Voice Cloning

Record or upload a short audio sample and Kolbo builds a reusable custom voice from it. Use that cloned voice in any Text-to-Speech session, or attach it directly to a Visual DNA character so the right voice is always pre-selected.

Overview

Voice Cloning copies the qualities of a specific real voice from a sample and saves it as a named custom voice in your library. It is the right choice when you have a particular reference voice in mind and a clean sample to work from.

What You Need

A clean audio sample. 10 to 30 seconds of clear speech is ideal, though a few seconds will work. The sample can be:

• A voice memo or recording you made specifically for this
• A clip from a podcast, voiceover, or video you have the rights to use
• A short sample you record directly in the browser from the voice creation panel

Kolbo processes the sample and builds a voice model from it. The clone is saved to your custom voice library under a name you choose.

How It Works

1. Open Text-to-Speech in Audio Tools and open the voice creation panel
2. Choose "Clone a voice"
3. Upload an audio sample, or record one directly in the browser
4. Name the voice and let Kolbo build the model
5. Select the cloned voice in any Text-to-Speech session

Assign a Voice to a Character

This is where cloning becomes most useful for character-driven work. Open any Visual DNA character and assign the cloned voice to it directly. Once it is attached:

• Text-to-speech sessions pre-select the right voice automatically when that character is active
• No manual re-selecting the voice each time you start a new session
• The voice assignment travels with the character across every project that uses it

If you work with a recurring cast, a narrator, a brand voice, or a set of fictional characters, clone or design a voice for each one and assign it. The workflow becomes: pick the character, write the script, generate. The voice is already there.

Clone vs. Design

Kolbo has two ways to create a custom voice:

• Voice Cloning copies a specific real voice from a sample. Use it when you have a reference voice and a clean recording.
• Voice Design creates a brand-new voice from a text description. Use it when the voice only exists in your head.

Both methods produce a named custom voice that lives in the same library and works identically downstream.

Tips

1. Use a clean sample - record in a quiet space with consistent tone for the best clone quality
2. Assign to a character once so the voice follows that character across every project
3. Name voices clearly so they are easy to find later in your custom voice library

Related