Feature request: PCM Sample import for Hive 2/Zebra 2/3

[vertibration]

Why, yes. I must clarify though, it won't happen for Zebra2 anymore.

But it might be common knowledge that - for me - the redesign of the waveform editor in Zebra is the main motivation to do version 3. As it happens, this is what I'm working on right now (and the reason KVR and Facebook haven't seen much of me lately). Furthermore a way to import and convert samples/.wav is part of that concept. I can't yet say how well this will work because I haven't fully implemented it yet: I can convert hand drawn waveforms to Zebra3 waveform geometry, and also anything formula based, but I'm waiting for a colleague to implement .wav drag'n'drop before I start testing that. And maybe another colleague has to figure out pitch detection in order to get it right.

Serum does this, and while it has some good results sometimes, other times require a lot of editing in the editor to smooth out the waveforms. Do you believe your new venture into this will allow a smoother replication of .wav into Zebra 3? I mean, do you intend it to fix some of the shortcomings that other editors have with how a sample is converted?

[EvilDragon]

No, sample will just become a wavetable. That's all.

[Lbdunequest]

So nice to read all this

So hows Zebra 3 development going ? what chances are to see it coming out this year ?

[pdxindy]

I might add, I'm still not keen on sample playback itself. Instead, we hope that the things we've planned for Zebra3 render sample import more or less moot.

Besides converting samples to wavetables... like in Alchemy there is Spectral and Additive resynthesis too. The resynthesis is quite good in Alchemy.

Also, PPG Infinite Pro had something interesting going on. Analyzing a wav file it also obtained an atonal/noise component which regular wavetables don't.

I'm glad you are not keen on regular sample playback. There is still a lot to explore with straight synthesis!

[Urs]

So nice to read all this

So hows Zebra 3 development going ? what chances are to see it coming out this year ?

The oscillator is going rather well. It's the core piece of Z3 development since it's all about the editors to create your own wavetables, and in turn multi stage envelopes.

From my current perspective, the first thing that'll come out is Zebralette 3. It's also supposed to act as a wavetable editor for Hive, which is something that's requested frequently. This is looking very good for this year (2022).

Zebra3 is not likely to be released this year. But this year we'll most certainly evolve Zebra2 into its next stage ("Zebra Legacy"), probably around the same time we hope to release Zebralette 3. I'll have to write something about that, because it changes how we plan to deal with upgrades (everyone who has Zebra2 and/or TDZ licenses today gets the upgrade privileges we laid a decade ago, but we won't offer these entitlements once we sell Zebra2 for a reduced price).

It's a lot of work ahead, but this time I don't think I'll let anything interfere.

[KBSoundSmith]

This all sounds very promising, and it seems to me that the development priorities are right where I was hoping they'd be (particularly regarding the waveform/MSEG/etc editors).

And FWIW, I'm 100% in agreement on the "no samples" stance. Without knowing all of the rationale for u-he's stance, I've my own list of reasons why I'd prefer not to see samples.

[Lbdunequest]

Ow nice. Will Zebralette gonna be free?

[Urs]

Ow nice. Will Zebralette gonna be free?

I'm pretty sure that's a yes.

[Shiek927]

Excited to hear! hopefully we'll all be playing with Z3 in the near future! do you think any of the drum modules from the Physical Modeling prototype will make it in or any percussive/drum-type sounds will have to be made purely with synthesis (e.g. comb filters)

[Urs]

Excited to hear! hopefully we'll all be playing with Z3 in the near future! do you think any of the drum modules from the Physical Modeling prototype will make it in or any percussive/drum-type sounds will have to be made purely with synthesis (e.g. comb filters)

Can't say for sure yet. The drum stuff is very well documented, but I have not yet had a field day with it. Once I move to comb filters - which are also ready in their basic form - I'll try to cross-reference with the drum modules. I have an inkling that both can pollinate each other.

Anyhow, there is so much stuff. For instance, I spent the past week writing an alternative rendering module for Z3 oscillators which uses 1000+ sine oscillators instead of wavetables. I got it down to about half the CPU of a plug-in with a similar method which I like a lot. This method allows for some elaborate and dynamic drums and mallets, all without any physical model at all.

We'll see. I don't think Z3 needs to be "overly mature" from the beginning. It's actually better to keep some headroom for future developments, and have some organic growth in features.

[Shiek927]

Excited to hear! hopefully we'll all be playing with Z3 in the near future! do you think any of the drum modules from the Physical Modeling prototype will make it in or any percussive/drum-type sounds will have to be made purely with synthesis (e.g. comb filters)

Can't say for sure yet. The drum stuff is very well documented, but I have not yet had a field day with it. Once I move to comb filters - which are also ready in their basic form - I'll try to cross-reference with the drum modules. I have an inkling that both can pollinate each other.

Anyhow, there is so much stuff. For instance, I spent the past week writing an alternative rendering module for Z3 oscillators which uses 1000+ sine oscillators instead of wavetables. I got it down to about half the CPU of a plug-in with a similar method which I like a lot. This method allows for some elaborate and dynamic drums and mallets, all without any physical model at all.

We'll see. I don't think Z3 needs to be "overly mature" from the beginning. It's actually better to keep some headroom for future developments, and have some organic growth in features.

No worries! if Z3 is going to be anything like the previous u-he synths, it's gonna be around for a long time to come. It makes sense to let it naturally grow it's feature-set the same way that Z2 did.

[nilhartman]

For instance, I spent the past week writing an alternative rendering module for Z3 oscillators which uses 1000+ sine oscillators instead of wavetables. I got it down to about half the CPU of a plug-in with a similar method which I like a lot. This method allows for some elaborate and dynamic drums and mallets, all without any physical model at all.

Does it mean Zebra could also dial super sharp, hyper detailed timbres à la Razor / Lazerbass / Harmor ? So much can be done with all these partials, so much more than brostep wobble basslines anyway… and with u-he’s UX, it could be marvelous !

[j wazza]

From my current perspective, the first thing that'll come out is Zebralette 3. This is looking very good for this year (2022).
...

thanks for the update, really excited to hear this

for my uses wavetable import would be great and i wouldnt need a sampler as when i use samples in synths its usually samples of analog synths which works well for making wavetables, much better than instrument samples

im curious will it just be chopping the samples to wavetables or will it also do it the fft way like how serum has both options?

it sounds like you dont want to make it a sampler is because providing the samples would be a big task, which is fair, but id love a sample player just for my own samples, i often run samples through zebrify but being doing that polyphonically and with fm would be great, but not needed

on that note, will there be a zebrify 3?

still though, id never complain about uhe, it would be blasphemy, it would be like getting a ferrari and complaining that it doesnt have cupholders, appreciate what it does have, i trust urs' vision and think he knows whats best for zebra

[nilhartman]

an alternative rendering module for Z3 oscillators which uses 1000+ sine oscillators instead of wavetables

As I've been fantasizing about how gloriously sharp it could sound, I got to ask the obvious : now that it's coded, will that new rendering module make the cut to Zebra / Zebralette 3 ? Please ?

[Urs]

an alternative rendering module for Z3 oscillators which uses 1000+ sine oscillators instead of wavetables

As I've been fantasizing about how gloriously sharp it could sound, I got to ask the obvious : now that it's coded, will that new rendering module make the cut to Zebra / Zebralette 3 ? Please ?

It looks pretty good for this. I can't promise that it'll rival specialised engines like Harmor, Alchemy or Loom though.

Funnily, there was an example mp3 for Hive a few years back before we decided on wavetables. That example was rendered with a predecessor of this Z3 rendering module. People didn't like what the example did back then, so we dropped it. I hope that this new one will be more likeable.