Prompt lab

Seed Audio 1.0 Prompting Guide — Write Better Prompts, Get Cinematic Audio

A great prompt is not a sentence — it is a creative brief for sound. The more clearly you describe the scene, voices, dialogue, music, sound effects and emotion, the easier it is to generate audio that feels polished, cinematic and ready to use. This guide shows you the exact structure, formulas and prompt examples that consistently produce broadcast-ready output from Seed Audio 1.0.

Try a Sample Prompt →Jump to Examples

Audio placeholder: #1 — Hero opening demo

Type: Comprehensive cinematic showcase (radio drama opening)

Duration: ~20 seconds

Display: Inline audio player with waveform visualization

Caption shown above the player: "Generated from the full prompt below — one pass, no post-production."

Caption shown below the player: "Click to expand the prompt that created this audio ↓"

Prompt used (collapsible accordion under the player): "Create a cinematic radio drama scene opening. Setting: a stormy night inside an old coastal lighthouse. Heavy rain hits the windows, distant thunder rolls over the ocean, and the lighthouse lamp rotates with a low mechanical hum. Music: subtle cinematic score with deep strings, soft piano, low ambient drones. Mood: mysterious, emotional, suspenseful. Narrator: clear audiobook narration, calm but tense: 'On the night the lighthouse went dark, Clara found the letter her father had hidden for twenty years.' End with rising strings and a single foghorn."

Prompt system

Create a [type of audio] for [scenario].
Setting: [place + atmosphere].
Mood: [emotion].
Music: [style + instruments + intensity].
SFX: [key sounds].
Characters: [roles + performance direction].
Dialogue: [natural lines].
Ending: [final sound or emotional beat].

Structure

9 prompt elements

References

@Audio 1, @Audio 2...

Examples

6 ready-to-use prompts

Output

cinematic audio

What Makes a Great Seed Audio 1.0 Prompt?

Most audio AI tools treat your prompt as a single line of text that needs to be voiced. Seed Audio 1.0 is different — it reads your prompt the way a film director reads a treatment. That means your prompt should describe a scene, not just a sentence.

A strong Seed Audio 1.0 prompt has nine layered elements that work together to give the model creative direction without overloading it: audio type, setting, mood, music, sound effects, characters, dialogue, pacing and ending. Master these nine elements and you can move from "AI voice generator" output to genuinely broadcast-grade productions.

Weak prompt

Make an audio story.

Vague. No scene. No characters. No mood.

Strong Seed Audio 1.0 prompt

Create a cinematic radio drama scene for a serialized audiobook.
Setting: a stormy night inside an old coastal lighthouse.
Music: deep strings, soft piano, low ambient drones.
Narrator, calm but tense: "On the night the lighthouse went dark,
Clara found the letter her father had hidden for twenty years."
End with rising strings and a foghorn.

Specific scene. Layered direction. Cinematic outcome.

The Quick Seed Audio 1.0 Prompt Formula

Whenever you don't know where to start, use this formula. It works for every audio type — radio drama, ad, podcast, video dubbing, voice companion, game audio.

Create a [type of audio] for [scenario].
Setting: [place + atmosphere].
Mood: [emotion].
Music: [style + instruments + intensity].
SFX: [key sounds].
Characters: [roles + performance direction].
Dialogue: [natural lines].
Ending: [final sound or emotional beat].

That's it. Fill each line. Skip what you don't need. The model handles timing, transitions and mixing automatically.

Audio placeholder: #2 — Formula in action

Type: A short 60-second audio generated by literally filling in the formula above

Duration: ~60 seconds

Display: Inline player with the filled-in formula shown line-by-line next to the player

Caption: "Here’s the same formula filled in for a 30-second coffee brand ad — listen below."

The 9 Elements of a Strong Seed Audio 1.0 Prompt

Below is the full structure that consistently produces high-quality output. You don't need to use every element — but knowing what each one controls makes your prompts dramatically better.

1. Audio Format — Tell the Model What You're Making

Start by stating exactly what kind of audio you want. Common Seed Audio 1.0 audio formats:

A cinematic radio drama scene
A podcast conversation between two hosts
A 30-second brand advertisement
A video dubbing track for a sci-fi scene
An immersive game soundscape
A personal AI voice companion conversation
A serialized audiobook opening scene

Weak: Make an audio story.
Strong: Create a cinematic radio drama scene for a mystery audiobook.

2. Setting — Give the Audio a Place to Live

The setting defines the listener's space. Include location, time of day, weather, room tone, background ambience, distance and movement.

Weak: It is raining somewhere.
Strong: A stormy night inside an old coastal lighthouse. Heavy rain hits the windows, distant thunder rolls over the ocean, and the lighthouse lamp rotates with a low mechanical hum.

Audio placeholder: #3 — Setting Before/After A

Type: 15-second clip generated from the WEAK prompt above

Duration: ~15 seconds

Display: Left side of a two-column Before/After audio comparison component

Caption: "Weak setting prompt — flat, unspecific"

Audio placeholder: #4 — Setting Before/After B

Type: 15-second clip generated from the STRONG prompt above

Duration: ~15 seconds

Display: Right side of a two-column Before/After audio comparison component

Caption: "Strong setting prompt — spatial, cinematic, specific"

3. Mood & Emotional Direction — Guide the Performance

Mood tells Seed Audio 1.0 how the whole scene should feel, and how that feeling should change.

calm and intimate
tense and urgent
playful and warm
lonely and cinematic
premium and confident

Mood: calm and intimate at first, then gradually tense and dangerous as the hidden door opens.

