Live-coded set arc: pacing, edit cadence, and energy design

The question every live coder faces after the first few patterns is not “what note next?” but “what is this set doing over time?” This module teaches the architecture of a set — how energy moves, when to hold back, what kind of edit belongs at which moment — so that individual code changes add up to a shape the audience can feel. A livecoding set that maximises energy from bar one and stays there is exhausting; one that treats silence and restraint as active arrangement choices earns the room’s trust.

The arc of the module mirrors a well-designed set. Start by internalising the mental model that a file save is a performance action, not a developer commit: the moment you press save, Strudel hot-swaps gaplessly and the audience hears the change. That shifts how you plan edits — each one is a musical gesture with a timing, a size, and a consequence. From there, work through the set-arc principles in order of a performance’s phases: establish a groove for a full phrase before developing it; build by adding voices and raising intensity while keeping one voice in reserve to subtract at the drop; reserve the single highest-energy move for the boundary you have built toward; and after the peak, drop energy to a genuine low so the following build has real contrast.

Alongside the arc principles, the edit-craft atoms give you the execution grammar. The four-tier diff-size ladder (micro param change / pattern edit / voice swap / section rewrite) maps onto the set phases: micro changes work inside a stable section, section rewrites are only warranted at a boundary when the target shares few voices. Edit cadence must track the grid — a two-bar save arriving on bar 1.5 sounds like a stumble, the same edit on bar 3 sounds intentional. One concept-id per edit keeps the diff legible and attributable to both the audience and your post-performance review.

The capstone integrates all of it: plan an arc on paper first (note where the peak goes and what is held in reserve), then execute it live under the constraint that every save is grid-anchored and every edit carries one concept. The post-performance brief is a self-audit: find each major energy move in the recording and name the principle it applied — this is how craft becomes repeatable rather than accidental.

Required atoms gate the capstone directly. Supporting atoms extend the picture: livecoding-no-offscreen-batch enforces that the arc must be built in real time in the running file (not pre-written and pasted), livecoding-imperfection-as-fuel reframes the moments when a build goes ragged as performance material rather than failure, and livecoding-copilot-vs-autonomous-cadence reminds agent-assisted performers to default one tier smaller on the diff-size ladder when uncertain about the moment.