Zebra 3 and Zebra Legacy

[Urs]

(all graphical user interfaces are, well, visual. Waveforms, envelopes and spectra are depicted visually, on hardware, in user guides, on graphical user interfaces. There obviously is an intuitively understood relationship between a visual representation of a waveform, a spectrum or an envelope and their aural experience. Hence, there is totally no reason to think that exploiting the experiences and advances of graphic design software is a bad thing for audio software. It isn't. It's a pretty cool thing. Pretty much every DAW does it, many synths and fx do it, and we simply do it in the uttermost depth and detail that we think is appropriate, if not necessary for what we want to achieve here)

[Etienne1973]

Sorry, yes, I'd say it's musical. Excuse my noodling

https://www.youtube.com/watch?v=O9yvGlfI2jo

Fascinating! Extended (Zebra2) sound design but still THAT musical soul.

[KBSoundSmith]

Sorry, yes, I'd say it's musical. Excuse my noodling

https://www.youtube.com/watch?v=O9yvGlfI2jo

Can't wait for the release. Looking like z3 is going to be special -- just the one oscillator looks to be more powerful than many complete synths.

[nilhartman]

Sorry, yes, I'd say it's musical. Excuse my noodling

https://www.youtube.com/watch?v=O9yvGlfI2jo

Can't wait for the release. Looking like z3 is going to be special -- just the one oscillator looks to be more powerful than many complete synths.

Exactly my thoughts. I feel like I could already write a full album with current version.

[Urs]

Yeah, hehehe.

Even the sceptics in my company are slowly coming around.

It's not that difficult. It's more like "oh, I didn't know one could do that".

[bmrzycki]

Sorry, yes, I'd say it's musical. Excuse my noodling

https://www.youtube.com/watch?v=O9yvGlfI2jo

That’s gorgeous sounding! It’s reminiscent of 90s Sierra game soundtracks but at full bit depth. I’m really looking forward to Z3.

[Fannon]

Yeah, hehehe.

Even the sceptics in my company are slowly coming around.

It's not that difficult. It's more like "oh, I didn't know one could do that".

It's nice that you have sceptics in your own company. Means you also have people to challenge you, which is really important.

Really looking forward to Zebralette 3! I wonder what the new Zebra 3 capabilities would allow when it comes to more typically "physically modeled" sounds.

[OdoSendaidokai]

Yeah, hehehe.

Even the sceptics in my company are slowly coming around.

It's not that difficult. It's more like "oh, I didn't know one could do that".

I was blown away when I saw the Zebralette Demo from superbooth and the graphical tools palette. The fact that the waveforms are spline based opens up so much more possibilities. Did you think about amount driven event parameters on curve points?

I mean something like you could define e.g. an upper point of a curve as an event, where one could define e.g. if the velocity (or any other modulation source) exceeds an amount value it starts another modulation, that wouldn't be started if that value wasn't exceeded. Or the other way downwards. That would be like playing expressively on synthesizers on steroids. Even without MPE For Sounddesigners a whole new universe how a sound / preset can develop in time and expression.

[Urs]

I am anticipating questions for a lot of crazy things. Such as making the tools of the editor available as modulation targets (which I would have to decline, if only for my own mental sanity).

I'd be happy to look at examples where other synths allow for curve altering modulations, but I'm not sure what the tradeoffs are in return for whatever it is, and whether or not an OscFX might accomplish the same thing.

What complicates matters is that a lot of the actions on curves and interactions between them are not happening in realtime. What looks snappy on the UI still has a lot of math behind it, which would use tenfold of Diva Filter CPU if rendered thousands of times per second. In fact, the oscillators and envelopes store the curves with quite a bit of pre-calculated data to optimise CPU. If the curves (or even single points) were subject to modulation somehow, I'm not sure as to how much could be pre-calculated.

As for expressive modulation and stuff, we've already started quite a bit of this in Hive, where e.g. the Function Generators can initiate some envelope action in response to, say, Polypressure. We're also in love with our Osmose and we want to work with Expressive E and others to get the most out of MPE. This should happen in parallel to Z3 development.

[OdoSendaidokai]

What complicates matters is that a lot of the actions on curves and interactions between them are not happening in realtime.

