A dream hardware controller for soft synths?

[Funkybot's Evil Twin]

Sounds like an interesting product. I completed the survey. I'd be very interested in something like this if it offered a ton of knobs/faders/buttons, and was housed in a good keyboard enclosure. I think there's already a fair number of controllers on the budget end of the spectrum but the best options at the higher end of the spectrum are synthesizers (Roland System-8, Novation Summit, both would make great controllers). Then the question becomes: at what price point does your product come in where it's more attractive than buying a similarly configured synth? I don't think you'll be able to create a controller with zero compromises for end users (it's not possible), so how will you have fewer compromises than competing synthesizers/controllers or offer a better value proposition? For instance, looking at that filter mock-up, I can see "cutoff", resonance as one knob, Envelope as another knob, and KBD tracking, and then maybe a filter type knob, but then I'm out of knobs. My System-8 has 8 knobs in the filter section. So right there, I'm looking at that mock-up thinking "that's not better than what I currently have."

I'm also not clear on how I'd assemble these modules in a space efficient enclosure. Let's say I want: 2 LFO's, 3 OSC's, a synth mixer module, 2 or 3 ADSRs, and some Effect modules to assemble into a synth layout...how much space will that take up? A space-efficient design/enclosure is key. Desktop space is at a premium (and a module like this would go on my desktop). Could you sell doubled up modules? Like one space efficient module that's 2 LFOs (no banking please) and 3 analog styles oscillators? Might save more space than buying 3 osc models. Also, I'd want a keyboard for it to feel like an instrument. I'd personally sacrifice some modularity for convenience and size, even if it were more expensive.

Also, how will mapping be handled? I'm assuming some kind of editor software will be needed versus hardwired MIDI CC's. But I haven't seen any reference of the process to do this.

Anyway, this is just feedback from an interested customer. I'd suggest you open up a new thread if you haven't already.

[noiseboyuk]

The idea is to design a MIDI controller fo software synthesizers, which will be built of modules.

Thanks for posting this Magdalena, it looks really interesting, and potentially the best solution out there to date. However, honestly I don't think it would do it for me. The more I thought about it, the more I figured it had to be screen-based for dynamic labelling with physical controls overlaid. That's the only way it can truly cover a wide variety of synths. I think a modular solution for most synths would become extremely big (and expensive) to be useful. Something like Omnisphere with a 4 layer architecture would be insurmountable.

But it might well work well for basic VA synths, and there could be a market for it. It's just not the droid I'm looking for...

[Funkybot's Evil Twin]

Thanks for posting this Magdalena, it looks really interesting, and potentially the best solution out there to date. However, honestly I don't think it would do it for me. The more I thought about it, the more I figured it had to be screen-based for dynamic labelling with physical controls overlaid. That's the only way it can truly cover a wide variety of synths. I think a modular solution for most synths would become extremely big (and expensive) to be useful. Something like Omnisphere with a 4 layer architecture would be insurmountable.

But it might well work well for basic VA synths, and there could be a market for it. It's just not the droid I'm looking for...

What about a system that wasn't entirely screen-based, but rather one that just had rows of LEDs for labels? The OLED screen on my Avid Artist mix looks sharp, is viewable from a bunch of angles...imagine a controller with several rows of similar strips above/below each hardware parameter. Then include a software editor to do the assigning/labeling/sharing mappings. They could still lay it out like a synth with tons of knobs, buttons and faders. If it had good OLED labels, then you could even offer multiple banks for the various sections so you switch to Bank B for your additional Osc controls for example. But design it with enough controls to keep banking at a minimum.

Basically, imagine if NI took NKS, combined it with the display strips of a control surface, and the form factor of a full blown synthesizer. I think that would be the killer controller. If NI put out a "Komplete Control Synth" like that, I think it would be wildly successful at even 2k.

[noiseboyuk]

What about a system that wasn't entirely screen-based, but rather one that just had rows of LEDs for labels?

Yes, it could work-ish, but you'd be quite limited in terms of flexibility of layout. With an underlay screen you can colour in terms of blocks so be much more dynamic and responsive - allocate 3, 4 or 6 knobs for the envelope as required, and purpose the others for something else. You could sort of do this with OLEDs, but visually it would be far more clumsy. Also I've always thought you can nod to the interface and colours of the synths you control, and simply zip through your different skins as you change synths. (maybe even possible to automatically do this somehow).

