The Sound Design Revolution in Modern Horror

Ask what changed in horror after 2010 and you will get answers about A24, about elevated horror, about trauma metaphors and slow burns and the return of the folk. All of that is true and most of it is about content. The change I would defend as the significant one is technical and almost invisible in a written description of any of these films: the genre worked out that the mix could do things the camera cannot, and it hired people who knew how.

The evidence is in the credits. Horror in the 1980s treated sound as delivery — the score arrived, the sting landed, the mix was a shelf you put things on. Horror since roughly 2013 treats it as the primary instrument of dread, with a specificity that would have been financially unthinkable when the genre’s budgets were built around latex. Some of this is Dolby Atmos, which rolled out from 2012 and put speakers overhead, giving a horror film a ceiling for the first time. Most of it is that a generation of directors grew up understanding that the audience’s ears are less defended than their eyes. You can close your eyes in a cinema. Everyone learns that at about seven. Nobody can close their ears.

Subtraction is the technique

The first move of the revolution was removal, and it is worth being precise about what gets removed.

Every recorded space has room tone — the ambient noise floor of a location, which an editor lays under dialogue so that the cuts do not click. It is inaudible by design. What modern horror discovered is that you can take it out, and the audience will feel the absence in their body without being able to name it. Silence in a film is the presence of a sound that should be there and is not.

A Quiet Place (2018) built a franchise on the principle. John Krasinski’s film is structurally a gimmick — a family that cannot make noise — and the reason it plays rather than curdles is the sound team: Ethan Van der Ryn and Erik Aadahl designed a mix in which the film keeps dropping to a floor so low that a single grain of sand on a wooden board becomes an event. It was the film’s only Academy Award nomination, for sound editing, which is the Academy accidentally identifying the film’s actual author. The other clever thing it does is switch perspective into the deaf daughter’s point of hearing — total silence, no room tone, no score — so that the audience is deprived of the sense the entire film has trained them to rely on.

Hush got there in 2016 with a fraction of the money and the same insight, and it is the better demonstration of the principle because it has no monster to distract you. When the film goes into its heroine’s silence, you notice how loud the rest of a normal thriller is.

The low end

The second move is the one that operates below the level of hearing.

Gaspar Noé used a low-frequency drone under the first half hour of Irréversible (2002) and has said openly that the intention was physiological — a sustained tone in the region where sound stops being heard and starts being felt in the chest and the gut. Audiences walked out of that film in numbers, and while the content did most of that work, the tone did some of it. It is the clearest statement of intent in the whole history of the technique: sound aimed at the body rather than the ear.

The idea has a folklore around it — the “fear frequency”, usually cited at about 19 Hz, traceable to Vic Tandy’s 1998 paper about a laboratory that seemed haunted until he found a fan producing a standing wave. The science is thinner than the anecdote and I would not build an argument on it. What is not in dispute is that a mix which sits information down at 20–40 Hz produces unease in a room with a subwoofer, and that horror has been quietly building its climaxes there for fifteen years. Watch the ending of a modern possession film with a good system and then again on laptop speakers. It is a different film. Half of it is missing, and the half that is missing is the half that was frightening you.

Somebody built an instrument for this

The single best fact in this argument is Mark Korven’s.

Korven scored Robert Eggers’s The Witch (2015) and wanted sounds that had no cultural associations — nothing an audience could identify as a violin doing a horror thing, because an identified instrument is an instrument the audience can defend against. Rather than reach for a synthesiser, he had the Toronto luthier Tony Duggan-Smith build him a physical object: the Apprehension Engine, a wooden box strung with rods, springs, a ruler and a rotating rubber wheel, designed to produce scraping, moaning and shrieking sounds that do not resemble music. It exists. It has been in a room with people. That is how seriously this era of the genre takes its sound: it commissioned new hardware.

The parallel run is what the composers started doing to acoustic instruments. Mica Levi’s score for Under the Skin (2013) uses strings that slide, detune and swarm — recognisably a viola and simultaneously wrong — which is the exact sensation the film is about. Colin Stetson’s work on Hereditary (2018) is a saxophone recorded with contact microphones on the player’s throat, so you hear the breathing and the keys as much as the note. That is a body making a sound in distress, which is the film’s subject rendered as a waveform.

Then there is Disasterpeace’s synth score for It Follows (2014), which is the hinge. Rich Vreeland came from games, and the score is loud, sixteen-bit and unembarrassed, and it made a generation of horror directors realise that Carpenter’s method was still live — that a synthesiser plus a good idea beats an orchestra plus a bad one, which was Carpenter’s argument in 1978 when he scored his own films because he could not afford anyone else’s.

Anatomy of a modern horror sequence

Take a scene apart and the architecture is legible.

A well-built dread sequence in a 2020s horror film runs roughly four layers. Underneath everything sits the drone — a sustained tone or a bed of processed noise, often pitched low enough to be felt rather than identified, and usually faded in so gradually across ninety seconds that no audience member can name the moment it arrived. Above that is the room: the ambience of the location, which the mixer is either thickening or stripping. Above that is the Foley — the specific, close-miked, unnaturally detailed sound of a hand on a bannister, a fabric shift, a swallow. And on top, if the film has any nerve, almost nothing.

The move that separates the good ones from the rest is what happens to the drone. It stops just before the scare, and the stop is the actual event — the audience has spent a minute and a half being tensioned by a sound they never consciously registered, and its removal produces a lurch that no sting could buy. The stab, when it comes, is landing on a nervous system that has already been dropped once.

Ari Aster’s films are ruthless about this, and so is James Wan’s, though Wan gets less credit because his branding is populist. Joseph Bishara’s Conjuring scores — Bishara also plays the demon, which is the sort of detail that tells you how integrated the department is on those productions — build and evacuate with the precision of a Swiss watch. And Skinamarink (2022) is the reductio: Kyle Edward Ball made a film in which the image is a corner of a dark ceiling for minutes at a time, so the entire text has migrated into the audio, and the thing is genuinely unbearable to sit with. It cost almost nothing. Sound was the whole budget’s worth of effect. If you want the field mapped by composer rather than by technique, the strongest work the genre has scored is a decent map of who was solving this and when.

The prosecution

I should not pretend the revolution has been all gain. Two objections, both fair.

The first is that a better mix has made the cheap shock more effective, not less common. If your system can deliver 6 dB more headroom, the temptation is to spend it on the stab. The loudness available to a modern horror film means a lazy director can produce a physical startle in an audience with no craft whatsoever, and many do. The instrument that lets Krasinski build a silence also lets a streaming horror film hit you with a violin at 90 dB because a cupboard opened.

The second is that it is not new. Peter Strickland made a whole film about this in 2012 — a film in which the horror is entirely the process of manufacturing horror sound, watermelons and all — and its subject was Italian genre cinema of the 1970s, an industry that was already doing everything I have just credited to the 2010s. Goblin’s work for Argento was doing whispered vocals, prog percussion and a mix that arrives before the image forty years before anyone said “sound design” in a marketing deck. The revolution is a rediscovery with better tools.

That is a genuine limit on my claim, and I will keep it. The point stands anyway: the tools got good enough that the rediscovery is now permanent, and the craft has moved from a small number of eccentrics to the default assumption of the genre. If you want to see the whole apparatus laid out, what a score does to an audience’s nervous system is the companion argument, and the Foley stage remains the most underrated room in horror — a place where a man snaps celery and a thousand people flinch.

Watch these films properly. Headphones, or a room with a sub, and the lights off. Half the genre’s last decade has been happening in a register your television is throwing away.