Zebra 2.3.3 beta

[kodama]

Nice! I really like the sound of foldback stuff, so that sounds awesome as well

[pdxindy]

"XMF" is the same as in the current XMF, "analogue" is the same as in Bazille

The forthcoming XMF sounds FANTASTIC! Especially "analogue" mode.

Looking forward to trying the new XMF

[Urs]

Hey, but don't expect too much... even though "having a 48dB filter" sounds like a great thing, its usefulness is rather questionable.

For reasons I'm not quite sure about it sounds massively different from the old XMF. Like in the Bazille filter, self oscillation produces a pretty clean sine with less harmonics than e.g. a Moog filter. The Overload of the old XMF did that trick, but I havn't yet got a clue how to reproduce it in the corresponding setting of the new XMF.

[Kyran]

so you mean it's currently not possible to make the new xmf sound like the old one?

I love the growl, overload and moogy sound of the xmf. Please don't remove that!!

[Urs]

so you mean it's currently not possible to make the new xmf sound like the old one?

I love the growl, overload and moogy sound of the xmf. Please don't remove that!!

Well, I think I just sussed that out...

Five distortion modes:

• XMF pretty close to old XMF, but 24dB less loud on full Overload
  
• analogue almost same as Bazille
  
• digital 8-Bit lofi
  
• eco close to analogue, but using a cheaper waveshaper (saves 30% cpu on each XMF)
  
• folded rather ill type of distortion

And of course, the resonance is tamed at frequencies higher than 8kHz... no more tinnitus-like experiences

[MitchK1989]

Holy crap urs, you're like a soft synth super hero!

[bagginz]

XMF pretty close to old XMF, but 24dB less loud on full Overload

Yes, the gain staging needed taming.

analogue almost same as Bazille

Cool. Haven't heard Bazille filters yet, but it's reputation is beginning to precede it.

digital 8-Bit lofi

Bah...

8 bit business is currently handled in the shaper mod.

I was hoping for an extreme saturation algo like the "digital" distortion mode in the Viri.

Also 8 bit is not that useful in my experience. Often when fellow producers talk about applying bit crushers, they are referring to Sample Reduction, which currently isn't covered in Zebra.

Still Urs, I don't want to appear ungrateful, or to diminish your extraordinary efforts in improving the already massively superb Zebra even further. The dual filter serial/parallel routing + filter offset alone is a major feature improvement in my books.

eco close to analogue, but using a cheaper waveshaper (saves 30% cpu on each XMF)

Useful

folded rather ill type of distortion

Hmmm. Will it useable?

cheers,
bagginz

[Echoes in the Attic]

That all sounds like some great XMF improvements! I'm not sure I'd ever load up the regular VCF!

On a side note, what came of that poll where you asked about Zebra's biggest weaknesses? It seemed that the waveshapers got the most votes. Are you tinkering with those at all for the update? Just curious really...

[Urs]

That all sounds like some great XMF improvements! I'm not sure I'd ever load up the regular VCF!

Well, the XMF got slightly more expensive... and the normal filters get their overhaul too. LP Allround has just seen some major improvements

On a side note, what came of that poll where you asked about Zebra's biggest weaknesses? It seemed that the waveshapers got the most votes. Are you tinkering with those at all for the update? Just curious really...

Yes and no.

I have a full new distortion module laying around which was coded by no-one less than George "Gasm" Bezerianos (remember TriDirt?). I initially wanted to merge that into the Shaper modules, but all controls are too different. So I have to add a new type of module for it. And I have to go through the algorithms and apply my latest trickery for speed. Thus, hmm, I'm not quite sure if everybody and myself is patient enough for this to come in 2.5. It may have to wait for 2.6

Urs

P.S:: Bagginz, I agree that the bit-crusher isn't all that great. Havn't sussed out the digital mode of the Virus yet though... I think the problem here is that I'm not just driving the filter input, but quite a few "transistors" within the whole filter circuit. I guess for the Viral Digitalness one needs to do a single flip-flop in one place and all the normal stuff in the other places.

[heavensdaw]

2.6

Hd

[bagginz]

P.S:: Havn't sussed out the digital mode of the Virus yet though...

Urs,
not that I think that the Virus' digital distortion mode is the holy grail of distortion, ( but it does the job o.k.) Just that the Virus "Digital" mode is the capable of the most saturation and I know that you have a Virus to use as a reference point.

Currently Zebra doesn't quite have this.

Closest I can get is by using a Shaper module set to Shaper mode and turning up Depth and Edge to max. And by using LowPass Allround with it's fiery resonance to make it scream a bit...

Putting another shaper module in series doesn't give the desired result either. Nor does various combinations of XMF filter distortion and shapers in series.

I'd send you a Zebra patch which is pretty close just so you can get an idea of what I am babbling about - if I knew how to attach it to this post. Got an address I can send it?

I think the problem here is that I'm not just driving the filter input, but quite a few "transistors" within the whole filter circuit. I guess for the Viral Digitalness one needs to do a single flip-flop in one place and all the normal stuff in the other places.

Well, if you say so, I'll take your word for it

I'm a musician not a programmer, I know nothing about how this stuff works...

Cheers,
bagginz

[pdxindy]

I have a full new distortion module laying around which was coded by no-one less than George "Gasm" Bezerianos (remember TriDirt?). I initially wanted to merge that into the Shaper modules, but all controls are too different. So I have to add a new type of module for it. And I have to go through the algorithms and apply my latest trickery for speed. Thus, hmm, I'm not quite sure if everybody and myself is patient enough for this to come in 2.5. It may have to wait for 2.6

Urs

Looking forward to trying that... 2.5 or 2.6.

One other thing for the future... Would it be possible to have a switch in Zebra so that when selected it rendered the same as if I put the audio interface to 96khz? It is a pain to run at 96khz resource-wise, but some of my Zebra presets need 96. Overall, I would rather run at 48khz and be able to selectively set Zebra to 96 as desired. I don't know if that is easy or hard to do for you and I can get by with my current workflow of bouncing Zebra at 96 but its a clumsy way.

Thanks for all your work. After 2.6 I guess I would not care if you did any further updates to Zebra. Recently I used some other synths mostly working with audio and last night I used Zebra for the first time in a couple weeks and it sounds amazing.

[Urs]

I'd send you a Zebra patch which is pretty close just so you can get an idea of what I am babbling about - if I knew how to attach it to this post. Got an address I can send it?

urs@u-he.com

Yesterday I managed to tame the resonance of LP Allround. At the same time it adds the pleasant capability of quite some nice distortion - it's quite a law that the amount of distortion is responsible for the resonance being behaved or not. Same accounts for HP12, RezBand and LP12 which are based on the same topology.

Today I'm going to look into the "difficult filters", which are LP Xcite, MidDrive, OldDrive, QBand, HP24 and the notch one. These are based on a topology that took me ages to distort at all. Their problem is that they sound instable on fast filter sweeps, and the yresonance becomes exceptionally loud and uncontrolled in the bass department.

It's the same topology that was used in FilterscapeVA's first lowpasses later on (FSVA filters were coded years after Zebra's original ones). There I'm using a spiffy trick to compensate for massive volume jumps (which is probably why the Hip Hop and Soul scene and loves that thing). Maybe I can somehow port that back into Zebra...

[Urs]

some of my Zebra presets need 96

Can you send me a bunch so that I can see which modules you mean? I'm currently oversampling some modules and algorithms anyway, so maybe it's right there already?

[GMusic]

what are the chances of a preset search/filter function?

would be so useful for something like finding a 303 sound: type in 303 and all presets with 303 in the name come up