Zebra3 Info

[wilx]

Zebra is so much more capable than most synths, it would be impossible to implement visual feedback in the same way. But all I'm hoping for is a few common-sense, well designed bits of visual feedback to make programming more efficient and intuitive.

BTW for me, the ideal modulation matrix would look very similar to a file browser. Thoughts?

Beyond the obvious synths, I quite like the way the mod matrix is handled in Sugar Bytes Factory. As mentioned, I think Logic's Alchemy handles modulation brilliantly. I can't imagine how a file browser-esque implementation would look, or how it would be better than Zebra 2's current matrix. Intrigued!

But it sounds like we're broadly in agreement re no unnecessary pyrotechnics, just some subtle tweaks for ease of use and intuitive programming.

Re the example about the slot1 vs slot 16 OSC modulation scenario - have I completely missed something there, or am I correct in thinking it would make more sense to see the waveform at position 16 if you're hearing it at position 16 - regardless of whether you're using the mouse or a modulation source to select that position?

[Guenon]

Re the example about the slot1 vs slot 16 OSC modulation scenario - have I completely missed something there, or am I correct in thinking it would make more sense to see the waveform at position 16 if you're hearing it at position 16 - regardless of whether you're using the mouse or a modulation source to select that position?

Only if you aren't currently actively editing a particular slot, and playing the sound repeatedly while doing it. It would get quite annoying if the slot kept changing based on modulation even though I just wanted to play the sound as a reference for the change I'm making . But yep, it could of course be smart like that, as in, when you have such an element in focus, it's in "editing mode" and isn't a moving target while editing it.

[wilx]

Re the example about the slot1 vs slot 16 OSC modulation scenario - have I completely missed something there, or am I correct in thinking it would make more sense to see the waveform at position 16 if you're hearing it at position 16 - regardless of whether you're using the mouse or a modulation source to select that position?

Only if you aren't currently actively editing a particular slot, and playing the sound repeatedly while doing it. It would get quite annoying if the slot kept changing based on modulation even though I just wanted to play the sound as a reference for the change I'm making . But yep, it could of course be smart like that, as in, when you have such an element in focus, it's in "editing mode" and isn't a moving target while editing it.

Good point - the visualisation would work better as a subtle highlight of the bottom preview rectangles, rather than having the editor window itself change. That way you could see exactly what is being modulated, and to what value, but it wouldn't get in the way of anything you're doing with the synth.

[Guenon]

I'd like subtle visualizations like that quite a lot, actually, to track the sound formation of a running patch, yeah. Especially as things can get pretty complex with a lot of overlapping modulation and so on.

Could be turned on/off with a little switch for saving that last bit of CPU, and allowing it to be used only when actually relevant.

[Urs]

It is not an easy task. Most synths that have such kind of visualization have a limited set of modulation targets. With Massive it's quite obvious by the design of the knobs. With more complex examples the visualization actually starts to be an impediment to sound design. All good examples are very limited in their options - in contrast to Zebra:

Our engine allows for the modulation of virtually any knob. Not 20 or 50, but a number somewhere in the vicinity of 1000 in Zebra2. It furthermore allows for N:M associations where any modulator can be assigned to any such knob, and any knob can receive modulation from any such source.

The house keeping under the hood is done with a few very simple mechanisms that are based on a fundamental principle of object orientated programming, which is, they know nothing about each other. It's a dumb system, it does not require sophistication and it does not require developer effort. This ensures that if it works in one case, it will work in any other - it allows us to develop burden-free, it just works. I'm pretty sure that any such synth who has complex visualization has some sophisticated, similarly complex mechanism under the hood - one which deploys full awareness of any parameter bearing entity with any parameter modulating one. That would be opposite of what we do. I don't know for sure, but I reckon this is why they don't (can't?) do what we do - and, in terms of visualization, vice versa.

Now, we can of course come up with a similarly simple way to visualize parameter states in real time. For each parameter we simply need a feedback parameter for the UI. So for 1000 or so modulation target parameters we need a corresponding feedback parameter. Sounds overkill? - It surely is, and it surely costs considerable CPU.

What I can imagine instead is showing modulation depths "on parameter". So when a ModMatrix is set to a target, the depth of this very matrix is visually represented on the target parameter. But then we can do only one at a time, simply also because I won't have people stack 10 or more rings of visualization around each knob, if you take into account N:M assignments.

However, if it's one matrix at a time for just a visualization of the modulation depth, I think it becomes moot to even think about it

Thing is, the human brain is fabulous. It like those visual hints, but Zebra's preset design history IMHO has shown that it doesn't really need them.

[wilx]

I'd like subtle visualizations like that quite a lot, actually, to track the sound formation of a running patch, yeah. Especially as things can get pretty complex with a lot of overlapping modulation and so on.

Could be turned on/off with a little switch for saving that last bit of CPU, and allowing it to be used only when actually relevant.

