I like the sounds of these guys and I'm thinking about finally buying Zebra2. I'm curious how the other synths relate to Zebra2 though. I know that a lot of the code is shared, but is Zebra2 a superset of Filterscape VA the same way it is a superset of ZebraCM? What about TripleCheese?
I know ZebraCM is basically two zebra oscillators and a few filters/effects in a more fixed configuration. Is triplecheese similar but with the comb filters?
Also, a complete side question, is there some reason that almost no synths have 18db filters? Aren't 18db filters what helped make the 303 sound? Since so many people love the 303, why wouldn't more soft synths implement 18db filters?
how do Filterscape VA, Zebra, and TripleCheese relate?
-
- KVRist
- 140 posts since 28 Apr, 2004 from switzerland
to the 18db filters: i think i heave read that the tb 303 has 24db filters! but this filters are very different from machine to machine. the elctric parts of the 303 are making nice natural sound because they are not perfect stable.
look this for 303 emus with differnt synths :
http://www.deepsonic.ch/deep/htm/acid.php
... i prefer zebra2 + mfm2
look this for 303 emus with differnt synths :
http://www.deepsonic.ch/deep/htm/acid.php
... i prefer zebra2 + mfm2
-
- KVRAF
- 9217 posts since 23 Jul, 2002 from Pequot Lakes, MN
Nope, they're 18 dB/octave.psycox wrote:to the 18db filters: i think i heave read that the tb 303 has 24db filters! but this filters are very different from machine to machine. the elctric parts of the 303 are making nice natural sound because they are not perfect stable.
ew
A spectral heretic...
-
- u-he
- 30213 posts since 8 Aug, 2002 from Berlin
Can't say much about 18dB filters... I never felled compelled to implement one... but Zebra's XMF will eventually get an 18 dB setting once my (borrowed) Matrix 12 comes back from repair...
Shared code is mostly along the logical bits such as LFOs and ADSR envelopes. If you get used to parameter scaling etc. in one plugin, you'll be familiar with the others.
Modules that actually process audio though are a different story. FilterscapeVA for instance makes use of State Variable Filters and Biquad filters with a separate waveshaper. Zebra's filters all have non-linearities built into the algorithm, which often is more subtle but also more organic sounding. Zebra's XMF has both non-linearities within the algorithm as well as an output distortion stage. ZebraCM of course hasn't got the XMF...
Filterscape features FM between aribtrary waveforms - which was possible cpu-wise due to the hardwired concept. Zebra doesn't, because it isn't hardwired and the cpu hunger would be immense.
Triple Cheese and Zebra share the fundamental delay principle (allpass interpolated delay line), but while Zebra's comb filters have a couple of fixed routings that are optimized for certain tasks, Triple Cheese is a more open system, based on slightly less complex elements. Zebra's comb filter modules use up 4 delay lines each, 3C's use up to 2. Zebra has two of these modules, 3C has three.
The heart of FilterscapeVA is the modulatable equalizer. This is very easy to use. Recreating the possible effects in Zebra (or any other synth) is not trivial. It's basically achieved by a certain user interface which I can't fit into Zebra in a sufficient way...
So, there are always similarities between things, especially in terms of usage. Soundwise each plugin has got its own algorithms and specialities...
Cheers,
Urs
Shared code is mostly along the logical bits such as LFOs and ADSR envelopes. If you get used to parameter scaling etc. in one plugin, you'll be familiar with the others.
Modules that actually process audio though are a different story. FilterscapeVA for instance makes use of State Variable Filters and Biquad filters with a separate waveshaper. Zebra's filters all have non-linearities built into the algorithm, which often is more subtle but also more organic sounding. Zebra's XMF has both non-linearities within the algorithm as well as an output distortion stage. ZebraCM of course hasn't got the XMF...
Filterscape features FM between aribtrary waveforms - which was possible cpu-wise due to the hardwired concept. Zebra doesn't, because it isn't hardwired and the cpu hunger would be immense.
