U-he Zebra 3 (Alpha)

[ATS]

From the screenshot, it looks like....Zebra2, but with an updated UI. Not much to get excited about yet IMO, although I'm sure there's much more to come.

yeah 10 years of work on a UI lol

[Urs]

The expectations are so high, I'm sure that many people will be disappointed or overwhelmed at first.

Some people will find it too complicated. The waveform and MSEG editor is of course the core of everything. Yes, it is very complex, but even if one uses it just in its basic functionality, it offers everything that Zebra 2 has, and more, in a consistent and more usable fashion. With some experience one can do things easily with the editors that seem impossible to achieve elsewhere.

Some people will be sad that we do not initially include an arpeggiator or sequencer. I hope they'll eventually see why.

Some people will be sad that Zebra 3 is not a classic wavetable synthesiser. It does something similar, but the basis are not Hive/Serum-style wavetable files.

Likewise, some people will be sad that Zebra 3 does its own form of Additive Synthesis, but one can't load samples and do resynthesis with it. The same goes for Modal Synthesis. You can't draw your own partials, but at least the selection of Modal Profiles we have is extensible.

Some people will find that it uses too much CPU. Every module is rewritten from scratch and offers better sound quality than their counterpart in Zebra 2. I reckon, as a rule of thumb, every module in Zebra 3 uses twice the CPU of its equivalent in Zebra 2, simply because most of them are now oversampled 2x. The Diva style filters use less CPU than their equivalents in ZebraHZ though. Except maybe the Wasp-y one.

But anyhow, if people expect that a Zebra 3 patch with 10 modules is as efficient as the same patch in Zebra 2, that is not going to happen. It's more like a Zebra 2 patch with 20 modules then, and the reason should be fairly apparent when comparing the modules.

So while I'm as excited as anyone here, I'm also bracing for impact, and much of the next few weeks and days will be in preparation of the criticism we anticipate.

[Caine123]

i have to admit, I still learn Zebra 2, it is very complex for me, and I dunno what Zebra 3 really makes me wow so far, I cannot judge it, but yeah I'm sure I will find out once it is out .

Urs, I still wonder, would it be possible to add the random function from Filterscape? i love this function!

[Lbdunequest]

The expectations are so high, I'm sure that many people will be disappointed or overwhelmed at first.

Some people will find it too complicated. The waveform and MSEG editor is of course the core of everything. Yes, it is very complex, but even if one uses it just in its basic functionality, it offers everything that Zebra 2 has, and more, in a consistent and more usable fashion. With some experience one can do things easily with the editors that seem impossible to achieve elsewhere.

Some people will be sad that we do not initially include an arpeggiator or sequencer. I hope they'll eventually see why.

Some people will be sad that Zebra 3 is not a classic wavetable synthesiser. It does something similar, but the basis are not Hive/Serum-style wavetable files.

Likewise, some people will be sad that Zebra 3 does its own form of Additive Synthesis, but one can't load samples and do resynthesis with it. The same goes for Modal Synthesis. You can't draw your own partials, but at least the selection of Modal Profiles we have is extensible.

Some people will find that it uses too much CPU. Every module is rewritten from scratch and offers better sound quality than their counterpart in Zebra 2. I reckon, as a rule of thumb, every module in Zebra 3 uses twice the CPU of its equivalent in Zebra 2, simply because most of them are now oversampled 2x. The Diva style filters use less CPU than their equivalents in ZebraHZ though. Except maybe the Wasp-y one.

But anyhow, if people expect that a Zebra 3 patch with 10 modules is as efficient as the same patch in Zebra 2, that is not going to happen. It's more like a Zebra 2 patch with 20 modules then, and the reason should be fairly apparent when comparing the modules.

So while I'm as excited as anyone here, I'm also bracing for impact, and much of the next few weeks and days will be in preparation of the criticism we anticipate.

I remember sometime ago you mentioned about possibility not directly playback samples but import them in a form of wavetable. Is it still on the table or was it only thought out loud?

[Yahnee]

So basically what Urs is saying is that this is similar to an Omni 2 → Omni 3 type of update. The workflow stays pretty much the same, but the sound quality gets a major upgrade — and as a result, it’ll chew through CPUs for breakfast.

[Funky40]

- My CPU is ready to get f*cked -


- - - - - - - - - - - - - - - - - -

[Urs]

I remember sometime ago you mentioned about possibility not directly playback samples but import them in a form of wavetable. Is it still on the table or was it only thought out loud?

The original concept was to additionally have wavetables for the oscillators. But that felt like a bit too much, after it was clear that we'd use the curves as modifiers in the OscFX. It would have to be done so that wavetables *and* curves have to coexist at the same time. This would have turned the oscillator into an exponentially more complicated unit... feature creep, really.

To be honest, having the curves in the OscFX is probably a much better thing and more valuable than just another synth with wavetable import.

One can still import single cycle waveforms from .wav into the editor and vectorise the waveform into a curve. That works well. One can also import multiple cycles, but this can take very long and spin forever it seems, if the file is too large. That's one thing I'm gonna work on in January, see if I can fix it. However, even if multiple waveforms can be imported at once, one has to manually work on the transitions, and it's - like I've said before - a bit hit and miss.