[wilx]

It is not an easy task. Most synths that have such kind of visualization have a limited set of modulation targets. With Massive it's quite obvious by the design of the knobs. With more complex examples the visualization actually starts to be an impediment to sound design.

Thing is, the human brain is fabulous. It like those visual hints, but Zebra's preset design history IMHO has shown that it doesn't really need them.

Thanks for the reply Urs. I totally understand - as I've said before. I'm just trying to figure out of there's a way of making Zebra 3 a tiny bit more intuitive for people who aren't as adept at programming such a beast. Certain things - like my example of seeing the wave warp appear to be at position 1 when it's actually at position 16 - seem totally counter-intuitive to me, and it would be good to see a design which acknowledges and improves that sort of visual feedback.

I have a feeling many people would be happy to see a small amount more intuitive visual feedback in the zebra 3 GUI. But at the same time I understand Zebra is a different kind of synth and (a subtler version of) Alchemy-style visualisation wouldn't necessarily translate.

[Guenon]

Thing is, the human brain is fabulous. It like those visual hints, but Zebra's preset design history IMHO has shown that it doesn't really need them.

I definitely tend to agree

Instead of battling with an all-encompassing visualization design that tries to show visual hints for all modulation, I think some carefully chosen key elements of Zebra could benefit from this, though. Indeed showing a visual cue for the currently playing wavetable frame, for example.

[EvilDragon]

Adding an oscilloscope to which you could send any source (be it a module like oscillator, filter, ringmod, or a modulator, like LFO, a particular mod matrix slot, etc.) should be quite helpful.

[wilx]

Adding an oscilloscope to which you could send any source (be it a module like oscillator, filter, ringmod, or a modulator, like LFO, a particular mod matrix slot, etc.) should be quite helpful.

An oscilloscope would be nice, definitely. But it wouldn't necessarily help with those times when you're hearing a timbral shift in the patch but can't tell exactly what parameter is being modulated by what modulater, and by how much. Especially if the visual feedback is misleading, like the wave warp example.

[Funkybot's Evil Twin]

Adding an oscilloscope to which you could send any source (be it a module like oscillator, filter, ringmod, or a modulator, like LFO, a particular mod matrix slot, etc.) should be quite helpful.

An oscilloscope would be nice, definitely. But it wouldn't necessarily help with those times when you're hearing a timbral shift in the patch but can't tell exactly what parameter is being modulated by what modulater, and by how much. Especially if the visual feedback is misleading, like the wave warp example.

We need an AI digital assistant for these kinds of tasks. "Ursula, which modulation is making that wub-wub-wub-ker-RRRAAANGGG sound?"

Not sure if these even makes sense, but what if the Matrix view was expanded to show all the modulations in patch and their positions? I'm thinking of a cross between Zebra2's current Matrix and Omnisphere's
Mod Matrix Zoom pages. So rather than a Massive style system where you're looking at the knob for feedback, you'd go looking on the matrix page and see one row for every modulator in a patch. And like Omnisphere, Z3 could even show a visual animation of the speed/position of the modulation to kind of help clue you in as to what you're hearing. Wouldn't be sexy, but could be effective.

[wilx]

What I can imagine instead is showing modulation depths "on parameter". So when a ModMatrix is set to a target, the depth of this very matrix is visually represented on the target parameter. But then we can do only one at a time, simply also because I won't have people stack 10 or more rings of visualization around each knob, if you take into account N:M assignments.

However, if it's one matrix at a time for just a visualization of the modulation depth, I think it becomes moot to even think about it

Reading back, I actually think this would be pretty useful - at least for constructing or deconstructing a patch, since you only tend to concentrate on one parameter at a time.

Funkybot's idea of having the visual feedback in the matrix itself (therefore easy to hide away) could work also.

Anyway, I'll stop banging on about this and wait to see what wonderful stuff Zebra 3 eventually brings!

[RobGee]

I’ve never had an issue with it but surely the easiest and least cpu intensive way to show all modulation assignments (those assigned within modules and also the modulation matrix) is to have a separate page/tab listing them.

[Funkybot's Evil Twin]

I’ve never had an issue with it but surely the easiest and least cpu intensive way to show all modulation assignments (those assigned within modules and also the modulation matrix) is to have a separate page/tab listing them.

A modulation matrix with everything listed out could also allow for easily muting, soloing, and inverting each modulation. Like, "hey, is this beating I'm hearing this particular assignment, let me mute it...no, not it...maybe it's this one...let me hold down a note and solo each modulation until I find it."

[Urs]

I’ve never had an issue with it but surely the easiest and least cpu intensive way to show all modulation assignments (those assigned within modules and also the modulation matrix) is to have a separate page/tab listing them.

What do you think of the modulation target display on the expanded MSEG page in Zebra 2.8?

This control is our answer to frequent requests for "a way to show me what this LFO/Env/... targets", testing it with the MSEGs. It's being discussed controversially internally as "overengineered", "great stuff", "not necessary" or "not as great as expected". However, we barely have any feedback from people out there.