September 19, 2026

Underscore Music: What It Is, How It Differs From a Score, and How Editors Use It

Underscore music is the instrumental music that plays beneath dialogue, narration and action to shape how a scene feels without drawing attention to itself. Unlike a featured song, it is designed to sit under other sound, follow the emotional arc of the scene, and stop or change exactly when the picture does.

Underscore Music, Explained for Editors

Most guides to underscore are written for composers. They talk about motifs, orchestration and spotting sessions, and they leave the person who actually places the music, the editor, with nothing to use. This page takes the other route. It defines underscore plainly, separates it from the terms it gets confused with, explains the four jobs it does in a scene, and then gets practical: how loud it should sit, why it must share a key with everything around it, and why finished library songs make such poor underscore in the first place.

Underscore vs Score vs Soundtrack vs Source Music

Four words get used as if they meant the same thing. They do not, and the differences matter when you are briefing a client or searching a library.

Score is all the original music written for a picture: title music, themes, end credits and everything under the scenes. It is the whole body of work.
Underscore is the part of the score that plays under scenes, especially under dialogue. It is the largest part of most scores by running time and the least noticed.
Soundtrack, in everyday use, means the songs associated with a film, usually with vocals and named artists, often sold as an album. The underscore is rarely on it.
Source music (also called diegetic music) is music the characters can hear: the radio in the kitchen, the band at the wedding. Underscore is the opposite; it exists only for the audience. Film theory calls this non-diegetic sound.

A useful test: if a character could turn it off, it is source music. If it would appear on the album, it is probably soundtrack. If nobody in the audience could hum it afterwards but the scene would fall flat without it, it is underscore. A recognisable song placed in a scene on purpose sits at the other end of the scale, and it is licensed very differently — here is how needle drop music is licensed and counted.

The Four Jobs Underscore Does in a Scene

Underscore earns its place by doing work the picture and dialogue cannot do alone. In practice that work falls into four jobs, and a single cue often does two or three of them at once.

Bed. A low, steady layer that gives the scene a floor. A sustained pad under an interview stops the room from feeling dead and makes cuts between takes less audible.
Pacing. A pulse or rhythmic loop that sets the scene’s heartbeat. The same three shots feel urgent over a fast pulse and reflective over a slow one.
Subtext. Music that tells the audience what a character is not saying. A calm conversation over an uneasy drone reads as a conversation about to go wrong.
Transition. A swell, riser or change of layer that carries the audience from one scene or mood into the next without a hard edge.

Notice that none of these jobs is “play a good song”. Each one is a specific, time-bound task tied to what the picture is doing at that moment. That is the reason underscore and library songs so often fight each other, and it is worth understanding before you open a music search.

Why Finished Songs Make Poor Underscore

A finished library track is written as a complete emotional journey: an intro, a build, a drop, a resolution, usually over three or four minutes. That shape is right for a music video and wrong for underscore, because a scene has its own shape and it is almost never the same one. The interview runs 48 seconds; the track peaks at 1:10. The reveal lands on a cut; the song’s drop lands two bars later. So editors loop sections that were never written to loop, fade out in awkward places, and stack two songs hoping they share a key.

Underscore needs the opposite property. It has to hold one feeling for exactly as long as the scene needs, then move when the scene moves, on the editor’s timing rather than the composer’s. Composers achieve that by writing to picture. Editors can achieve it by working with the parts of the music instead of the finished whole, which is where the layered approach comes in.

A song is a finished journey you cut to fit. A kit is layers you stack and reshape on your terms.

How to Build Underscore From Layers

Layered underscore is built from a small number of stems that share one key. A typical stack has three parts: a foundation (a drone, texture or pad that gives the scene its floor), a mid layer (a rhythmic loop or slow chord movement that sets the pacing) and a top layer (a melodic phrase or signature sound used sparingly, where the scene turns). Bonus elements such as tonal hits, risers and transitions punctuate the edit between sections.

Because the layers share a key, they stack and cross-fade without clashing, so the editor decides when each one enters and leaves. Under a long interview you might run the foundation alone for a minute, bring the mid layer in as the story tightens, and save the top layer for the single line that matters. This is what the guide to tonal sounds for video editing calls scoring with building blocks rather than songs, and a Soundtrack Kit is simply a cue delivered together with those blocks.

Every layer you’re mixing here is pulled straight from our Moods and Emotional Ambiances series — finished cues you can also break down into stems. Press play, then hit New mix to reshuffle; it all stays in one key, so the layers never clash.

