Hi all,
I’m working on a small Eurorack hardware project for controlling visual parameters in TouchDesigner. I am principally an artist but I am comfortable with a soldering iron and microcontrollers and have done some pcb runs for fun in the past. I have a pretty clear idea of what I want to build but have some questions.
The gap I’m trying to address is hands-on, CV-based visual control from inside the modular rack: envelopes, followers, sequencers, logic, triggers, buttons, toggles, and knobs that live with the patch rather than in the TD network or on desk midi controllers. I have been working with StreamDecks, midi controllers, an ES-10, and TD-networks but it requires a lot of screen-work to set up and work with, and there’s no physical relationship between the controls and the eurorack patch. I love patching in the rack and I want my controls to be patchable and arranged for visual parameter control rather than audio conventions.
The basic idea is a generic CV acquisition board that makes it easy to design input modules with various combinations of CV inputs and physical controls using a standardized header. I’m designing for a minimum width of 6HP.
The spec is up to 8 channels of CV/controller input, up to 8 channels of digital input for buttons and toggles, power for pots and sensors, and I2C for an LED controller. I designed one standardized header for an input board and another for an MCU board, so it’s straightforward to design different modules with different input combinations and to design a different MCU board for different use cases. I’m currently working with a Seeed XIAO SAMD21 - it’s tiny and cheap with native USB support. The firmware will output either binary serial over USB or USB MIDI.
Early binary serial tests show approximately one frame of latency at 60fps in Touch Designer with a simple binary frame, which works for visual parameter control. I am decidedly not trying to make an audio rate or percision Expert Sleepers, Boardbrain or Befaco module - I am going for dumb(ish), economical hardware state reporting, targeting up to 120Hz per channel. I tried OSC from an ESP32 over Ethernet but the latency was 4-7 frames. Bummer. I’ll also test 7bit and 14bit MIDI over USB.
From what I see, rack controllers tend to focus on audio conventions like fader rails (eg sweet 16, VCMC). For performance and patching, I want something designed for visual control. For example, a module with three concentric pots for XYZ offset and range, toggles for invert, CV jacks for modulation from the rack, and a small OLED screen with an encoder for labels. I would use the crap out of that.
The main design question I’m trying to settle before locking the hardware is about voltage range. My current preference is 0 to +5V. A lot of the sources I use operate unipolar: envelopes, sequencers, followers, gates, triggers, logic, and Hagiwo-MOD1 style 0-5V modulation sources. Bipolar LFOs can be scaled and offset upstream when needed with things like an MIA or Triplatt module - or just use Expert Sleepers for that.
The -5V to +5V bipolar range is more modular-native for LFOs and other bipolar CV, but it adds a lot of electrical conditioning complexity and I’m not sure it’s actually better for visual control, where the midpoint is a mapping decision rather than a fixed electrical requirement. This is not meant to be a precision pitch or audio interface. It’s more of a rack-native visual control surface and CV-to-parameter bridge.
Does this seem useful? 0-5V or -5V to +5V range for visual control? Any other things worth considering?
I’d appreciate any feedback.