GHOST (beta) - New 6 osc wavetable FM synth

[fmr]

Assign another envelope (for example ENV 2) to the volume of oscillator A, and give it a very slow attack. You will notice that you will get sound immediately, despite the fact that you should hear a very slow volume crescendo.

I have to try that tonight. I would classify that kind of behaviour as a bug or design flaw.

The design flaw is having ENV 1 hardwired to the overall volume of ALL oscillators. It can't be that way, not in a FM synth. Each oscillator needs to be controlled by it's own envelope (and ONLY by it).

[gentleclockdivider]

Envelopes which determines the FM amount over time are the most important aspects of an FM synth .
Doesn't matter if it's FM or Ph.M.
Hardwiring all operators to one envelope will severely restrict the possibilities in sound creation , there is no justificaton in doing that .

[gentleclockdivider]

Each oscillator needs to be controlled by it's own envelope (and ONLY by it).

ANd perhaps additional LFO's if desired .

[fmr]

Another thing: I copied my User wavetable folder from Serum to Ghost (I have a quite large collection). Some wavetables open, other don't. So, what exactly are the specs for a wavetable to be recognized in Ghost?

EDIT: I think I found the problem. Many of the wavetables were not in Serum format. Since Serum is capable os reading them no matter the format, I didn't bother to convert them. But after I resaved them with Serum they became readable in Ghost.

[SeBaer]

You can't. Just try it.

I did, and I think I got it. You have to use the FM-Matrix. The setup is pretty unique, a little weird, but once you wrap your head around, quite powerful.

FM-Matrix.PNG

In this setup I have VCA of OSC2 output modulated by ENV2, OSC2-->OSC1 FM Modulated by ENV3 and feedback of OSC modulated by LFO1.

If you give it some thought, you must realize it's a stroke of genius to think of such an elaborate modulation concept. Also the fact, that you can flip the order of filter and distortion is really cool.
But I really would like to be able to resize the interface. It's all a little small and fiddly.

[SeBaer]

A graphical bug:
When you instanciate a reverb in the effects section, there is a graphical glitch in the EQ. Once you touch the high frequency, the graphic snaps back to normal, i.e. you see the entire EQ-curve.

Reverb.PNG

[gentleclockdivider]

You can't. Just try it.

I did, and I think I got it. You have to use the FM-Matrix. The setup is pretty unique, a little weird, but once you wrap your head around, quite powerful.
FM-Matrix.PNG
In this setup I have VCA of OSC2 output modulated by ENV2, OSC2-->OSC1 FM Modulated by ENV3 and feedback of OSC modulated by LFO1.

If you give it some thought, you must realize it's a stroke of genius to think of such an elaborate modulation concept. Also the fact, that you can flip the order of filter and distortion is really cool.
But I really would like to be able to resize the interface. It's all a little small and fiddly.

Yes , but env 1 is still the master envelope controlling all operators

[SeBaer]

Yes , but env 1 is still the master envelope controlling all operators

Of course it is, or else you would have a free running oscillator, which would put out sound, whether a key is pressed or not. And even you couldn't want that, could you.

[fmr]

You can't. Just try it.

I did, and I think I got it. You have to use the FM-Matrix. The setup is pretty unique, a little weird, but once you wrap your head around, quite powerful.
FM-Matrix.PNG
In this setup I have VCA of OSC2 output modulated by ENV2, OSC2-->OSC1 FM Modulated by ENV3 and feedback of OSC modulated by LFO1.

If you give it some thought, you must realize it's a stroke of genius to think of such an elaborate modulation concept. Also the fact, that you can flip the order of filter and distortion is really cool.
But I really would like to be able to resize the interface. It's all a little small and fiddly.

That's not what I was talking about. I'm aware of the FM Matrix. Actually, it's the section I've been pending more time on. I have no problems in assigning envelopes, LFOs or whatever. What I have problems with is controlling the volume output with another envelope INDEPENDENTLY OF ENV 1. If you ever used an FM synth you'll know what I mean easily

Just do what I described earlier. That's nothing of what you were describing. What you came up with doesn't deal with different oscillator outputs. In your setup, Osc B is set to 0, so it shouldn't even be sounding, no matter what you've assigned to control the output (which is NOT a VCA, BTW). You don't need to modulate anything, just route two or three osc illators to the output, set them differently enough so you can listen which one is sounding, and try to have one with a fast attack and decay, and another with a slow attack, that rises during the decay of the other and ends sounding ALONE after the other faded away.

Forget Vital, forget Serum. This is a completely different perspective.

[fmr]

Yes , but env 1 is still the master envelope controlling all operators

Of course it is, or else you would have a free running oscillator, which would put out sound, whether a key is pressed or not. And even you couldn't want that, could you.

WTF are you talking about? Did you EVER use an FM synth?

[SeBaer]

What I have problems with is controlling the volume output with another envelope INDEPENDENTLY OF ENV 1.

How so? I exactly showed you. If you want it parallel to another oscillator, route it to output, set its volume to zero and modulate it with an envelope. In its initial state ENV1 just acts as a gate. It controls just the output volume. For controlling everything else, you have to set up further envelopes.

