Make a Music Spotting Sheet for the Actual Commercial Sample

A spotting sheet says where music goes, what it does while it's there, and where it stops. If it doesn't name the picture it was written against, it isn't a plan. It's a set of numbers that will be wrong by Thursday.

So: the file, not the memory. Not the cut your collaborator watched on a laptop in a hallway, not the version that played in the room before the client asked for one more beat of the reaction shot. The actual sample, at a stated length, read through a stated clock — and every cue in the sheet tied to that.

Two things get called a cue sheet in this business. One is the working list a director and composer build while the music is still being decided. The other reports music that was actually used: what it was, how long it played, who owns it. That second document is a different artifact with a different reader and a different job. This is about the first one. Nothing below prepares you to file the second.

Start with the file, not the music

Before a single time goes on the page, the sheet needs a header, and the header needs to be specific enough that a stranger could tell whether they're looking at the same movie you are.

• Picture: the filename, exactly as it appears, including the version suffix. bench_bowl_A_v4.mov, not "the bowl cut."
• Duration: the running length of that file. If your sheet implies 24 seconds and the file runs 26, someone will eventually notice.
• Sound state: whether the file carries temp music, a scratch track, or nothing. A composer who assumes the file is silent and then hears a temp cue will spot very differently.
• Time base: the clock the times are read from, and its frame rate and start time.
• Date and status: when this sheet was written, and whether it is a proposal or a record of something agreed.

That fourth item causes more confusion than the other three combined. On a short commercial sample there may be no embedded timecode at all — the numbers in the corner are just elapsed time from whatever player is open. That's workable, and often the honest choice for short-form, but only if everyone reads the same player. Two applications can display different values for the same frame, and a time copied from one into notes read against the other points at the wrong moment. Frame rate matters for the same reason, and if the project runs at 29.97 the drop-frame question decides whether an hour-long offset accumulates. State the convention in the header, once, in plain words. Then never wonder again.

You don't have to use timecode. You don't have to use anything in particular. You have to say what you used, and mean it.

And when the picture changes — because it will — keep the old file. Don't drop a new export onto the old filename and leave the existing notes sitting there, apparently describing a movie nobody has seen. The old sheet keeps the old picture's name and date; the new sheet gets a new header. If you overwrite the file behind an existing set of notes, you have destroyed the only thing that made those notes checkable.

Watch for the job, including the gaps

Spotting is not a hunt for pretty moments. It's a pass through the picture asking one question repeatedly: where would music change what the viewer expects, or notices, or feels about the relationship between what happens and how someone reacts to it?

Three passes tend to work better than one long sitting. The first pass, watch with your hands off the keyboard. No notes. You're learning the shape. The second pass, write in plain language what each moment is doing — "this is where we start to wonder whether she'll hand it over" — with no times at all, because a time will make you think you've decided something. The third pass is where numbers appear, and by then you know what you're putting numbers next to.

Two habits are worth building early.

The first is noting absence. Where music stops, or where it never starts, is a decision as real as any entry. An unaccompanied stretch after a cue ends can make the room feel bigger, colder, more exposed. If you don't write that down, someone downstream will assume you forgot.

The second is refusing the automatic accent. A cut is a place where a musical event could land. It is not a place where one must. Dialogue beats, scene changes, and shot boundaries all look like invitations, and a score that hits all of them sounds like a metronome with a melody. The question isn't "is there a cut here?" but "does anything change for the viewer here?"

Say what the music should do. Don't reach for a reference track. "Something like the theme from that thing" tells a composer what you hummed in a meeting; it doesn't tell them what job the music has. Describing the job — this should feel like a question with no answer yet — gives them something they can actually solve.

Here is a picture to work on. It's invented for this piece, so there's no client and no confidential anything; it's just short enough to hold in your head.

A night workshop. Twenty-four seconds, silent, no temp score.

• 0:00–0:06 — Wide. A maker lifts a cooled bowl off a shelf and sets it on a bench.
• 0:06–0:11 — Close on her hands turning the bowl. At 0:11 her hands settle and she begins to lift her head.
• 0:11–0:14 — She looks up, off-frame left. At 0:12 her gaze leaves the bowl.
• 0:14–0:20 — A buyer enters frame left, crosses to the bench, picks up the bowl, turns it over, inspects it.
• 0:20–0:24 — Wide. The buyer exits frame right carrying the bowl. The maker watches. The picture ends at 0:24.

The plain-language pass gives you three things to write down. The close shot on the hands is where we're asked to care about an object. The look up is where the film changes its mind about what matters — the object becomes less interesting than whoever is about to walk in. And the inspection is where the film asks us to hold still and watch someone make up their mind, which is not the same experience as the first two moments and probably shouldn't borrow their music.

Now those get times.

Give every cue an identity it can survive on

A cue row needs more than a start and a stop. Six fields will carry you a long way.

