Before you could earn anything, you would have spent thousands of dollars.

 

Last edited: Dec 31, 2021

Yes I did but I missed the point that you were only asking for assistance in building such devices. I wanted to state that there are two approaches already (Joue and Sensel) and they did not have that much success.

 

Last edited: Dec 31, 2021

using a camera is in many aspects something completely different than using a glove.

 

you´d usually program that kind of stuff yourself.

building a "drehbank" rackmount unit with full 128 controllers based on doepfer DYI kits is on my to do list since many, many years.

and making the software is really easy - compared to building a custom metal enclosure and getting all the money you will need for 128 proper encoder knobs.

 

As long as you can earn money with it, there will also be new innovations. The future will be very complex.
Tesla is building its first house robot. I hope the robot can do the dishes. But I don't want a robot sitting at my desk.

 

KI

Rory Kenny: From my experience of developing AI music and the challenges we face, I think we are extremely far from ever reaching the beauty of the human ability to make music. AI powered music will serve a different purpose. I would never aim to do that. I don't compare AI music to human music. That is a wrong comparison. Human music has history, context and experience, it is clever. By definition, AI has no history or context and it has no emotions. So it's important to separate AI music and human composed music.

How does your AI work when generating music? First of all, AI has to learn something and be trained. Our special AI is trained to sequence audio loops. As a producer, it is now common practice to buy audio loops and samples from the Internet or to create your own. You then sequence that into a techno or house track and this sequence is an attempt by humans to create a song structure. We have 8 million project files across 10 genres - one project file has 8 audio channels filled with audio loops.

If we give that to the AI, it knows what each loop is - a drum loop, a base loop, a synth loop - and it learns how we created the song structure. And we taught her that she understands beginning, middle and end in terms of song structure. It couldn't be easier: the AI learns which sounds it gets and in which order they should appear, and that's it. Other AIs create or compose a sound from scratch. The idea is to create inspiring pieces that you can then modify later.

www.musikexpress.de/wo-und-wie-kuenstliche-intelligenz-bald-musikerinnen-ersetzen-koennte-1827125/

 

True, but current software should be able to recognise at least hands, if not a complete body. If the CPU is powerful enough...
However, the MiMu Gloves are nowhere to buy now and were very expensive.
The Leap Controller with Geco seems to be another approach.
But this is off-topic. I would rather prefer an own thread for such a discussion.

Edit: The Mimu Gloves are only available in the MiMu store (about $1300 ?) and currently out of stock due to chip shortage.

 

Last edited: Dec 31, 2021

It depends on what exactly do you need a MIDI controller for. For loop artists and maybe for performances yes that glove could be a cool thing, but for say a producer that does only beat after beat that would slow him down.

DAW level controllers are nice if the implementation itself is what you need. Sure you can do scripting for most DAW's like FL,Ableton but that would take some time to do or maybe wait on someone to do it for you (sucks if you already have a controller and that one doesn't have a script).

I think for the most part, except the ones that are made specifically for a DAW and it works great, midi controllers are just a novelty, maybe a old school way to "feel" the knobs turning that you would've done with hardware back in the day (or now if you have hardware).

 

Creativity / laser technology

Intel Lazer Harp at UCD’s Evolution Garden

Jean-Michel Jarre - The Time Machine Live (Laser Harp)

 

Sure, I don't speak german, but I found a few videos that were similar to the ones you posted about how to make midi controllers out of an arduino and some basic components. My question was mainly about how to manufacture the physical controller itself. You need a plastic housing, where do you get that? 3D printing? Laser cutting? Where would you do that? Assuming I'm not the only one who doesn't have a laser cutter sitting in my garage. You also need the raw materials, buttons, lights, etc. It all has to be sourced, so where is the best place to get it?

I couldn't disagree more. How do you play chords with a mouse? You can't click three notes at once without some kind of auto-chord midi effect. It would be way faster just to reach down to a keyboard and play the notes.

Yes, my intention from posting this was to design, build and sell the controllers myself.

THIS, is actually closer to what I was thinking of creating. I was thinking of doing it a bit differently, but more or less a similar idea. I might even try to buy one of these if I can find one.

To create a full-scale company that sells these, I agree it would be necessary to have all the things you mention.

For now, however, I think I could make and sell a handful of controllers all by myself. And if it turns into a good business with a large market, then I can hirepeople and find investors and such.

 

It is best to get together with other people and share your knowledge. One can solder, the other can do this and the other can do that ....! You are more successful in the tribe when you go hunting together, you can also set traps, but you are only successful if you act intelligently.