I didn't point out that I hadn't the oscillators in mind, instead the MSEG envelopes I was thinking about. Sorry about that. I can imagine that the CPU costs for the eventually only subtle effects would be a massive imbalance.

[pierb]

I didn't point out that I hadn't the oscillators in mind, instead the MSEG envelopes I was thinking about. Sorry about that. I can imagine that the CPU costs for the eventually only subtle effects would be a massive imbalance.

Same. I've always wanted to be able to morph or modulate MSEGs in Zebra. This would open up so many possibilities.

[axb312]

I am anticipating questions for a lot of crazy things. Such as making the tools of the editor available as modulation targets (which I would have to decline, if only for my own mental sanity).

I'd be happy to look at examples where other synths allow for curve altering modulations, but I'm not sure what the tradeoffs are in return for whatever it is, and whether or not an OscFX might accomplish the same thing.

What complicates matters is that a lot of the actions on curves and interactions between them are not happening in realtime. What looks snappy on the UI still has a lot of math behind it, which would use tenfold of Diva Filter CPU if rendered thousands of times per second. In fact, the oscillators and envelopes store the curves with quite a bit of pre-calculated data to optimise CPU. If the curves (or even single points) were subject to modulation somehow, I'm not sure as to how much could be pre-calculated.

As for expressive modulation and stuff, we've already started quite a bit of this in Hive, where e.g. the Function Generators can initiate some envelope action in response to, say, Polypressure. We're also in love with our Osmose and we want to work with Expressive E and others to get the most out of MPE. This should happen in parallel to Z3 development.

Are there any plans to make Zebra 3 Multi-layered? Multiple voices with their own oscillators, LFO, Modulations and FX ie.? Would be greatly appreciated...

[Urs]

Are there any plans to make Zebra 3 Multi-layered? Multiple voices with their own oscillators, LFO, Modulations and FX ie.? Would be greatly appreciated...

Layering is a bit of a pet peeve of mine.

Why should we prevent an LFO from Layer 1 to be reused in Layer 2 and Layer 3 and Layer4? Why shouldn't one be able to send a portion of the audio of Layer 1 to the reverb on Layer 2?

My designs have always been developed with efficiency in mind, and to me the concept of layers seems more like a restriction than an offering.

I understand that layering allows for quick experimentation by combining simple sources in an easy manner. But even that is more difficult to achieve in a single instance of a virtual instrument than in two or more instances in a DAW, where each preset browser can be open at the same time. As in: For quick multi layering you'd need two separate kinds of presets, multis and singles. This drives cost and complexity. I believe it makes the very thing harder to achieve that the user may be seeking.

As for sound design efficiency in Zebra, we are experimenting with UI refinements that let people group modules and show/hide those to various ways to keep focus on what could be seen as a layer - even if it isn't necessarily a layer in the classical sense.

[axb312]

Are there any plans to make Zebra 3 Multi-layered? Multiple voices with their own oscillators, LFO, Modulations and FX ie.? Would be greatly appreciated...

Layering is a bit of a pet peeve of mine.

Why should we prevent an LFO from Layer 1 to be reused in Layer 2 and Layer 3 and Layer4? Why shouldn't one be able to send a portion of the audio of Layer 1 to the reverb on Layer 2?

My designs have always been developed with efficiency in mind, and to me the concept of layers seems more like a restriction than an offering.

I understand that layering allows for quick experimentation by combining simple sources in an easy manner. But even that is more difficult to achieve in a single instance of a virtual instrument than in two or more instances in a DAW, where each preset browser can be open at the same time. As in: For quick multi layering you'd need two separate kinds of presets, multis and singles. This drives cost and complexity. I believe it makes the very thing harder to achieve that the user may be seeking.

As for sound design efficiency in Zebra, we are experimenting with UI refinements that let people group modules and show/hide those to various ways to keep focus on what could be seen as a layer - even if it isn't necessarily a layer in the classical sense.

Thank you for the response.

[operator]

[...]

