A Daisy Seed synth voice using deepnote
This repo uses CMake external pojects to manage external dependencies. These external repos will be cloned and built on your first build. For build and device programming info see daisy-seed-vscode-template.
I used the Synthux Simple Fix as a hardware platform. It's nice since it includes a Daisy Seed and a few pots and switches. Here's how I have it configured:
Control values are read via ADC channels in the main loop. A set of global state variables are updated in the main loop, and read from to update subvoice parameters in the audio callback.
The complete voice is comprised of 5 "duo" subvoices and 4 "trio" subvoices. Each duo subvoice has 2 oscillators and each trio subvoice has 3 oscillators. Subvoice oscillators can be detuned.
Volume - simple volume control where full CCW is off and full CW is max volume. Raw control values are linearly mapped to the range 0.f to 1.f.
Target Chord Root - sets the root note of the target chord. Really you're changing the row index in the frequency table. Knob rotation is divided into 12 evenly spaced indexes from 1 to 12. One for each note in the chromatic scale starting with 1 = C. See Target Chords and Frequency Table below.
Detune - sets the amount of detune (Hz) between the oscillators in a two or three ocscillator "sub-voice". Raw control values are linearly mapped to the range 0.f to 5.f.
Reset Toggle - every time you togle the switch from one position to the other that start frequency of all sub-voices is reset to a random frequency between 200Hz and 400Hz.
Mono Audio Out - yep, mono only
The frequency each subvoice should be targeting at any point in time is obtained via frequency table lookup. Conceptually the frequency table is a 2D array of frequencies. There is a column for each subvoice, and a row for each chord the voice can play. The values in the frequency table were computed using tools/freqtable-builder.py. It takes a primative chord description and generates output containing the frequency per note (one note per column or subvoice) in the chord transposed to all 12 keys (one chord per row).