Ask someone to hum the theme from a horror film and, more often than not, they can't — not because the score wasn't effective, but because it usually isn't built to be hummed. Where an adventure theme or a love theme lives in melody, an effective horror or suspense cue tends to live somewhere else entirely: in a curve that won't settle. A filter that opens but never fully arrives. A drone that drifts a quarter-tone flat and stays there. A pulse that almost, but doesn't quite, lock to the edit.

That's not an absence of composition. It's a different compositional target — and it's the same target a modular synthesist is chasing every time they patch a slow LFO into a filter cutoff instead of playing a note.

Tension Is a Rate of Change, Not a Chord

Standard harmonic tension resolves. A dominant seventh wants to resolve to the tonic; an unresolved suspension eventually releases. Audiences have decades of trained expectation for how and when that release arrives, which is exactly why horror scoring tends to avoid it. The goal isn't a chord that's dissonant — dissonance alone reads as "aggressive" or "chaotic," not dread. The goal is a curve that keeps promising resolution and keeps not delivering it.

That's most reliably built with modulation, not harmony: an LFO on filter cutoff set slow enough that a listener can't consciously track it, only feel that something is shifting. A slow pitch drift — a few cents, applied to a sustained drone — that never lands on a clean interval. An amplitude pulse just slightly off the tempo grid, so it reads as almost-rhythmic without ever confirming it. None of these are notes. All of them are curves, and all of them are doing the emotional work a melody would do in a different genre.

Borrowing the Patch, Not Just the Sound

This is where modular synthesis becomes a genuinely useful reference model for scoring dread, independent of whether the source is a Eurorack rack or a software instrument's own modulation matrix. A modular patch is defined by its routing — which slow-moving source (an LFO, an envelope, a random generator) is controlling which parameter, and at what rate. Horror scoring benefits from thinking the same way:

Pick one slow-moving modulation source per cue — not three, not five. A single unresolved curve, sustained, reads as more unsettling than a pile of competing ones.

Set the rate below conscious tracking speed. If a listener can count along with the modulation, it reads as rhythm, not dread. Below roughly one cycle every four to eight seconds, it tends to register as atmosphere instead.

Route it to something structural, not decorative — filter cutoff, pitch, or stereo width, rather than a reverb send that barely registers.

The instrument doesn't need to be exotic. A single sustained pad under a slow filter LFO does more for a suspense scene than a full orchestral cluster chord, because the cluster resolves the moment it stops sounding surprising, and the filter curve simply keeps going.

Closing the Loop

This is the same argument this week's blog opened with, narrowed to a single genre: the curve is doing compositional work, not decorative work. Applied generally, it's the difference between a static cue and one that moves. Applied specifically to horror and suspense, it's the entire mechanism by which dread gets built — not through what note plays, but through what refuses to resolve.

Sample Pack - Rapture Cinematic Hits is a thunderous collection of high-impact sounds built for trailers, films, and games that need the moment tension finally breaks. Explore it at sonalsystem.com

More on composition, sound design, and the business of music in the Frequencies archive.