The Unda Legio ambient chain
The 16 firmwares in our recent Legioverload release came from a wide range of inspirations, from pure experimentation to specific patches we’d always wanted to do.
One excellent example of this is Unda Legio. Back in December, in preparation for a holiday vacation, Stephen wanted to put together a 64 HP Case of Steel system for ambient patches. He had a few specific goals for the case: he wanted to do sequencing with Multi Repetitor and Gamut Repetitor (a seriously powerful combo for four-part generative sequencing), and he wanted four complete voices as well as a Desmodus Versio for reverb. And he wanted it versatile enough to keep him occupied for the whole trip (in this case, to a farm in San Luis Obispo where the dogs learned that horses exist). This left things somewhat space constrained, so he decided to make a Legio firmware that could handle the synthesis side of things.
Goals, constraints, and mixing
There were three design constraints for the firmware: it had to be a complete synth voice, it had to sound nice in multi-part harmonies, and it needed to be chainable for use without a mixer.
Mixing was easy: the audio inputs on Legio were repurposed to pass through external signals and mix it with any sound generated internally. This meant that any number of Legio voices could be chained together for multivoice patching without requiring a mixer module – perfect for a compact system.
The synth voice itself was a bit more complicated. With just two switches and two pots (and an encoder, but that’s reserved for pitch) choices were somewhat limited, so Stephen decided to go for a two-operator phase modulation synth, due to its wide range and relatively simple control structure. One switch offsets the modulator’s octave, and one pot controls the modulation amount. The beauty of phase modulation is that this still gives a very wide range of timbres, especially once we get some envelope modulation going.
Speaking of, the other two controls were delegated to envelope duty. The switch selects the number of envelope stages (2, 3, or 4), and the pot controls the envelope timing of all segments. Since the voice was developed specifically for slow, evolving melodies, Stephen dialed in the timing relationships of the different segments to sound good for that style of voice across the range of the parameter.
Sometimes, the beauty of modular is working within constraints. The resulting firmware was a surprisingly expressive voice despite its compact size, and the results speak for themselves: Stephen ended up making 13 different patch recordings with his Unda-based ambient case! You can watch them all in the playlist here.
