Speech-to-Speech Voiceovers

Speech-to-speech (STS) voiceovers let you reVoice an existing audio recording — converting it from its original voice into one of your custom Euryka voices. This is useful when you have recorded content (a rough cut, a CEO recording, an interview) that you want to present in a consistent branded voice.

How speech-to-speech works

1. You upload a source audio file — the original recording in the speaker's voice.
2. You select one of your custom Euryka voices as the target voice.
3. Euryka converts the audio: the words and delivery from the original are retained, but the voice itself is replaced with your chosen custom voice.

Creating a speech-to-speech voiceover

1. Click Voiceovers in the sidebar.
2. Click New Voiceover Project.
3. In the voiceover creation form, select the Speech-to-Speech mode (or look for an STS toggle/option).
4. Configure:
   • Project name
   • Source audio — upload the original recording (MP3, WAV, M4A, etc.)
   • Target voice — select the custom voice to convert into
   • Brand / Project (optional)
5. Click Generate.

Reviewing the output

Once complete, preview the converted audio and compare it to the source. If the result isn't quite right:

• Adjust the source audio (trim silence, improve audio quality) and regenerate
• Try a different voice from your library

Best practices

• Clean source audio produces the best STS results — remove background noise, avoid overlapping speakers
• Consistent pacing in the source helps the model convert more naturally
• STS works best for single-speaker recordings

Related articles

• Creating a custom voice
• Text-to-speech voiceovers
• Managing voiceover projects