When is Urs going to finally release that Diva MIDI controller? Kickstarter?

[knowix]

Hadn't seen that Powermate thing before. Definitely going to investigate it further.

[xh3rv]

Instead of U-he being required to code something, every single hardware maker should recode their controllers with additional logic to be able to handle a condition..

I know this was meant sarcastically, but I actually agree

I'd argue the MIDI mapping scheme for DIVA/ACE/Bazille is innovative - not magically, but practically and plausibly. It doesn't use a proprietary or arbitrary transport layer like AutoMap, or some host-specific hardware, or anything that's OSC but not MIDI. Rather it could be said to implement a higher level abstraction on top of the ubiquitous MIDI, using it as transport layer. There's a sort of logical type system for CC assignments (more types are imaginable, but the design pattern is the point) and paging. Both of those features benefit from being inside the plug-in. Relying on VST/AU parameters means only seeing their parameter types, which is limiting. Some GUI cues for paging are also possible inside the plug. There are compelling features here, from a scheme that could applied in any plug.

As a user I want MIDI controller hardware to get cooler and more capable along these lines. It'd be awesome!

[PAK]

As a user I want MIDI controller hardware to get cooler and more capable along these lines. It'd be awesome!

Yep, agreed. You make some decent points too. Though "higher level abstraction on top of the ubiquitous MIDI" pretty much describes what stuff like Automap does

Yes, there's definitely advantages to a plugin going that route itself, and removing the need for these 3rd party layers. Some issues are removed, and those which remain are at least put more directly under the control of the plugin dev. I think it's good Urs is obviously thinking down these lines about how to solve things for U-He's plugins.

The problem, of course, is that it takes two to tango. U-He can provide all these things, but it still takes hardware to make the most of it. And, much as though it could be said "it's all just MIDI" it doesn't change the fact that important basics, like defining the resolution of controls in a flexible manner, still needs hardware cooperation.

So who will provide it and when? If anyone expects the "If we build it, they will come" philosophy to work then maybe you've not been paying much attention to the tendencies of companies operating in this area. Plus, even if you gain support, it's still an U-He universe limited to U-He plugins.

Ok.. It's just MIDI.. Other companies are free to do their own solutions.. But are they going to? Does that question even have to be asked? That's why systems like Automap will still be required going forward, to deliver these features to a wider range of plugins..

Automap has flaws.. The dll wrapping, the page system, and poor resolution of non-rotary controls. But fix those things and it's doing most of what's required to deliver an excellent experience. Similarily, the Nektar stuff seems like it might be a very good solution too..

The appeal of these solutions is that they're not just trying to provide a solution for U-He plugins, but for everything. That's why I think it's important to try to keep host automation based control well supported (Yes, the shocking truth comes out.. the principle of good automation support is what's important, and not so much an Octave issue )

People still have to live in the reality of what's around and available to them for now. Diva has over 300 controls to automate (though it chooses not to expose 10% of these to host Automation). What's the best solution to automate all that? People already won't program stuff like Automap, and so we're supposed to believe they're now going to deal with setting it all up over MIDI? But, hey, maybe I under-estimate the number of Sysex masochists out there.

Of course, the real issue is MIDI itself has outlived it's welcome by a good 20 years. Much as though Dave Smith might like to think it stands testament to the good job he and Roland made, and how the joy of parameter smoothing means you'll never need more than 7 bits for anything, it's an even bigger testament to an industry full of companies utterly unable to cooperate with each other.

Until that day (n)ever comes, I'd like to keep good support for host automation. Companies like U-He should do their own things, by all means. Just provide good host automation experiences too, please.

[Uncle E]

So who will provide it and when? If anyone expects the "If we build it, they will come" philosophy to work then maybe you've not been paying much attention to the tendencies of companies operating in this area. Plus, even if you gain support, it's still an U-He universe limited to U-He plugins.

I know the CEO of CME personally and could probably work with him to make u-he specific controllers. As long as we're the ones doing all the work and taking all the financial responsibility, I think he'd be fine with doing something like that. My only concern is how feasible it really is given how modular Zebra and DIVA are. It seems to me that an iPad app would be better suited for these.

I can imagine a hardware ACE controller working out well and I bet people would get a kick out of using the patch cables. It'd be even better if the cables could send out control voltage to analog modular synths.

[ChromeGhost]

So who will provide it and when? If anyone expects the "If we build it, they will come" philosophy to work then maybe you've not been paying much attention to the tendencies of companies operating in this area. Plus, even if you gain support, it's still an U-He universe limited to U-He plugins.

I know the CEO of CME personally and could probably work with him to make u-he specific controllers. As long as we're the ones doing all the work and taking all the financial responsibility, I think he'd be fine with doing something like that. My only concern is how feasible it really is given how modular Zebra and DIVA are. It seems to me that an iPad app would be better suited for these.

I can imagine a hardware ACE controller working out well and I bet people would get a kick out of using the patch cables. It'd be even better if the cables could send out control voltage to analog modular synths.

I definitely would! If I could send out CV with it to analog it would be golden. What do you think about the controller at synthprojects?

[Urs]

Hehehe, you can't send out CVs from cables that route a polyphonic synth. Unless you use 16-core cables for 16 voices.