I'm aware of the FM Matrix. Actually, it's the section I've been studying more.

That is good. Keep studying and eventually you will get it, it's a little tricky. This is not a pure FM-synth, so keep an open mind.

Just do what I described earlier.

You can't. Just try it. Assign another envelope (for example ENV 2) to the volume of oscillator A, and give it a very slow attack. You will notice that you will get sound immediately, despite the fact that you should hear a very slow volume crescendo. Now do the opposite: configure ENV 1 for a very slow attack, and give a very fast attack to ENV 2. This time you will get a very slow crescende. despite the fact that ENV 2 has a vey fast attack. Don't know about you, but this isn't the way I want the envelopes to behave. This will prevent to take the most out of the FM engine.

In Vital, each oscillator has the envelope you choose to assign to it, and it is summed to the effect of envelope 1. That way, although you start to get sound immediately (and that's smething I don0t like, to be honest) you can still hgear the other envelope effect in the sound. That's not the case in Ghost. And Vital is a plain wavetable synth, as is Serum, with three oscillators. Ghost is quite different, and was thought with FM in mind, I think.

I just tried it.

That's nothing of what you were describing. That doesn't deal with different oscillator outputs. In your setup, Osc B is set to 0, so it shouldn't even be sounding, no matter what you've assigned to control de output (which is NOT a VCA, BTW).

I'm trying to understand what you want to say. But I guess you did not fully understand, what I tried to say.
The oscillators don't have an output volume knob, so you need to assign the envelope in the FM-Matrix. If you route an oscillator directly to output you can assign a value. That's the amount of output. If you assign an envelope to this, the previously assigned value is the base line to which the envelope adds.
Try it out: set output to zero, assign an evelope with positve power.
And now set the output to 100, and assign the envelope with negative power.
You will get exactly the result of what you discribed as desired behaviour.

And brillantly, you can modulate the output of an oscillator indepently from the FM-amount it has on other oscillators.

Forget Vital, forget Serum. This is a completely different perspective.

It's not completely different. The usage of ENV1 is the same. It's hardwired to the output amplification, not of the indicidual oscillators, but the sum of all signals, pre-FX. And also this is not a pure FM-synth, it also capable of subtractive synthesis.

[SeBaer]

WTF are you talking about? Did you EVER use an FM synth?

I won't even dignify this with an answer. Don't make me go 386.

[gentleclockdivider]

WTF are you talking about? Did you EVER use an FM synth?

I won't even dignify this with an answer. Don't make me go 386.

Take the following scenario
OP3 goes into OP2 goes into OP1 , al serially connected , and OP1 as carrier
OP3 Needs to have and instant attack , and very short decay
Op2 also needs a instant attack and verr short decay
Op1 ( the carrier ) needs a medium slow attack
NOw , this is doable in GHost when ENV 1 (with a slow attack) is NOT used for OP1 because otherwise it would also affect OP3 and 2 and we don't want that .
The only workaround is to assign another envelope t op OP1 that is not the master envelope , right , problem solved ?
Noppes , now lets double the operators to 6 so we need 6 independent envelopes , see were I am going to ?
You'll gonna have to use env 1 one way or the othe since there ae only 6 env's available .
I am also not sure if the current implementation of modulation is the way to go , if you modulate the fm amount by drag dropping on the matrix , fi we want to modulate from full to zero we need to set the matrix amount to zero and adjust the mod amount in the mod window .
This way we need to constantly switch to the mod page to see the mod amount .

[SeBaer]

Take the following scenario
OP3 goes into OP2 goes into OP1 , al serially connected , and OP1 as carrier
OP3 Needs to have and instant attack , and very short decay
Op2 also needs a instant attack and verr short decay
Op1 ( the carrier ) needs a medium slow attack
NOw , this is doable in GHost when ENV 1 (with a slow attack) is NOT used for OP1 because otherwise it would also affect OP3 and 2 and we don't want that .
The only workaround is to assign another envelope t op OP1 that is not the master envelope , right , problem solved ?
Noppes , now lets double the operators to 6 so we need 6 independent envelopes , see were I am going to ?

I can guess what you're getting at, but the patch you're proposing does not make much sense. When the carrier (OP1) has a medium slow attack, but both modulators (OP2 & OP3) have a very short decay, the effect of the modulators would have worn off, before the carrier is audible.
Also, in my understanding ENV1 does not affect any OSC individually, but rather the whole audio signal after the FM-section, before the FX-section.
That aside, the more envelopes the better. I agree on that. Also more LFOs (which you can use as MSEGs, by the way...)

[SeBaer]

I am also not sure if the current implementation of modulation is the way to go , if you modulate the fm amount by drag dropping on the matrix , fi we want to modulate from full to zero we need to set the matrix amount to zero and adjust the mod amount in the mod window .
This way we need to constantly switch to the mod page to see the mod amount .

You do realize, that you can adjust the modulation amount in the matrix by clicking and dragging the coloured dot, don't you? Positively an negatively. Sure in the modulation tab you can adjust more precisely, but basic adjustments you can do in the matrix. And the mod amount is represented by a small bar below the modulated parameter in the matrix. And here we come back to resizing the GUI. It's reeeeaaaaally small, and very hard to see.