Bazille - Modulation of fractal resonance according to played note possible?

[SamDi]

I intend to simulate a sync. How can I modulate the fractal resonance amount in a way, that the resulting envelope signal tone follows the played note? Is it somehow possible? Try to get it since 2 hours, but until now fail

[Viktor [TUC]]

Any modulation that depends on the played note would most likely include KeyFollow as a modulation source, did you already try connecting a "KeyF" source to the Fractalize mod depth input?

[Ah_Dziz]

If you need envelope control and keyfollow then you can combine them in one of the mixers.

[Scrubbing Monkeys]

I would also think this could be accomplished using the mapping generators. Each key would have its own value that you can assign.

[Howard]

I intend to simulate a sync. How can I modulate the fractal resonance amount in a way, that the resulting envelope signal tone follows the played note? Is it somehow possible? Try to get it since 2 hours, but until now fail

I don't understand "that the resulting envelope signal tone follows the played note". Could you explain?

[SamDi]

I intend to simulate a sync. How can I modulate the fractal resonance amount in a way, that the resulting envelope signal tone follows the played note? Is it somehow possible? Try to get it since 2 hours, but until now fail

I don't understand "that the resulting envelope signal tone follows the played note". Could you explain?

OK, I understand that the term "envelope signal" (inspired by ring modulation) is misleading. I meant these "overtones" introduced by typical sync (I guess that term is technically incorrect too).

I got it wrong:
When I first started to explore the fractal function it seemed to me that these "overtones" or as I called it "envelope signal" didn't follow the fundamental frequency and would be fixed instead, according to the amount of the Fractalize knob. So The Unshushable Coktor got my problem right. The problem of his hint was, that KeyF just would sent positive values from MIDI note #52 ("E below middle C") but I would have needed a kind of "unipolar" signal. Then I messed around with adding 5V to the KeyF signal. But even then it didn't work. Then I tried around with scaling the signal linear, make it sqare, etc. but nothing worked right.

As some days later, I tried again, I recognized, that my base assumption was wrong: these "overtones" do follow the base frequencey. I cannot tell you, why I heard it different the first time. Thus the problem is kind of "solved".

Nevertheless there result some further question from my side:
1.) Adding 5V to KeyF does it make it unipolar, means it adds a positive value starting from MIDI note #0 instead of #52 or just shifts to another MIDI note as zero point?
2.) Is it possible to make an exponential signal out of a linear? I mean with multiplexer I can nake it square, but how can I get it really exponential?

[Sound Author]

I would use a Mapping Generator instead of the Multiplex.

Make sure you have a table of 128 values selected in a Mapping Generator and change the Mode to "Key". Then make sure the "Source" is set to "None". Then right-click inside the Mapping Generator window to open the Context Menu and select "Shapes>Ramp". If you want to create a unipolar Ramp, you can re-open the Context Menu and select the option "Make Unipolar". If you want to create an upward exponential slope, you can instead select "Shapes>Quadric" and again select the "Make Unipolar" option. This is my preferred method of doing things with KeyFollow without ever having to mess with the Multiplex.

[SamDi]

OK the maper seems to be the gold standard for such problems.

If you want to create an upward exponential slope, you can instead select "Shapes>Quadric" and again select the "Make Unipolar" option.

But "quadric" doesn't make it exponential

Seems one has to put exponential values by hand then or is there another smart solution?

[Sound Author]

I don't know the math very well but, yes, after consulting Google, it appears that quadratic and exponential functions are different.

[SamDi]

I don't know the math very well but, yes, after consulting Google, it appears that quadratic and exponential functions are different.

It's easily explained:
While x^2 grows quadric as 1,2,4,9,16,25,36,49,64,81,100... (for x = 1-10)
an exponential function e.g. 2^x grows much faster as 2,4,8,16,32,64,128,256,512,1024

In nature you have very often exponential behaviours as e.g. populations of bacterias double in a certain time.
In the audio world you have also many exponential relations, as getting a tone for one octave higher means doubling the frequency for each octave. The same with volume, where increasing the volume for 3 dB, means doubling the power.

In a certain range around 1 the differences seems not to be that drastic (as you may see for e.g. x=5 it's still pretty similar), but for higher values the difference of values from the different grow rates gets much more obvious..

In practice this means, that for envelope slopes quadric or exponential curves have a similar quality and in most cases it should be very difficult to say, which is which.

But on the other side, where you would need an exackt exponential function (imagine an OSC frequency, where you control the frequency linearly in Hertz but you have a keyfollow value also linear) a quadric function would not be sufficient, as for each 12 tones you would need the signal to control OSC frequency with double amount.

That's all. But I think your proposal to go over the mapping modulator is the golden way for that. Just a pity that values cannot get imported from another source, because typing them manually in is some PITA.

[Howard]

Exponentiall-ish in Bazille: Scale the mod signal by itself (very easy to do with a multiplex).

[pierb]

Exponentiall-ish in Bazille: Scale the mod signal by itself (very easy to do with a multiplex).

I do this all the time in Zebra

Eg: modhweel sidechained by the modwheel.