Since cables can be MIDI learned in ACE/Bazille, it's possible to do a controller. The controller logic however requires a bit of Arduino competence, one must use demultiplexers to periodically evaluate a matrix of outputs vs. inputs. One could also build a Diva controller using patch cables instead of the not-so-analogue modulation source drop downs.

- Urs

[Uncle E]

What do you think about the controller at synthprojects?

It looks amazing to me. I want it badly. If he completes it, I'd love to talk with him about making a (relatively) mass market version of it.

[Uncle E]

Hehehe, you can't send out CVs from cables that route a polyphonic synth. Unless you use 16-core cables for 16 voices.

Hah! Good point. Perhaps I could convince you to make a mono synth version for such a project, eh?

One could also build a Diva controller using patch cables instead of the not-so-analogue modulation source drop downs.

Or maybe I could convince to make a patch cable DIVA? People would go nuts over that.

[Urs]

Or maybe I could convince to make a patch cable DIVA? People would go nuts over that.

Well, the first beta had a patch field in the Modify tab. I'm not sure how people took it, it looked a bit messy.

A hardware Diva with patch cord would be great, but I can't really divert any of my time to such a project atm.

[Uncle E]

A hardware Diva with patch cord would be great, but I can't really divert any of my time to such a project atm.

Yeah, I know. It's fun to dream of these things but I'm not sure the market's there for them. A u-he controller app would certainly be both easier and more successful, despite being far less cool.

[PAK]

My only concern is how feasible it really is given how modular Zebra and DIVA are. It seems to me that an iPad app would be better suited for these.

Zebra? What's HZ using for it lately? Even a tablet seems a bit small for that purpose. Larger 4k display formats might help, but they're pretty expensive for now. The thing is, you could do all that, but I'm not sure you'd end up with something significantly better than the mouse experience at the other side of it.

Touch screens are good for things like X/Y control. But, IMO, they're often overrated even just in terms of replacing a mouse, and certainly for replacing physical controls. You're still interfacing with a slab of glass in the end, and it lacks tactility. Maybe one day the haptic feedback will get good enough to fake things convincingly

Synths like Diva could definitely be done. But, as said previously, it'd take a large surface to do well. Large being at least a dozen sliders, upwards of 40 knobs, and a lot of buttons and logic.

Unless you're going to go down the path of CV modulation I fail to see the appeal of actual cable patch points. A virtual button patch system, using full-colour LED back-lights for the many buttons, would work better for the purpose. The bonus being it'd also provide major bling factor ala Maschine's pads

The biggest issue is both cost and market size. A market is there, but many seem unprepared to spend money for a good hardware controller experience. Yet some will pay far more money for other hardware.

I don't get why computer users aren't more realistic about that stuff. Look at DSI announcing their Pro Two just yesterday. It must have, what, about 50 knobs on it?! And that's a dedicated interface which had the benefit of being designed as an integral part of the development process. You just can't provide the amount of controls these generic controllers are providing and expect a good experience..

[Echoes in the Attic]

PAK I agree with pretty much everything you're saying. It's great that Urs has a midi system that works well with midi, but the truth, hardware control is going more and more dependent on host automation IDs rather than midi cc's. And it's much better that way. Automap, Kore, Maschine, Nektar Panorama, Ableton Devices like the APC40 and Push or any other instant mapping controller, Bitwig controllers. They all map based on automation ID. And they all work simultaneously mapping to the same automation IDs in two way communication without any assigning anything. For example I can have a Novation SL controlling a plug-in that's loaded in Maschine. Knobs on both hardware devices could control parameters that update the other as well as a Push or other device controlling the same parameters via Ableton or Bitwig device being selected. There no chance of accidentally controlling two things because the same midi cc is assigned or any of that nonsense. Things are just mapped to parameters by their IDs that the developer assigned to them. You automate the control directly by just moving it. No midi CC layer in between.

The perfect system already exists, but unfortunately it's split between two systems - Novation Automap and Nektar Panorama. Automap has great control over the signal (step, range etc.) that is sent but has to be wrapped to work in any host and has no page names or way of getting to pages quickly (only page up and down). Nektar on the other hand has a nice page selector with page names for different control groups so you can scroll to a page quick and know what's going to be on the page, and also doesn't need a wrapper because they support hosts individually. Unfortunately that relies on the host capabilities (for example Bitwig currently not being able to use the whole controler, just 8 encoders for plaug-ins) and there is less flexibility in the user assigning behavior of a knob. If these two systems were merged there would be nothing left to improve in terms of hardware mapping and it wouldn't matter how big a plug-in is. But in theory with existing methods, everything we need is there using automation IDs (as long as developers actually expose their parameters!).

[aMUSEd]

Well you've pretty much described Kore but NI killed it off. It has the control over the signal including range and multi assignments to the same knob or button (each with different ranges) and easy page naming and movement between them using the controller.

[Echoes in the Attic]

Well you've pretty much described Kore but NI killed it off.

You're right, Kore was good. The SL in practice for large plug-ins is much better though because 24 continuous controls with 24 buttons is much more useful than 8 knobs and 8 buttons. Reduces page scrolling a hell of a lot.

[rod_zero]

The market is really missing a dedicated midi controller for soft synths, with a knob/fader layour that makes sense: an osc, filter, envelopes and modulation sections.

Rob Papen posted this on FB this year:



Somthing like it would be great, specially with led rings or led as those in the Softtube Console 1 (great concepto too, but price ....)

Really I wonder why nobody has done something like this.