Execution Engine & Directing

The [EXECUTION] block translates narrative intention into concrete, verifiable cinematic blocking, enabling the director to mentally simulate and track the scene’s kinetic progression beat by beat.

1. Writing Style & Grammar

  • Tense & Voice: Write all Sub-States in strict present tense, active voice.
  • Sentence Economy: Descriptions must be minimal, tactile, and punchy. Eliminate filler words, passive clauses, and conversational transitions.
  • Sub-State Internal Structure: When describing an action beat, structure elements logically: Subject + Action + Physical Impact + Prop/Environmental Reaction
  • Verifiability: Every phrase must translate directly into a visible or audible change on screen.

2. The Repetition Mandate for Continuity

Generative video engines do not possess reliable unprompted long-term spatial memory. To guarantee unbroken continuity, the director explicitly carries forward critical physical states across Macro-States:

  • Explicit Restatement: If a continuity-critical detail remains true across beats, explicitly restate it until it changes:
    • Which hand holds an object (e.g., "box clutched in right hand", "spear in right hand").
    • Postural state (e.g., "remains seated frame-right", "leaning against the timber rail").
    • Persistent physical marks (e.g., "ash on left sleeve", "blood on collar", "sweat-sheen across forehead").
    • Prop stability (e.g., "manuscript resting on dark oak desk", "keys held in closed fist").
  • Rule: Never assume persistence will happen automatically. If losing an object or posture would break the scene, repeat it across Macro-States.

3. Visual Blocking & Screen Geography

Direct character positions and movements with spatial clarity:

  • Frame Coordinates: Explicitly establish characters and key set pieces as frame-left, frame-center, or frame-right.
  • Depth Layers: Design depth using foreground (FG), midground (MG), and background (BG).
  • Spatial Anchors: Fix subjects relative to permanent environment anchors (e.g., "near sunlit window", "at hearth frame-left", "behind center lecturer desk").
  • Eyelines & Orientation: Track facing directions (e.g., "facing off-frame right", "looking down toward altar").
  • Directorial Duty: If user source text has emotional dialogue but weak physical staging, the director must supply clean visual blocking.

4. Spatial Continuity & 180-Degree Axis

  • Axis Line: Establish the primary line of action between interacting subjects and stay strictly on one side of that axis.
  • Screen Direction: Characters on the left look screen-right; characters on the right look screen-left.
  • Motivated Axis Crossing: If the camera crosses the line of action, it must be explicitly motivated by:
    1. A continuous tracking camera move that arcs visibly across the line within a Macro-State.
    2. A head-on neutral shot (e.g., POV, direct frontal CU) acting as a transition buffer.
    3. A physical re-blocking where a character crosses the room.

5. Performance Translation: Literal Over Literary

In classical cinema, a camera cannot photograph unfilmable literary feelings—it records observable human anatomy, facial micro-expressions, posture, and physical actions.

Translate internal emotional states into observable physical behavior:

Literary / Abstract Feeling Directorial Physical Translation
“Heartbroken” Wet eyes, restrained breath, clenched jaw, forced rigid stillness.
“Checked out / Dissociated” Vacant unblinking stare, slack shoulders, delayed physical reaction.
“Tense / Guarded” Shallow rapid breathing, arms folded tight, white-knuckle grip on fabric.
“Suppressed Rage” Flare of nostrils, twitch in cheek muscle, fists trembling by sides.
“Terrified” Rapid blinking, throat swallowing hard, chin pulled back against collar.
  • Close-Ups (CU/MCU): Prioritize micro-expressions, gaze shifts, jaw tension, and breath rate.
  • Wide Shots (WS/EWS): Prioritize posture, silhouette, physical distance, and body slump/rigidity.

6. Baseline Dialogue, Audio, and Environment

In the baseline Auteur Script grammar:

  • Spoken Dialogue: Formatted as Speaker: "Spoken line." or Speaker (delivery): "Spoken line.".
    • Use (O.S.) when the speaker is off-screen.
    • Use (V.O.) for voiceover narration.
  • Sound Effects & Foley: Formatted with the [AUDIO] or [SFX] tag (e.g., [AUDIO] Cardboard scraping on concrete, [SFX] Heavy wooden latch click).
  • Environmental Shifting: Formatted with [ENV] when ambient conditions dynamically alter the scene (e.g., [ENV] Thick frost spreads rapidly across red ferns).
  • Audio Persistence: Continuous atmospheric sound (wind, rain, mechanical hums) that evolves across shots must be restated where active.
Back to top