Diegetic Sound

What is diegetic sound?

Diegetic sound is sound whose source belongs to the story world of the film. The simplest test: if a character standing in that space could hear it, it is diegetic. It is one of the first distinctions to learn in film sound, and film-music scholars such as Claudia Gorbman made the pair "diegetic" and "nondiegetic" the standard way to sort a soundtrack in the 1980s.

Diegetic sound falls into three families that every sound department recognizes:

• Dialogue — what the characters say to each other, on screen or through a door, a phone or a PA system.
• Effects and ambience — footsteps, doors, cars, weather, crowd murmur and the steady bed of a location. Much of it is re-created in post by Foley artists and sound editors, but it still counts as diegetic because it represents sounds of the world, not commentary on it.
• Source music — music with a source in the scene: a band on stage, a jukebox, a car radio, a character humming, headphones leaking a song.

Diegetic sound does not have to be visible. On-screen diegetic sound has its source in the frame: we see the trumpet player blow. Off-screen diegetic sound comes from the world just beyond the frame edge — a dog barking in the yard, a siren passing outside, a voice from the next room. Off-screen sound is one of the cheapest ways to make a set feel like a real place that keeps going past the frame line. Michel Chion calls a sound whose source we hear but don't see acousmatic, and a hidden voice can hold enormous power precisely because we can't see who is speaking.

Note what does not decide the category: the recording method. A line re-recorded in ADR and a punch performed on a Foley stage are both diegetic, because the story says they happen in the scene. Diegesis is about where a sound lives in the story, not where the microphone was.

What is non-diegetic sound?

Non-diegetic sound (also written nondiegetic, sometimes called extradiegetic) has no source in the story world. The characters cannot hear it; it exists to steer the audience. The main kinds:

• Score — the composed film score, from a full orchestra to a single synth drone. The two-note shark motif in Jaws (1975) is heard by us, never by the swimmers.
• Narration — a voice-over addressed to the viewer, like the dead screenwriter who narrates Sunset Boulevard (1950) or the unnamed narrator of Fight Club (1999).
• Songs laid over picture — the Pixies' “Where Is My Mind?” over the ending of Fight Club is not playing in that room.
• Stylized effects — whooshes on title cards, risers into a hard cut, a comic stinger, or music that imitates every movement on screen (Mickey Mousing).

The practical consequence: non-diegetic sound can do things diegetic sound can't. It can swell without anyone turning a knob, run continuously across cuts in time and space, anticipate what is coming and tell us how to feel before the characters know anything. That freedom is also why it can feel manipulative when it is overused — and why pulling it out entirely is such a strong choice.

Why do filmmakers use diegetic sound?

Because it carries information the audience trusts. Music from the score is the film talking to us; music from a radio is the world talking to the characters. That difference gives directors several distinct tools.

Grounding and realism

A scene built only from dialogue, effects and ambience feels observed rather than told. No Country for Old Men (2007) runs long stretches with almost no audible score, so the tension lives in wind, footsteps in a motel corridor and the rising beep of a tracking receiver. The audience leans in because nothing tells them how to feel.

World-building

Source music tells us what a place sounds like and who lives there. The cantina band in Star Wars (1977) makes Mos Eisley feel like a real bar with a house band before the trouble starts. The music at Rick's in Casablanca (1942) is part of the plot: Sam's piano carries “As Time Goes By,” and when Laszlo has the house band strike up “La Marseillaise” to drown out German officers singing “Die Wacht am Rhein”, the room itself becomes the battlefield.

Character

What a character chooses to play is characterization. Mia puts on Urge Overkill's “Girl, You'll Be a Woman Soon” at home in Pulp Fiction (1994), and the twist contest at Jack Rabbit Slim's is danced to Chuck Berry's “You Never Can Tell” playing in the restaurant. Radio Raheem's boombox blasting Public Enemy's “Fight the Power” in Do the Right Thing (1989) is his voice in the neighborhood — and the sound at the center of the final conflict.

Irony and counterpoint

A diegetic song doesn't care what happens on screen; it just keeps playing. Chion calls this anempathetic sound. In Reservoir Dogs (1992) Mr. Blonde tunes the radio to “Stuck in the Middle with You” and dances before the torture, and the breezy tune makes the scene colder precisely because it belongs to the room. When he walks out to his car, the song drops away; when he comes back inside, it returns. The soundtrack obeys the geography, and that makes the violence feel real.

Suspense

When sound is a threat inside the story, every diegetic noise becomes a plot point. In A Quiet Place (2018) a dropped object or a creaking floorboard can get someone killed, so the audience starts listening the way the characters do.

Diegetic music: source cues and songs the characters hear

Source music, often called source cues, is the most visible use of diegetic sound and the one people search for most. A few patterns cover most of it:

• The mixtape film. In Guardians of the Galaxy (2014) Peter Quill's Walkman and his “Awesome Mix Vol. 1” make 1970s pop part of the plot; the opening, where he dances through a ruin to Redbone's “Come and Get Your Love” on his headphones, sets the tone of the whole film. Baby Driver (2017) goes further: the hero plays music nonstop to drown out his tinnitus, and the action and cutting are timed to the songs on his iPods.
• The radio film. American Graffiti (1973) runs on a night of rock-and-roll radio heard from car radios all over town. To make it sound like it really came from cars and parking lots, sound designer Walter Murch played the music back through speakers in real spaces and re-recorded it — a method he called worldizing.
• The performance. A band on stage, a singer at a wedding, a piano at a party: the musicians are part of the blocking, and the music can set the scene's rhythm and pace.
• Source to score. A cue starts as a thin radio in a kitchen, then opens up to full-range stereo as the scene turns emotional or the film cuts to a montage. The song has crossed from the world into the score; audiences rarely notice the moment, but they feel it.

