Fabfilter Twin 3 --Released Feb. 7th

[Biome_Digital]

Next (related to previous point), I take it the only way to get a note-on unique random modulation value is to use an XLFO of one step where that single step is a random value, triggered by note-on, frequency 0. A lot of work for something so simple, but that's what presets are for I guess. Am I missing something more basic to accomplish the same thing?

The solution just described works as desired and expected. In the Clean preset, I set the per-note-random-value XLFO to modulate Pan in the single oscillator, and I get a nice distribution of L-to-R placement. I set it to modulate a second arbitrary target (Delay amount) and they both visually are in sync. The problem is that the audio panning and the visual feedback on the pan control have nothing in common. I read a post here about suspect visual feedback on modulation. Is this an instance of that problem? [Edit] Oh yeah, and I also tried modulating osc pitch and saw/heard the same thing. Nice random sequence of pitches but no relation to the visual modulation feedback.

yes, same here.

[SamDi]

One strange thing though, the "Main Output" is BEFORE the compressor. So there is no gain after compression and moving the main output affects the compression amount!!! What?!

or have I missed something here??

No, you didn't miss anything. That's how it is.

.. and yes, it's weird and silly in my opinion and it's something I did give feedback about during beta. Unfortunately it's not been changed.

It seems, you gave partly some really valid feedback, which was completely ignored, which makes them look like, ehm stubborn...

[dmbaer]

The XLFO can be bi-polar or uni-polar and anything in between. You just need to set the steps as you want them. For instance, if you got a simple sine going, just move step 1 to a value of zero. It defaults to -1 for the first step and +1 for the 2nd step, aka bi-polar. You also have full control over the phase start (and it can be modulated).

Yes, until you want to introduce randomness. There's no way I can see to ask for a non-negative random value for a step.

[bmanic]

The XLFO can be bi-polar or uni-polar and anything in between. You just need to set the steps as you want them. For instance, if you got a simple sine going, just move step 1 to a value of zero. It defaults to -1 for the first step and +1 for the 2nd step, aka bi-polar. You also have full control over the phase start (and it can be modulated).

Yes, until you want to introduce randomness. There's no way I can see to ask for a non-negative random value for a step.

Luckily there is a "hack" to do just that. Simply assign a random LFO to a target. Then assign that same random LFO to the previous random LFO modulation slot, move that first slot's value to zero and use the 2nd LFO to do the actual modulation. Boom. "Instant" unipolar modulation.

[pierb]

Luckily there is a "hack" to do just that. Simply assign a random LFO to a target. Then assign that same random LFO to the previous random LFO modulation slot, move that first slot's value to zero and use the 2nd LFO to do the actual modulation. Boom. Instant unipolar modulation.

Jesus... that's really bad UX. Couldn't they just add a button to make an LFO unipolar?

Personally I think unipolar LFOs should be the default but of course one should be able to chose without resorting to hacks.

[bmanic]

Luckily there is a "hack" to do just that. Simply assign a random LFO to a target. Then assign that same random LFO to the previous random LFO modulation slot, move that first slot's value to zero and use the 2nd LFO to do the actual modulation. Boom. Instant unipolar modulation.

Jesus... that's really bad UX. Couldn't they just add a button to make an LFO unipolar?

Personally I think unipolar LFOs should be the default but of course one should be able to chose without resorting to hacks.

I disagree. Bi-polar LFO is the norm as it's what you most commonly use for such basic things as vibrato (hence most common as the "default" on most synths).

Though I do agree about the button! Anything that simplifies things would be welcome!

I also think the LFO should have a built in delay/attack ramp for ease of use. Heck, I was hoping they would finally re-write the whole XLFO and make it spline based like pretty much everybody else in the business does.

[pierb]

I also think the LFO should have a built in delay/attack ramp for ease of use. Heck, I was hoping they would finally re-write the whole XLFO and make it spline based like pretty much everybody else in the business does.

Yeah absolutely!

Honestly after using PhasePlant all LFOs seem quite outdated in comparison.

[dmbaer]

Couldn't they just add a button to make an LFO unipolar?

I'd settle for a Random 0 to 1 option. I confess, I'm not seeing how bmanic's suggested work-around delivers.

[bmanic]

Couldn't they just add a button to make an LFO unipolar?

I'd settle for a Random 0 to 1 option. I confess, I'm not seeing how bmanic's suggested work-around delivers.

It's very simple: The first LFO target is bi-polar but by setting it to zero, the next LFO that is modulating the original LFO target will only modulate the positive (in this example), thus converting it to uni-polar. Want to make it go in the other direction (negative)? Simply set the first LFO target to be negative and keep the first one positive. It all makes kind of sense if you think about it.. but yeah, I'd still call it a bit of a "hack".