Grab the packs the mixer is pulling from:

Moods Vol. 3

View pack →

Emotional Ambiances Vol. 2

View pack →

Moods Vol. 2

View pack →

Placing Underscore Under Dialogue: Level, Key and Space

Three things decide whether underscore supports a voice or fights it. Level: keep the music well below dialogue and ride it up only in the gaps between lines; many mixers work 15 to 25 dB under the voice and check on small speakers, because if you strain to hear a word the music is too loud. Frequency: the human voice lives roughly between 200 Hz and 4 kHz, so a foundation layer that sits low (drones, sub pads) leaves that space free, while a busy mid-range melody competes with every consonant. Key: everything tonal in the scene, including risers, hits and ambience beds with a pitch, should share the key of the underscore, or the audience hears something subtly wrong without being able to name it.

Space is the fourth habit worth building. Underscore does not have to be continuous. Dropping the music out entirely for one exchange, then bringing the foundation back, is often the strongest move in the scene. The sidechain and ducking techniques in the guide to mixing layered music under dialogue handle the rest.

Underscore Music Examples by Scene Type

The same layered stack produces very different underscore depending on which parts you use and when. A few common editorial situations:

Documentary interview: foundation only, very low, entering a few seconds after the first answer so the audience never notices it starting. Add the mid layer when the subject reaches the difficult part of the story.
Brand film voice-over: a warm pad plus a slow pulse for pacing, with a top-layer phrase reserved for the product reveal or the final line.
Wedding film, vows: an emotional pad under the words, nothing else, then a rising layer into the kiss and a full stack for the exit.
Tension or reveal: a low drone that sits under a calm conversation and tells the audience it is not calm, resolved with a tonal hit on the cut. The tension building sound effects guide covers that arc in detail.
Montage: mid and top layers together for energy, foundation removed, with transitions between the sections of the montage.

Kits organised by mood make this practical, because editors think in feelings, not technical terms. The packs below are the ones editors reach for most often when the job is underscore rather than a featured track: emotional beds and moods for dialogue-heavy work, and the cinematic kit for anything that needs to grow.

Reach for these kits to build it:

Emotional Ambiances Vol. 2

Where to Get Underscore Music You Can Shape

When you search for an underscore music library, check what you actually receive. A finished stereo file gives you one shape and a licence. A kit gives you the cue and its layers in one key, so the underscore can follow your scene instead of the other way round. Every layer is a plain WAV or MP3 that drags into Premiere, Resolve or Final Cut, and the free desktop app tags each one by key and tempo so you can audition combinations before you commit. Duende Sounds packs are a one-time purchase: you buy the kit once and keep using it, in client work included, without a subscription. The sound glossary defines the layer names used above if any are new to you.

Frequently asked questions

What is underscore music?

Underscore music is instrumental music placed beneath dialogue, narration or action to shape how a scene feels without calling attention to itself. It is non-diegetic: the characters cannot hear it. Good underscore follows the emotional arc of the scene and changes or stops exactly where the picture does, rather than running as a complete song.

What is the difference between underscore and score?

The score is all the original music written for a film, including main themes, title music and end credits. Underscore is the part of the score that plays under scenes, especially under dialogue. Every underscore cue is part of the score, but not every score cue is underscore: a main title theme played over black is score, not underscore.

What is the difference between underscore and soundtrack?

A soundtrack, in everyday use, is the collection of songs associated with a film, often released as an album, and usually featuring vocals and recognisable artists. Underscore is the instrumental music that sits under scenes and is rarely released on its own. A film can have a famous soundtrack and an underscore almost nobody consciously notices.

How loud should underscore music be under dialogue?

Low enough that every word is clear. Most mixers keep underscore well below dialogue, often 15 to 25 dB down, and then ride it up in the gaps between lines. The practical test is to listen on small speakers: if you strain to hear a word, the music is too loud or occupies the same frequency range as the voice.

Can I use royalty-free music as underscore?

Yes, but finished library songs are built as three or four minute journeys with an intro, a build and a drop, so they rarely sit still under a scene. Underscore works better when you have the building blocks of the music, such as a drone, a pad and a rhythmic loop in one key, so you can hold a feeling for exactly as long as the scene needs.

Do I need a composer to get underscore music for my video?

Not for most editorial work. A composer is the right call when a project needs bespoke themes. For interviews, documentaries, brand films and short-form pieces, editors build underscore themselves by layering key-matched stems, keeping a foundation layer under the dialogue and adding movement only where the scene turns.

Underscore is the music nobody is meant to notice and everybody feels. Build it from parts in one key, keep it under the voice, and let the scene decide when it moves. If you want to start with the building blocks rather than another finished song, browse the Soundtrack Kits and try the mixer above first.