AI Music for Podcast Intros: Length, Mood, Loudness
Podcast intros should run 6 to 12 seconds for a solo show and 8 to 15 seconds for a multi-host or narrative show, land at about -14 to -16 LUFS integrated, and open with a short musical motif rather than a cold full mix. The usual mistake is dropping a 30-second generated track at the top; most listeners skip it before the host speaks. The framework below turns that into a repeatable brief: pick a length by show type, set a loudness target that matches the body, choose a mood that fits the host voice, and hand a single paragraph to any AI music generator.
The length framework
Length is the dial that most affects whether people stay. A short intro signals confidence; a long one reads as padding. Use the show type to set a ceiling, then cut until the music stops right as the host begins.
| Show type | Recommended intro | Why this range |
|---|---|---|
| Solo commentary | 6 to 10 seconds | One voice, no need to set a scene; get to the point |
| Two-host or panel | 8 to 12 seconds | Room for a small motif that frames the topic |
| Narrative or documentary | 10 to 15 seconds | Needs a beat to pull the listener into the story |
| Branded or corporate | 8 to 12 seconds | Must feel produced without stealing attention |
| Trailer or teaser | 15 to 20 seconds | Sells the whole show, so it can breathe |
Anything past 20 seconds should be a trailer, not a recurring intro. If your generated clip runs long, ask the tool for an exact length in the structure line and fade the last second.
Loudness targets
Loudness is measured in LUFS, not in raw volume. Match the intro to the body so the handoff to speech is smooth. Intros can sit a touch brighter and slightly louder than the body bed, but not so loud that the host has to fight it.
| Element | Integrated LUFS | True peak | Notes |
|---|---|---|---|
| Intro sting | -14 to -16 | -1.0 dBTP | Slightly brighter than the body, short |
| Body bed | -16 to -18 | -2.0 dBTP | Matches platform norms, stays behind voice |
| Ad or sponsor bed | -14 | -1.0 dBTP | Stands out just enough to read as a break |
| Outro | -16 | -2.0 dBTP | Fade rather than cut |
If you do not have a meter, the practical test is simpler: the music should feel like a door opening, not a wall. When the host starts, the music should already be stepping aside.
Mood by show type
Mood does the quiet work of telling the listener what kind of show this is. Match it to the host voice before you match it to the topic. A calm host over an aggressive track feels like a mismatch even when the genre is right.
| Show type | Mood | Tempo | Instrumentation |
|---|---|---|---|
| Solo commentary | Warm, calm | 70 to 90 BPM | Soft piano plus pad |
| Interview | Neutral, clear | 90 to 100 BPM | Light guitar plus shaker |
| True crime | Tense, low | 60 to 75 BPM | Drone plus sub bass |
| Comedy | Bright, bouncy | 110 to 125 BPM | Pluck synth plus claps |
| Business | Confident, clean | 100 to 112 BPM | Piano plus strings |
| Health or meditation | Airy, open | 50 to 65 BPM | Pad plus soft bell |
The copy-paste brief
This one paragraph works in any AI music generator. Change only the bracketed parts; keep the structure line so the tool knows exactly when to stop.
Podcast intro sting, 10 seconds, [mood] [genre], [tempo] BPM, no vocals, [instruments], structure: 2-second rising motif, hold, clean end at 0:10, exclude: lyrics, exclude: long reverb tail, exclude: vocals, exclude: sudden drop
Run it twice with the same wording before judging it. Generators include randomness, so the first take is a sample, not a verdict. Keep the version that landed closest to your length and reuse it as your house sound.
Where the intro meets the voice
The handoff matters more than the clip itself. A clean intro still fails if it crashes into the first sentence. Three habits fix most intros:
- Leave a gap. Stop the music 0.4 to 0.6 seconds before the host speaks so the ear has a breath.
- Duck, do not kill. If the intro carries under the opening line, drop it to about -22 LUFS so the voice stays on top.
- Fade the tail. End on a short fade rather than a hard cut, which reads as accidental.
Many editors place the full intro once, then a much quieter version under the recap at the close. That ties the show together without repeating the cold open.
Common questions
How long should a podcast intro be?
For most shows, 8 to 12 seconds is the safe band. Solo commentary can go shorter, narrative work can go longer, and anything past 20 seconds belongs in a trailer. Listeners decide fast, so the music should open the door and step back.
What LUFS should podcast music be?
Target -14 to -16 LUFS for the intro and -16 to -18 LUFS for any bed under speech, with true peaks near -1.0 to -2.0 dBTP. The goal is a smooth handoff to the voice, not maximum loudness. If the host sounds buried, lower the music rather than raising the voice.
Should podcast intro music have vocals?
Usually no. Vocals compete with the host from the first second and make the mix harder to duck. An instrumental motif is easier to place, easier to loop, and easier to fade. Save sung elements for a separate theme or a trailer.
How do I keep the intro from overpowering my voice?
Set the intro to end just before the host starts, duck any music that continues under speech to about -22 LUFS, and fade the final tail. If you still hear a fight, the intro is probably too long or too bright for the voice register.
Bottom line: write the length first, set the loudness to match the body, pick a mood that fits the host, and paste the brief above into any AI music generator. One short, well-placed sting does more than a long track nobody waits for.