Diegetic music also cuts differently. A score cue can be edited or rewritten to fit the picture; a song playing in the room has to stay continuous across cuts, so an editor who trims three seconds from a shot must hide the jump in the music. That is why dialogue scenes with a band or a jukebox are almost always shot without the music playing, with the track laid in later. And a song heard on a radio still needs the same licenses as any other song in the soundtrack.

When the line blurs: trans-diegetic and internal sound

The categories are tools, not laws, and some of the best moments in film sound come from crossing them on purpose.

Trans-diegetic sound

Sound that switches category in the middle of a scene. The famous gag is in Blazing Saddles (1974): swinging big-band jazz plays over Sheriff Bart riding across the desert, and we assume it is score — until he rides past Count Basie and his orchestra playing in the middle of nowhere. Mad Max: Fury Road (2015) plays the same idea with a straight face: the Doof Warrior, a guitarist strapped to a truck stacked with speakers, plays a flame-throwing guitar during the chase, so the pounding music we take for score has a performer in the convoy, and Junkie XL's orchestral score weaves in and out around it.

Internal (subjective) diegetic sound

Sound from inside a character's head: thoughts, memories, a ringing ear. It belongs to the story world, because it happens to the character, but only one character hears it. On Omaha Beach in Saving Private Ryan (1998), after a nearby blast, the battle collapses into a muffled roar and a high ring, and for a few seconds we hear the war the way Captain Miller does. A Quiet Place drops to near silence in scenes told from the point of view of Regan, the deaf daughter, then snaps back to the full room when the film cuts away from her.

Character narrators

A narrator who is also a character, like Joe Gillis in Sunset Boulevard, comes from inside the story, but speaks to no one in the scene. Theorists argue over what to call it; in practice it is mixed like non-diegetic sound: dry, close and independent of the room on screen.

Displaced sound

A sound from the story world can also play over a different time or place: a phone ringing that bleeds into the next scene, a line of dialogue that begins before the cut. Those techniques have their own pages, sound bridge and asynchronous sound. The sound stays diegetic; it is just not locked to the shot we are watching.

How to plan, record and mix diegetic sound

Getting diegetic sound right starts in pre-production, long before the mix.

1. Mark every cue in the script. For each piece of music and each key sound, write down whether it is SOURCE (in the world) or SCORE (for the audience), what the source is and whether it is on or off screen. If the source is on screen, the art department has to put it there: a radio, a jukebox, a band whose instruments match the recording.
2. Keep the set quiet. Location sound should capture clean dialogue; the jukebox, the party music and most background chatter are added in post, so the editor can cut freely. Actors dance or talk over silence, or to a click or low thump in an earpiece for timing.
3. Play back for performance. When actors sing or dance to a track on camera, shoot to playback so tempo, breathing and lip movement match the final song.
4. Record the space. Capture 30–60 seconds of room tone and wild ambiences at every location. They are the glue that makes added diegetic sound feel like it belongs in that room.
5. Futz the source. In the mix, shape diegetic music to its source: a phone call is band-limited to roughly 300 Hz–3.4 kHz, a small radio loses its low end, music through a wall loses its highs. Add the room's reverb and match perspective to the shot — louder and brighter in the close-up, quieter and duller as the camera moves away or a door closes.
6. Decide the crossover points. If a song will move from source to score, choose the cut or camera move where it opens up. Sound mixing can hide the change or make it a statement, but both work better when the picture offers a clean moment.

In FlashBoards, keep this map next to the picture: put each storyboard frame on the board with a note saying which sounds are source and which are score, and pin reference frames of the on-screen sources (the radio, the stage, the car stereo) beside the scene, so the art department, editor and composer all read the same plan.

Diegetic vs. non-diegetic sound

The one-question test: could anyone in the scene hear it? Yes means diegetic, no means non-diegetic. When the answer is "only one character," it is internal diegetic sound.

• Dialogue between characters — diegetic, on screen or off.
• A narrator speaking to the audience — non-diegetic.
• A song on a car radio — diegetic (source music).
• The same song swelling over a montage — non-diegetic.
• Footsteps re-created on a Foley stage — diegetic; the recording method doesn't matter.
• A composed orchestral cue — non-diegetic, unless an orchestra is playing in the scene.
• A character's remembered voice in their head — internal diegetic.

Diegetic sound vs. film score

The score comments on the story; diegetic sound is part of it. A composer can use a leitmotif to follow a character anywhere, across continents and decades; a jukebox can only play in the diner. The score can also know things the characters don't, which is why a single low note can make an empty hallway frightening while the characters notice nothing.

Diegetic dialogue vs. voice-over

Voice-over sits outside the scene's acoustics and speaks to us; diegetic dialogue carries the room with it — distance, reverb, the chance of being interrupted or overheard. Mixers treat that difference literally: diegetic dialogue gets the space it was spoken in, while narration stays dry and front and center.

Further reading

Claudia Gorbman, Unheard Melodies: Narrative Film Music (1987). Michel Chion, Audio-Vision: Sound on Screen (1994). David Bordwell and Kristin Thompson, Film Art: An Introduction.