HyperQbe

So, what would you recommend? To start with, I thought of getting myself some stuff like the "core" (whatever that's gonna be), an LC display, some buttons, encoders and LEDs for the interface and some basic testing, and maybe some unit for SD cards... of course I'd need some I/O, too, but I think USB should do well...

So, well... Maybe, someone can help me with the most important thing. I looked through the hardeware section at uCapps.de, and there are 3 core modules described — at least, if I get that right. To me it seems, the STM32F4 based core would be the simplemost to use, for there's no soldering neccessary. But somehow I have a feeling that it can't work as a "brain" all on its own... Exactly what parts would I need to buy and which steps would I need to do to get a very basic setup from which I can move on extending and somehow test my ideas in the first place?

I think, that may be a proper solution. But since regular MIDI I/O are so BIG, and the (main) box will rather be somewhat small, I tend to use mini DIN connectors or even stereo audio jacks for chaining. I'm thinking of granting 4 inputs to the main box and putting each one in and out on every extension box. So you'd have a max of 4 extensions connected to thew main box, that cann be arranged and positioned just as you need it. And thinking of the approximate size these boxes are likely to become, 4 of them plus the main would take up enough workspace, so you won't want any more! ;) I've done my homework, so I know, things would cost a significant amount of money compared to ready-made controllers like the Behringers! But since I'm NOT Mr Oizo, by saying "affordable" I mean I just cannot spend some hundreds of €uros to build one huge complete controller al at once, just to test if my concept would work out. And that's one reason why I came to the modular idea, for I can then spend much less money on each box and build up my system step by step. (I know the total cost would be much higher, because I need lots of extra parts and some of them multiple times, compared to one single box. It's more about what I can afford at once!) Programming is not my problem. As long as I know how the language works and what it's capable of, I can somehow get my stuff together. Right now I'm learning to write proper codecs that make it possible for Reason to work with controllers... I've already tried merging the data streams of my Behringers. They're actualy build to be daisy-chained if needed, and I also tried a MIDI merger / splitter to achieve some positive results. But in this case, the existing codecs for these devices in Reason "have grown old", and they don't seem to be able to work properly if you have multiple chains of the same devices connected. Unluckily, there's no way cou could change those codecs to your needs, so writing a new and modern one is the only chance to get it done...

Well, that merging variant seems quite appealing to me! Would it then also be possible to make some basic settings of the chained modules from the "master box" and also store it there? I mean, it should be possible to recognize connected modules via MIDI and make neccessary adjustments to them as well as to the master itself, during boot-up, doesn't it? At least, this solution would not limit the number of modules, as long as I'd work out how to separate their messages in a way that any receiver doesn't get identical messages for different purposes... Another questiion is, whether the master box can also be placed in the middle of such a chain -- in case, someone likes to not place all his modules to one side of the master box, but maybe 2 on the one side and one on the other...? (see attached picture) In the picture I provided earlier, the "master section" is the center part aswell, making it more accessible if the controller should be placed more to one end of the workplace. Maybe, someone could recommend one of the described brains / core modules from the main website (ucapps.de), so I could start by building some basic controlls around it and see how it would work with my system. Thus I would know if it makes sense to do the "big thing" at all! ;) Just to get that right: I'd like to have that display-and-buttons-thing just ONCE to set up the whole modular system. If that's impossible, even with each module having its own core, and I'd have to put this part in each controller, then I'll skip the modular idea and just try and do the thing as shown in my first graphic...

Twisted pair sounds a lot like RJ45 network cable... But I guess I'm wrong with that, for I'm not that much into english terms! That's also the reason why I ask that much stuff I could as well read on the uCapps pages — I basically understand what is written there, but I just can't make any useful sense out of it... By 256 digital inputs, I guess they mean 256 INDIVIDUAL digital inputs... Taking those LED rings into account (for I haven't found some proper mini displays yet that could show the momentary value), that would mean that I could connect no more than 1 or 2 modules to the main section. So the controller would look not much different from the draft I provided... :/ Could you then perhaps give me some tips what's best to start with when developing such devices (besides designing a layout as I've done already)...? :)

Ummhh...I think I'd go for #1 then. Are those the same cables as already used by MIDI? :) That's a nice idea, for there are quite short cables available on the market.. But the question is still, would one brain be capable of getting it all together, even if someone decides to, let's say, connect 4 or 5 controller modules - and which one would do best fulfilling that task? Just for consideration: The "rotary" module would technically have 64 encoders plus buttons, the "apc matrix" module could have 64 of those pads (maybe velocity sensitive) plus additional controls... I think that's quite a lot, isn't it? :) Oh! AND: Would it be possible to daisy-chain those modules, or would one have to connect every on to the main module, separately? If the latter, that would at least limit the number of modules to connect. ;)

Well, thanks for your interest and for your quick answers, guys! @Phatline: Nice idea! But I did some research, and there already are controllers in similar cases out there. They're modular in the same way as"classic" modular synths. But that's not quite the way I meant it... As I mentioned, my design bases on the BCR/BCF2000 controllers from Behringer. I've attacheda pic showing one of the latest drafts of this design. As you can see, it is devided in three parts: the "BCR" part on the left, the "BCF" part on the right and some kind of "master section" in the middle... Now, my idea of making it "modular" was to separate these three parts, each in its own casing, but leaving the "brain" as well as all connections (MIDI, USB, power) to that master section. You could then go and buy the parts you need (the rotaries, the faders, maybe some kind of APC-like button matrix and so on...) and just connect them together to form one single controller that perfectly suits your needs! :) Well... Maybe ther'd be some need for extra power supply, so I would throw out the power supply from the main section and give each part its own USB power supply... :) Now, what's your opinions? Would this be feasible or would the "brain" simply tilt if you connected too many modules?

Hi there! I've got this idea spinning around in my head... It all started out with a bunch of Behringer's BCR-/BCF2000 which I considered to be the best controllers I could get (for the money) to control Propellerheads' Reason with. The only problem was that I just couldn't get all 4 of them running together, controlling what I wanted them to and giving me the feedback I was craving for... Then someone crossed my way and helped me build a codec (a file that tells Reason all it needs to know about the controller) which handles every 2 of them as one device, which got me the control I wanted, but still lacked proper feedback, since I wans't yet able to get access to the BCs' original functionality... Then I came up with the idea to create my very own MIDI controller, based of an improved BC-like design, and built on top of the MIDIbox platform! :D My latest spin on this idea is this: I think it would be a lot more affordable for me to build some smaller controller units one by one, than doing the whole big thing all at once! So... would it be possible to, let's say, make a controller that's just a MIDI brain with some basic/general controls, and then make some additional "modules" (e.g. one with lots of encoders, one with a bunch of faders, perhaps an additional one covered with buttons, etc...) and then connect them all to the one basic controller, which in turn provides data and power connections? I've already done some research, but although there are already some more or less modular controller platforms out there, there doesn't seem to be something like this... yet! :) I'll appreciate any kind of response, and, maybe some time, a little (more) help in development of such a system -- if only to see it it would be feasible... Thanks!