Generates Comfort Noise after a specified time delay when silence is detected.
Parameters
name |
description | default |
|---|---|---|
| Wave Audio Base64 | The base64 value of a WAV file to be played when silence is detected. | Base64 value of a keyboard recording |
| Required Silence (MSec) | Required Silence duration before comfort noise is played in milliseconds. | 400 |
| Comfort Noise Audio Volume Percentage | Percentage of the volume to adjust the given base64 wave. | 100 |
Inputs
Audio
Accepts audio from a single channel
Events
| name | description | known nodes that generate this event |
|---|---|---|
| Speech Started | Used for deciding when to stop adding comfort noise. | Vad |
| Speech Ended | Used for deciding when to start adding comfort noise. | Vad |
| Merge Audio Started | Signals the start of audio merging. Used for deciding when to stop adding comfort noise. | Batch To Stream Audio Merger |
| Merge Audio Ended | Signals the end of audio merging. Used for deciding when to start adding comfort noise. | Batch To Stream Audio Merger |
Outputs
Audio
Mono Streaming Audio
Events
None
Remarks
Requirements
- This node needs consistent streaming audio input, works as pass-through when comfort noise is not added.
- Needs to be connected in series to audio input-output
- Only Needs the event output from VAD Nodes.
Project Structure
A simple project can be built as such:
The project allows the user to speak concurrently with TTS Voices, Batch to stream audio merger uses its barge-in feature to make sure the actual voice and TTS voice don't overlap. Once there is no active speech or TTS voice, it plays a comfort noise.
Supported flow types
Stream