieT8n18h0

https://www.youtube.com/watch?v=Wf1mEHvZo8E

https://www.youtube.com/watch?v=1cJo4yS_Nyg

https://www.youtube.com/watch?v=vQZbJ_MjOzc

Just the beginning of the synth side of things, it should go a lot faster than the theoretical and processor side of things...

Give me your constructive criticism TW, I promise I won't bite (much)! 

Posted: 12/23/2017 11:37:23 PM
xtheremin8

From: züriCH

Joined: 3/15/2014

thank you for the clearing words about the wobblies. i was worried but now i'm not.

yes, colours are worrying and the rather lame eye get's easily fooled. the synesthetic thing is also a bit more complex. i don't know if you understand german : http://www.farblicht.ch/farblichtfluegel  . good thing  when you have quantised/ stepped pitches, it can also be usefull to slightly trigger external effects.

UI's are really crazy sometimes, as said vcv-rack was my eye/ear opener for this year. still beta. only limit is cpu. and when i look at some hw modular synth modules...i would like to know what drugs these people are on while developing their products.  or the theremini, good fact on that thing: there's so much obviously wrong with it, for example the double feature of the knobs. one needs exclusive pitch and volume knobs, dedicated by size or placement or both. all the rest is mostly labeled or after two three times of use already learned. as more complex it get's the more infos one needs. olé oleds..... and the most confusing manuals trophy goes to : yamaha, no roland is worse.

 

 

Posted: 12/24/2017 12:01:32 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

The Theremini needs more knobs.  Gobs of them are trivial when you've got the pins and logic galore of an FPGA, but they're more problematic with processor GPIO.   IMO the drugs these guys are on is "get it out the door / get the kids thru college" capitalism, and basic lack of background (gotta wear a lot of hats with this stuff).  Theremin himself was stupendously well suited for what he accomplished.

Posted: 12/24/2017 1:25:56 AM
xtheremin8

From: züriCH

Joined: 3/15/2014

the country i live in missed the chance of giving every citizen a basic income, so one could decide more freely what to do with the given time. you're right, to know on witch side the soldering iron gets hot is far not enough. oh, already..  merry christmas..

Posted: 12/24/2017 7:36:33 AM
gerd

From: Germany (Black Forest)

Joined: 11/25/2017

Does she have a basic income? I hope so ...

https://f1.blick.ch/img/incoming/origs4819380/644253663-w1280-h960/stock-image-fail-soldering-iron-bob-byron-1.jpg

 

Posted: 12/24/2017 8:32:13 AM
Oilspill

From: France

Joined: 11/13/2017

"Does she have a basic income? I hope so ... https://f1.blick.ch/img/incoming/origs4819380/644253663-w1280-h960/stock-image-fail-soldering-iron-bob-byron-1.jpg"

 

sealed not necessary for her, she is married....sealed

@dewster: i like the "Tupperware" cases.. sometimes i also use kitchen-things for my prototypes, my wife hates me sometimes for thattongue-out

Cool project

 

Posted: 12/24/2017 1:09:36 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

Yeah, that's gonna leave a mark...  and there isn't a lot on that side of the motherboard that would need soldering.  To be fair, I know zero about modeling, and the best high-end irons allow one to grip it closer to the action (as god intended).

Much as I think it's one of the greatest ideas ever, basic income will never happen because humans are hard-wired to enslave everything in sight - plants, animals, us - and they really don't care how much damage and suffering it causes as long as their bank account gets another nickel.  The ice caps are melting and the insects are dying off, but, hey, I got mine.  The way we're going one guy will end up owning the entire earth any day now, the redistribution of wealth vs. time curve is directionally incorrect (like a mofo).  I'm so old I remember when they were predicting we'd all be living the life of Riley at this point.  I didn't even get the shirt.

Oops, I meant: ho ho ho!

Posted: 12/24/2017 9:24:18 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

Volume Side Nailed

Believe it or not I've been losing sleep thinking about the volume side processing.  I've wanted something that works equally well for either far or near louder, with parameters that don't interact too much.  What's been confounding me is the transition from float to UINT (unsigned 32 bit integer) after the LOG2, multiply, and EXP2 (to do the linearizing variable fractional negative power).  Up until today I was variably gaining up the float to UINT conversion so as to maximize UINT resolution, then manipulating things afterward, but that isn't ideal. This morning it hit me that I should offset the last float stage with a parameter before gaining up the conversion.  Making the float negative allows us to use a positive number to set the zero crossing point in the field, which is more intuitive.  And if the float is converted to a signed INT rather than a UINT, then converting the resulting signed int to unsigned is a simple matter of inverting the MSb, which centers the whole shebang in the middle of the final UINT.  This works like a champ!  It works so well I've applied it to both the volume and pitch sides.  On the pitch side it gives one more parameter to fiddle with, but you don't really have to fiddle with it once it's set to something that works over the range you're using, and it gives you finer pitch vs. hand position setup control vs. only having octave bank switching.

So the volume zone size (operating width) and zone location (height above the antenna where things start happening) are much more independent, and switching the sense doesn't require a lot of fishing around for the right settings.  Setting the zone size is a matter of placing your hand at the point where you want the zone to start or end and adjusting the knob until you hear the desired zone edge kicking in.  

So that's a very basic thing finally put to bed!  It's weird, implementing these things once there's a game plan goes pretty quick, but sizing adjustment range, gain, and response of the new parameters can take hours.

Next up I suppose will be experimenting with high pass filtering of the volume signal mixed in to produce faster attacks and the like. And I need to dig out a microphone and start analyzing my own dynamic glottal waveforms.

=========

I'm using a Logitech C920 webcam to take my movies, and the included recording tool gives you a surprising level of manual control over the internal settings of the camera.  But for some reason it really messes with the sound, even when I'm using direct input from my motherboard rather than the built-in stereo mic.  High frequencies alias all over the place and I hear a lot of "hash" associated with them.  I suppose they're doing some serious audio compression or something.  I wish I could turn that off.  So I'm using a hodge-podge of tools to record the audio separately, strip the WMV audio, edit the audio to match the video, then reinsert the audio.  What a pain...

Posted: 12/25/2017 8:13:28 PM
oldtemecula

From: 60 Miles North of San Diego, CA

Joined: 10/1/2014

dewster said: “Volume Side Nailed”

Can you feed a familiar audio signal through your volume control circuit to demonstrate response and quality of signal pass through?

I do this in my sample which also demonstrates another feature I achieved by using my method of PWM. The outer volume field is more responsive/aggressive than the inner near the loop. This allows for a very gentle, yet wider, soft control window near the loop before off.

You get a freaky daily visitor count on your thread.

I like to think it is several competitive manufactures waiting for a moment to buy out your research with a perfect offer you cannot refuse. That end packaging scares me for you.

The journey has been long for both of us.

Merry Christmas

Christopher

Ran a little test with the little girl pic, this thread is huge while being cataloged in the search engines (google). Most hits are landing on other pages, not the most recent or this one. A visit is still a visit. Would be nice you get a dollar for every hit, they all see that nice theremin ad at the top of the page.

You must be logged in to post a reply. Please log in or register for a new account.