Learn the vocabulary, control the result
A dolly and a zoom look superficially similar but feel completely different: a dolly physically travels, so foreground and background shift against each other (parallax); a zoom just magnifies, flattening the scene. That's why DOLLY IN feels like approaching a person and CRASH ZOOM feels like a jolt of attention. VERTIGO ZOOM does both in opposite directions at once — the Hitchcock effect where the background warps while the subject holds still, pure unease.
Orbits (ORBIT LEFT, FULL ORBIT) add dimensionality and are the most reliable way to make a subject feel three-dimensional. Crane moves (CRANE UP, DRONE PULLBACK) change scale relationships — rising from a face to a skyline turns a portrait into a story about a place.
Match the move to the clip length
Percussive moves read best short — which is why Nidhogg caps presets like CRASH ZOOM, WHIP PAN, ZOOM PUNCH, and ELEVATOR DROP at 5 seconds. A crash zoom stretched across ten seconds isn't a crash, it's a slow push with anxiety. Gentle moves (SLOW ZOOM, STEADICAM GLIDE, SETTLE IN) do the opposite: they need duration to breathe.
One move per shot is the professional default. Stacking an orbit onto a zoom onto a tilt reads as chaos; a single committed move reads as intent. If you need complexity, cut between two clips with one clean move each.
Pairing moves with subjects and models
Static subjects want moving cameras (orbit a parked car, crane up a building); moving subjects often want steadier frames (LOCKED OFF or STEADICAM GLIDE lets a dancer's motion carry the shot). LOW TRACK flatters vehicles; WORM EYE makes anything monumental; HANDHELD RUN injects documentary urgency into chase and action prompts.
For model choice: Kling 3 is Nidhogg's flagship for cinematic coherence on long graceful moves; Seedance Pro iterates fast when you're auditioning several moves against one subject before committing your best render.