A cue ID that belongs to the cue and not to its position in the film. SP-01 stays SP-01 even when it moves. If you renumber cues every time the picture shifts, you've made a list that only works on the day you wrote it.

An entry and an exit, each with a note about what kind of boundary it is.

Any consequential internal change — a moment inside the cue where something is supposed to happen. A shift in register, a change of instrument, a held note that releases. If the cue spans four seconds and something matters at 0:12, that's not a detail; that's a second boundary that just happens to sit inside the first.

A short statement of function, in a sentence a composer can disagree with.

A stop type, because "ends at 0:14" is not enough information. An abrupt stop and a fade to nothing do different things to a viewer, and a pulled-down ending is a third thing again. Write which one you mean.

A status, so that a decided boundary and a guess don't look identical on the page. Agreed, provisional, open. The precision of a timestamp is not evidence that anyone agreed to it; 0:12 looks authoritative whether or not a human being has ever considered it.

One more distinction worth making in the sheet, because it catches people: the end of a cue is not the end of the file. If the music runs out at the final frame, say so deliberately, so nobody mistakes an unfinished cue for a planned one.

Here's the sheet for the twenty-four-second version. Times are written to the second, because in this example no boundary needs frame precision. Where a boundary has to land on an accent, the row should carry the frame value and a name for the event, so a reader can look at the picture and see immediately whether the number still describes anything.

Spotting sheet — version A Picture: bench_bowl_A_v4.mov — 0:24 — silent, no temp score Time base: player display, 25 fps, start 0:00 Status: proposal, not yet reviewed

Note that SP-01's exit and SP-02's entry are the same moment. Two cues sharing a boundary is fine, and the sheet should say so out loud — otherwise the composer can't tell a deliberate handoff from two people who each assumed the other had written something there. It also means that if the picture ever moves that moment, both cues move together.

Check the sheet against the actual playback

A sheet isn't finished when the page stops. It's finished when someone has read it while watching the named file.

That read-through has a specific purpose: sorting each boundary into one of three kinds, because they get treated differently.

• A fixed picture event — the buyer's frame entry, a door closing, a hand leaving a surface. The boundary belongs to the picture. If the picture moves, the boundary moves with it, and the music has to be built to stop there.
• A desired experience — "the figure should not resolve," "the silence should feel too long." The exact moment is negotiable; the effect isn't.
• A suggestion open to composition — a rough placement offered because a conversation needed somewhere to start. Marking these as suggestions is not weakness. It's the difference between a brief and a cage.

Do this with the composer, or whoever owns the musical decision, watching the same file. Write down what was settled and what wasn't, in separate places. The temptation is to flatten both into one clean-looking column, and a clean-looking column is how an open question turns into an assumed decision that nobody ever made.

This is where the composer keeps authorship. The sheet's job is to make the questions specific enough to answer — this is the moment we stop caring about the bowl; make that audible — not to specify the notes. A sheet that has fully determined the music has stopped being a plan and started being a score, badly.

One limit to keep in view. There is a widely-read composer's account of this workflow — Heather Fenoughty's guide to spotting cues, on her site — that describes deciding where music is needed, noting cue boundaries, and keeping a usable cue list as the film changes. That's the mechanism: need, boundaries, a list that survives revision. It isn't a fixed standard, and it says nothing about which times or formats any particular production should use. The account was read in September 2025; a later attempt to reach the page returned a verification screen rather than the article, so the description above rests on that earlier reading and hasn't been re-confirmed since. Treat it as one experienced composer's description of a practice, which is what it is. The sheet you build still has to fit your own production.

And still nobody has reviewed the invented sheet above. That review — a composer reading it back against the picture and saying whether the question is answerable — is the step that turns a tidy table into a usable document.

Re-spot the new cut without erasing the old one

The edit arrives. Say the only change is that the look-up in the third shot is held two seconds longer after her gaze leaves the bowl, pushing everything after it later.

Version B — bench_bowl_B_v5.mov, 0:26. Shot one and shot two are untouched. Shot three now runs 0:11–0:16. The buyer enters at 0:16, the inspection runs 0:16–0:22, and the final wide runs 0:22–0:26.

Now, do not add two seconds to every time in the sheet. That would move SP-01 for no reason — it would appear to start at 0:08, and the cut into the close shot is still at 0:06. An edit that changes the picture internally is not an offset. It's a change to one part of the movie, and the parts before it are exactly where they were.

The reissued sheet keeps the version A sheet intact, under its own filename and date, and gets a new header.

Spotting sheet — version B Picture: bench_bowl_B_v5.mov — 0:26 — silent, no temp score Time base: player display, 25 fps, start 0:00 Status: reissue of the version A sheet. Version A retained.

Three things happened here, and only one of them is interesting.

SP-01 didn't move, because nothing it touches moved. The cue's identity survived and so did its boundaries.

