Zebra 3 Public Beta (final beta)

[aMUSEd]

Already curious how many 3rd party presets in the future will accidentally have a burner module in it..

haha, yes.

+1, renaming to "CPU Burner" would be appropriate, imho.

Already a thing behind the scenes

Screenshot 2025-12-20 at 17.14.08.png

[jrwaltb]

Wow. So I kind of understand like 5% of this finally and its really good.

[buffalo roam]

4-voice polyphony is absolutely possible, by modulating each of the 4 pitch centers. And then assigning each to an oscillator of its own or something. That's different from playing multiple notes (each of which might use all 4 oscillators) at once of course, I get that. Just saying.

Personally I'm more of a "the DAW/keyboard is how I do that" kind of person anyway. But the 4 pitch thingies are really clever I find. I love how they let you do mono-, duo, trio- or quadra-phony. Really fun stuff.

Pretty much any modulator can be set to a scale and that can be assigned to anything that takes pitch. So the four main oscillators, the FM Oscillators and modal filters. The combs and the filters also can be pitched.

The four pitch modules can be set to quantize those scales to certain pitches keys. So up to 10 note polyphony is possible with the osillators. 4 X Osc, 4 X FM and 2 X modal. The mappers are well suited to pitch sequencing.

I respect your technical ability to achieve polyphony in the way you describe, and if you actually compose music that way with Zebra 3, more power to you. But for creating a preset with conventional harmony and counterpoint in it, Zebra 3’s structure just isn’t composer-friendly the way Serum 2 and Nexus 5 are. You could argue that I should just use the piano roll in my daw for that, and well, OK. But the thing is, at least some of Zebra 3’s appeal is that it presents presets that have movement more interesting than just four-on-the-floor pulses, and note-specific rather than just providing evolving drones. It’s kind of a tease.

[SamDi]

We have some open tickets for this. It's one of the areas we're hoping for feedback for.

But we'll surely add visual feedback to ABCD, and we very likely also make them automatable.

Yes, please. Having only 4 macro-like parameters exclusively controlled via MIDI and hence not accessible by DAW's modulations or DAW's macro controllers is odd.

Any macro-system planned as in HIVE2?

[Andreya_Autumn]

4-voice polyphony is absolutely possible, by modulating each of the 4 pitch centers. And then assigning each to an oscillator of its own or something. That's different from playing multiple notes (each of which might use all 4 oscillators) at once of course, I get that. Just saying.

Personally I'm more of a "the DAW/keyboard is how I do that" kind of person anyway. But the 4 pitch thingies are really clever I find. I love how they let you do mono-, duo, trio- or quadra-phony. Really fun stuff.

Pretty much any modulator can be set to a scale and that can be assigned to anything that takes pitch. So the four main oscillators, the FM Oscillators and modal filters. The combs and the filters also can be pitched.

The four pitch modules can be set to quantize those scales to certain pitches keys. So up to 10 note polyphony is possible with the osillators. 4 X Osc, 4 X FM and 2 X modal. The mappers are well suited to pitch sequencing.

I respect your technical ability to achieve polyphony in the way you describe, and if you actually compose music that way with Zebra 3, more power to you. But for creating a preset with conventional harmony and counterpoint in it, Zebra 3’s structure just isn’t composer-friendly the way Serum 2 and Nexus 5 are. You could argue that I should just use the piano roll in my daw for that, and well, OK. But the thing is, at least some of Zebra 3’s appeal is that it presents presets that have movement more interesting than just four-on-the-floor pulses, and note-specific rather than just providing evolving drones. It’s kind of a tease.

Oh yeah make no mistake, I am not likely to use the internal polyphony that way very often. I set up some chord patches and sequence patches to try it here, and it really does go surprisingly deep.

But yeah if you want my totally honest opinion, for my own music making I'm kinda conservative
about this stuff. When working on actual music that I care about I use DAW/MIDI features 100% of the time. That's a more powerful sequencer/arpeggiator than any plugin dev has ever pieced together, by a long shot. Seems like people like the fact that it's more simple and built-in, but for doing "actual work/music" I dont' really care tbh.

Setting up a synth patch that "plays itself" is something I only really do for fun and without a specific goal in mind. And doing so in Zebra3 already is really fun IMO!

[Urs]

I respect your technical ability to achieve polyphony in the way you describe, and if you actually compose music that way with Zebra 3, more power to you. But for creating a preset with conventional harmony and counterpoint in it, Zebra 3’s structure just isn’t composer-friendly the way Serum 2 and Nexus 5 are. You could argue that I should just use the piano roll in my daw for that, and well, OK. But the thing is, at least some of Zebra 3’s appeal is that it presents presets that have movement more interesting than just four-on-the-floor pulses, and note-specific rather than just providing evolving drones. It’s kind of a tease.

So, obviously, Pitch is an important aspect in Zebra 3. You can already create polyphonic sequences that stay in any chromatic or diatonic scale. You can use the Pitches to add harmonic elements that stay on 1, 5 and 7 while others roam free on chromatic scale and yet others can be confined to pentatonic, or whatever you want. This goes far beyond anything any of the frequently mentioned plug-ins do, it'll even be difficult or tedious to set up in modular environments. Yet, in Zebra 3 it is done easily, in a few seconds.

Apparently, this power would be wasted if we just slapped a simple sequencer or arpeggiator in front of it. It might need something that can interface with the Pitch modules. Not sure yet how.