The thing that gets me is that the screen side of it should be relatively cheap, I don't think it need be that expensive. Underneath it a Linux tablet, I'd have thought, or possibly Android.

[Dasheesh]

I've always thought Novation makes the best controllers for today's soft instruments, but the MK 3 is a TANK. it's too large for my set up. hopefully the mk 4 will be thinner and lighter. their keyboard is one of the best on the market tho, even if I prefer a FATTER key along the lines of the old waldorfs. those were the best paying keyboards i ever felt.

[sacer]

Touchable would be my dreamcontroller if it would Store customised Controller presets to a ableton live set.
It‘s easy to design or rebuild a software synth on ipad or or a touchnotebook

[Jace-BeOS]

I’ve not read this thread, but I’m assuming someone already mentioned things like high-resolution MIDI 2.0 support on all controls, as well as ribbons, and MPE, right?

[rhearhino]

I like the idea of modules (I better had, since I brought it up earlier), but I really, really want proper DAW integration. That means a protocol that just works off the bat and shows the name of the parameter you are controlling on the controller (LCD or something). Ideally there'd be a bit of smart logic behind the protocol too, so that if you had a envelope controller/template loaded, there'd be a way to communicate with the VSTs. The controller says "Hi, I control envelopes, here are my ADSR controller IDs" and the VST boots up with a default mapping that is already set up for that. Of course, you should be allowed to define you own mappings if you want, but it's awfully tedious having to do all of this stuff manually for every single VST when they tend to share a lot of common architecture and parameters that you'd like to access... ADSR, filter resonance and cut off etc. etc.

This is hopefully something that will happen if and when MIDI 2 happens. Until then, everything just feels like clunky (and proprietary) workarounds. I really think this needs to be baked into the DAW itself to work properly.

Thanks for the feedback!

Our idea initially assumed the development of a universal controller layout based on many different VST plugins – so each knob / slider would be permanently labeled, e.g. by screen printing. So the mentioned ADSR would always be on the same module, in the same place. This would also allow users to use the VSTs without even learning them – they would have a muscle memory from the controller, where the Filter Cutoff is always in the same place, the Oscillator Pitch in the same, etc. This was the assumption that would require no additional integration with the DAW, except for reading the parameter value and other obvious things. Of course – as you write – we would like our controller to map automatically to each uploaded plugin.

MIDI 2.0 will definitely help in this, but we cannot rely on it yet, which is why we are looking for different solutions.

What do you think about that?

Sounds like an interesting product. I completed the survey. I'd be very interested in something like this if it offered a ton of knobs/faders/buttons, and was housed in a good keyboard enclosure. I think there's already a fair number of controllers on the budget end of the spectrum but the best options at the higher end of the spectrum are synthesizers (Roland System-8, Novation Summit, both would make great controllers). Then the question becomes: at what price point does your product come in where it's more attractive than buying a similarly configured synth? I don't think you'll be able to create a controller with zero compromises for end users (it's not possible), so how will you have fewer compromises than competing synthesizers/controllers or offer a better value proposition? For instance, looking at that filter mock-up, I can see "cutoff", resonance as one knob, Envelope as another knob, and KBD tracking, and then maybe a filter type knob, but then I'm out of knobs. My System-8 has 8 knobs in the filter section. So right there, I'm looking at that mock-up thinking "that's not better than what I currently have."

I'm also not clear on how I'd assemble these modules in a space efficient enclosure. Let's say I want: 2 LFO's, 3 OSC's, a synth mixer module, 2 or 3 ADSRs, and some Effect modules to assemble into a synth layout...how much space will that take up? A space-efficient design/enclosure is key. Desktop space is at a premium (and a module like this would go on my desktop). Could you sell doubled up modules? Like one space efficient module that's 2 LFOs (no banking please) and 3 analog styles oscillators? Might save more space than buying 3 osc models. Also, I'd want a keyboard for it to feel like an instrument. I'd personally sacrifice some modularity for convenience and size, even if it were more expensive.

Also, how will mapping be handled? I'm assuming some kind of editor software will be needed versus hardwired MIDI CC's. But I haven't seen any reference of the process to do this.

