MEET SEEDAUDIO
SeedAudio
Bring dialogue, ambience, music and sound effects into one creative brief. Listen to our original examples, learn how to prepare prompts and references, then open the generator with a clear sound in mind.
From a scene idea to a sound you can hear
Set the direction, then add optional media
Or one image; the two cannot be combined
Model limit; review the actual result
START BY LISTENING
Explore foley, music, multilingual dialogue, voices and product ads. Understand the input, listen for the important details, then open the full prompt to see how the scene was described.
PRODUCTION BRIEF
Create a 12-second storm: huge breaking surf, powerful water spray, and rolling thunder. No dialogue or music.
GENERATED AUDIO
PRODUCTION BRIEF
Compose a 20-second original cello cue over soft sustained strings. Use a warm, reflective mood, build at eight seconds, and resolve with a natural decay. Instrumental only.
GENERATED AUDIO
PRODUCTION BRIEF
On a rainy night train, create a 10-second two-character exchange in Italian and Chinese. Keep quick natural responses, clear pronunciation, soft carriage ambience, and no music.
LANGUAGE VERSIONS
VOICE COMPARISON
Input reference
NEW SCRIPT
Use @Audio1 as the authorized voice reference. Read: “Every new chapter begins with one clear breath. Take your time. Your story is ready to be heard.”
New speech
Only clone voices you own or have explicit permission to use.
PRODUCTION BRIEF
A 12-second sparkling-drink ad: a crisp can opening, ice and fizz, bright English voiceover, and an upbeat musical finish.
GENERATED AD
A CLOSER LOOK AT REFERENCES
Take a closer look at Sixteen Seconds to Escape. Nora uses @Audio1 and Ethan uses @Audio2. Hear the references first, then follow their exchange inside the space capsule.
@Audio1 · INPUT
@Audio2 · INPUT
GENERATED AUDIO
Compare the two characters, their urgent reactions, and how the hatch and engine sounds support the dialogue. References guide the result; review each performance.
UNDERSTAND THE SCENE
Basic text-to-speech focuses on reading a script. Scene-level audio also considers who is speaking, where the action takes place and when each sound happens. Describe those relationships in one prompt, then review the mixed result.
The output is a mixed track. This is a planning diagram, not separately exportable stems.
CAPABILITY MAP
Explain the task in text. Add one image or up to three audio clips when useful; choose one reference type.
Check input limitsName the characters, language, emotion and conversational pace. Compare the train scene in two languages.
Hear the train dialogueDescribe foreground and background layers, then place the important sound events in order.
Hear the product adUse a voice reference to guide recurring character traits. Compare the voice and delivery after changing the script.
Compare input and outputBEFORE YOU CREATE
WHERE IT FITS
Build the depth of a storm with surf, rolling thunder and a natural fade. Hear the breakwater scene.
Listen to this exampleWrite a clear exchange between characters. Hear the same night-train scene in Italian and Chinese.
Listen to this exampleGuide two voices through an urgent space-capsule scene, with hatch and engine sounds supporting their exchange.
Listen to this exampleUse a voice reference with a new script. Compare Sophie’s reference with the generated performance.
Listen to this exampleWRITE YOUR FIRST PROMPT
Name the format, place, action and mood.
Define speakers, language and delivery. Quote the spoken lines.
Add ambience, music and key effects. Set their priorities.
Check speech, pauses and the busiest layer. Change one thing at a time.
This is a writing summary. Use “Full prompt” on the example to read the actual generation instructions.
Hear the train dialogueScene: a short exchange on a rainy night train. Voice: two adults responding naturally in Chinese. Layers: soft wheels and rain, with no music. Review: are the words clear and the responses natural?
YOUR NEXT STEP