Questions about Zebra Structure
-
- KVRist
- 457 posts since 30 Jun, 2003 from cinci, oh
Now, suppose I want to mix varying amounts of osc1, osc2, and ring-mod osc1*osc2 prior to the filter. What would that look like ? I cannot get the ring- mod thingy.
-
- KVRist
- 457 posts since 30 Jun, 2003 from cinci, oh
Interesting find; really thru me for a loop...
The max morph speed of the Spectromorph is limited by its resolution setting. So, if you have spectromorph resolution cranked way done (to reduce CPU hit), then placing an LFO on OSC "wave" does nothing. An ADSR on "wave" still works, but decay rate is then limited by "resolution" and not ADSR "decay".
Is it supposed to work this way?
The max morph speed of the Spectromorph is limited by its resolution setting. So, if you have spectromorph resolution cranked way done (to reduce CPU hit), then placing an LFO on OSC "wave" does nothing. An ADSR on "wave" still works, but decay rate is then limited by "resolution" and not ADSR "decay".
Is it supposed to work this way?
- u-he
- 30208 posts since 8 Aug, 2002 from Berlin
Most likely, yes.pummel wrote:Interesting find; really thru me for a loop...
The max morph speed of the Spectromorph is limited by its resolution setting. So, if you have spectromorph resolution cranked way done (to reduce CPU hit), then placing an LFO on OSC "wave" does nothing. An ADSR on "wave" still works, but decay rate is then limited by "resolution" and not ADSR "decay".
Is it supposed to work this way?
Don't forget that LFOs are bipolar modulators. Thus you have to set osc waveform and modulation depth to 7 or so to grab its modulation on two sides.
Another thing, putting resolution to minimum means that the waveform modulations by envelope or LFO are sampled only in intervals of a bit more than 4 seconds or so. So, if the LFO is set to anything like 1/4, hence faster than the osc needs it, it can happen that the LFO is always sampled at close to same values or at non-predictable phases (this is actually the same effect as aliasing).
Low resolution settings are considered, uhm, cpu savers for static sounds that need no modulation or for anything that benefits from the slowliness. You'd normally never use values below 4. However, I got some interesting results by modulating the resolution in the ModMatrix
Cheers,
- u-he
- 30208 posts since 8 Aug, 2002 from Berlin
Oh. That's tricky, indeed. And you have a good point here!pummel wrote:Now, suppose I want to mix varying amounts of osc1, osc2, and ring-mod osc1*osc2 prior to the filter. What would that look like ? I cannot get the ring- mod thingy.
The main problem is, the volume of the ringmod depends on the volume of the oscillators. Hence, to have independent control over RM volume, both oscillators must have static gain. So, there's gotta be a setup in the Grid that allows for independent control over oscillators and ringmodulation.
Now, look at this:

OSCs 1 + 2 run in parallel and have a fixed volume. Ringmodulator takes input from OSC1 and sidechain from OSC2.
ChannelMix1 mixes OSC1 against OSC2.
ChannelMix2 mixes the pure oscillators against ringmodulation.
Depending on what you want to do, this might be a sufficient and elegant setup already. Due to usage of ChannelMix modules, the output level should be pretty constant, regardless of mixer settings.
Unfortunately you (currently) have to waste two ModMatrix slots to modulate the ChannelMixes' Mix parameter.
If this is not what you want to do, i.e. if you want to exactly recreate the structure of a Virus, I might have to think about some stuff. In that case there might have to be a mode for the ChannelMix modules that does not mix input *against* sidechain, but add the sidechain in adjustable level to the input.
Cheers,
- KVRAF
- 8700 posts since 9 Jan, 2004 from leroyaumeuni
so..
how do I patch osc1 and osc2 to make the zebra scream like a bastard?

how do I patch osc1 and osc2 to make the zebra scream like a bastard?
My other host is Bruce Forsyth
- u-he
- 30208 posts since 8 Aug, 2002 from Berlin
Route stereo detuned OSCs into a filter using "LP Allround" with Resonance upped quite a bit and a waveshaper with Edge set to 90.spaceman wrote:so..
how do I patch osc1 and osc2 to make the zebra scream like a bastard?