Anyway, this is just feedback from an interested customer. I'd suggest you open up a new thread if you haven't already.

Thanks a lot for completing the survey!

Our solution would be at the intermediate price level – at least that is the assumption at the moment. We want it to be really durable and made of good quality parts, but we want the basic set-up to be much cheaper than hardware synthesizers with a lot of knobs (like the Roland System-8 mentioned). For the drawings – as I pointed out at the beginning, it's just a teaser – there will be much, much more knobs and various controls on one module. We just didn't want to make detailed drawings or renders right now, so as not to attach to details that are not yet fully developed in their final form So don't worry – the filter module will have much more possibilities than the one drawn!

As for the dimensions – I don't want to discuss it now because we are before choosing the best. For sure, it will also be possible to control more than one oscillator / envelope / LFO from one module. That will depend on the buyer – whether he prefers to have, for example, one module, where he can control e.g. 9 LFOs (by Shift, like the "banking" you mentioned), or to have more modules with three LFOs each. Etc. But I don't want to get into the discussion about dimensions before more accurate renderings or prototypes – certainly the issue of size is one of the most urgent and we will try to do everything as small as possible (while maintaining users convenience of course). But it certainly will not be like: one module / one oscillator.

There is this problem with the keyboard that you have to do it really well because there are plenty of them on the market. I have a dilemma – why make a new one? How big would it be? Most people already have a keyboard controller, some 25 keys, some 81. We will definitely consider the keyboard, but probably not in the first, most important part.

Mapping – as I mentioned, we are at such an early stage that I'm not yet able to explain how exactly it will work. We have several ideas that we will test and choose the best one. This applies to MIDI 1.0 and MIDI 2.0 as well.

We also have our own thread on the forum, there are a lot of comments But I decided that not everyone who is interested will see it and I have added information in several places. So thanks for feedback once again!

The idea is to design a MIDI controller fo software synthesizers, which will be built of modules.

Thanks for posting this Magdalena, it looks really interesting, and potentially the best solution out there to date. However, honestly I don't think it would do it for me. The more I thought about it, the more I figured it had to be screen-based for dynamic labelling with physical controls overlaid. That's the only way it can truly cover a wide variety of synths. I think a modular solution for most synths would become extremely big (and expensive) to be useful. Something like Omnisphere with a 4 layer architecture would be insurmountable.

But it might well work well for basic VA synths, and there could be a market for it. It's just not the droid I'm looking for...

Thanks for the feedback!

I wrote back to you about the screen issue in the another thread on the forum We rather don't aim at this type of controller, but we are thinking about various options with screens.

[sjm]

Thanks for the feedback!

Our idea initially assumed the development of a universal controller layout based on many different VST plugins – so each knob / slider would be permanently labeled, e.g. by screen printing. So the mentioned ADSR would always be on the same module, in the same place. [...]
What do you think about that?

I think that pre-packaging something exclusively as "ADSR", is somewhat limiting. I mean, an ADSR is really just 4 controllers that are grouped together thematically. But that same physical layout is just as suited to something like "delay time, delay feedback, mod depth, mod rate" for a delay (just as an example). From a physical standpoint there's actually no difference between the two. The difference is purely what you expect the controls to affect in your software.

I think that the ultimate controller would be one that is just as suited to controlling Synth A and B as it is to controlling FX A and B. That's why I *personally* am not sold on pre-printed modules. I'd much rather have an LCD display that tells me what I am currently controlling - a la Automap.

Instead, what I think I want - and I haven't fully thought this through - is a series of modules that I can arrange as I want to - sliders, knobs, XY pads, joysticks, wheels etc. - and some sort of smart interface that allows me to assign certain modules to certain "macros" (for want of a better word). I realise that the tech isn't necessarily here yet. But bear with me.

So instead of having pre-conveived layouts, I'd set up my own "templates" for certain use cases. I might have one template for subtractive synths, say. And one for delays.

In the subtractive template, I would like to be able to say that the first module with 4 knobs is the amp ADSR, the second one is filter ADSR. Then have the controller communicate with the VST and negotiate which of the VST parameters are amp/filter envelope parameters, and automatically set up a mapping for me. And if I have the same template loaded and switch to another subtractive VST, I'd want the mapping to be the same without me having to do anything.

