Zebra overrated? (+ feature request)

[jjensson33]

Once upon a time there was a thread called "the most overrated softsynths" or something similar. I suggested Zebra, because of all the audio samples i had heard (available at that time). A bit later i decided to try it myself and compared it to some of my favorite softsynths and... i was impressed! It where simply the audio samples that showed Zebra in a strange light, but i found out that all the other sounds i craved are possible with Zebra - it can replace a Moog Modular, a Tera, a Cube, for the sounds that i wanted to make anyway... And the sound quality is superb! Congratulations, Urs!

Needless to say, i'm planning to byu this Zebra-Monster in the next months and have some requests to make, wich could make Zebra a better workhorse, in my opinion.

1) Better handling of partials (identification & selecting) + partial detuning (?) --- It's extremely difficult to tune selected partials to exact values. Say i need the 3rd partial to 88%, the 5th to 52%, the 6th to 25%, the 9th to 64% etc (when i have an exact additive wave to program) - i had to use the integrated WindowsXP magnifying glass (sic!) to manipulate the partials. And each time adjusting i had to painfully count the bars...

If it's too much work to change the GUI, why not make a possibility to write the partial values with the text editor directly into the h2p-file? As far as i see, wave data is encrypted ATM, but it should be easily possible, am i right?

Partial detuning would be nice for ensemble-like or inharmonic sounds. But i don't know if it's not too hard to implement.

2) 128 pole filter w/ free curve-drawing (formant filter?) - like in Virsyn Tera. That's a feature i'm missing very much. Not only one can make interesting formant effects, but also simulate that crazy digital wavetable sweeps!

3) Missing modulation sources (sorry, i spotted these in Tera too, and i think they can be very helpful): Key alternate (can be used to get a different sound on every second note - for natural-instruments-like articulations like bowing or plucking/picking etc) and Key random (creates a random control signal on every note - OK, this is for fun )

OK, Urs, i know you are working your a** off, sorry for more requests, but it would be absolutely great if these features could be implemented in future Zebra updates (maybe Zebra3?). I would like to hear your quick opinion and hope that others would find these features as useful as i do. Everyone is welcome to comment.

Thanks for taking the time to read this,
jj

[Bubbamusic]

You just can't make music with it the way it is? It's already one of the best. Make your own "dreamsynth".

Peace: bubba

[botkiller]

Well, developers have to listen to their users either way, weather or not they take their input into consideration is up to them.

[zircon]

You can already do the wavetable sweeps by just drawing in (or loading) the 16-wave table of your choice, increasing the resolution, and modulating the Wave knob on the osc of your choice. With 4 different oscs that's a lot of potential for wavetable fun. There's always the formant filter as well.

I agree a key alternate modulation would be cool. I have a couple acoustic emulations in my preset banks that would probably sound better with that feature.

[Bubbamusic]

No offense intended - it just seems to me that Zebra already may be the most versatile soft-synth that there is. Most synths specialize in one thing or do a few things very well. JJenson33 has described Cube's specialties pretty well. I guess that the better you do your job, the more special requests you will get.

Peace: bubba

[ATS]

so you said zebra was overrated even before you tried it?

[djsubject]

so you said zebra was overrated even before you tried it?

Welcome to KvR

[Urs]

Yo,

I mentioned it somewhere... the waveform/spectrum editor will be improved...

You can already create partials from text... http://www.kvraudio.com/forum/viewtopic.php?t=166487 - Instead of the binary representation one can create (or modify) each waveform/spectrum programatically. It's just not documented due to some stuff that's higher priority at the moment...

Partial detune is not possible with the way the osc engine works. I have no clue how Harry and Ben did it in an efficient way...


Regarding a 128 pole filter... sure, there were ideas, but there wasn't anything convincing. Instead I went for the osc fx which are easier to handle and IMHO offer more dramatic results. I think that formant filters mostly *thin out* the sound... however, quite dramatic sweeps can be done in the "SpectroMorph" oscillator mode.

A spectral filter as an extra module is naturally impossible as it had to be based upon FFT and thus introduce too much latency...


However, it's an extensible system. We'll see what happens...


For stuff like Key alternate etc... d'oh... well, Key Random is on the list. Other stuff can probably be added once the scripting engine can be used to construct modulators...

Urs

[Urs]

Well, developers have to listen to their users either way, weather or not they take their input into consideration is up to them.

?

[jjensson33]

You can already do the wavetable sweeps by just drawing in (or loading) the 16-wave table of your choice, increasing the resolution, and modulating the Wave knob on the osc of your choice. With 4 different oscs that's a lot of potential for wavetable fun. There's always the formant filter as well.

Yeah, that will work of course, i just don't know if the results will be as dramatic as in Tera. I have to make some experiments again. It's over 3 months ago i tried Zebra...
jj

[jjensson33]

Yo,

I mentioned it somewhere... the waveform/spectrum editor will be improved...

You can already create partials from text... http://www.kvraudio.com/forum/viewtopic.php?t=166487 - Instead of the binary representation one can create (or modify) each waveform/spectrum programatically. It's just not documented due to some stuff that's higher priority at the moment...

Partial detune is not possible with the way the osc engine works. I have no clue how Harry and Ben did it in an efficient way...


Regarding a 128 pole filter... sure, there were ideas, but there wasn't anything convincing. Instead I went for the osc fx which are easier to handle and IMHO offer more dramatic results. I think that formant filters mostly *thin out* the sound... however, quite dramatic sweeps can be done in the "SpectroMorph" oscillator mode.

A spectral filter as an extra module is naturally impossible as it had to be based upon FFT and thus introduce too much latency...


However, it's an extensible system. We'll see what happens...


For stuff like Key alternate etc... d'oh... well, Key Random is on the list. Other stuff can probably be added once the scripting engine can be used to construct modulators...

Urs

Thank you, Urs. That was very informative. And i'm happy to hear about the waveform/spectrum editor. Will try your suggestions later.

Best regards,
jj

[Howard]

so you said zebra was overrated even before you tried it?

Maybe JJ's not the only one who does that. If so, it would mean Z2 could do with a few really good audio demos on the website!

[Urs]

so you said zebra was overrated even before you tried it?

Maybe JJ's not the only one who does that. If so, it would mean Z2 could do with a few really good audio demos on the website!

I like the demos. We have to live with that.

[wrench45us]

It's over 3 months ago i tried Zebra...

Will try our suggestions later.

I would anticipate a time lag here of approximately 3 months.

[jjensson33]

so you said zebra was overrated even before you tried it?

Maybe JJ's not the only one who does that. If so, it would mean Z2 could do with a few really good audio demos on the website!

I like the demos. We have to live with that.

It's not the demos per se, but more the range of sounds they represent. At the time i was disappointed by Zebra i never heard the powerful ("druckvoll") sounds i did myself with the demo. But wait a sec, later i heard very good sound-demos, like that new "old school" sound library (forgot the name of the author).
jj