Nodes/Difforum/Difforum · Camera → Prompt
ComfyUI Node

Difforum · Camera → Prompt

The camera, in words - for prompt-driven video models (MiniMax H3, LTX, Seedance, Veo...). Uses the Director's blocks when available, or reads any camera track and names its moves.

By chillithebillis·Created 3 months ago·Updated about 7 hours ago· 8
Difforum · Camera → Prompt
  • direction
  • params
  • camera
  • text
◄formatsentence►
◄prefix►
◄include_looktrue►
◄h3_modereference (ref2va / guides)►
◄soundscape►
◄music►
◄cutsfalse►
CategoryDifforum/5 · Video model bridges

Inputs (10)

NameTypeDefaultDescription
formatCOMBOsentencesentence: one line of camera direction (+ look) to append to any prompt. timed lines: [0.0s-2.5s] camera ... per block. prompt suffix: 'Camera: ...'. H3 structured: the whole Director timeline (scenes, camera, keys, look) written in MiniMax H3's native prompt format.
directionoptDIFFORUM_DIRECTIONThe Director's wire: camera, energy curve, scene prompts, keys and look in one cable.
paramsoptDIFFORUM_PARAMSFrom Setup: frame count, fps, width and height. Plug the same Setup everywhere.
cameraoptDIFFORUM_CAMERAA camera track from Camera (keys), Camera (expressions) or Camera Import. Overrides the Director's camera.
prefixoptSTRINGWhat is on screen. sentence / suffix: put before the camera text. H3 structured: the opening composition ([Shot 1]); in reference mode, what <Picture 1> shows.
include_lookoptBOOLEANtrueAppend the Director's look sentence (deforum morph, stop-motion...).
h3_modeoptCOMBOreference (ref2va / guides)H3 structured only. Which H3 task the prompt is for: sets the alignment line (I2VA / FL2VA) or the six full-reference sections (ref2va with H3 Guides, <Picture 1> as the first frame).
soundscapeoptSTRINGH3 structured: overall_soundscape - ambience and physical sounds only (wind, footsteps, rain). Empty = soft natural ambience.
musicoptSTRINGH3 structured: non_diegetic_music - instruments, tempo, dynamics, no mood words. Empty = N/A (no score).
cutsoptBOOLEANfalseH3 structured: off = one continuous take (scene changes become transformations). On = every scene starts a new [Shot N] at its cut time.

Outputs (1)

NameTypeDescription
textSTRING—