Now when I switch to another type of VST, e.g. a delay or a piano or whatever, those controller assignments don't necessarily make sense any more. So I'd want to use my delay template, where those 4 ADSR knobs are now automatically controlling delay time, feedback amount, mod depth and mod rate (as an example). And I'd want the parameters I'm controlling to be obvious from the controller. A D S R labelling just doesn't work when you are controlling something that isn't an envelope.

Without that, there's no real advantage over existing controller. In fact, I own a controller that comes with various overlays for different synths etc. But printing 1 overlay per synth is simply not realistic nor efficient, and a real chore when switching between VSTs to have to manually switch your overlay. Having dynamic labelling solves this issue.

What I want is one controller to rule them all. And ideally, it'd be smart enough to identify the current VST type, and select an appropriate template (that you can override if you want to of course). So any VST identified as a subtractive synth uses the "subtractive" template; anything identified as a delay uses the "delay" template out of the box.


The thing is, that you can already label controllers. It's the lack of flexibility that annoys me about this approach. The labelling should be dynamic and reflect what it is the knob is *actually* linked to in the software, not what the designer of the controller thought it should be used for. That's an artificial restriction that doesn't really have any benefits.

The problem I see is that the tech simply isn't here yet, and it'll require the DAW manufacturers and plugin devs to work together to find something that just works OOTB without needing to faff around for ages setting up each individual plugin to work with your controller.

[Echoes in the Attic]

The technology for a dream controller already exists but is split into two different products. The Novation Remote SL (mk1 and mk2) and Native instruments Maschine/Maschine Jam.

The layout of the original Novation SL was perfect. Screens above each of two sides. Knobs on the left, sliders on the right and a whole bunch of buttons. The problem was that the screens back then weren't the best and there are knobs and sliders that need a pickup mode. Start with the original Novation SL layout, but have both rows of knobs be the endless style of the SL mkII with light rings around them (or like Push or Maschine or whatever other good quality encoders but keep light around them) and then replace the sliders with the Maschine Jam style touch strips. Then you put higher quality Maschine/Komplete/Push 2 style displays on both sides (but keep them small). And you keep the buttons which can be illuminated of course like they all are now. And make sure to have a page selector where you can jump right to a parameter page. This is a huge issue with both the NI controllers and the Novation SLs where you have to press left or right page, which is terrible for large instruments. It doesn't matter how many parameters you have if you can jump directly to a page. Oh and also allow the left and right side to go to their own pages. This means you can have a wider combination of parameters at any given time. That in my opinion is the perfect controller. Every control is endless and updates with new pages including sliders and a huge number of parameters is not an issue.

Of course with this kind of thing you need a layer in between. Maschine and Komplete Kontrol software acts as that layer/wrapper. In the case of Novation the automap server did the same thing, but it was totally transparent. And it worked really well. Unlike Maschine/Komplete, you could remove those wrapped plugin version (for VSTs) any time and the regular VSTs continue to work fine. You swap them in and out, no issues. There are huge benefits to the automap style wrapper, like setting specific start and end values and the resolution of a control. IT's the only way to get a proper hardware experience.

So all the tech is there to make a pretty much perfect hardware experience now with existing plugins, just not in one product and it's really annoying. I'm fairly happy using automap with a combination of an SL61 mk1 keyboard and an mk2 SL Zero above it, but best where the plugin is not too big.

[acYm]

^ you say that because Komplete Kontrol sucks really bad as a sub-host, indeed, but still, it's better to have one. There's no vst host nearly as comfortable as Kore, all the DAW's are really far back.
inserting a midi-out sequencer or arp in the slot just before your instrument and having it just work like magic isn't something you tire of, nor is having all your patches imported, sorted and tagged. Everything is easier and quicker in there.

[noiseboyuk]

The layout of the original Novation SL was perfect.

i sure wouldn't call it that, though it was better than most. It still fundamentally doesn't resemble a typical synth layout, its more resembles a mixer. You could do more with it with a proper screen underneath it all with dynamic labelling / colour coding, but I wouldn't lay it out like the Novation.

I still have yet to hear any convincing argument why a good size screen with translucent strips and physical knobs isn't the most versatile and cost-effective solution. Pretty much every objection seems to be based on a misunderstanding of the concept.

