ComfyUI Node
LTX Director Motion Brush V2 Director Guide
A ComfyUI node in sd with 18 inputs and 5 outputs.
LTX Director Motion Brush V2 Director Guide
- positive
- negative
- vae
- latent
- guide_data
- motion_guide_data
- model
- positive
- negative
- latent
- model
- latent_downscale_factor
◄ic_lora_nameNone►
◄ic_lora_strength1.00►
◄scale_by1.00►
◄upscale_methodbicubic►
◄image_attention_strength1.00►
◄cropcenter►
◄auto_snap_ic_gridtrue►
◄use_tiled_encodefalse►
◄tile_size256►
◄tile_overlap64►
◄retake_modefalse►
Categorysd
Inputs (18)
| Name | Type | Default | Description |
|---|---|---|---|
| positive | CONDITIONING | Positive conditioning to add guide keyframe info to. | |
| negative | CONDITIONING | Negative conditioning to add guide keyframe info to. | |
| vae | VAE | Video VAE used to encode the guide images. | |
| latent | LATENT | Video latent — guides are inserted into this latent. | |
| guide_data | GUIDE_DATA | Guide data produced by Prompt Relay Encode (Timeline). | |
| motion_guide_dataopt | MOTION_GUIDE_DATA | Connect motion guide data from the timeline node to use IC-LoRA video guidance. | |
| modelopt | MODEL | Connect model if using IC-LoRA for motion guidance. | |
| ic_lora_nameopt | COMBO | None | Select the IC-LoRA model to use for motion guidance. |
| ic_lora_strengthopt | FLOAT | 1.00-100–100 | — |
| scale_byopt | FLOAT | 1.000.01–8 | Scale the latent by this factor. |
| upscale_methodopt | COMBO | bicubic | Method used to upscale/downscale the latent. |
| image_attention_strengthopt | FLOAT | 1.000–1 | — |
| cropopt | COMBO | center | 2 options: disabled, center |
| auto_snap_ic_gridopt | BOOLEAN | true | — |
| use_tiled_encodeopt | BOOLEAN | false | — |
| tile_sizeopt | INT | 25664–512 | — |
| tile_overlapopt | INT | 6416–256 | — |
| retake_modeopt | BOOLEAN | false | Force Retake Mode. If false, it will still auto-detect Retake Mode from the timeline data. |
Outputs (5)
| Name | Type | Description |
|---|---|---|
| positive | CONDITIONING | — |
| negative | CONDITIONING | — |
| latent | LATENT | — |
| model | MODEL | — |
| latent_downscale_factor | FLOAT | — |