ComfyUI Node
LTX Set Audio Latent Binary Masks
Converts a binary mask image or mask video into an audio latent noise mask by reducing each frame to active/inactive, resizing only along time, and broadcasting that timeline across the audio latent.
LTX Set Audio Latent Binary Masks
- audio_latent
- masks
- audio_latent
◄threshold0.50►
◄resize_modenearest►
◄existing_mask_modeoverwrite►
Categorylatent/audio
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| audio_latent | LATENT | Audio latent whose noise_mask will be set. | |
| masks | MASK | Binary mask image or mask video. Only the temporal activity of each frame is used; spatial dimensions are ignored after thresholding. | |
| threshold | FLOAT | 0.500–1 | Per-pixel threshold used when deciding whether a mask frame is active. If any pixel in a frame meets this threshold, that frame activates audio masking. |
| resize_mode | COMBO | nearest | How to resize the derived binary timeline to the audio latent length. 'nearest' preserves hard ranges; 'linear' smooths transitions before the final binary threshold. |
| existing_mask_mode | COMBO | overwrite | How to combine with an existing audio noise mask. 'overwrite' replaces it, 'add' takes the max, and 'subtract' clears masked regions. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| audio_latent | LATENT | — |