Actions13
- Speech Actions
- Voice Actions
- History Actions
- User Actions
Speech → Speech to Speech
AI-generatedOverview
The "Speech to Speech" operation of the ElevenLabs node converts input speech audio into a new speech audio output using a selected voice model. This transformation can be used to change the speaker's voice, apply stylistic effects, or improve clarity and quality while preserving the original spoken content.
Common scenarios include:
- Voice dubbing or localization where the original speech is converted into another voice.
- Creating personalized voice messages or announcements with a specific voice style.
- Enhancing audio content by applying voice stability and similarity tuning.
- Generating synthetic speech outputs for accessibility or entertainment purposes.
For example, you could input a recorded message and convert it into a celebrity voice or a custom AI-generated voice with specific style and stability settings.
Properties
| Name | Meaning |
|---|---|
| Voice ID | The voice to use for generating the output speech. Can be selected from a searchable list of available voices or specified directly by an ID. |
| Binary Name | (Optional) Custom name for the binary output data. Defaults to "data". |
| File Name | (Optional) Custom file name for the generated audio file. Defaults to "voice". |
| Streaming Latency | (Optional) Enables latency optimizations at some cost to audio quality. Accepts values from 0 (no optimization) to 4 (maximum optimization). |
| Output Format | (Optional) Format of the generated audio output, e.g., "mp3_44100_128". |
| Model Name or ID | (Optional) Identifier of the voice synthesis model to use. Can be chosen from a list or specified via expression. |
| Stability | (Optional) Controls how stable the voice sounds, ranging from 0 (unstable) to 1 (very stable). |
| Similarity Boost | (Optional) Adjusts how closely the output voice matches the selected voice, from 0 (low similarity) to 1 (high similarity). |
| Style | (Optional) Exaggerates the voice style, from 0 (none) to 1 (maximum exaggeration). |
| Speaker Boost | (Optional) Boolean flag to activate speaker boost, enhancing certain voice characteristics. |
| Seed | (Optional) Fixed seed number to produce consistent voice output across runs. |
| Stitching | (Optional) Boolean flag to enable stitching, which provides context by passing previous text to the model, improving continuity in speech generation. Defaults to true. |
Output
The node outputs the generated speech audio as binary data. The binary data contains the synthesized audio file in the specified output format (e.g., MP3). The output binary property name defaults to "data" but can be customized. The file name of the output audio can also be set via properties.
The JSON output field typically includes metadata about the request and response, such as the voice ID used, model details, and any relevant status information returned by the API.
Dependencies
- Requires an API key credential for ElevenLabs API authentication.
- The node makes HTTP POST requests to the ElevenLabs API endpoint
https://api.elevenlabs.io/v1. - The user must have access to the ElevenLabs service and appropriate permissions to use voice models.
- Optional: To select voices and models dynamically, the node uses internal methods to list available voices and models.
Troubleshooting
- Invalid Voice ID: If the voice ID is incorrect or not found, the API will return an error. Verify the voice ID by listing available voices.
- Authentication Errors: Missing or invalid API key credentials will cause authentication failures. Ensure the API key is correctly configured in n8n credentials.
- Unsupported Output Format: Specifying an unsupported audio format may result in errors or no output. Use supported formats like "mp3_44100_128".
- Latency Optimization Quality Tradeoff: Enabling streaming latency optimizations may degrade audio quality; adjust the level according to your needs.
- Model Not Found: If a model ID is specified but does not exist, the request will fail. Use the provided listModels method to find valid model IDs.
- Stitching Disabled Issues: Disabling stitching might cause less coherent speech if the input is long or segmented. Enable stitching for better context handling.