Screams more like Zaphod Beeblebrox then
- KVRAF
- 8700 posts since 9 Jan, 2004 from leroyaumeuni
My other host is Bruce Forsyth
-
- KVRist
- 457 posts since 30 Jun, 2003 from cinci, oh
Thanks for the mod matrix diagram !!
But how did you get two inputs into the ringmod widget; thats what I am missing. On the ringmod's RMB droplist, I can pick "input1" or "input2", not both. I saw some words in your description about a sidechain; where is that ?
But how did you get two inputs into the ringmod widget; thats what I am missing. On the ringmod's RMB droplist, I can pick "input1" or "input2", not both. I saw some words in your description about a sidechain; where is that ?
- u-he
- 30208 posts since 8 Aug, 2002 from Berlin
That's one of those stupid bugs. Once you load a preset, there'll be options for "sidechain", which is just a secondary input...pummel wrote:Thanks for the mod matrix diagram !!
But how did you get two inputs into the ringmod widget; thats what I am missing. On the ringmod's RMB droplist, I can pick "input1" or "input2", not both. I saw some words in your description about a sidechain; where is that ?
Cheers,
-
- KVRist
- 321 posts since 26 Dec, 2002 from Alaska
Hello Zebra2 enthusiasts,
(1) Can the Zebra2 be programmed to do Totem sounds, described here:
http://www.kvraudio.com/forum/viewtopic ... ight=totem
(If yes, example please)
(2) Is there a way to assign an LFO to the pan? So I can get panning sweeps.
Thanks in advance,
YBaCuO
(1) Can the Zebra2 be programmed to do Totem sounds, described here:
http://www.kvraudio.com/forum/viewtopic ... ight=totem
(If yes, example please)
(2) Is there a way to assign an LFO to the pan? So I can get panning sweeps.
Thanks in advance,
YBaCuO
- u-he
- 30208 posts since 8 Aug, 2002 from Berlin
(1) Not exactly the way Tim described it. But if we had an audio example I'm pretty sure someone could cook the same thing in Z2.
(2) Yes. Each oscillator, noise generator, fm oscillator and comb filter has its own modulatable pan. In addition there are mixing units that also do stereo balance and of course modulatable pans in the main mixer. You can use LFOs or whatever you want to create pan sweeps.
Cheers,
Urs
(2) Yes. Each oscillator, noise generator, fm oscillator and comb filter has its own modulatable pan. In addition there are mixing units that also do stereo balance and of course modulatable pans in the main mixer. You can use LFOs or whatever you want to create pan sweeps.
Cheers,
-
- KVRAF
- 2875 posts since 28 Jan, 2004 from Da Nang, Vietnam
I'm having trouble with this too. Do you mean that I have to load a preset that already has sidechain enabled somewhere or just any preset? I've tried loading a few simple presets but that doesn't seem to make any difference.Urs wrote:That's one of those stupid bugs. Once you load a preset, there'll be options for "sidechain", which is just a secondary input...pummel wrote:Thanks for the mod matrix diagram !!
But how did you get two inputs into the ringmod widget; thats what I am missing. On the ringmod's RMB droplist, I can pick "input1" or "input2", not both. I saw some words in your description about a sidechain; where is that ?
-
- KVRian
- 697 posts since 2 Dec, 2002 from northern germany
- u-he
- 30208 posts since 8 Aug, 2002 from Berlin
Yep!luCiPHer wrote:load a preset that has a mixer and/or ringmod in it, that will enable sidechaining.
stupid bug, but i'm sure urs will work it out soon
Loading the "XY FM Synth1" preset will always enable side chaining in the Grid.
It's really a stupid bug that needs a fix soon.
-
- KVRist
- 457 posts since 30 Jun, 2003 from cinci, oh
I found one of luciphers patches with ringmod enabled. Using that as the "initial-eyes" patch, I can create instances of objects with "dual" inputs.
BTW, why are we limited to two inputs per object? There are four buffers; why not 0...4 inputs per entity?
BTW, why are we limited to two inputs per object? There are four buffers; why not 0...4 inputs per entity?