REQUEST / VISION:
Dear Urs, would it be possible in Zebra3 to somehow incorporate a "class-based system" like in HTML/CSS please (but visually like in some Wordpress “Class-First-Page-Builders” e.g. Bricks). This would tremendously speed up the sound-design-workflow by allowing convolution/multiplication of presets and settings inside the synth (globaly or only on certain modules). I don´t like to use Presets in my music-creation workflow and therefore create all sounds from scratch WHICH often is VERY repetitive and kills the creative moment (100%, every time). I know most of all music-creators are using presets heavily because of this reason so a request like that mike not come up very often BUT after 10+ years of music production (never used a preset ever – can´t even create music with a preset-based workflow) I really think this could be an amazing feature for sound-designers like me (Zebra Core User Base). This would allow us to almost work at the speed of thought (after initially setting up standard/favorite classes or even creating a simple "utility class framework").

PROBLEM:
For example when having a simple setup like "Osc + Filter + ModEnvelope = FilterPluck" and then trying to inject some musicality by randomization (analog vibe) on A,D,S,R, Filter, Mod-depth - It just takes too much time for me to keep the flow-state active. Also I am loosing the mental-model of the sound I am working on while i am trying to rush the drag`n`drop to 7-10 parameters just to give them a little randomization.

SOLUTION:
If we could create for example our own utility classes (framework) with their custom properties and assigned functions to it, we could work very fast by just adding the “class-tag” on the Synth/Module while still keeping the mental-model of the sound in the mind. This would also allow to have the DRY-principal in Zebra --> don´t repeat yourself. So when you created for example a “.randomEnvAll” class while working on Env1 you could then easily transfer it to whichever Envelope you would like (in Zebra even to OSCs because they are all the same) without the need to add the same setting to other Envelopes --> this also has better scalability when using a lot of modules/settings. (just one usecas out of MANY)

BENEFIT:
So music producers can interact with their presets/classes in an additive-manner rather than subtractive-manner like almost all producers do it right now (--> open a preset, remove FX, remove some bloat settings AND then lastly get a sense what the preset is doing). On the other hand, if we could create our own final-sound by just adding 3-5 classes to the Init-preset (or whatever preset -> thats the beauty ) THIS would open up a more goal-oriented workflow AND maybe finally we get those lazy music-producers away from overusing presets and can direct them rather to (Zebra-)Frameworks. --> The standard Framework which is then shipped with Zebra could be called: “Equisetum” --> aka. “Horse-Tail” (you now because of Tailwind-CSS Framework )

CONCEPT:
The UI and UX I imagine it rather similar to a software I am using with Wordpress. It is a visual “class-first-PageBuilder” called Bricks.

How to create a Class:
1.) In Zebra we would have an Input-Field at the top where we can add class-Tags by typing in some custom Text, we can have multiple tags in a row 2.) When we had previous Parameter Tweaks in the synth THEN adding a “class-Tag” which is “active/highlighted” changes the synth temporarly to Init-state 3.)Then while the class is active we can dial in some values on whichever module we like --> which then is applied only to the class 4.)After deselecting the “class-tag” at the top the synth displays the previous synth-parameter-tweaks again with the class-values added by a new colored mod-depth indicator. So the class system could be seen as highly complex "Mod-Source via Constant" modulator ontop of all.

How to add Class: Just slap it on the synth or module by typing in the class name.


This would allow to multiply/convolute different Zebra settings (e.g. add lots of randomization at once). ALSO it would allow those “preset-using-music-producers” to have a simple way of interacting with the “complex-scary-presets/module-structure” they have infront of them --> e.g. instead of searching the preset/synth for unnecessary FX just add the utility-class “.fxNone” and all effects are turned of no matter what the synth has activated.


Examples of simple class-architectures which could be added to the Init-Preset (or to any other Preset):

Add to Init Preset: ".oscSuper", ".filterPluck", ".randomFilterAll", ".randomPitch", “.fxCrushVerb”
Add to Init Preset: “.oscSqr”, “.oscPWM20”, “.8Bit”, “.envPluck”



I am realizing right now this might fit better to a fixed synth architecture because of the logic going on in the back BUT on the other hand Zebra3 would benefit the most from it due to its fluid/modular architecture. But maybe this might be better suited for Hive, yet Zebra could also be seen as a synth-framework already... so those controls might make sense in that way.