Easiest is to just try it and see for yourself. It works.

[vitocorleone123]

Just took it for my first test drive, and I'm in the sounds-fabulous camp. Nice job, bmanic - well done indeed!

However, some questions. I take it the XLFOs are only bi-polar. The EGs can be uni- or bi-polar, a feature of very limited value, IMO. But the XLFOs, especially in their guise as MSEGs, would really benefit from a uni-polar option. Am I overlooking something?

Next (related to previous point), I take it the only way to get a note-on unique random modulation value is to use an XLFO of one step where that single step is a random value, triggered by note-on, frequency 0. A lot of work for something so simple, but that's what presets are for I guess. Am I missing something more basic to accomplish the same thing?

The solution just described works as desired and expected. In the Clean preset, I set the per-note-random-value XLFO to modulate Pan in the single oscillator, and I get a nice distribution of L-to-R placement. I set it to modulate a second arbitrary target (Delay amount) and they both visually are in sync. The problem is that the audio panning and the visual feedback on the pan control have nothing in common. I read a post here about suspect visual feedback on modulation. Is this an instance of that problem? [Edit] Oh yeah, and I also tried modulating osc pitch and saw/heard the same thing. Nice random sequence of pitches but no relation to the visual modulation feedback.

One last thing. I was delighted to read that CC64 was now hard-wired to sustain. Except that it doesn't seem to be. It did not work in any patch I tried. Is there some global setting that needs to happen to get this to work? Anyone else experiencing the same thing?

Thanks for the tip on the voice panning.

While I wouldn't mind some of this stuff saved as presets (which I'm doing), I like how all the building blocks are there to create something unique.

[Biome_Digital]

The XLFO can be bi-polar or uni-polar and anything in between. You just need to set the steps as you want them. For instance, if you got a simple sine going, just move step 1 to a value of zero. It defaults to -1 for the first step and +1 for the 2nd step, aka bi-polar. You also have full control over the phase start (and it can be modulated).

Yes, until you want to introduce randomness. There's no way I can see to ask for a non-negative random value for a step.

Luckily there is a "hack" to do just that. Simply assign a random LFO to a target. Then assign that same random LFO to the previous random LFO modulation slot, move that first slot's value to zero and use the 2nd LFO to do the actual modulation. Boom. Instant unipolar modulation.

Yep, super instant!

[bmanic]

The XLFO can be bi-polar or uni-polar and anything in between. You just need to set the steps as you want them. For instance, if you got a simple sine going, just move step 1 to a value of zero. It defaults to -1 for the first step and +1 for the 2nd step, aka bi-polar. You also have full control over the phase start (and it can be modulated).

Yes, until you want to introduce randomness. There's no way I can see to ask for a non-negative random value for a step.

Luckily there is a "hack" to do just that. Simply assign a random LFO to a target. Then assign that same random LFO to the previous random LFO modulation slot, move that first slot's value to zero and use the 2nd LFO to do the actual modulation. Boom. Instant unipolar modulation.

Yep, super instant!

Touché!

Yeah, one mouse click versus.. lets see, click + drag, click + drag + one extra drag for removing LFO 1 default amount of modulation. Yeah, not optimal!! But at least it can be done.

.. fixed my original post and the quote!

[Biome_Digital]

The XLFO can be bi-polar or uni-polar and anything in between. You just need to set the steps as you want them. For instance, if you got a simple sine going, just move step 1 to a value of zero. It defaults to -1 for the first step and +1 for the 2nd step, aka bi-polar. You also have full control over the phase start (and it can be modulated).

Yes, until you want to introduce randomness. There's no way I can see to ask for a non-negative random value for a step.

Luckily there is a "hack" to do just that. Simply assign a random LFO to a target. Then assign that same random LFO to the previous random LFO modulation slot, move that first slot's value to zero and use the 2nd LFO to do the actual modulation. Boom. Instant unipolar modulation.

Yep, super instant!

Touché!

Yeah, one mouse click versus.. lets see, click + drag, click + drag + one extra drag for removing LFO 1 default amount of modulation. Yeah, not optimal!! But at least it can be done.

.. fixed my original post and the quote!

I was only joking of course. It is a cool hack.

[Harry_HH]

Stupid moaning.

[dmbaer]

One last thing. I was delighted to read that CC64 was now hard-wired to sustain. Except that it doesn't seem to be. It did not work in any patch I tried. Is there some global setting that needs to happen to get this to work? Anyone else experiencing the same thing?

This problem was encountered when using Blue Cat Audio Patchwork as a the VST (3) host. In Cubase, Twin 3 sustain works as expected. Patchwork successfully hosts any number of other VST 3 instruments for which sustain works just fine; only Twin 3 seems to have this problem.