Solutions
Engine
Core Pipeline
Intelligence Modules
ENGINE MODULE
Lens Orchestration
No random camera. No repetition across cuts.
“Camera grammar is distributed by film style profile. No random assignment. No repetition across cuts.”
LO enforces story-wide camera movement variety and distributes camera grammar according to the film style profile. It ensures that static shots, push-ins, tracks, and handheld movements are distributed across your entire story — matching the distribution a real cinematographer would plan for the chosen visual style.
LO prevents the most common flaw in AI-generated video sequences: identical adjacent camera movements. If Shot 5 uses a push-in, Shot 6 will not. LO also assigns per-shot camera specifications: focal length (in millimeters), stability rating (1-10), movement type (static, track, push-in, handheld), movement intensity (1-10), and movement complexity (1-10).
The distribution is governed by the Film Grammar Profile — a documentary style gets more handheld and observational movement, while a cinematic style gets more dolly and crane-equivalent movements. Every decision is traceable to the profile selection.
LO — Camera Grammar Distribution ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Story Camera Baseline: Lens Bias: 35-50mm (standard dramatic range) Stability Variance: 6-9/10 (controlled handheld) Distribution (19 shots): Static: 14 shots (74%) Track: 1 shot (5%) Push-in: 3 shots (16%) Handheld: 1 shot (5%) Motion Contrast Check: Shot 5 (push-in) → Shot 6 (static) ✓ Different Shot 6 (static) → Shot 7 (push-in) ✓ Different Shot 11 (static) → Shot 12 (handheld) ✓ Different No adjacent camera movement repetition detected. ✓ Per-Shot Example: Shot 7: 42mm | Handheld stability 6/10 | Push-in 5/10 Shot 8: 50mm | Static stability 9/10 | None
Actual engine output from a StoryDirector story compilation.
Director's Notes
A real DP plans camera movement for an entire shooting day — they don't randomly pick a movement for each shot. LO does the same planning for AI video: distributing movements across the story so the visual rhythm feels intentional, not generated. The motion contrast enforcement alone prevents one of the most common tells that content is AI-generated — the uncanny repetition of identical camera movements from shot to shot.
Classification
Fully Deterministic
Introduced
Engine 5.1.0 (Cinematic Evolution), expanded in 6.0.0
Dependencies
MOMA™ (camera ceilings per act), Film Grammar Profile, Shot role classifications
Outputs to
DirectorLogic™ (per-shot camera specs), Prompt Assembly V2 (CINEMA line)
Determinism
100% — distribution is deterministic from profile + shot roles
Experience LO™ in action.