[Funkybot's Evil Twin]

The layout of the original Novation SL was perfect.

i sure wouldn't call it that, though it was better than most. It still fundamentally doesn't resemble a typical synth layout, its more resembles a mixer.

I still have and still use the original Remote Zero SL (including the editor software to make custom templates) and agree. It's a great for FX, but not inspiring as a synth controller. It's also terrible for mixing, but not because of the layout but just because the knobs/faders suck.

But the original RemoteZero SL wins my vote for the best VST effect controller to this day because of the combination of dual screens, lots of buttons, knobs and faders, along with the editor software allowing you to create proper labels/templates for various effect configurations. I have various templates for reverbs, compressors, EQs, and just map common VST parameter controls to the template layout so I can get all the primary controls for just about any plugin up and running. No automap. I use the DAW to learn them once. It's not leaving my desktop until someone comes up with a better FX controller with labels!!!

I still have yet to hear any convincing argument why a good size screen with translucent strips and physical knobs isn't the most versatile and cost-effective solution. Pretty much every objection seems to be based on a misunderstanding of the concept.

If money were no object: I'd love to see something with a tablet like surface area for labeling/graphics, with physical knobs, buttons, faders, built into a 5-octave keyboard with pitch+modwheels and poly aftertouch. Good editor software would be key, because nothing like Automap would work well enough (maybe MIDI 2.0 could change that - talking right now). But if users could create and share their own templates, that'd be good enough. In my head, this dream products looks like a Roland JD-XA but the plastic panel is replaced with a tablet-like surface, TouchOSC-like vector graphics, but hardware MIDI controllers. Probably prohibitively expensive to build.

Next best thing: a similar synth+effect type layout, with OLED strips for customizable labels for each parameters. Similar to the OLED display bank of an Avid Artist Mix (which looks great), but you'd have rows of these. Editor software still required for labeling the strips. This is probably doable without being crazy expensive.

Third best thing: start with Native Instruments Komplete Kontrol S49 MKII, and add a crap ton of buttons, knobs, and faders in a synth-type layout. Still supports NKS, and users could create their own NKS mappings if needed.

I don't think any of these imaginary products would start for under $1,500 (which I'm fine with).

What I'd settle for: a new version of the Novation Remote Zero SL (MK 3) with better screens and higher quality knobs and faders (along with updated software).

[Ploki]

https://special-waves.com

what about this? Has this been mentioned already?

[noiseboyuk]

I still have yet to hear any convincing argument why a good size screen with translucent strips and physical knobs isn't the most versatile and cost-effective solution. Pretty much every objection seems to be based on a misunderstanding of the concept.

If money were no object: I'd love to see something with a tablet like surface area for labeling/graphics, with physical knobs, buttons, faders, built into a 5-octave keyboard with pitch+modwheels and poly aftertouch. Good editor software would be key, because nothing like Automap would work well enough (maybe MIDI 2.0 could change that - talking right now). But if users could create and share their own templates, that'd be good enough. In my head, this dream products looks like a Roland JD-XA but the plastic panel is replaced with a tablet-like surface, TouchOSC-like vector graphics, but hardware MIDI controllers. Probably prohibitively expensive to build.

Next best thing: a similar synth+effect type layout, with OLED strips for customizable labels for each parameters. Similar to the OLED display bank of an Avid Artist Mix (which looks great), but you'd have rows of these. Editor software still required for labeling the strips. This is probably doable without being crazy expensive.

Third best thing: start with Native Instruments Komplete Kontrol S49 MKII, and add a crap ton of buttons, knobs, and faders in a synth-type layout. Still supports NKS, and users could create their own NKS mappings if needed.

I don't think any of these imaginary products would start for under $1,500 (which I'm fine with).

What I'd settle for: a new version of the Novation Remote Zero SL (MK 3) with better screens and higher quality knobs and faders (along with updated software).

Integrating a keyboard is a whole new world of pain. Everyone would want a different keyboard, and many wouldn't want any.

I'd have thought my dream design - single screen across the whole surface for labelling - translucent strips with mounted knobs / faders / switches could be done for circa $500. It would need some simple brains a la an android tablet or linux running a mapping app, or alternatively run the app from the PC / Mac and communicate via USB.