A new competitor to Uhe Diva has arived!!!

[xh3rv]

Technically I don't think 0df *has* to model nonlinearities, but it's very different in theory and computation from moving the non-linearities inside the feedback loop as necessary to model analog hardware.

What I meant was that a system in which the output is fed back into the input is nonlinear, and that's what's happening with the 0df filter, no?

Feedback is nonlinear.

I'm probably tempting fate since a lot of this is Greek to me, but ...

Vadim Zavalishin's The Art of VA Filter Design includes a section on a zero delay feedback filter with no non-linearities at the end of chapter 3. There is some variance and nuance in what 'linear' means between filter design versus some other scientific or mathematical contexts, and I can't quite formulate but sort of get the sensibility that feedback is non-linear. But sticking to convention, I think this is strong evidence. Interestingly Vadim notes a 'smoothing' element should be in place in his linear 0df example, and later in ladder & diode chapters uses non-linearities as an alternative - I think this is where the mojo happens, it's really complex, chaotic behavior here.

(Link to Vadim's text) http://ay-kedi.narod2.ru/VAFilterDesign.pdf

[hakey]

it's really complex, chaotic behavior here.

Exactly.

And a system that is mathematically chaotic is nonlinear, though I think this is a different case of nonlinearity than the distortion that Urs mentioned.

[hakey]

perhaps feedback is non-linear, although a tap delay will decay in linear fashion

True (if it were a simple digital delay with nothing happening in the feedback path), but I suspect that, in the case of a resonant filter, feeding the output back into the input is likely to produce chaotic behaviour.

That the 0df implementation in Diva uses a clever iterative best guess algo, instead of a straight linear calculation, rather suggests this is so.

[AnalogGuy1]

"Linearity" just means two things:
* if you scale the input the output similarly scales.
* if you know that filter input x1 causes filter output y1, and similarly for x2 and y2, then if you feed a linear filter x1+x2 it will produce y1+y2.

Feedback doesn't preclude linearity. Runaway positive feedback will cause saturation, and that is non-linear, but feedback doesn't imply non-linearity.

[Sendy]

Ah I didn't know Monarch is locked up.. oh well. I presume that I can still use the filter and osc blocks for my own synth designs?

Cheers!
bManic

If parts of it are locked I think it's most probable that the filter and/or oscillator are candidates. Whether "locked" means you can't look inside or you can't move/copy/paste as a whole remains to be discovered (on my end at least).

[hakey]

"Linearity" just means two things:
* if you scale the input the output similarly scales.
* if you know that filter input x1 causes filter output y1, and similarly for x2 and y2, then if you feed a linear filter x1+x2 it will produce y1+y2.

Feedback doesn't preclude linearity. Runaway positive feedback will cause saturation, and that is non-linear, but feedback doesn't imply non-linearity.

Yeah, been thinking it over and reading up and I'm slowly getting my head around it.

Thanks for the explanation.

Presumably a 0df filter with no nonlinearities would be (relatively) trivial to implement. That it hasn't been done (afaik) might suggest that it's not worthwhile, that some nonlinearity is necessary to produce the behaviour that is considered desirable in a filter.

[Sendy]

For a utility filter no distortion might be a good thing, but I think the nonlinearities in a synth filter can really help to bind the oscillators together (a bit like you might use a saturator on a drum buss).

Given that 0df filters are mainly about improving the resonance's response to movement in cutoff, I'm guessing the applications for 0df in utility filters are limited.

[EvilDragon]

Ah I didn't know Monarch is locked up.. oh well. I presume that I can still use the filter and osc blocks for my own synth designs?

Cheers!
bManic

If parts of it are locked I think it's most probable that the filter and/or oscillator are candidates. Whether "locked" means you can't look inside or you can't move/copy/paste as a whole remains to be discovered (on my end at least).

