Zebra 3 Public Beta Revision 20399

[Synthosynthony]

NAD

This will be off-topic, so briefly. I'm not sure you're right about everything. The LHC creators at CERN deliberately simplified things about antimatter and micro black holes at launch to attract the world's attention to the project. We all know about ATLAS thanks to sensational statements from a respected astrophysicist claiming (allegedly) "aliens were there". Nothing like that happened with Oumuamua, and that's why few people know about it, even though there was plenty of reason to. A bright, juicy informational backdrop will allow a lot to be done in the future in terms of observing interstellar comets. NASA gives Twitter accounts to its Mars rovers etc. I'm sure everyone has "manuals," but if it's about loudly announcing yourself—especially after making something really cool... sometimes you have to simplify and use something bright before the manual. Maybe it's better to do it before TikTok bloggers simplify. And if they didn't understand, you can help them understand too. Why not? Is that snobbery or what? I'd be thrilled if the internet was buzzing with discussions about Zebra 3 instead of "new colorful sample packs that will change your life". That's why the free beta is a great move. But it doesn't cancel out the fact that familiar attention-grabbing mechanisms kick in at final version launch.

[Synthosynthony]

Your initial post was not about how to use Zebra3 at all, you basically asked for external proof why it even exists because you couldn't find any on YouTube.

You seem to know Zebra2 so you should be aware of Zebra's approach and as others have mentioned, U-He provide the beta for free for you to test it and see if it fits your shoe.

No, that's just your understanding of my post. I was talking about the fact that I still can't figure out how to use vectors and what I can get from it. The concept of creating a waveform this way in a vacuum, without a goal, is still abstract to me, and it would be interesting to see videos on what the advantage is of creating waves this way (beyond the quality of transforming one shape into another thanks to vector graphics).

I've seen a lot about Zebralette—I have it—and I see the convenience of drawing with a pencil. But I don't understand how to apply it or why I need to draw.

For example, I can explain why granular synthesis is needed and why, say, it's great to have control over each granule. I can understand what spectral synthesis is for. But in my explanation, I would suggest first listening to a specific sound example and then, based on it, understanding technologically how such a sound was achieved.
My original post wasn't about "why does this even exist" in a negative sense, but "help me figure it out—are there videos from the final timbre to the drawing?" From the assembled puzzle to its pieces.

Oh so actually, you're not familiar with making your own waveform by drawing it, is that right? If that's the case then I understand the confusion here.

Yes, that's why I still don't understand why this is needed, and it would be interesting to start from the result. To see and hear the difference from standard wavetable synthesizers.

[jtsterays]

Yes, that's why I still don't understand why this is needed, and it would be interesting to start from the result. To see and hear the difference from standard wavetable synthesizers.

Okay, I see. I thought you were talking about something else, so I accidentally shifted the main point (sorry for that).
Let's just remove the word vector, it doesnt really matter here.

All the other wavetable synths u know like Serum, Phaseplant, etc... They can actually draw waveforms from scratch too. But most of the times those synths users will sample other waveforms (eg vintage synths, etc...) or use premade wavetables, like for example Serum with the huge factory wavetables. Why not draw? because those synths have a very subpar set of tools to make them manually, so it's really slow and tedious to draw complex waveforms. Also due the nature of their wavetable system.

So in Z3, it uses a new system that graphic design software use, with the included tools from them that makes the process of making waveform (they call it curve) from scratch way faster, especially for really complex curves. It could be 2,3 times faster than Serum, etc...

Ofc you dont have to draw from scratch, you can just use from others in the future (it's beta so there's not alot built in rn)

[Q-Bass]

I already wrote it elsewhere.. I see the difference between normal wavetable synths and Zebra 3 similiar to difference between pixel and vector graphic. Or - because the waveforms are not meant to be static - like difference between stop motion animation and CGI.

[jtsterays]

Yes, it's bitmap vs vector, but its not the core point here: He didnt know that previous wavetable synths can draw waveforms too so he thought Z3 is the first that does it, so hes questioning it's usefulness.

[stippenstoh]

NAD

This will be off-topic, so briefly. I'm not sure you're right about everything. The LHC creators at CERN deliberately simplified things about antimatter and micro black holes at launch to attract the world's attention to the project. We all know about ATLAS thanks to sensational statements ....

I can't remember anyone else here who struggled with the "why, oh why" of vector drawing. It's not nuclear research, whatever questions haunt you might get answered if you work the beta.

[Urs]

(beyond the quality of transforming one shape into another thanks to vector graphics)

But that is the point, and the minimum effort it takes.

I like to bring up a thought experiment, which is based on an extreme example, which might be enlightening. Take a wavetable synth that is *not* vector based but lets you draw your own waveforms. Serum and Vital come to mind, probably a few others.

First, create a wavetable that mimics PWM - from scratch. It's a very simple task. Draw 256 frames, each time moving the step in the pulse wave a little further from one end to the other. This is pretty doable, probably takes half an hour or so.

Second, instead of PWM, draw a wavetable from scratch that smoothly transitions from sawtooth to triangle. This will take hours, and it has the potential to get totally messed up depending on the drawing grid or whatever tools are available.

