Canonical Shot Object

What it is

Every shot Cadenza plans is captured as a Canonical Shot Object: the director's full intent, independent of any model's prompt format.

• Narrative and emotional purpose: why this shot exists in the arc.
• Camera: lens, movement, angle, plus computed physics (field of view, subject distance, depth of field, motion-blur risk).
• Framing, blocking, lighting: the visual grammar from the Director Mode.
• Action beats and performance detail: what happens, externalized physically.
• Negative constraints: what must not appear.
• Continuity references: which locked references this shot inherits.
• Research enrichment: cinematic idiom, edit timeline position, frame-rate feel, depth layers, color intent, artifact risk score.

Why the surplus depth

Different models want different depth. Per-model encoders compress the canonical object into each model's optimal prompt shape, and the surplus is what makes that compression possible without loss of intent. A new model is a new encoder, never a change to the canonical object.

Where you see it

The studio shot list, the plan_director_prompt tool, and public directed pages all render views of the same object. See plan_director_prompt to get it raw, with per-model prompts, for free.