Improvements!

[bagginz]

Hello Urs,
some more feedback and ideas for improvements:

LFO polarity is an issue, along with time-based LFO rates rather than beat based.

No big deal, as I mentioned, one can get around it using the USER shape.

Here comes a downer: Separation for the existing filters is not doable, or it would not provide for the anticipated results. This is because there are no identical algorithms used in serial filter stages. I.e. A resonant SVF lowpass might be followed by a non-resonant SVF lowpass. The whole formant-ish thing of filter separation will only work if two or more serial stages are resonant... which unfortunately (and as far as I remember) in Z2 never is the case...

Ok, but not really a downer. I'm just being lazy really. I'll just have to continued using 2 x filter modules instead. It's actually not difficult at all in Zebra.

Incidentally, the formant/vocally thing always works better for me with parallel filters...

Eq module in synth section would be cool, but I have already done that in FilterscapeVA with a highly optimized mono Eq... I don't wanna put a stereo version of this in Zebra right now...

Fair enough.

I am actually using the "EQ Peaking" quite a lot. Though it is a nice bonus to have it in Zebra,(I can't think why a Parametric isn't a standard filter in most synth filter sections as it's much more useful than straight notch) However, I do think that it could be better.

For example; the (resonance) Q goes way too narrow and doesn't add anything useful beyond the first 35% of it's travel. After that it goes out of the sweet spot and just whistles like a Low pass on maximum rez. Most of the useful action seems to be going on in the very first part of the knob's travel.


Perhaps the knob should be more exponential to allow better access to the more useful areas?

The boost could be greater too...

Both band pass filters are similarly un-optimized (mainly on the resonance controls) as in they have large parameter areas which aren't very useful - if at all.

THe BPQBand is much quieter than the other filters. OK, Sure it's a band-pass as so passes less signal, but it's quite unbalanced compared to the other filters. There's also a low pass that is much louder than the others. (can't remember which one now...) I just had a look: AllRound and LP12

However... the normal filters feature Peaking/Shelving modes...

The peaking mode is good and as I mentioned, is getting caned The shelving filters are not much use (to me) as a dynamic filter. Statically, yes useful. But probably I'd use the EQ module for that job instead.

I'm thinking of an OSC mode where the points don't change their order when you drag them. That'd just do what you want

Sounds interesting. Just one modulatable point would be enough
for me...

Agreed, more shaper modes would be cool

IMO the shaper module is one area that could do with some serious improvement, which could expand the sonic possibilities considerably.

Maybe because I am also a guitarist and have spent half my life evaluating how different combinations of valve amps, overdrive pedals and fuzz boxes distort a signal, but I'm actually not than keen on the sound of the available modes.

They sound like the distorted signal is in parallel with the dry signal. They're ok, but just that: ok.

IMO they're punching well below their weight and not really up to the general high standard of the rest of the instrument.

Apart from a good sample redux mode as previously discussed, a couple of (or 3) more bog-standard clipping distortion modes would be really useful. But do make sure to have an extreme mode...

However, most of this has to wait until the general improvements come up...

Yeah, no problem. Bottom line is that I am happy with it as it is...

Lastly, I can't think what advantage there is in the tuning going up in half semitone steps (50 cent intervals) It just makes tuning oscillators more difficult because:

a) When I tune an oscillator most of the time I want it to be in semi-tones or octaves.

b) If I want to tune into the cracks between I'll use <shift> to fine tune.

Can we lose that function? It serves no purpose other than to make Zebra more difficult to use.

Again it's laziness on my part but the semitone tuning is already quite sensitive and difficult to set quickly. Perhaps ask some others to see if they feel that same?

Again, thanks for a great synth.

Cheers,
bagginz

[mkastrup]

About the semitone tuning thing, hmm well i'm mostly for we keep the .50 increment cause it helps me spot tunings for percussion like bongo, conga, toms, bells, cymbals, DX E-Piano etc much faster + all the weird side effects from Ring Mod and Sync when not directly on the note.

/Michael

[dmacintyre]

It would be nice to be able to scroll over the arpeggiator Step, Length, Gate, Voices labels themselves and adjust all steps simultaneously. E.g., if I want all the gates to be maximum, I have to go over all of them and set each one individually.

Would save a bit of time when doing arp "mod only" situations.

David

[bagginz]

I've seen it mentioned before in some other thread so this is another vote for having a seperate arpeggiator on/off button so that we get the option to select mono legato mode (non retriggering envelopes) while the arpeggiator is on.

Cheers,
bagginz

[Kebmaster]

Hi Baggins!!!! Met you at Zogg that time. Downloaded sum of your bass patches!!!! Glad youve discovered Zebra-Its my fave VSTi out there!!!

[Urs]

I've seen it mentioned before in some other thread so this is another vote for having a seperate arpeggiator on/off button so that we get the option to select mono legato mode (non retriggering envelopes) while the arpeggiator is on.

Hey, I'm not quite sure what you mean...

Do you mean being able to have the 2 step sequencers (ArpMod 1 & 2) while you can normally play poly/legato/retrigger?

Or do you mean having the Arpeggiator run normally, but keep the envelopes in Single Trigger so that the notes are "legatoed" over the sequence? - If so, there's a "Gate" parameter in the sequence which can be set to 6, looking like []> which is exactly that...

