Will Zebra 3 be able to load samples?

[Urs]

The script is called "DeCrackle.zip" but you might have to read the posts leading up to it to get the context.

[Urs]

Yeah. This plane ain't gonna fly. Tried that years ago already on this very forum. Though I too think a sample playback would be welcome addition (for just a bit of hybrid synthesis), it's been clearly stated several times that it's not Zebra's goal. ...sad, but it's a descision u-he is fully entitled to. Zebra's still awesome. <3 That said, I'm happy to own Falcon for manipulation-style synthesis. Alchemy was may be even better in that regard, until Apple hijacked it.

I wished that people would simply let go of the idea that "everything u-he touches gotta be good". We can only excel in a small corridor of expertise, and that's hammered out quite well in our portfolio. If we start diverging from it, the quality of stuff will decline in that area and also in our core business.

Apparently our strengths are in synthesis technologies, workflow concepts and modular economy. While we have acquired great competence in content management, making a sample based device on the same level as our other products requires a whole lot more manpower, which is a) hard to find and b) hard to pay for.

Therefore I wished people could simply settle with the fact that we don't do sampling and that is that. Enjoy our synths and effects. They're good. They only can be good because we don't try to do things we can't.

[neuromod]

I also really want constraints, and am really happy to put my trust in Urs and his team to decide some of my constraints for me. Some of the suggestions on any synth/software board tends towards "it should be able to do everything", which is most definitely not what I want. I don't mean to single out this suggestion for criticism - it's good for us to have different priorities! But there are ... I think literally zero companies I trust like I do this one.

[gondii]

I also think that was a great and perfectly logical answer. Personally I‘m not missing samples at all in u-he synths. I think it‘s good that there are a few „can-do-anything“-synths out there, but I have come to appreciate the „fewer-features-very-well-thought-through“-variety a lot more.

I‘m looking forward to Z3 whatever it‘s going to be. I‘m not hoping for or even wanting a synth to replace all my other synths, I‘m hoping for one that complements them.

[recursive one]

I also think that was a great and perfectly logical answer. Personally I‘m not missing samples at all in u-he synths. I think it‘s good that there are a few „can-do-anything“-synths out there, but I have come to appreciate the „fewer-features-very-well-thought-through“-variety a lot more.

Big +1

It seems that many people want every synth to have every imaginable feature, but what will be the difference between them then?

Now I mostly enjoy synths with limited feature set and focused worklow that are somehow pre-adjusted by the developer to sound good. Hive2 is a great example of such synth (well, it does load 3rd party content but now I mostly explore what is possible with the factory wavetables - it seems that they were selected with great care to provide sonically good results. Aside from few 3rd party FM wavetables and the Virus TI WT pack made by Adam Szabo, the factory stuff is what I actually use the most). Another synth I enjoy using a lot is Massive X, it doesn't load wavetables and I'm somewhat glad that it doesn't. Instead of browsing billions of WTs of questionable quality I just pick something among the available ones and bend it into the direction I want.

I'm sure if/when Zerba 3 is released it will have enough synthesis options to go without sample import

[Howard]

Instead of browsing billions of WTs of questionable quality…

I hate that too - why do they do it?

[Cepheus]

I think “In der Beschränkung zeigt sich erst der Meister” applies here. For the non german speaking: less is more (sort of….). Too many options, wavetables, samples, presets often limits creativity or the appreciation of what is already there. I don’t need another synth, effect or soundbank….back to the Cookbook and quality time to procrastinate.

[MorpherX]

My feature request in this regard was and is "Resynthesis" of samples to wavetables NOT a sample player.
As mentioned before the base of wavetable synth like Hive and Zebra ( 2 + 3) is still a single-cycle and / or a wavetable.
With resynthesis you are able to extend the sources, no matter if it is done via script or GUI based.
Even if U-He sees the filters, modules and their routings as the main strength, it's all based on that waves in the tables. Otherwise it would be something like Diva or Repro 5 etc.
Therefore I see that resynthesis as a little but important feature in modern wavetable synths.

[Urs]

I see the strength of our wavetable stuff for instance in the brilliance and universal usefulness of our .uhm scripted wavetables. I see future Zebra as an extension to this with visual/interactive means to create equally as precise and "sharp" wavetables.

I totally get that people want to drop a flute sample (<- for example) on such an editor and hope to harvest something that sounds alike. I just think that crafting a flute sound from an oscillator, a noise generator and a set of resonators/bodies will lead to a much better outcome. In other words, I'm most certain that in the majority of cases a converted sample is not the best source material for the sound it represents. It would be ridiculous to take a drum sample to create a drum sound.