Triple Cheese and Zebra share the fundamental delay principle (allpass interpolated delay line), but while Zebra's comb filters have a couple of fixed routings that are optimized for certain tasks, Triple Cheese is a more open system, based on slightly less complex elements. Zebra's comb filter modules use up 4 delay lines each, 3C's use up to 2. Zebra has two of these modules, 3C has three.
The heart of FilterscapeVA is the modulatable equalizer. This is very easy to use. Recreating the possible effects in Zebra (or any other synth) is not trivial. It's basically achieved by a certain user interface which I can't fit into Zebra in a sufficient way...
So, there are always similarities between things, especially in terms of usage. Soundwise each plugin has got its own algorithms and specialities...
Cheers,
Urs
-
- u-he
- 30213 posts since 8 Aug, 2002 from Berlin
kodama wrote:But Zebra will become "Modules that actually process audio though" at some point, right?
Well, in my thinking modules are divided into 3 kinds:
- modules that are modulators
- modules that process audio
- envelope followers (they are mdoulators that process audio)
As such, everything that's on Zebra's Grids (oscillators, reverb, filters, delays, combs) are modules that process audio. Everything that's not on a Grid (envelopes, lfos, msegs, arpeggiator etc.) are modulators. Envelope followers are only built into Filterscape and FilterscapeQ6.
But of course I know what you mean. And of course I'm working on that. I'm even as far as to streamline my build/package/upload process so to also offer a VST fx version of Zebra
It's only the fact that building an fx version is really not elegant. Or maybe, that an fx version shouldn't be polyphonic but should feature a "balance of signals", whatever that would be.
That said, I'd *love* to do it. But I'm also an idealist. It must be done right & a bliss to use, not just a workaround.
Later,
Urs-
- KVRAF
- 1529 posts since 12 Jun, 2004 from Portland, OR
Urs wrote:
It's only the fact that building an fx version is really not elegant. Or maybe, that an fx version shouldn't be polyphonic but should feature a "balance of signals", whatever that would be.
That said, I'd *love* to do it. But I'm also an idealist. It must be done right & a bliss to use, not just a workaround.
Later,
Well, many hosts support inputs for instruments.
I guess I don't see the downside to instruments that already exist that have inputs.
I know part of it is that you are unhappy with the way VST (maybe AU) implement certain features, but I don't think most of us mind those limitations so much?
-
- KVRAF
- 1628 posts since 3 Oct, 2001 from Thailand
I think I've heard this somewhere, but is it true that the 18dB/Oct filter on 303 actually doesn't behave like 18dB/Oct all through the frequency range?
Peace, my friends. I'm not seeking arguments here.
-
- u-he
- 30213 posts since 8 Aug, 2002 from Berlin
In VSTland one simply has to add an fx version for each instrument. They differ in the unique plugin id and a single function call. The difference is close to non-existant, but the effort for compilation, packaging, testing, documentation and support is hilarious.kodama wrote:I know part of it is that you are unhappy with the way VST (maybe AU) implement certain features, but I don't think most of us mind those limitations so much?
Seriously, nothing speaks against input in normal VST 2.x instruments. It's just that historically no host supported it.
Now, what I could do is to unleash audio input in the VSTi and say "won't work in many hosts, maybe only in Live 7". That won't help much I guess...
I think we can meanwhile also agree that VST 3.0 won't happen within a couple of years, if at all.
Thus, my idea is to at one point figure out what an fx version of Zebra would need, and which parts could be left away. It certainly doesn't need exaggerated polyphony and too many oscillators. Instead, maybe it needs a monophonic design with both synthesis and effects in a single Grid. A monophonic design has the advantage that a lot of hassle that comes with lfo and envelope considerations would be gone. It would be a Zebra that lends itself more towards uses like "vintage modular system". Without voices, a ton of new possibilities arise because some tight dependencies can be broken up.
Hmmm...
Urs-
- KVRist
- 221 posts since 2 Oct, 2006
If you're considering that, you could at least unleash the multi-in AU, which would work on a variety of hosts, if no windows onesUrs wrote:Now, what I could do is to unleash audio input in the VSTi and say "won't work in many hosts, maybe only in Live 7". That won't help much I guess...