Massed tone and a wall of sound
The Branca Series is named for a tradition rather than a person's gear: massed instruments, tuned against each other, played as one. Many guitars, one sound. That idea decides almost everything about how these nineteen modules work.
The gesture is the whole mass
Most modular oscillators give you one voice and expect you to build density by patching more of them. That works, but it means density is a patching decision, fixed at the point you make it, and changing it means re-patching.
These modules invert that. Density is a knob.
String Mass Core gives you up to sixteen voices per polyphonic channel and MASS decides how many are running. DroneClone gives you eight per channel, so sixteen channels of V/OCT drive 128 oscillators from one 22 HP module. Wall Conductor's DENSITY sweeps four channels in and out one at a time, so you can assemble and dismantle an arrangement with one hand while the other does something else.
The sum is 1/√N normalised throughout, which matters more than it sounds: adding voices changes the texture without changing the level. You can go from four voices to sixteen mid-performance without touching a fader.
Tuning is a mode, not a patch
The other thing a massed instrument needs is a way to decide how the voices relate in pitch, and that is a musical decision that deserves a control rather than a cable.
String Mass Core has four: everything in unison; distributed across odd-harmonic sections; across Ptolemaic just-intonation ratios; or all on the fundamental with each voice given its own slow vibrato rate so they never lock.
Harmonic Pressure takes it further and emits the harmonic series itself as polyphonic V/OCT — one channel per partial, calculated from the ratios rather than quantised to a scale. There is no scale table in it. Feed it into String Mass Core in HARM mode and every partial carries its own bank of voices.
Sixty-four oscillators, all tuned to each other by physics rather than by ear. That patch is what the series is for.
Collapse is a first-class idea
A wall of sound has one obvious problem: it is difficult to end, and difficult to make dynamic. Sustained density gets boring quickly.
So several modules share a COLLAPSE gate — an instant drop to silence with a shaped return. Wall Conductor pairs it with RECOVERY, which sets how long the sound takes to come back: 50 milliseconds for a rhythmic stutter, ten seconds for a structural event you use once in a piece.
Collapse Saturator inverts it — its COLLAPSE ramps drive to maximum and then eases back. Same gesture, opposite direction: one takes everything away, the other adds too much of it.
Both are performance controls. Neither is an effect you set and forget.
Feedback you can play
Feedback in a modular patch is usually either too timid to be interesting or one knob-turn from destroying your monitors. Feedback Governor exists to make it an instrument instead of a hazard: the loop level is hard-capped below unity, there is a lowpass on the return path so each pass gets darker the way real room feedback does, and DECAY attenuates every pass so the loop dies away instead of building.
KILL zeroes the whole thing instantly. That is the control you actually reach for on stage.
Send approaches it differently, with a one-sample-delayed internal bus that lets a signal feed itself without the loop closing inside a single sample. It self-oscillates happily and it cannot run away.
The rules every module follows
- CV and an attenuverter on every knob. Anything you can turn, you can modulate, bipolarly, with the attenuation next to the jack.
- V/OCT pass-through everywhere. Modules chain without a multiple.
- Bounded outputs. Every audio path ends in a saturator or a limiter. Nothing in the set can produce a spike that damages anything downstream.
- Parameter order never changes. New parameters are appended, never inserted — VCV Rack serialises by position, and inserting one would silently corrupt every saved patch.
What arrived recently
The last three additions each cover a gap rather than adding another oscillator.
Street Grid Clock is the first clock in the set — before it, a patch built only from these modules could not generate its own time. Sitar Grid brings modal sequencing and a physical string model with a jawari bridge, three sequencers running at independent lengths so it stops repeating in any obvious period. Swarm Core plays recorded insects, which turns out to be the same massed-voice problem solved by evolution instead of by DSP.
Practicalities
Nineteen modules, 302 HP, MIT-licensed, free. The Windows build is a 248 MB download — almost all of it Swarm Core's sample bank. macOS and Linux build from source; cross-building is not set up, so there is no binary rather than a broken one.
It is not in the VCV Library, so installation is manual. The Installation page covers it, including the one mistake everybody makes: the plugin folder has to be named amplified-futures, and if it is not, the modules never appear and Rack says nothing about why.