Macros "Filter/Amplifier" and "Envelope Generators" are locked.


http://www.native-instruments.com/forum ... tcount=125

[aMUSEd]

Ah I didn't know Monarch is locked up.. oh well. I presume that I can still use the filter and osc blocks for my own synth designs?

Cheers!
bManic

If parts of it are locked I think it's most probable that the filter and/or oscillator are candidates. Whether "locked" means you can't look inside or you can't move/copy/paste as a whole remains to be discovered (on my end at least).

Macros "Filter/Amplifier" and "Envelope Generators" are locked.


http://www.native-instruments.com/forum ... tcount=125

What it doesn't say is whether they can still be copied and used as whole macros in other ensembles. After all that wouldn't be that different from any off the shelf modular synth unit even if you can't get inside it and take it apart.

[chilly7]

Ah I didn't know Monarch is locked up.. oh well. I presume that I can still use the filter and osc blocks for my own synth designs?

Cheers!
bManic

If parts of it are locked I think it's most probable that the filter and/or oscillator are candidates. Whether "locked" means you can't look inside or you can't move/copy/paste as a whole remains to be discovered (on my end at least).

Macros "Filter/Amplifier" and "Envelope Generators" are locked.


http://www.native-instruments.com/forum ... tcount=125

Nnnnnnoooooooooooooooooooooooooooooooooooooooooooooooooooooo

[EvilDragon]

Ah I didn't know Monarch is locked up.. oh well. I presume that I can still use the filter and osc blocks for my own synth designs?

Cheers!
bManic

If parts of it are locked I think it's most probable that the filter and/or oscillator are candidates. Whether "locked" means you can't look inside or you can't move/copy/paste as a whole remains to be discovered (on my end at least).

Macros "Filter/Amplifier" and "Envelope Generators" are locked.


http://www.native-instruments.com/forum ... tcount=125

What it doesn't say is whether they can still be copied and used as whole macros in other ensembles. After all that wouldn't be that different from any off the shelf modular synth unit even if you can't get inside it and take it apart.

I'm pretty sure you can use those elements in your own ensembles. I'm not sure if you're allowed to upload such ensembles to the user library, though. Probably not.

[Ingonator]

deleted

[pdxindy]

Diva = Filter FM
Monark = No filter FM

In Monark you could assign Osc 3 to Oscillator modulation (= FM) and also to Filter modulation (= Filter FM) using the corresponding switches (same way as in the real Mini). You are not limited to using Osc 3 only in "LO" mode (= use as LFO).
You could also switch keytracking for Osc 3 off using the "K.T." switch (= "Osc 3 control" in the real Mini).
With the "Modulation Mix" knob (like in the Mini) you could mix Osc 3 and the Noise source for being used with the modulation.

There is also no need to deactivate Osc 3 and/or turn the volume to 0 when used for modulation, especially when used in the audible range. If used as a LFO usually the volume is set to 0 and/or the OSC deactivated.

The modulation amount is adjusted with the Modwheel. In Monark this setting could be saved with the patch.

Diva = multiple envelope types including release control
Monark = no separate envelope release control

Monark got 2 Release switches (= "REL"), one for each envelope.


About velocity and aftertouch:
With the full Reaktor version you could add the corresponding modules yourself.
Connecting the MW module and the Aftertouch module to a multiplier module i was able to use modulation by aftertouch (e.g. for controlling vibrato) and adjust the maximum modulation amount with the Modwheel.



Ingo

I edited out the reference to filter FM in my original comparison post...

Monark has no separate release control, just the switch tied to decay, which was my original point so I left that one.

The overall point remains... Monark is not a competitor to Diva because the feature set is hugely different. You can compare one small subset of Diva to Monark, but Diva is so much more. And the difference will soon be growing significantly larger with Diva getting a step sequencer, an arpeggiator, a new osc model referencing the JP osc with its supersaw and feedback osc types (plus more), and a new filter model.

[Ingonator]

deleted

[ObsoleteAcc99]

[DELETED]