Spiral Synthesis - possible in SM?

Sk · 29 Apr 2005

I read it fast, but I'm going to read it with mroe attention. It seems very interesting, since I'm into that kind of studies. Still I didn't get the point in using those waveforms to produce sounds. What would they sound like? And, how would you manage a complex signal to produce sound?

NothanUmber · 29 Apr 2005

Thanks for the link, that's really interesting. Here a SynthMaker-conversion of the given c-code to experiment with:

http://home.in.tum.de/~strixner/SpiralSynthesis.osm

(perhaps someone can host that at another place, I'm not at uni anymore so I don't know how long my account and therefore webspace will continue to exist...)

NothanUmber

malc · 29 Apr 2005

No problem:

[old link]

Haven't had a chance to read about this yet - sorry. Sounds intriguing.

NothanUmber · 30 Apr 2005

Today I tried to get some sound out of the oscillator. Althought the frequency should be in the hearable range I hear - nothing. I've no clue if that could have to do something with it but I realized the following strange behaviour:

* Decouple the "Number of Samples" input of the analyzer from the "300" float * create a new float module and connect it with the analyzer * enter the number "500" into it (now you should see a curve - but there is only a straight line) * now change the number to "501" (the curve is there now) * if you change back to "500", everything works as expected this time

Could it be that I do something that the analyzer (and the sound out) doesn't expect or is there simply a bug hidden somewhere?

NothanUmber

P.S.: The analyzer bug seems to be solved in SynthMaker 0.9.3e. That leaves the question of the sound (or the lack of...)

NothanUmber · 30 Apr 2005

Another toy to play with:

http://home.in.tum.de/~strixner/SpiralSynthesis2.osm

(this file contains a slightly modified spiral oscillator that is connected to a sound destination and a midi in that modifies the angular speed and the impulse force in order to change pitch and velocity. The scaling modules could be improved, e.g. one could automatically change the impulse force corresponding to the angular speed - the sound softens atm for higher pitches... In order to produce some (potentially) interesting sounds one could perhaps connect some modifier to the pole position input in order to let the sound become louder at each recursion step and then swell down for pole positions smaller one again. Or one could decouple the modulations for the resonator point and carrier angle and apply two completely different modulation functions to each of them. Feel free to experiment 🙂 )

Btw.: Does anybody know how to make sources that produce sound without an incoming signal (that is not constant), is that possible? (E.g. as tried in the previous oscillator in SpiralSynthesis.osm above)

NothanUmber

P.S.: Feel free to use the frequency instead of the pitch for note pitch, should make more sense... 😉

NothanUmber · 1 May 2005

Here the current snapshot:

http://home.in.tum.de/~strixner/SpiralSynthesis3.osm

(Now with separate modulators for carrier angle and resonance point, that way one has much more possibilities to shape the sound. Also changed from pitch to frequency for pitch (was a little bit late last night... 😉 ). The velocity rescaling is still far from being optimal (the current code is not systematically thought out, just the result of some short "trial and error hacking". The real dependency between the impulse and the modulating parameters is nontrivial, some more thoughts needed here.))

Have to go back to other nonsynthmaker related projects in the next time, but I'll most probably be back in the not too far future again. (Perhaps in order to start working on a Spiral Synthesizer, who knows... 😉 )

NothanUmber

Shifrin · 1 May 2005

This sounds very interesting, I'm getting the .osm you posted now 😀

Shif.

Tom7777 · 1 May 2005

I think it would be cool if there was some sort of display module in SM that made a 3d graph of the sound like Fig. 3.1 from the link skOre said about. NothanUmber thanks for the .osm examples, great work 🙂

Sk · 1 May 2005

Use Matlab instead! Lol

Tom7777 · 1 May 2005

no i mean like a realtime updating one that you could use on an actual synth.

skOre · 2 May 2005

Tom7777 wrote: I think it would be cool if there was some sort of display module in SM that made a 3d graph of the sound like Fig. 3.1 from the link skOre said about. NothanUmber thanks for the .osm examples, great work 🙂

Working on a 2D/3D Coordinate System based interface, currently at 35% Dev Status of the 2D type, so hold your breath 😉

NothamUmber: Great! Thanks for your work, started figuring these out, but it is indeed a hell of a value mess. Guess it will be most likely heavy key dependant (making the keys to dynamically change other values than just the pitch and velocity) to get this thing to play nice. But then I think that it will be a very versatile instrument!

Shall we couple up on this one NU ?

Tom7777 · 2 May 2005

Working on a 2D/3D Coordinate System based interface, currently at 35% Dev Status of the 2D type, so hold your breath

Cool, sounds good 🙂

NothanUmber · 2 May 2005

skOre wrote: Shall we couple up on this one NU ?

Sure, I'll gladly try to help out if you have questions and could apply some small improvements. (As posted earlier I'm a little bit short on time atm. but I'll see what I can do.)

bis demnächst, NothanUmber

skOre · 2 May 2005

I could probably just work on it (figure it out I should say) and hand it over to you to have a look at it. Just don't feel well about stealing this away from you... 😉

NothanUmber · 2 May 2005

Here another small update:

http://home.in.tum.de/~strixner/SpiralSynthesis4.osm

(Just gave up on using the resonance point modulation for pitch adjustment and use the rotation in the generating plane instead. That way the "hack" in the velocity scaler is unneccessary - now the impulse can be modified by the velocity in a linear way.) The "stuttering" at the beginning of the sound has nothing to do with the oscillator itself, it seems to be caused by the displays (go e.g. into the oscillator component subview and play there, now the giggles should be gone.)

NothanUmber

sKore wrote: I could probably just work on it (figure it out I should say) and hand it over to you to have a look at it. Just don't feel well about stealing this away from you... Wink

Take whatever you can use 🙂 In the next days I'll try to at least paint some pictures of what I think could be a nice concept to handle such a spiral synth in a more intuitive way than "adjust the divisor for the angle speed of the sinusoid that modulates the resonance point to 17.6... 😉 ) Perhaps some people jump on the train so that the effort can be handled even if everybody works on the projects just for a few hours a week.