Anyhow, as I said many times before, it won't be a problem to load .wav files for individual waveforms. I just want to avoid that people expect something like Alchemy from us.

[MorpherX]

I see the strength of our wavetable stuff for instance in the brilliance and universal usefulness of our .uhm scripted wavetables. I see future Zebra as an extension to this with visual/interactive means to create equally as precise and "sharp" wavetables.

I totally get that people want to drop a flute sample (<- for example) on such an editor and hope to harvest something that sounds alike. I just think that crafting a flute sound from an oscillator, a noise generator and a set of resonators/bodies will lead to a much better outcome. In other words, I'm most certain that in the majority of cases a converted sample is not the best source material for the sound it represents. It would be ridiculous to take a drum sample to create a drum sound.

Anyhow, as I said many times before, it won't be a problem to load .wav files for individual waveforms. I just want to avoid that people expect something like Alchemy from us.

I completely understand your arguments and that's the strenght of Zebra and in this field it's the campion.
But on the other side there are many people, including me, who often want the quality of Zebra but don't fidel and tweaking for hours to get a Flute-Like or Trombone-Like sound but simply convert such a sample to wavetable and then tweaking it only a bit with filters or modules, but in a fast way.
Not every musican is a sound-design guru !

[Urs]

simply convert such a sample to wavetable and then tweaking it only a bit with filters or modules, but in a fast way.

Sure, but maybe that's not going to be Zebra's purpose.

It's flattering that people think that we could just say "let's add this feature" and somehow a week later, there it is. But this is not how this works.

The by far best solution would be an external tool. This keeps the complexity within Zebra moderate and allows for the same economical approach as Hive. Here's why:

I have to weigh two factors against each other:

- complexity of editors
- complexity of data

These two need to be in an equilibrium. If data gets too complex, simple editors won't be appropriate. If editors get too complex, even simple data is hard to manage. We aim for really good tools to edit up to 16 slots, just like Zebra2. File based wavetables (.uhm, .wav) will allow for more frames, but won't be editable.

Sample import always involves highly complex data. Therefore it's desirable to not make it editable, i.e. have the samples ready on the hardrive and then not touch them anymore.

Sample import for wavetables is however possible in two ways:

- automatic (requires extensive research and development)
- with human interaction (requires highly complex editors)

As we've seen, we would like to avoid having to have sample editing tools just to make wavetables, and we want to avoid having large editable wavetables that need to be saved with preset. A strong argument here: Writing the tools to visually port samples to wavetables might take months, if not way beyond a year. There are complicated matters involved, e.g. we'd have to add a layer of abstraction into our codebase to be able to pre-listen to results while bypassing the rest of the synth engine. Is that worth it for a result of, say, 16 waveform slots in an editable wavetable such as in Zebra?

This leaves us with some kind fo "automatic" sample to wavetable conversion. The pro side: Data can be complex (many wavetable frames, i.e. more than 16) and all that needs to be saved with preset is the file reference, like Hive does. Con: People have done this before and they spent years on this technology. I do not see myself doing that and I would not know where to acquire the competence.

Therefore, we'll keep waveform import simple, stupid. Please do not expect a full blown resynthesis option, it simply is beyond our means and beyond Zebra's scope.

[pdxindy]

simply convert such a sample to wavetable and then tweaking it only a bit with filters or modules, but in a fast way.

Sure, but maybe that's not going to be Zebra's purpose.

It's flattering that people think that we could just say "let's add this feature" and somehow a week later, there it is. But this is not how this works.

The by far best solution would be an external tool. This keeps the complexity within Zebra moderate and allows for the same economical approach as Hive. Here's why:

I have to weigh two factors against each other:

- complexity of editors
- complexity of data

These two need to be in an equilibrium. If data gets too complex, simple editors won't be appropriate. If editors get too complex, even simple data is hard to manage. We aim for really good tools to edit up to 16 slots, just like Zebra2. File based wavetables (.uhm, .wav) will allow for more frames, but won't be editable.

That is perfect... I have no need of an editor in Zebra. And there are plenty of resynthesis options. I just want to be able to import existing wavetables into Zebra. They will make great transient attacks for the comb modules!

Oh, and Uhm scripts with OscFX!

And Hive's improved unison... plus voice per channel... go Zebras 3!

[recursive one]

So Zebra 3 will load Uhm wavetables?

[Urs]

So Zebra 3 will load Uhm wavetables?

But of course!

Apart from "Morph" and "Blend" there'll be something like "wav/uhm" as an option. Unlike former it will not offer editors, because like Hive, we don't want to store megabytes of wavetables in the presets.

[The Nerdy Music Guy]

But the Zebra 3 equivalent of the OscFX would still work on those? That would be amazing!