Prompt for a music generator: a 30-second instrumental for a short
A brief in the shape music generation tools read: structure by seconds, BPM, three instruments, no vocals, a drop that lands on your hook cut, and an ending that stops instead of fading.
Instrumental, 30 seconds, {BPM} BPM, {key} {major or minor}. No vocals, no lyrics, no vocal chops, no spoken words. Genre: {genre}. Instruments: {instrument 1}, {instrument 2}, {instrument 3}. One texture only: {e.g. 'tape hiss' / 'vinyl crackle' / 'clean'}. Structure, by seconds: 0-2 s: a single hit or a held chord, no drums (the first frame of the video needs air). 2-{drop} s: build, sparse, no kick; a rising element that stops exactly at {drop} s. {drop} s: the drop, full drums and bass, first downbeat exactly here. {drop}-{break start} s: full groove, one chord change every 2 bars. {break start}-{break end} s: breakdown, drums out, chords only. {break end}-28 s: the groove returns, same chords as the first groove. 28-30 s: a final hit that rings out and stops dead by 30 s. No fade-out. Mood as harmony: {e.g. 'major sevenths, warm' / 'minor with an added sixth, tense'}. Even dynamics; no risers, no cinematic hits, no reverse cymbals. Loop point: the audio at 30 s must match the audio at 2 s so the video can loop. Master to -16 LUFS integrated, true peak -1.5 dBTP, stereo, 48 kHz WAV.
Paste into your assistant as the first message of a new chat; fill the braces first.
Steps
- Set {drop} to the exact second of the cut right after your hook; read it from the timeline, not from memory.
- Generate 3-4 takes; pick by whether the first downbeat lands within one frame of your cut. Nudge the track in the editor by a few frames rather than regenerating.
- If the tool added something you did not ask for (a vocal, a riser), regenerate with the 'no' lines moved to the top of the prompt.
Tips
- Tools follow the structure by seconds loosely. Leave a one-beat hold on the shot at the drop so a small miss is absorbed.
- Ask for the drums as a separate stem if the tool offers it: the edit can mute them under speech and keep the chords.
- Where the video has live sound, the generated track gets a hole there, not a duck.
| works with | Any LLM: Model-agnostic: Gemini, Mistral, a local model. · Claude: Paste into claude.ai or the Claude app. · ChatGPT: Paste into ChatGPT; works with any GPT model. · By hand: No tool needed: a checklist or a technique. |
| cost | Free. No API, no key. |
| status | Unverified: written from the technique, not yet run end to end by us. |
| license | CC BY 4.0 for the text here; skills are MIT. |
| updated | 2026-11-01 |
| for agents | generated-music-30s-track-brief-prompt.md |
Related
Music brief: describe the track you need before you search for it
A fill-in brief that names mood as harmony, a BPM range, genre, length and where the drop must land relative to your hook, so a library search or a composer returns something that fits the cut.
Beat-sync cutting: cut on the beat, not on every beat
A rule and an agent prompt for cutting to music from a list of beat timestamps: cuts land on beats, phrases get room to breathe, and the best shot is held through the hook.
Score with a hole, not ducking, where something plays live
Cut the score to silence under live music, a laugh or a knock instead of ducking it; ducking is audible, a hole is not. The rule, the measured levels and the ffmpeg line.
Number hook: a count the viewer can track, and the one item that justifies it
Template for the hook built on a number (3 mistakes, 5 tools, 7 days), with the rule that the number must be small enough to finish and the last item must be named as the reason to stay.
Question hook: the opener the viewer answers in their head before you do
Template and rules for the question hook, with the test that separates a question that holds (the viewer has an answer and wants it checked) from one that is skipped (nobody cares either way).
Hooks for a product with nothing to show: text and voice carry the first 3 s
For an API, a CLI, a service, a habit, anything without a screen worth filming: five hook shapes built on a typed line, a number and a voice, and what goes on screen instead of the product.