This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHubThis node converts a compressed audio representation (latent samples) back into an audio waveform using a Variational Autoencoder (VAE). It processes the data in smaller, overlapping sections (tiles) to manage memory usage, making it suitable for handling longer audio sequences.
Inputs
| Parameter | Data Type | Required | Range | Description |
|---|---|---|---|---|
samples | LATENT | Yes | N/A | The compressed latent representation of the audio to be decoded. |
vae | VAE | Yes | N/A | The Variational Autoencoder model used to perform the decoding. |
tile_size | INT | No | 32 to 8192 | The size of each processing tile. The audio is decoded in sections of this length to conserve memory (default: 512). |
overlap | INT | No | 0 to 1024 | The number of samples that adjacent tiles overlap. This helps to reduce artifacts at the boundaries between tiles (default: 64). |
Outputs
| Output Name | Data Type | Description |
|---|---|---|
output | AUDIO | The decoded audio waveform. |