These are simple things that are easy to do in Zebra 3 but very time consuming in the "standard" synths for this kind of user-editable stuff.

But then, in Zebra 3 you can as easily create smooth transitions with very complex waveforms that one would never even think of doing in classic, sample based waveform editors. Sure, some of these editors have high level manipulation tools that aide a few basic transition types, such as symmetry, some warping some crossfading... but those are very limited in their applications, they can't remotely compete with the vast set of possibilities in Zebra 3.

[Synthosynthony]

Yes, it's bitmap vs vector, but its not the core point here: He didnt know that previous wavetable synths can draw waveforms too so he thought Z3 is the first that does it, so hes questioning it's usefulness.

Not quite. I know that other wavetable synthesizers allow you to draw the waveform, but in terms of labor intensity, it looks like redrawing the movement of a character frame by frame from video just to replicate that movement in 3D graphics. That is, you need to move every bone in 24 consecutive frames, orienting yourself by the video image. This way, in Serum, it's practically impossible or too hard to replicate a complex timbre. We know some geek examples but I have never tried this myself. Are you saying that in Zebra3, thanks to vector graphics, all this is faster? Well, then we're back to wanting a video example of replicating any complex sound—from the result to the method

[ThoughtExperiment]

If you really have 'scoured the entire internet' then you must have spotted the ton of videos for Zebralette 3 which explain how to do exactly that...

[Urs]

Like this really old one...

[Synthosynthony]

If you really have 'scoured the entire internet' then you must have spotted the ton of videos for Zebralette 3 which explain how to do exactly that...

I'd be happy if you share a links to the ton of videos where we're not just spinning a waveform in a vacuum, but recreate a real complex instruments

[canum lupus]

If Zebra were to adopt a similar approach, one possible solution would be to always show dimmed modulation dots at all time.

I know. Hence our internal builds already have it:

Wow — exactly this!
Yep, that’s precisely what I meant. Looks like I didn’t need to say anything

Hehehe, no worries... the visualisation of "what modulates what" is probably the most discussed theme of this thread, and the whole beta period. More so than the absence of a sequencer or the issues of CPU load.

It took me by surprise a bit, so we're implementing an initial set of visual clues, just like you suggest, and we'll probably put more focus on it in future work.

A bit late... but, I would like to have a sequencer Just mention it!
Yes I know, there are different possibility's, but sequencer is a sequencer with, I hope, additional modulation opportunity's.

And Yes, Zebra 3 OSC is not easy at the first glance, but after a while, next to the excellent sound of course, it's just an amazing synth!

The Distortion FX is trying to kill my CPU ...poly presets!

[pummel]

Goal is to make using knobs/sliders faster, by [optionally] using only the mouse keeping the other hand free for the key bed...

KNOB USAGE:
current:
click knob perimeter, drag vertical to adjust base level
click knob center, drag vertical to adjust modulation

proposed: (as an option in settings)
click knob anywhere, drag vertical to adjust base level
click knob anywhere, drag HORIZONTAL to adjust modulation
...implemented so one can click once, then adjust both by dragging in different directions.
...with 'direction toggle' along the 45 degree and 135 degree lines

KNOB ADJUSTMENT RATE: (as an option in settings)
Current:
coarse, click drag
fine, ctrl click drag
Proposed:
continuously variable; mouse drag RATE determines very-fine <--> course
...functions similar to key velocity; perhaps with a user definable mouse rate curve

[neuromod]

I still can't get over the "where did you get your physical modelling code" question. This thread is like the mélange; addictive glimpses of other worlds, too much is fatal. A few more weird responses and I will have visions of Urs leading armies against the Imperium.

[Urs]

Goal is to make using knobs/sliders faster, by [optionally] using only the mouse keeping the other hand free for the key bed...

KNOB USAGE:
current:
click knob perimeter, drag vertical to adjust base level
click knob center, drag vertical to adjust modulation

proposed: (as an option in settings)
click knob anywhere, drag vertical to adjust base level
click knob anywhere, drag HORIZONTAL to adjust modulation
...implemented so one can click once, then adjust both by dragging in different directions.
...with 'direction toggle' along the 45 degree and 135 degree lines

KNOB ADJUSTMENT RATE: (as an option in settings)
Current:
coarse, click drag
fine, ctrl click drag
Proposed:
continuously variable; mouse drag RATE determines very-fine <--> course
...functions similar to key velocity; perhaps with a user definable mouse rate curve

These are really interesting ideas, but my initial feeling is, I don't think they'll work well. Nevertheless, I'll be happy to try them! - Just can't say when and in which context.

So here's why I don't think they'll work well:

Setting Mod Depth independent of Parameter value will be an acrobatic exercise with the mouse. We'll also have to display 2 values at a time... it's gonna be more strenuous than having to click and drag twice most of the time. Gotta admit, it may have its uses when weighing Param value vs. Mod value. But what about params that have 2 or more modulations going?

Mouse acceleration... a pet peeve of mine. This will force people to do tasks they know how to do quickly in slow motion. Adds an extra degree of consciousness for otherwise routine tasks.

Anyhow... we can try... might take time to do so, but we'll see...