ComfyUI Node
LTXVAddRefMultiFc
Add multiple reference images at automatically assigned negative frame indices, uses DynamicCombo which requires ComfyUI 0.8.1 and frontend 1.33.4 or later.
LTXVAddRefMultiFc
- positive
- negative
- vae
- latent
- positive
- negative
- latent
- processed_images
- frame_step
- info
◄step_multiplier2►
◄mask_modeconstant►
◄ramp_frames1►
◄upscale_methodnearest-exact►
◄negative_frame_modeallow_cross_zero►
◄num_refs▾►
CategoryKJNodes/ltxv
Inputs (10)
| Name | Type | Default | Description |
|---|---|---|---|
| positive | CONDITIONING | Positive conditioning to which guide keyframe info will be added | |
| negative | CONDITIONING | Negative conditioning to which guide keyframe info will be added | |
| vae | VAE | Video VAE used to encode the guide images | |
| latent | LATENT | Video latent, guides are added to the end of this latent | |
| step_multiplier | INT | 21–100 | Multiply frame_step to space reference frames. |
| mask_mode | COMBO | constant | Select how to build the guide noise mask. |
| ramp_frames | INT | 11–64 | Expand single-frame refs to this many frames when using ramp. |
| upscale_method | COMBO | nearest-exact | Resize method used before preprocessing. |
| negative_frame_mode | COMBO | allow_cross_zero | How to place negative frame indices when refs span multiple frames. |
| num_refs | COMBO | Select how many reference images to use |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| positive | CONDITIONING | — |
| negative | CONDITIONING | — |
| latent | LATENT | Video latent with added references |
| processed_images | IMAGE | Batch of processed reference images |
| frame_step | INT | Frame step (vae.downscale_index_formula[0]) |
| info | STRING | Reference index mapping info |