A track playing in a bar, a song heard through a car stereo and original score added in post do not occupy the same dramatic space. They often enter production through separate workflows anyway. Editorial places a temp song, clearance begins elsewhere, the composer develops underscore, and the dub eventually discovers that all three are making different claims about the scene.
I prefer to map that boundary before writing. The decision is not source music versus score. It is what each layer knows, who can hear it and how the film moves from one listening position to another.
Define the job before choosing the track
Source music belongs to the physical world of the scene. Characters may hear it even when the speaker or performer is not shown. It can establish place, period, social context or ownership. It may cover an awkward silence or create deliberate contrast with the action.
Underscore has a different freedom. It can follow interior point of view, bridge distant scenes or anticipate information unavailable to the characters. It does not need a plausible source in the location.
Before discussing genre, I ask what the music proves. “This song tells us the character planned the party” is more useful than “we need something fun.” “The score begins when he stops listening to the room and retreats inward” gives the transition a directorial purpose.
Build one music map
A playlist is not a music plan. Production needs a map that distinguishes licensed songs, custom source cues, score and the moments when one category transforms into another.
For each use, I want the picture version, timecode, intended source, dramatic function, clearance status and decision owner. That prevents one slot from being treated simultaneously as a song to license and a cue to compose.
The map also reveals repetition. If several scenes use radios or venues to establish period, the score may not need to reproduce every stylistic marker. It can create continuity at another level.
On a for-hire production, authority matters as much as taste. The director may choose the dramatic direction, but a producer still needs visibility into approval, cost exposure and replacement deadlines. A map makes those dependencies legible before they become emergencies.
Treat the temp as a hypothesis
A song can become structural very quickly. Its pulse shapes cuts, its lyrics colour performance and its artist brings an identity from outside the film. Even when everyone calls it temporary, picture starts depending on it.
I need to know whether the temp is a creative reference, a serious licensing target or simply an editorial tool. Those cases require different responses. A reference can be translated into dramatic function. A licensing target needs a replacement plan. An editorial tool should not quietly become a promise to the filmmaker.
If the track goes away, the team should identify what the cut loses: duration, accents, a lyric idea, energy or cultural recognition. The score should not imitate the song. It should address the need that the song exposed.
Design transitions as storytelling
Moving from source music into underscore can be a major narrative gesture. A cue begins in a speaker, opens into full bandwidth and continues after the character leaves. In the opposite direction, score may narrow until it seems to come from an object in frame.
That change is not only a mixing trick. It affects arrangement, key, tempo, acoustic perspective and sometimes production playback. If actors perform or dance to a track, the set needs a stable approved version. If score will quote a motif from source music, the production needs both a musical plan and the appropriate rights.
I ask where the audience should understand that the source has changed. The transition can be hard, gradual or deliberately ambiguous. Without a choice, it often becomes a volume automation move rather than a shift in point of view.
Divide responsibility without dividing the story
Director, picture editor, music supervisor, composer, producer, music editor and re-recording mixer may all touch the sequence. The number of people is not the problem. The problem is each department solving its own task against a different version of the plan.
Music supervision can track searches, rights holders, territories, term and intended use. The composer develops the score architecture and any custom source material. Editorial tests placement and duration. Production controls schedule and financial authority. The music editor and dub team manage final integration.
One person still needs to own the current music map. A licensed track can change after score approval, altering the neighbouring key, tempo or duration. That change must reach the composer before final deliverables, not during playback on the dub stage.
Plan versions and stems early
A source cue rarely needs only one master. Production may require a full playback version, a dialogue-friendly reduction, an instrumental, stems or an alternate ending. A custom song might also need a separate album master.
I define those outputs before final delivery. If camera and characters move through several spaces, the mixer may need vocals, percussion, bass and harmonic material separately. If underscore enters beneath the song, useful stems or an instrumental make the transition controllable.
Every deliverable needs a common timing reference, stable version name and clear approval status. A folder of files named “final” is not a delivery system. It is a small suspense film made specifically for post-production.
Recheck point of view after picture lock
A transition that worked in an earlier cut may become unclear after a few seconds disappear. A visible source may move off-screen. A song may continue long enough to feel non-diegetic even though the story intended it to remain in the room.
At picture lock I review boundaries as well as cues. Who hears the music now? What does the audience understand? Does the change reveal something, or is it merely decorative? Can the score retain its identity after a strongly branded song?
This review can also remove music. If a song has already established energy and point of view, immediate underscore may be redundant. A short return to production sound can make the next score entrance more meaningful.
Clear boundaries create more freedom
Source music and score are not departments fighting for the soundtrack. They are two listening positions. The film can move freely between them when the transition has a dramatic purpose and the workflow supports it.
One shared music map, explicit clearance status, designed transitions and agreed deliverables prevent late collisions. The composer can write without tracing the temp, editorial keeps its precision, and production knows which decisions remain open. The boundary becomes part of the filmmaking rather than a problem discovered at the dub.