10 Amazing DIY MIDI Controllers Made with Arduino
https://all3dp.com/2/diy-midi-controller-with-arduino-project/

Spoiler: DAW controller DIY - build your own midi controller

DAW controller DIY - build your own midi controller
Posted by DAW Controller October 22, 2016 Category ( ies ): General , DAW Controller Technology Tags : Arduino , Ardurio Uno , DIY DAW Controller , DIY MIDI Controller , Micro Controller 4 Comments

What does DAW Controller DIY mean?

Digital Audio Workstatin Controller Di It Yourself … which means nothing more than building a DAW controller or midi controller yourself.

How this can work is described in the following article.

Why build a midi controller yourself?
The motivation for building your own midi controllers is certainly very different. In addition to the fun that creative building can be, there will certainly be a desire for a very individual midi controller .

The design, number, type and arrangement of the controllers usually play the decisive role.

Regardless of whether it is a rotary control, fader, button or joystick, whether only 1 potentiometer or a total of 50 - there are no limits to the design of your own midi controller when you build your own.

Daw Controller DIY - what is needed?
1. Rotary potentiometers, faders, buttons
The decisive motivation for the DIY midi controller is usually the individual number and arrangement. The individual components can be obtained easily (e.g. via the Internet) and are usually quite inexpensive.

Required components can be:

• Rotary control
• Fader
• Button
• Joysticks
• Displays
• other controller tools such as distance meters and others

2. The micro controller
A micro controller is required for measuring, testing and controlling. This is where Ardurio comes in.

Open source hardware was developed under this name. Corresponding controllers are available from various manufacturers. The original costs around 30 EUR, similar products are available from 10 EUR.

http://dawcontroller.de/daw-controller-diy/

Wooden MIDI Music Machine

Custom MIDI controller (started 3/2013
https://sites.google.com/site/uglybuckling/MIDI-controller

www.ucapps.de

 

Last edited: Dec 31, 2021

this keyboard-keyboard looks weird. what could it be good for? isnt it better to have two extra devices so that you can place them like you want? ;)

 

Housing:
The best solution for that would be
plastic injection molding in small series
for Example :
https://prototechasia.com/en/small-series-plastic-parts
I don't know where you're from
but If you search for the Keywords, you'll find something in your region .

But even something made of Metall (Alumium or Copper ) has it's own flare and can be made easy in CNC milling machines

or also very unique would be in wood (plywood) and can be made by hand if every other Method is too expensive


For nearly everything else, Ebay and Aliexpress are also a good choice

Resistors (ali)
LED (ali)
Rotary Resistors (ebay)
Micro Push Buttons (ali)

 

Last edited by a moderator: Dec 31, 2021

It would be good for someone who doesn't have a lot of space on their desk. If I build one, it would be smaller than the Prodikeys one and be even better.

 

I'm located in the U.S., so how would one go about getting either a plastic mold or a metal construction? I assume you would need a schematic, like some sort of auto-CAD file? Where do the raw materials come from?

 

See this whole thread. A controller from bare idea to finished product. It may be helpful to you:
https://forum.cockos.com/showthread.php?t=237950&page=21

 

Actually there is no need to build new midi controllers: existing ones can be hijacked by forcing them to perform actions other than those for which they were built
Example: I use custom firmware in the Launchpad to manage my external gear

Launchpad into Digitakt (no computer needed)
https://www.fenixsong.com/digipad/

Note: I’m not the author of this modified firmware

 

Yup, as others have already mentioned:

1. Program/develop working core controller feature set on rasberry Pi/arduino.
2. Prototype the external casing/controller parts in something like Autodesk Inventor/Fusion 360.
3. Send your prototype model parts either to your own 3D printer or to a commercial CNC/3D service. Protolabs are very good and, being american, have a large USA presence - they can render almost anything in plastic/metal etc.
4. Probably go through a number of modifications/refinements until it finally looks/feels/works like you thought it should...
5. Now it's biz BS time... to try to protect your intellectual property, along with commercial manufacture & marketing spend. Unfortunately it'll cost you in the region of $100,000 to patent your design/feature set across all territories - assuming that it even qualifies for a patent. Patents will need to be registered in the US/Europe/Asia & numerous individual countries, each with fees and differing criteria/beaurocracies to address. For manufacturing, you'd have to identify/contract a reliable chinese tech company to do the job, same as nearly every other western company does nowadays. Depending on the run size, complexity & build quality of the unit etc, this could cost you anywhere between $10,000-$100,0000++. As for marketing costs, how long is a piece of string?

Basically, unless you've got really deep pockets, you'll almost certainly end-up spending at least year or so doing a whole heap of hard work and - if it really is a novel, useful design - only to then see the essential concept ripped-off by a larger company... As others have already noted, the midi controller market is relatively small & accordingly represents a substantial financial risk for probably no/little net gain, indeed quite possibly the opposite.

Honestly, i'm not trying to dampen your enthusiasm but to alert you to the core issues up front.

 

Last edited: Jan 1, 2022