Scoring Session creative 1h 10m

W2D8Sc

The picture is locked. Can your sound make a stranger feel exactly what you choose β€” fear, awe, or grief β€” in 90 seconds?

Music + SFX Bed

β–Ά Enter Project

Context

Your locked trailer cut is finished and silent. A festival screening is one week away. The silent version tests dead with focus groups. Today you build the one thing that can save it: a single 60-second music cue plus timecoded SFX accents that lock the emotional read. The studio dub session is scheduled for tomorrow; your bed must be locked, versioned, and handoff-ready by end of day.

Mission

Compose and sync a single-cue emotional score to your locked trailer: a structured three-act music bed (tone bed β†’ riser β†’ BRAAAM hit β†’ button release) plus three named SFX accents (whoosh/hit/stinger), timecoded to picture frames and justified in a one-page score note that an external mixer can execute blind.

Finish Line

A timecoded Music + SFX Bed laid under your locked trailer, screened to the room with the lights down.

  • Music + SFX Bed

    lesson

    A timecoded audio plan - one music spine and its key SFX accents - mapped beat-by-beat to your locked trailer so an editor could drop it under picture tomorrow.

  • Conductor

    Sets the emotional target and holds it.

    • Before any generation, name ONE feeling the cue must land (dread, awe, defiance, grief, wonder β€” not multiple). Write it as: 'This bed must make the audience feel [feeling] because [one reason tied to the picture].' This is your veto lever for every take.
    • Reject any generated cue (no matter how 'cool') that doesn't serve that one feeling. You own the emotional brief; Composer generates against it, and you sign off when a take lands it.
    • At the sync stage, listen to the full bed + picture. If an SFX accent distorts the feeling (e.g., a cute whoosh in a dread bed), name it and veto. You're the arbiter of whether the bed does its one job.
  • Composer

    Architect of the three-act shape.

    • Once Conductor locks the target emotion and names the genre that carries it (e.g., 'low brass horror for dread'), write a generation prompt that specifies the exact shape: 'Create a 60-second piece: [0–20s] tone bed in [genre] establishing [feeling]; [20–40s] riser building to a peak; [40–45s] BRAAAM hit landing the climax; [45–60s] button stinger that releases tension.' Use the same prompt structure for every take so results are comparable.
    • Capture the output (URL, audio file, prompt used, generation timestamp) and name each take by version number and timestamp. Version 1.0 is the first take; Version 1.1 is a reiteration of the same prompt; Version 2.0 is a new prompt (new genre, pacing, or target emotion).
    • Judge each take against the prompt: does it have a tone bed, a riser, a distinct BRAAAM, and a button? If the output is a flat loop or single texture with no act structure, reject it and regenerate with the same prompt or iterate to Version 2.0.
  • Sound Designer

    Builder of the three impact accents.

    • Working from the Conductor's target emotion and Composer's locked cue, generate exactly three named sound effects in Suno.ai or ElevenLabs (if using voice layer): (1) the whoosh β€” enters 0.5s before the cut and sets up impact; (2) the hit β€” the loudest sound in the bed, lands on the exact cut frame; (3) the stinger β€” under the title card, releases the tension built by the hit. Each must be a separate layer, not merged into the music.
    • Record the name, purpose, and generation details (tool, prompt, timestamp, URL) for all three accents. The purpose is your design justification: whoosh earns the impact, hit lands the reveal as threat/victory/twist, stinger releases. If you can't name the purpose in one line, you're layering decoration, not accents.
    • Deliver the three accents as separate audio files so the Editor can layer them and mix levels. The hit must be the loudest element in the entire bed (never quieter than the music).
  • Editor

    Syncs picture to sound.

    • Layer the Composer's music cue and Sound Designer's three accents under the locked picture in your DAW (Audacity, Final Cut, Adobe Audition). Play it back and verify three sync points: (1) whoosh enters 0.5s before the cut; (2) hit lands on the exact cut frame (not the frame after); (3) button lands after the title card (within one frame). If any sync drifts, nudge the audio, not the picture.
    • Write the one-page score note with three lines, one per sync point: (1) Whoosh placement and why ('Whoosh 0.5s before cut to earn the impact'); (2) Hit placement and why ('BRAAAM hits cut frame 47 to land the reveal as [emotion]'); (3) Button placement and why ('Button stinger after title (frame 173) to release tension'). This note is what a studio mixer reads to reproduce the bed without watching your working file.
    • Once all three sync points are locked, export the full bed (music + accents mixed, no dialogue) as a stereo file and play it against picture one final time. If sync holds and the feeling lands as Conductor intended, the bed is locked.
  • Golden Trailer Awards

    Golden Trailer Awards

    The industry awards body for movie trailers β€” the exemplar bar for what a finished, professional-grade trailer looks and sounds like.

  • Runway AI Film Festival

    Runway

    The premier festival showcase of finished AI films β€” the "this is what pro AI filmmaking looks like" gallery. Complements the Golden Trailer Awards craft bar.