How to use Seed Audio 1.0 with tested prompts and playable results

Create a 20-second cinematic audio scene in English: two detectives whisper in a rain-soaked alley at midnight. Add distant traffic, soft thunder, wet footsteps, a low suspense pad, and a metal door slam at the end. Keep the dialogue clear and natural.

What this example shows

This is the strongest example for explaining Seed Audio as a scene generator instead of a plain text-to-speech tool.

• Name the setting, spoken style, ambience, music layer, and final sound event.
• Keep dialogue direction short so ambience has room to appear.
• A single ordered effect cue can work better than a long sound list.

Create a 20-second podcast welcome segment in English for a calm creativity show. Use one warm host voice, light studio room tone, a soft synth bed, and a clean broadcast-style ending. Keep the speech friendly, natural, and easy to understand.

What this example shows

The prompt works because it uses a simple role, a clear delivery style, and a narrow show format.

• Use one speaker when voice clarity matters.
• Mention room tone and music bed separately.
• Ask for a clean ending when the clip will be edited into a timeline.

Create a 20-second looping fantasy game ambience with no speech. Use night forest insects, soft wind through leaves, distant owls, faint magical shimmer, and one small branch snap near the middle. Keep the loop calm and usable in a game scene.

What this example shows

The no-speech constraint is useful when you want an ambience bed rather than a narrated scene.

• Say no speech when you want pure ambience.
• Describe the loop purpose so the output stays less dramatic.
• Place one small event in the middle to test timing control.

Create a 20-second audiobook narration in English with one clear narrator. The scene is a quiet train ride at sunrise. Add very subtle carriage room tone and soft rail rhythm, but keep the voice intimate, steady, and easy to hear.

What this example shows

This format is good for testing whether background texture competes with voice intelligibility.

• Use subtle for ambience when narration is the product.
• Give one physical scene cue, not a full plot.
• Ask for steady delivery when you do not want dramatic acting.

Create a 20-second educational explainer in English for students. A friendly teacher voice explains that sound is vibration moving through air, water, or solid objects. Add light classroom ambience and two gentle demonstration tones.

What this example shows

A clear instructional role plus two demonstration tones makes the output easier to evaluate.

• State the teaching role and audience level.
• Limit demo sounds to a small count.
• Use gentle if you do not want harsh effects.

Create a 20-second short-video audio scene in English about a small bakery opening before sunrise. Include a gentle narrator, espresso machine steam, trays sliding into an oven, soft morning street ambience, and an uplifting acoustic music cue.

What this example shows

This output came back closer to 30 seconds, which is useful evidence that duration instructions are approximate.

• Scene-rich prompts can extend beyond the requested length.
• Use fewer foley events if you need a tighter clip.
• Put the most important sound first.

Create a 20-second wellness app audio in English. Use a calm guide voice, soft breathing pace, quiet room tone, warm ambient pad, and a gentle bell at the end. Keep the result relaxing, minimal, and clean.

What this example shows

The minimal prompt shape helps prevent the ambience from becoming too cinematic.

• Use minimal and clean for app UI audio.
• Add one ending marker such as a bell.
• Avoid stacking too many music adjectives.

Create a 20-second museum audio guide in English. Use a clear narrator describing a restored ceramic vase in a quiet gallery. Add subtle gallery room tone, distant footsteps, and a refined calm delivery suitable for visitors.

What this example shows

Like the bakery sample, this returned a longer clip than requested, so it is a good example for duration planning.

• Visitor-facing prompts benefit from refined and calm delivery cues.
• Public-space ambience should be distant and subtle.
• Plan to trim the result when exact timing matters.