There is no reliable percentage of a film that should be scored. A feature can carry music through much of its running time and still feel restrained. Another can use a handful of cues and make them feel enormous. The useful production question is not how many minutes a film is supposed to contain. It is where music adds a point of view that picture, dialogue and sound do not already provide.
I like to settle that question before a cue count hardens into a promise. It keeps the team away from two familiar traps: filling every open space because it feels unfinished, or treating silence as automatically more serious.
For a director, producer or editor, music density is part of the storytelling design. It also affects the schedule, approval chain, recording plan and final deliverables.
Give each cue a job
I start by writing one active sentence for every proposed musical section. Hold back the reveal. Connect two locations. Keep us inside a character’s private fear. Carry an ellipse without making it feel rushed. If the sentence only says “add emotion,” the brief is not specific enough to guide writing or notes.
A strong scene does not automatically need to remain unscored. Music may oppose what is visible rather than reinforce it. A calm performance can sit against an unstable texture. A fast montage can gain weight from a slow musical span. The point is to define the relationship, not to reward a scene for being important.
When no clear job emerges, I test the sequence without score. That is not a refusal to compose. It is a way to hear what the film already owns.
Plan the whole arc, not isolated moments
Spotting one scene at a time can make every cue sound defensible. The problem appears when those cues are assembled. Similar entrances accumulate. A theme returns before the audience has had time to miss it. Transitions acquire the same musical shape and stop telling us anything.
I map the film from beginning to end. I mark dialogue-heavy passages, sound-led sequences, major changes of perspective and stretches where the pace shifts. I also mark long gaps without music. Those gaps can become structural if the return of score is earned.
Duration is only one measure. Ten minutes of sparse sustained material does not carry the same density as ten minutes of sharply articulated sync points. The number of starts matters too. A score that constantly enters and exits can feel more assertive than one long continuous cue.
This map can remain provisional before picture lock. It still gives editorial, sound and music a shared view of the intended arc.
Treat silence as material
No score does not mean no sound. A scene may contain voices, room tone, traffic, cloth and movement. Removing music can bring bodies and locations forward.
I find it useful to separate three kinds of absence. Neutral absence means the scene carries itself. Expressive absence creates a break after a scored passage. Strategic absence holds music back so that a later return has greater weight.
At the spotting session, I ask a practical question: what do we hear when the score is gone? The answer should involve the production track and the intended sound design, not an abstract idea of silence. This helps the team avoid scoring over material that will eventually do the dramatic work.
Silence also needs continuity. If music covers every transition, each entrance loses value. If the score disappears for a long stretch without a legible reason, its return may feel imported. Presence and absence belong to the same design.
Do not let the temp decide by default
A temp track can help an editor establish duration, momentum or contrast. It can also turn an experiment into an assumption. Once every key scene contains a reference, it becomes easy to believe that every one of those scenes requires score in exactly that place.
I ask to screen selected sequences both with and without the temp. We note the actual change. Does the scene lose clarity? Does it become usefully ambiguous? Does a cut feel longer? Does the performance gain space? This comparison separates a dramatic need from listening habit.
Placement and musical language should also remain separate. A reference may reveal that the sequence needs continuous motion without dictating harmony, instrumentation or melody. That distinction gives an original score room to exist.
Test systems, not twenty unrelated cues
I group scenes by dramatic function. Investigative passages may share one rule. Intimate scenes may withhold the main idea until a specific turning point. Location changes may use a related texture without repeating the same cue.
Then I write a few representative mockups. The purpose is not to score the entire film early. It is to test whether the system survives repetition. A palette that works brilliantly once may become heavy on its third appearance. Discovering that before full production saves revisions later.
This also improves feedback. The team can say that music enters too early across one family of scenes, or that a planned absence weakens a progression. We can revise an underlying rule instead of patching every cue separately.
Connect density to scope
The amount of score affects workload, but raw minutes do not define a fee or schedule on their own. A short cue with precise synchronization may require more iterations than a long stable passage. Live players, thematic transformations, alternate versions and detailed stems also change the scope.
After an initial spotting pass, I can build a more useful estimate: approximate cue count, duration, complexity, likely revisions and delivery requirements. If editorial is still moving, I identify which cues are safe to develop and which depend on picture lock.
On a constrained production, fewer cues can become a deliberate strength. The answer is not to remove minutes evenly. It is to concentrate resources where music changes the viewer’s understanding. A rare, prepared entrance can establish more identity than a continuous neutral bed.
Approve architecture before polish
I recommend two layers of approval. First approve the architecture: where score begins, where it stops, what it connects and when an idea returns. Then approve the writing: pulse, texture, harmony, orchestration and sync.
When those conversations are mixed, a note about instrumentation may hide a more basic disagreement about whether the cue belongs in the scene. Separating them makes revisions clearer and protects the schedule.
Before picture lock, the music map can use ranges and confidence levels. After lock, it becomes a production document tied to timecode, versions and deliverables. The shape should survive even if individual details evolve.
The right amount of music is not a target percentage. It is a sequence of intentional choices. I want every entrance to have a reason, every withdrawal to leave something useful behind, and the full arc to remain coherent even for a viewer who never consciously notices the plan.