H3 Auto Chain Audio

Splits one AUDIO input into sequential chunks and reports the chunk duration for H3 Reference To Video.

By Ltamann·Created 2 days ago·Updated about 14 hours ago· 18

H3 Auto Chain Audio

◄chain_idh3_auto_chain►

◄chunk_seconds20.0►

◄fps24.000►

◄trim_frames22►

◄final_tail_modeexact audio duration►

◄final_tail_frames24►

◄resettrue►

◄style_prompt►

◄clip_prompts►

◄start_clip1►

◄end_clip0►

Categoryconditioning/minimax

Inputs (12)

Name	Type	Default	Description
audio	AUDIO		Complete audio for the entire automatic chain. The node outputs one sequential chunk per run.
chain_id	STRING	h3_auto_chain	Unique name for this chain. It identifies the saved latents, clips, and final video.
chunk_seconds	FLOAT	20.01–600	Target duration of each clip. The final chunk may be shorter.
fps	FLOAT	24.0001–240	Frame rate of the generated video. Clip starts and audio cuts are calculated on this frame grid.
trim_frames	INT	220–4096	Leading frames removed by H3 Motion Context Trim from continuation clips. Match the Motion Context context length; 22 is about one second at 24 FPS.
final_tail_mode	COMBO	exact audio duration	Final partial clip output policy. Exact audio duration removes the silent tail. Audio plus tail keeps the configured silent tail after the source audio ends.
final_tail_frames	INT	240–4096	Silent frames retained after the source audio ends when audio plus tail is selected.
reset	BOOLEAN	true	Start a new chain and discard the in-memory position. Enable this for a new run or resume.
style_prompt	STRING		Shared style, character, lighting, camera, and appearance text added to every clip prompt.
clip_prompts	STRING		Write one tagged prompt per clip: [1] prompt for clip 1, [2] prompt for clip 2, and so on. Tags must start a line.
start_clip	INT	11–9999	First clip to process. To resume after clips 1 and 2, set this to 3.
end_clip	INT	00–9999	Last clip to process. 0 means continue until the audio ends.