Audio Logger Node. Stores the inputted audio to object storage, as a Wave file for PCM/ULaw encoding, or as an Ogg/Opus file for Opus encoding.
Parameters
name |
description | default |
|---|---|---|
| Channel Count | The channel count of the inputted audio. | 1 |
| Encoding | PCM for quality, Opus or ULaw for storage. |
Opus |
| Identifier Postfix | A unique postfix used to identify the saved audio files. Can only contain alphanumeric characters or '_'. | audio |
| Data Flow External Address | The externally reachable address of this Data Flow instance. It is used to build the audio-logs-v2 URL returned in this node's AudioLogReport Event Output. This needs to be changed in testing environments only. |
http://{InClusterManagementAddress}.sestek-core:{InClusterManagementPort}/ |
Channel Count must be left as 1 unless a Mono to Stereo Mixer is directly connected as the only streaming audio input to the Audio Logger. You can check the Project Structure section for more information.
Inputs
Audio
Accepts mono or stereo audio from a single channel.
Events
| name | description | known nodes that generate this event |
|---|---|---|
| Start of TTS Fragment | Signals the start of a TTS-generated audio segment. The node starts collecting a new "batch" audio file for this segment; if the event carries a Data.Text field, it is echoed back on the resulting AudioLogReport event as IncomingText. |
TTS Http, TTS Websocket, TTS Microsoft Http, TTS OpenAI Http, TTS ElevenLabs |
| End of TTS Fragment | Signals the end of the TTS-generated audio segment. The node finalizes and uploads the collected "batch" audio file. | TTS Http, TTS Websocket, TTS Microsoft Http, TTS OpenAI Http, TTS ElevenLabs |
Outputs
Audio
None
Events
| name | description |
|---|---|
| AudioLogReport | Emitted once an audio segment finishes uploading to object storage, successfully or not. |
Remarks
Requirements
MinIO Node Audio Logging must be enabled in the server configurations by the system administrator for this node to function properly.
Audio logging is also automatically disabled for the session if the tenant's audio log retention period is configured to zero.
Project Structure
This node can be given any audio connection to log a wave file.
An example Audio Logger node that collects and stores modified streaming audio.
The node above will collect audio through the session and log it as a single wave file.
An example Audio Logger node that collects and stores TTS Outputs is like.
The node above will collect each TTS output separately and log each output in a different wave file.
The node above will collect all the audio through the session that comes from Batch To Stream Audio Merger, and log it in a Stream wave file, and collect each of the TTS outputs and log them in separate Batch wave files.
Channel Count must be set to 2 if and only if a Mono to Stereo Mixer node is directly connected as the only streaming audio input.
Supported flow types
Batch, Stream