[Urs]

So basically what Urs is saying is that this is similar to an Omni 2 → Omni 3 type of update. The workflow stays pretty much the same, but the sound quality gets a major upgrade — and as a result, it’ll chew through CPUs for breakfast.

It depends. If you look at Zebra 2, hundreds of great presets have 1-5 modules in the main grid. Those will be absolutely playable in Zebra 3, even most presets with 6-10 modules may play easy.

But when people use 15+ modules to create complex sound scapes... those better not be pad sounds with long releases and high polyphony.

Also of course, people wanted Diva-quality filters in Zebra 3, and we gave them some incredibly good sounding filters. If someone uses 4 of them in a single voice, yeah, that's gonna be heavy. Make it a monophonic lead sound and all is well!

But there are also very good sounding filters that use a lot less CPU. Use those for your big poly pads.

[aMUSEd]

So basically what Urs is saying is that this is similar to an Omni 2 → Omni 3 type of update. The workflow stays pretty much the same, but the sound quality gets a major upgrade — and as a result, it’ll chew through CPUs for breakfast.

I think it's a much more substantial update than Omnisphere's was. For a start it has the new spline based OSCs and OSC FX with additive and spectral sound generation and processing (previewed in Zebralette 3) - this is a deep and powerful OSC that can do so much before you even get to filters and fx. Plus there are also new OSC types for physical modelling, and overhauled FM OSCs.

[wintoid]

So while I'm as excited as anyone here, I'm also bracing for impact, and much of the next few weeks and days will be in preparation of the criticism we anticipate.

The halls of KVR are littered with the bruised egos of other developers who have wrestled with this sort of thing. You guys seem so straightforward, and just nice people as well. I'm glad to read what you wrote here, and happy for your mental health that you can view it in such a pragmatic way

I will be cheering pretty much whatever happens.

[Yahnee]

Personally, I have 110% trust in what Urs and his team are doing with their products, and I have no doubt Zebra 3 will keep people busy for years to come. I still feel like I’ve only scratched the surface of Zebra 2. And these new powersynths that have come out recently are built not only for today, but also with future, more powerful CPUs in mind. And of course, I’ll be getting the new striped horse as soon as my wallet gives me the green light. But first, I’ll need to dig a bit deeper into my pockets next week during the sale to grab a couple more U-he synths for my collection.
Has there been any discussion about the price? I was stupid enough to sell my Zebra 2 license that was eligible for the €30 upgrade during a moment of craziness, but I was still happy to pay full price for the legacy version a few months ago. Well, I’m going to buy Diva for the third time next week, so if you want to talk about GAS, hit me up.

[GruvSyco]

i have to admit, I still learn Zebra 2, it is very complex for me, and I dunno what Zebra 3 really makes me wow so far, I cannot judge it, but yeah I'm sure I will find out once it is out .

Urs, I still wonder, would it be possible to add the random function from Filterscape? i love this function!

The thing with Zebra 2, there is so much there you could probably use it for years and years and still never touch on everything it can do. Zebra 3 will be the same.

[Caine123]

i have to admit, I still learn Zebra 2, it is very complex for me, and I dunno what Zebra 3 really makes me wow so far, I cannot judge it, but yeah I'm sure I will find out once it is out .

Urs, I still wonder, would it be possible to add the random function from Filterscape? i love this function!

The thing with Zebra 2, there is so much there you could probably use it for years and years and still never touch on everything it can do. Zebra 3 will be the same.

yeah, also i hope the ignoring from Urs means there are plans

[Urs]

Busy days... sorry...

Randomisers are tricky here. If it can get loud in Filterscape, imagine how loud it can get in Zebra... one of the things that keep me from adding randomisers is the possible liability when people get their gear or - worse - their ears blown up. So a randomiser would require a built in limiter and all that (which we did for Filterscape), and probably a lot of extra logic to make sure that whatever you hear, is compatible with human ears.

[Teksonik]

I would like to address a couple of posts from the subforum:

You can arp the synth in piano roll.

Not even remotely the same thing. The short answer is Arpeggiators can be played, a Piano Roll can not.

I've spent countless joyful hours playing Arpeggiators going all the way back to a Juno 60 in the 80's and they are an important part of a synth's architecture to me. So naturally I would eventually want an Arp in Zebra 3 or a viable alternative that can also be played in the same way.

If one does not own a midi keyboard then an Arp will not be of much use but for some of us who do it is a very useful and inspiring musical tool.

With all those arp/midi sequencing plugins available these days i don't see any need for a dedicated Arp section in a synth, tbh.

With all those Filter plugins available these days I don't see any need for a dedicated Filter module in a synth,tbh....Of course I'm being hyperbolic but you get the point.

A built in Arp becomes an integrated part of a synth's patches whereas a third party Arp even if it is more feature rich is still tacked on. So those who don't want or need a built in Arp will have to forgive those of us who do as we express our wishes.

The real question for me is will Zebra 3 be able to make sounds that I can't already create with the other synths in my extensive collection either on their own or in combination.

I have to assume it will but we'll all just have to wait to try the beta or demo and find out for ourselves. I'm looking forward to getting my hands on it as much as anyone so we'll see what the future holds not only for the initial release but subsequent updates as well.

Now I'm going to eat and drink far too much as is the custom in my country on this day....