ComfyUI Node
H3 Auto Chain Motion Context
A ComfyUI node in conditioning/minimax with 9 inputs and 2 outputs.
H3 Auto Chain Motion Context
- conditioning
- vae
- latent
- context_frames
- context_latent
- audio_vae
- context_audio
- conditioning
- trim_frames
◄context_length22►
◄audio_context_length24►
Categoryconditioning/minimax
Inputs (9)
| Name | Type | Default | Description |
|---|---|---|---|
| conditioning | CONDITIONING | — | |
| vae | VAE | — | |
| latent | LATENT | — | |
| context_length | COMBO | 22 | Frames of the previous clip's picture to carry over. Only these lengths are whole numbers of latent steps, so only these are offered. 5 is just barely fluid, 22 is nearly seamless. Longer windows pin more motion but come off the front of the delivered clip, so 56 spends 2.3 seconds of the render on frames you throw away. |
| audio_context_length | INT | 240–240 | Frames of tail audio to pin, independent of the picture window. 0 follows it. The window is END-aligned with the pinned video, so this only controls how far back the sound reaches. Multiples of 3 land exactly on the 40 Hz audio grid and multiples of 24 are whole seconds: 24 pins the last second. Off-grid values are widened to the nearest whole step. |
| context_framesopt | IMAGE | Decoded frames of the previous clip. Used when no context_latent is wired. When one is, the picture comes from it instead and this is ignored. | |
| context_latentopt | LATENT | Previous clip's SAMPLER OUTPUT latent (the same one you wire into the decode nodes). Supplies both picture and sound, sliced straight out, skipping the decode and re-encode that cost a little quality at every link of a chain. Must be the same resolution as the clip being generated. | |
| audio_vaeopt | VAE | H3 audio VAE. Supply with context_audio to carry the previous clip's tail sound across the join. Not needed when context_latent is wired. | |
| context_audioopt | AUDIO | Audio of the previous clip. The tail matching the pinned frames is encoded and pinned alongside them. Ignored when context_latent is wired. |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| conditioning | CONDITIONING | — |
| trim_frames | INT | — |