SP-02's exit moved, and it moved because a fixed picture event moved. The buyer still enters frame; the wait is just longer. The cue's reason for existing is intact, so the ID stays. What's genuinely new is the question sitting in the middle of the cue: two extra seconds of held look, after her gaze has already left the bowl and before anyone arrives. Does that stretch want the same material sustained, a delay before the next move, or something else entirely? The function is settled. The music for that gap is not. Writing "open" in the status column is the whole point of having a status column.

And the silence moved, because everything after an inserted two seconds moves. Same reason as before: the inspection should feel unaccompanied, and the film should still end unaccompanied. If you had shifted the sheet by a uniform offset, SP-01 would have been wrong, and if you had shifted nothing, SP-02 and the silences would have been.

One conditional worth naming, because it's the case that trips people up later. If the edit had gone the other way — if the look-up had been trimmed to nothing and the buyer walked in on the heels of the close shot — SP-02 wouldn't have a moved exit. It would have lost its reason. A cue that exists to make the viewer feel an approaching presence has nothing to do if the presence arrives instantly. In that case you retire the ID and write down why. You don't reuse SP-02 for whatever music now covers that stretch instead. The ID is the cue, not the slot.

What comes out of all this is a plan someone else can follow. It names the file it describes. It says which boundaries belong to the picture and which belong to taste. It keeps the earlier version around instead of quietly rewriting it, and it tells you, in the change column, exactly what moved and why.

Its authority doesn't come from the time column looking tidy. Anyone can produce tidy times. It comes from the fact that a composer, watching a specific file with a specific running length, can read the sheet and know what is being asked — and two seconds of held look, in a cut nobody has scored yet, is a question worth asking properly.

Frequently asked questions

What has to be in a spotting sheet's header?

The picture filename exactly as it appears, including the version suffix — not "the bowl cut" but bench_bowl_A_v4.mov; the running length of that file; the sound state, meaning whether it carries temp music, a scratch track or nothing, since a composer who assumes silence and then hears a temp cue will spot very differently; the time base with its frame rate and start time; and the date and status, saying whether the sheet is a proposal or a record of something agreed. If the sheet implies one duration and the file runs another, someone will eventually notice.

Why does the time base cause so much confusion?

On a short commercial sample there may be no embedded timecode at all — the numbers in the corner are just elapsed time from whatever player is open. That's workable, and often the honest choice for short-form, but only if everyone reads the same player, because two applications can display different values for the same frame and a time copied from one into notes read against the other points at the wrong moment. Frame rate matters for the same reason: at 29.97, the drop-frame question decides whether an hour-long offset accumulates. State the convention in plain words in the header, once, then never wonder again. You don't have to use any particular convention; you have to say what you used and mean it.

Should a musical event land on every cut?

No. A cut is a place where a musical event could land, not one where it must. Dialogue beats, scene changes and shot boundaries all look like invitations, and a score that hits all of them sounds like a metronome with a melody. The question isn't "is there a cut here?" but "does anything change for the viewer here?" The related habit worth building is noting absence: where music stops, or where it never starts, is a decision as real as any entry, and an unaccompanied stretch can make the room feel bigger, colder, more exposed. If you don't write that down, someone downstream will assume you forgot.

What do I do with the sheet when a new cut arrives?

Reissue it with a new header naming the new file, and keep the previous sheet intact under its own filename and date rather than rewriting it — and never drop a new export onto the old filename with the old notes still sitting there. Don't add a uniform offset to every time: an edit that changes the picture internally is a change to one part of the movie, and the parts before it are exactly where they were. In the example, SP-01 stayed put because nothing it touches moved; SP-02's exit moved because a fixed picture event moved; the silences moved because everything after the inserted seconds moves. New uncertainty inside a cue gets an "open" status. If a cue loses its reason — say the look-up is trimmed to nothing and the buyer arrives instantly — retire the ID and write down why rather than reusing it for whatever music now covers that stretch; the ID is the cue, not the slot.

How much weight should the referenced composer's guide carry?

It's a widely-read composer's account that describes the mechanism: deciding where music is needed, noting cue boundaries, and keeping a usable cue list as the film changes. That's need, boundaries, a list that survives revision. It isn't a fixed standard, and it says nothing about which times or formats any particular production should use. The account was read in September 2025; a later attempt to reach the page returned a verification screen rather than the article, so the description rests on that earlier reading and hasn't been re-confirmed. Treat it as one experienced composer's description of a practice. Similarly, the twenty-four-second workshop sheet used as the example is invented and hasn't been reviewed — the review, a composer reading it back against the picture and saying whether the question is answerable, is the step that makes it usable.

Further reading

• Fit Temporary Music to a Commercial Sample: Re-edit the Structure or Stretch the Time? — Go here once the sheet is agreed and a temp cue has to stop where the picture stops, choosing between a structural edit and a time change.
• Add Captions to a Film Pitch Clip and Deliver the Right Version — Useful after a recut, when you need the same discipline on another time-coded track: cues belong to one edit and land late once the picture moves.