Wireless sensors for TouchDesigner

I’d be keen to get some feedback from the TouchDesigner community about this demo where it’s shown how to interface an orientation sensor via BLE-MIDI (wireless MIDI) to TouchDesigner, see https://youtu.be/wGgp3Lzivgk. All the other I-CubeX sensors (and actuators) can be interfaced just as easily. Would you rather use another way and if so, what way and why ?

I am looking at using these sensors too! the I-cube-x systems look like they should be pretty straightforward in getting set up talking to Touch. If anyone out there has used them before do you know limitations that might arise scaling up to 50 - 200 sensors? :mrgreen:

This might be good for some people, I would not buy it.

The first thing I think about is how expensive this is for what you are getting.

It’s clear they put a lot of work into the [max/msp] interpreter software and the firmware on the controller, however for most projects I would probably buy discrete components and put it together myself.

for example, this ESP32 has BLE and WIFI built in for $21

This can connect to TouchDesigner Very easily without any middleware software.

You can also get an IMU here that will connect directly to the above chip.


Now, you have basically the same thing as a i-cube-x at less then 1/2 the cost, but you will need to program them yourself. Heres a link to some code for the chip:
github.com/sparkfun/SparkFun_LS … no_Library

The benefit of all this, besides huge cost reduction, is that you can program the components to work precisely for your project. When I build hardware for an install, I want to make sure it has firmware without any extras and requires the least amount of drivers/middleware to get working.

if you are just getting data to Touch and back, I would avoid Bluetooth (because it’s less reliable for long durations) and stick to WIFI. ( you can even get a cheaper chip if you aren’t using BLE sparkfun.com/products/13231 )

I would also use a different protocol like UDP/OSC messaging to communicate directly to TouchDesigner (no need for interpreter software to setup/configure) Your data rates should be faster.

Thanks so much for the comments.

The WiFi OSC feature was already on the todo list for the WiDig (OSC via USB is already available) and I am pleased to announce it’s now available too.

FutureEM, WiDigs can accommodate 8 (analog or digital) sensors (or actuators) each so realizing a project with 50-200 sensors would require 7-25 WiDigs. The WiDig hasn’t been tested this way yet, but I don’t see any limitation other than the limitations you’d encounter using Sparkfun’s Thing, while you’d have all the support of I-CubeX and wouldn’t need all the technical expertise Sparkfun’s Thing requires. If you’re interested in exploring this with WiDigs, don’t hesitate to PM me or contact me via Infusion Systems. I’d be keen to help you realize it.

Harveymoon, am I correct in saying that since you have the expertise to develop hardware yourself and have requirements that can’t be met by I-CubeX gear, the cost of your time is irrelevant ? Would you consider I-CubeX if your project requirements can be met by I-CubeX and you’d feel it’d save you time ?

Look, I don’t mean to be rude. You are running a business and I am sure lots of people would buy your stuff on here.

I have the perspective that everything should be as simple as possible and if I write it myself, I will know exactly how to fix it. When an entire project depends on the device working, I prefer not to be on the phone with support.

Learning arduino code is a lot easier than learning python or learning shaders.Knowing how to buy some stuff on sparkfun and edit the code they provide for free has been really important with a number of touchdesigner projects I have worked on.

I have a code-base for wireless sensors ( that I also posted on these forums somewhere ) that does all the necessary functions of getting a device to talk to Touch. It’s SUPER simple in 2019 to do this. If I need to swap out a library for a temp sensor or a relay or a potometer or whatever, its super easy and takes me 10 minutes to make. It will always be faster/easier for me to make something then to use i-cubex.

There are TONS of expensive PLC’s that you can buy if you prefer to have something industry standard, or made in germany or has good support. I prefer to buy really cheap so that I can afford extra backups.
If you want perfect engineering that does everything just go with beckhoff at that point.

I like to make things, so I prefer to.
Making wireless sensors today is really cheap and has never been as accessible to buy parts or learn how to do. I figured this community, with a programing environment like touchdesigner, could figure out some arduino code.

side note, Sparkfun is a proponent of microPython. So if you already know python its easier then you think. Heres a beginner tut for anyone brave enough:
learn.sparkfun.com/tutorials/mi … sp32-thing

Okay so little update / thoughts after testing both the I-cubex Widig setup and building my own with the sparkfun ESP32 board (thanks @harveymoon for the direction).

The WiDig was a little glitchy for me getting it connected ( couldnt get it setup on windows but could on OSX) but otherwise was pretty easy to get talking to touch via OSC.

However, for my situation (trying to have 50 wireless sensors stream data into touch) the sparkfun ESP32 board was a better option for a few reasons:

  1. Needing one sensor per person having to buy 50 WiDig’s would have been $$$$$.
  2. Power - The sparkfun board has usb + lipo battery power options meaning not only is the board much smaller but also battery options are much smaller and have the ability to charge through the usb.
  3. Networking - Being able to network my devices myself using the ESP32 IDF for Arduino gave me a lot more options of how to network the devices together and bring the data into touch. (there is a lot of good documentation out there on this too!)

All in all,
I would have to agree with @harveymoon on this one that simple was better and building a barebones version was not only significantly cheaper but ultimately better since writing my own code for the micro-controllers gave me a lot more control (and reliability) over how the devices were functioning and was pretty easy to get them talking to touch. :smiley:

V cool. Happy the project turned out and you learned some stuff. Would be cool to see some documentation of this project if you have any to share!

@Harveymoon defiantly learnt a lot! Show is this Friday. I will post some process and documentation after :slight_smile:

Anyone here have any experience sending midi via Wi-Fi using an esp32 board to IOS devices?