AI Voice Nodes

AI Voice Nodes are powerful tools for generating speech from text or transforming existing audio/video performances into different voices. They support Text-to-Speech (TTS), Voice Cloning, and Speech-to-Speech (STS) workflows.

Creating an AI Voice Node

To create an AI Voice Node, right-click on the canvas and select AI Voice Node from the context menu.

Capabilities & Inputs

The AI Voice Node has three main input zones for different workflows:

1. Script (Text-to-Speech)

Connect a Text Node to the AI Voice Node to provide the script.

• Drag from a Text Node to the Script input zone (main body).
• The AI will read the text using the selected voice.

2. Reference Audio (Voice Cloning)

Connect an Audio Node to the top input zone to clone a voice.

• Drag from an Audio Node containing a sample of the voice you want to clone.
• Drop onto the “Set Reference (Clone)” zone at the top of the node.
• Click “Save as New Voice” to add it to your library.

3. Performance Audio (Speech-to-Speech / Dubbing)

Connect an Audio Node or a Video Node (Scene) to drive the performance.

• Audio Node: Acts as a guide for the speed, emotion, and intonation, but changes the voice to the selected one.
• Video Node: Extracts the audio from the video, changes the voice while preserving the original timing/performance, and merges it back into the video (Dubbing).
• Drag to the “Set Performance (STS)” zone at the bottom of the node.

Using the Node

1. Select a Voice: Choose a voice from the dropdown menu. You can use pre-made voices or ones you’ve cloned.
2. Connect Inputs: Link the necessary Script, Reference, or Performance nodes as described above.
3. Generate: Click Generate Audio.
4. Result:
   • For Audio outputs, a player will appear.
   • For Video outputs (dubbing), a video player will appear with the new audio track merged.

Downloading Results

You can download the generated audio or video by clicking the Download icon in the node header or by right-clicking the node and selecting Download Media.