(I must confess that the UI for the Arp could be more comfortable tho... just like David posted before you)

Cheers,

Urs

[jupiter8]

Would'nt it be cool if you could convert a waveform from one mode (GeoBlend->SpectraBlend) ? Some conversions would of course be really hard to do as anything->GeoMorph but most others would'nt be so hard,would they ?

Or maybe you can already do that and i just have'nt found out how.

[jupiter8]

One of Kontakts killer features GUI wise is the visual feedback you get when you use modulation.



In this example if i were to modulate the volume for example the orange bar around the knobs moves accordingly to the modulation.
That way it is really easy to see what the modulation does. I sometimes struggle with hitting the endstops and this feature will show you in notime what the problem is. This probably be a downright bitch to implement but something to keep in mind should you ever do a GUI overhaul.

I hope the suggestions don't come across as nagging or whining. It's easy to take them that way. They're merely suggestions on how to make a really good product even really gooder.

[Urs]

Would'nt it be cool if you could convert a waveform from one mode (GeoBlend->SpectraBlend) ? Some conversions would of course be really hard to do as anything->GeoMorph but most others would'nt be so hard,would they ?

Or maybe you can already do that and i just have'nt found out how.

It's not 100% doable, that's why it's not in there.

The "Morph" waveforms have up to 1024 harmonics

SpectroBlend deals with 127 harmonics, Phase either 0° or 180°

GeoBlend has up to 63 harmonics, but variable phase

Latter two are upsampled with linear interpolation before bandlimiting so that additional harmonics might pop in.

One could render a Morph into a Blend, but not the other way round. One could convert between GeoBlend and SpectroBlend, but doing so forth and back will not result in the original. As this might occurr to some people as a bug then (which it is not), and lead to complaints, I'm reluctant to make it happen.

However, anyone wanna write a script that performs an FFT on GeoBlend and writes it back into the Wavetable?

Cheers,

Urs

[jupiter8]

Would'nt it be cool if you could convert a waveform from one mode (GeoBlend->SpectraBlend) ? Some conversions would of course be really hard to do as anything->GeoMorph but most others would'nt be so hard,would they ?

Or maybe you can already do that and i just have'nt found out how.

It's not 100% doable, that's why it's not in there.

The "Morph" waveforms have up to 1024 harmonics

SpectroBlend deals with 127 harmonics, Phase either 0° or 180°

GeoBlend has up to 63 harmonics, but variable phase

Latter two are upsampled with linear interpolation before bandlimiting so that additional harmonics might pop in.

One could render a Morph into a Blend, but not the other way round. One could convert between GeoBlend and SpectroBlend, but doing so forth and back will not result in the original. As this might occurr to some people as a bug then (which it is not), and lead to complaints, I'm reluctant to make it happen.

However, anyone wanna write a script that performs an FFT on GeoBlend and writes it back into the Wavetable?

Cheers,

Urs

Ok,i see. I thought they were just different representations of the same thing.Now i see they're not.

That scrpit would be a no brainer were it not for the compression/cryptation algorithm.
But writes it back into what ? A SpectroBlend ?

[Urs]

One of Kontakts killer features GUI wise is the visual feedback you get when you use modulation.

I've been thinking of this for a long time. AFAIK it's been done only in plugins that are cpu hogs. Which IMHO is no surprise...

I think that most people are used to hear what's going on, rather than having it visualized. The visuals would only work for slow modulations anyway, as the screen update rate can't deal with fast envelopes etc.

The best concept I've seen so far IMHO is the display in OhmForce's QuadFrohmage, where they have some sort of electro cardiogram that records the envelopes. I've been thinking of something like that, like a scope that can be configured to monitor changes within a modulation source. Or maybe up to 4 modulation sources.

However, with all modulations possible, I don't think that visual feedback on parameters will be any good. At least not in Zebra. Maybe in a simpler synth where you'd monitor like 3-5 "typical" modulation targets. Or, once everybody has QuadCore cpus and plugin guis can run on several cpus (which they currently can't).

Urs

[Urs]

That scrpit would be a no brainer were it not for the compression/cryptation algorithm.
But writes it back into what ? A SpectroBlend ?

Well, it would just be a .h2p preset that converts back and forth between the modes.

It would first read out which mode the osc is in:

2 - GeoBlend
3 - SpectroBlend

Then, upon GeoBlend it would perform an FFT. Could be done lazily by just multiplying with sine waves (just a DFT across all 64 possible harmonics), and then write out these values back into the wavetable. And set mode to 3.

Upon SpectroBlend it would just perform a couple of sin() operations multiplied by the volume of each slot. And set mode to 2.


I'll try myself. Why build in and debug an extra feature when you can just write a preset for it

[Dunc]

Couldn't you do the visual stuff on mouse over (the relevant control) only, or is that still too much of a hog ?

[Urs]

Couldn't you do the visual stuff on mouse over (the relevant control) only, or is that still too much of a hog ?

That might be possible...

Still, it would involve a *lot* of work. The gui does not only need a plain parameter channel to the dsp but also a secondary metering channel. All modules have to be modified to store intermediate values - after adding modulations. Some values then have to be converted back from scaled representations within the dsp part. To tell you the truth, it would be a shitload of work.

Well, I've been thinking about it for ages, but I havn't come up with a plan that's feasable to implement, neither cpuwise nor, uhm, effortwise...

[Dunc]

Oh well...maybe it's an improvement in time for Zebra 4