Reference Audio
Extract audio from YouTube videos or provide audio via upload/URL.
Overview
The Reference Audio node provides audio from multiple sources: extract from a YouTube URL, upload a file directly, or enter a direct URL. Includes audio preview playback after successful extraction or upload.
Configuration
| Field | Type | Default | Description |
|---|---|---|---|
| Source | Select | youtube | Source type: YouTube, Upload File, Direct URL |
Source: YouTube
| Field | Type | Description | |——-|——|————-| | YouTube URL | Text input | Video URL to extract audio from | | Extract Button | Action | Fetches metadata and extracts audio track |
Source: Upload File
| Field | Type | Description | |——-|——|————-| | File Input | File upload | Accepts: MP3, WAV, M4A, AAC |
Source: Direct URL
| Field | Type | Description | |——-|——|————-| | Audio URL | Text input | Direct link to audio file |
Inputs & Outputs
Inputs: None (this is a source node)
Outputs:
- Audio URL — extracted or provided audio
Best Practices
- YouTube extraction includes metadata fetching — verify the correct video before extracting
- Use WAV uploads for highest quality when file size isn’t a concern
- Audio preview lets you verify the correct content before running the workflow
Common Use Cases
- Extract background music from YouTube for reference or dubbing
- Provide audio for voice cloning workflows
- Source audio for Forced Alignment (word-level timestamps)
- Extract narration from video for transcription
Tips
- Extraction status shows: ready → extracting → complete (or failed)
- If YouTube extraction fails, try the direct URL method with an alternative source
- The audio preview player appears automatically after successful extraction