4. Background Music — Direct It, Don't Just Request It

Music works best when you describe style, instruments, intensity and when it should rise or fall.

Style: cinematic, lo-fi, acoustic, electronic, orchestral
Instruments: strings, piano, guitar, synth pads, percussion
Behavior: quiet under dialogue, rising before reveal, cut to silence at final line

Music: subtle cinematic score with deep strings, soft piano and low ambient drones. Keep it quiet under dialogue, then rise before the final reveal.

Audio placeholder: #5 — Music direction Before/After A

Type: Weak music prompt

Duration: ~15 seconds

Caption: "Weak: Add music"

Audio placeholder: #6 — Music direction Before/After B

Type: Strong music prompt

Duration: ~15 seconds

Caption: "Strong: cinematic score with clear instruments and timing"

5. Sound Effects — Place Them at the Right Moments

Sound effects should support narrative beats. Place them where they happen instead of listing them randomly.

door creaks as she enters
thunder crack after the final line
coffee machine steams under the narrator
radio static before the transmission cuts out

SFX: rain on glass throughout. Thunder crack as Clara opens the letter. Foghorn after Elias says the final line.

6. Characters — Describe by Role, Not by Identity

Define characters by role, personality and performance direction. Do not ask for exact imitation of real people.

Narrator: clear audiobook narration, calm but tense.
Detective Ray: tired, sharp, low voice.
Maya: whispering, afraid, trying to stay composed.

7. Dialogue — Write Lines People Would Actually Say

Natural spoken dialogue beats stiff written dialogue. Use contractions, short sentences, pauses and emotional beats.

Weak: I am afraid because this situation is dangerous.
Strong: Something’s wrong. We need to get out of here.

Audio placeholder: #7 — Dialogue Before/After A

Type: Weak dialogue prompt

Duration: ~15 seconds

Caption: "Weak dialogue — stiff and written"

Audio placeholder: #8 — Dialogue Before/After B

Type: Strong dialogue prompt

Duration: ~15 seconds

Caption: "Strong dialogue — spoken and natural"

8. Pacing & Timing — Control the Rhythm

Pacing controls how quickly the scene moves. Tell the model when to pause, when music should rise, and when a beat should land.

slow opening, then faster dialogue
pause before the final line
music drops before the reveal
quick overlapping reactions in a podcast

Pacing: slow and tense for the first 10 seconds, then dialogue becomes urgent. Pause before the final reveal.

9. Ending — Land the Scene

A strong ending tells the model how to close the audio emotionally and sonically.

End with rising strings and one final foghorn.
End with a soft brand chime and fading coffee shop ambience.
End with radio static cutting to silence.

Advanced: Using Multiple Reference Audios in Seed Audio 1.0 Prompts

Seed Audio 1.0 supports multiple reference audios in a single prompt — so you can assign different cloned voices to different speakers in one generation. This is what turns a single-narrator output into a full-cast radio drama, podcast or commercial.

How to assign reference audios

When writing your Seed Audio 1.0 prompt, clearly state which character should use which reference audio. The recommended syntax:

Character Name: [role, personality, speaking style], performed by @Audio 1:
"Dialogue line here."

Host A, performed by @Audio 1, speaks in a calm and curious tone:
"Today we're talking about whether household robots would actually make life better."

Host B, performed by @Audio 2, replies playfully:
"Helpful? Sure. But I don't need a robot judging my midnight snacks."

5 best practices for multi-reference prompts

Give each character a clear role name (Detective Ray, Host A, Narrator…).
Add a short personality or performance direction next to the name.
Assign a reference audio (@Audio 1, @Audio 2, …) — and keep it consistent for that character throughout the prompt.
Write natural, spoken dialogue — not narrated explanations.
Add per-line emotion or tone guidance when the performance shifts.

Keep reference assignments consistent

Once you bind a reference audio to a role, do not switch it mid-prompt unless the story requires a clear character transformation (e.g. possession, disguise, flashback). Switching the same character between @Audio 1 and @Audio 2 will produce inconsistent voice identity.

Separate voice identity from audio effects

Reference audio controls who is speaking. Prompt text controls what happens around the voice: music, ambience, room tone, SFX and emotional direction.

Multi-reference prompt template

Create a [audio format] with multiple speakers.
Use @Audio 1 as [Character A].
Use @Audio 2 as [Character B].
Use @Audio 3 as [Narrator].

[Character A], performed by @Audio 1, [emotion]:
"[dialogue]"

[Character B], performed by @Audio 2, [emotion]:
"[dialogue]"

Audio placeholder: #9 — Multi-reference demonstration

Type: Multi-character scene with 3 reference voices

Duration: ~45 seconds

Display: Inline player below the multi-reference prompt template

6 Ready-to-Use Seed Audio 1.0 Prompt Examples

Each example below is a full Seed Audio 1.0 prompt you can copy, paste and modify. Listen to the real generated output, then adapt the structure to your own scene.