So we see this as a process. With Zebra 2 we went from 2.0 to 2.9 over the years, and I see no reason to doubt that Zebra 3 will be any different.

[kraster]

4-voice polyphony is absolutely possible, by modulating each of the 4 pitch centers. And then assigning each to an oscillator of its own or something. That's different from playing multiple notes (each of which might use all 4 oscillators) at once of course, I get that. Just saying.

Personally I'm more of a "the DAW/keyboard is how I do that" kind of person anyway. But the 4 pitch thingies are really clever I find. I love how they let you do mono-, duo, trio- or quadra-phony. Really fun stuff.

Pretty much any modulator can be set to a scale and that can be assigned to anything that takes pitch. So the four main oscillators, the FM Oscillators and modal filters. The combs and the filters also can be pitched.

The four pitch modules can be set to quantize those scales to certain pitches keys. So up to 10 note polyphony is possible with the osillators. 4 X Osc, 4 X FM and 2 X modal. The mappers are well suited to pitch sequencing.

I respect your technical ability to achieve polyphony in the way you describe, and if you actually compose music that way with Zebra 3, more power to you. But for creating a preset with conventional harmony and counterpoint in it, Zebra 3’s structure just isn’t composer-friendly the way Serum 2 and Nexus 5 are. You could argue that I should just use the piano roll in my daw for that, and well, OK. But the thing is, at least some of Zebra 3’s appeal is that it presents presets that have movement more interesting than just four-on-the-floor pulses, and note-specific rather than just providing evolving drones. It’s kind of a tease.

I don't think it's some exotic esoteric thing and more of a regular function of any modulator.

It's built into every modulator at the bottom under "q". Next to it is where you set the polarity of the modulation and other functions.

Then the pitch modules can be applied scale quantization before OR after modulation.
As I said above the Mappers are particularly suited to pitch sequencing.

One of the big advantages for me in Zebra 3 is that pitch and trigger information can be completely decoupled if you want them to be.

It might be less immediate than a regular arp but pretty trivial to set up once it clicks and far more flexible.

[chagzuki]

I've only made a few simple patches in this so far, but I've found myself not enjoying the unison detune calibration. I'm generally aiming at values somewhere around 8 or so, on a dial that goes up to 100. So I'm shift-dragging the mouse to try to get precise values, which of course works but feels awkward.

[jtsterays]

I've only made a few simple patches in this so far, but I've found myself not enjoying the unison detune calibration. I'm generally aiming at values somewhere around 8 or so, on a dial that goes up to 100. So I'm shift-dragging the mouse to try to get precise values, which of course works but feels awkward.

Yes, scaling is quite bad on that, the most useful values are below ~10. Also I dont like the detuning law that much, supersaws feels kinda static, I hope they open this up like Serum or Dune.

[chagzuki]

I've had a good experience so far with the filters, some strike me as having some sort of 'high definition' quality, against which other synths sound a bit veiled. I'll have to test that some more. But some of the bread and butter stuff such as the basic detuning and envelopes don't cut it in 2025... but it appears the MSEGs are designed to act as fully-featured envelopes, clearly I need to dig deeper.

[Urs]

The envelopes do a lot more shifting than first meets the eye... The ability top be triggered not just by gate alone sets them apart from many other approaches... also, they actually Key Track (half time per octave up) when Key Scale is set to -100. Which other synth has that? - And it goes well together with Comb filter and Modal Resonator excitation.

As for unison... the detune law is almost exactly like Serum's. Their knob reaches "25" half way, whereas our's is scaled linearly, 25 after a quarter of the knob. But at 25, both sound pretty much identical. We can maybe improve that by scaling 25 towards the middle of the knob movement as well, but we have to check if that breaks any presets (if anyone modulates detune in specific ways).

Serum and others have additional parameters to blend the middle oscs(s) in and out. I usually don't think much of that as we have a lot more oscillators than they do, so why not save the knobs.

At some point, we might offer options for stereo, detune und volume distribution. I have explored so much of it when we did Hive, and never found this to be a huge source of variety, hence it kind of became less important during the development.

[jtsterays]

At some point, we might offer options for stereo, detune und volume distribution. I have explored so much of it when we did Hive, and never found this to be a huge source of variety, hence it kind of became less important during the development.

Idk, in Dune 3 there's a Swarm detune option where each voice have a seperate LFO (diff speeds & phase too i think) modulating the tuning which has a very unique sound, it's the only reason I still use it. Also in Ableton's wavetable and old Massive you can do wavetable positions unison too. And ofc more basic one like non linear detune - JP8000 supersaws. Personally for me they're quite important because not a lot of synths have them.

[Urs]

Don't get me wrong, I'm not against any of it. I just have to make some compromises on priorities if we want to finish this thing in some usable shape. There'll be a lot of headroom for future improvements.

So part of this beta is not just to get some feedback, but also do some expectation management, where we can say "yo, we'd love to do this, but have to do it later or else we'll never have anything to release"

[jtsterays]

Don't get me wrong, I'm not against any of it. I just have to make some compromises on priorities if we want to finish this thing in some usable shape. There'll be a lot of headroom for future improvements.

So part of this beta is not just to get some feedback, but also do some expectation management, where we can say "yo, we'd love to do this, but have to do it later or else we'll never have anything to release"

When you guys eventually do this, pleaseeee just straight up steal the Swarm mode, I dont care about ethics, that mode is just pure fire, AFAIK only Dune has it, it's one of its selling points.

[Urs]

Hive uses independent LFOs for the osc unison, too...