---
title: "Beat-synced soundtrack kit by Scenario — voice & voiceover tutorial and best practices for AI video"
description: "Music at an exact BPM and a 12-sound SFX kit from Scenario MCP, a beat-grid fit, and a -14 LUFS mix on the beat. Install in Claude Code, Codex and Grok with one command."
url: https://varg.ai/how-to/skills/scenario-beat-synced-soundtrack
updated: 2026-10-30
---

# Beat-synced soundtrack kit by Scenario — guide skill for voice & voiceover

> **For AI agents:** find the right skill in one call — `curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>"` returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add `.md` to its path. Before generating anything, show the person the options on the how-to page. From a terminal: `curl -fsSL https://varg.ai/install | sh` installs the vargai CLI (https://varg.ai/how-to/cli.md): `vargai search` and `vargai install <slug>` with no key, every varg model after `vargai login`.

Music at an exact BPM and a 12-sound SFX kit from Scenario MCP, a beat-grid fit, and a -14 LUFS mix on the beat.

- **Type:** [Guide](https://varg.ai/how-to/guides.md) · [Voice & voiceover](https://varg.ai/how-to/browse/skills/audio.md)
- **Slug:** `scenario-beat-synced-soundtrack`
- **Status:** live
- **Publisher:** Scenario (vendor)
- **License:** Scenario: free download, no license file in the ZIP; fonts SIL OFL
- **Source:** https://www.scenario.com/explorations/kinetic-film/
- **Works in:** Claude Code, Codex, Cursor
- **Requires:** Claude Code or another agent that reads SKILL.md, Scenario MCP (music, sound effects, pictures), Node 18+ with Playwright (or Google Chrome), ffmpeg and ffprobe, Python 3 with numpy and opencv-python-headless
- **Cost:** Scenario credits for the music and every sound (each second of sound costs), priced with a dry run before each run; the beat grid and the mix are local and free.
- **Runtime:** minutes: two music takes and about 15 sounds generate in parallel; beat_grid.py and mix.py run in seconds

## What it does

Part of Scenario's Kinetic Film skill (https://www.scenario.com/explorations/kinetic-film/): the sound half of a kinetic launch film. Through Scenario MCP it picks a music model with a tempo input (recommend, model_schema_get, a priced dry run), generates two instrumental takes at the film's exact BPM, and generates a kit of one-shot sounds (whooshes, impacts, a big impact, hit, slash, key clicks, pop, tick, riser, glitch, logo hit) plus custom sounds for the film's own joke. beat_grid.py measures each take (tempo, first strong downbeat, downbeats, drop candidates with their rise, grid strength) so you choose the offset that lands the drop on the planned beat. mix.py places every cue from the film's own cue list, aligned on each sound's attack, ducks the music under big hits, cuts the music dead on the logo beat, masters to -14 LUFS / -1 dBTP in two passes, and lists any hit quieter than the music.

## Use when

- Music for a promo or teaser that must hit exact cut points (a drop on beat 8)
- A consistent sound-effect kit for motion graphics: whooshes on whips, hits on cuts, key clicks on typing
- Measure the tempo, downbeats and drop of a generated or licensed track before editing to it
- Mix music and sound effects for a social video to streaming loudness

## Not for

- Songs with vocals or lyrics
- Voiceover and dialogue (use a speech model)
- Licensing commercial music

## Install

- Claude Code: `curl -fsSL -o /tmp/scenario-kinetic-film.zip "https://iieqkinwjorxrzjp.public.blob.vercel-storage.com/explorations/2217ab5dbd452185a5c3c1509994da46b1ac20d2ebeaa888536244372ce73ebf/scenario-kinetic-film-skill.zip?download=1" && unzip -oq /tmp/scenario-kinetic-film.zip -d ~/.claude/skills && claude mcp add --transport http scenario https://mcp.scenario.com/mcp`
- Codex: `curl -fsSL -o /tmp/scenario-kinetic-film.zip "https://iieqkinwjorxrzjp.public.blob.vercel-storage.com/explorations/2217ab5dbd452185a5c3c1509994da46b1ac20d2ebeaa888536244372ce73ebf/scenario-kinetic-film-skill.zip?download=1" && unzip -oq /tmp/scenario-kinetic-film.zip -d ~/.codex/skills && codex mcp add scenario --url https://mcp.scenario.com/mcp && codex mcp login scenario`
- Any MCP client: `https://mcp.scenario.com/mcp`
- Any agent, with the vargai CLI: `vargai install scenario-beat-synced-soundtrack`

## Example output

https://s3.varg.ai/files/4325c578-378d-4c43-ae53-3cc3218df1ea/file_acb97730770a.mp4 — "Just code" with its soundtrack drawn under the picture: the envelope of the real mix and the 174 BPM beat grid scrolling past the playhead (beats 66-77 and 133-143). (credit: Scenario, "Just code" (Kinetic Film); waveform and beat-grid overlay drawn by varg, https://www.scenario.com/explorations/kinetic-film/)

## More voice & voiceover skills

- [ElevenLabs Text to Speech](https://varg.ai/how-to/skills/elevenlabs-text-to-speech.md): Turn scripts into natural voiceover in 90+ languages with ElevenLabs voices, streaming and stitching.
- [ElevenLabs Voice Changer](https://varg.ai/how-to/skills/elevenlabs-voice-changer.md): Re-voice a recording in a different voice while keeping its emotion, timing and delivery.
- [AI Voice Cloning](https://varg.ai/how-to/skills/inference-sh-ai-voice-cloning.md): Generate expressive voices in 100+ languages with Inworld TTS-2, ElevenLabs and more, from one CLI.
- [Dialogue Audio](https://varg.ai/how-to/skills/inference-sh-dialogue-audio.md): Produce multi-speaker dialogue audio with speaker tags, emotion and pacing for podcasts and explainers.
- [ElevenLabs Dialogue](https://varg.ai/how-to/skills/inference-sh-elevenlabs-dialogue.md): Generate a whole multi-voice conversation as one audio file with ElevenLabs, from a script.
- [ElevenLabs TTS (inference.sh)](https://varg.ai/how-to/skills/inference-sh-elevenlabs-tts.md): Voice scripts with ElevenLabs v4, v4 Turbo, v3, Multilingual v2 and Flash, via inference.sh.