OPENING — First-Frame Canvas

Block Header: [OPENING]
Parent Pillar: AESTHETIC (The World & Anchor)
Production Phase: Phase 1: STAGING (Pre-Production Setup)
Placement: Block 4 (The First-Frame Canvas Anchor \(S_0\))

1. Purpose & Functional Role: The First Canvas (\(S_0\))

[OPENING] locks the initial composed still frame (\(S_0\)) of the scene before any temporal motion begins. It acts as the first visual canvas that the video generation model uses as its physical anchor to reality:

  • The Anchor Principle: If the opening canvas is geometrically grounded with clear subject coordinates and environment anchors, the entire kinetic timeline holds continuity. If [OPENING] is vague, floating, or broken, downstream generation drifts and hallucinates.
  • Still-Frame Mandate: [OPENING] is a photograph taken at \(t = 0\). It captures a frozen moment of physical reality, not narrative progression or unfolding action.

2. The “Frame of Film” Law (Strict On-Screen Visibility)

[OPENING] describes strictly what is visually inside the initial camera frame:

  • No Off-Screen Characters: If the overall scene contains 4 characters, but the opening shot is a tight Medium Close-Up focused on 2 characters, never describe the off-screen characters in [OPENING].
    • Why: Describing off-screen entities in [OPENING] forces the video model to cram all characters into the initial tight composition, breaking shot geometry.
    • Resolution: Staging for characters outside the opening frame is introduced dynamically later during [EXECUTION] via spatial tags ([BLOCK], [ACT]) when the camera pans, tracks, or cuts to them.
  • On-Camera Scope Only: Describe only the visible subjects, immediate physical set pieces, and lighting falling within that specific opening lens angle.

3. The Directorial Stance: Cinematographer & Production Design Lingo

While [EXECUTION] is written from the stance of a film director calling action, [OPENING] is composed from the stance of a Director of Photography (Cinematographer / Still Photographer) collaborating with the Gaffer / Lighting Lead, Production Designer, and Costume Designer:

  • Cinematographic & Optical Lingo: Use photographic terms describing framing, lens perspective, depth planes (foreground, midground, background), focal falloff, and composition balance.
  • Assembled Practical Lighting & Shadow (Gaffer / DP): Describe how light physically assembles onto the subjects and set in this specific frame—where key lights strike, where rim lights catch hair/shoulders, how practical fixtures illuminate props, and where shadows fall. Do not repeat general lighting formulas from [AESTHETIC]; describe its concrete manifestation on this canvas.
  • Production Design & Set Dressing: Ground the frame in physical reality—describe how furniture, walls, practical lamps, architectural surfaces, and background set pieces frame the subjects.
  • Costume Drape & Physical State: Describe how garments physically sit on the body in this frozen posture (e.g., oversized sleeves falling past knuckles, damp coat collar pulled up against neck, creased denim across knees).

4. Formatting & Writing Style

  • Header: [OPENING] on its own line.
  • Structure: Written as one single, rich, natural descriptive paragraph (2–5 dense sentences).
  • Strict Zero-Tag Mandate: [OPENING] must NEVER contain square-bracketed tags (no [CAM], [ACT], [BLOCK], [AUDIO]). Sub-State tags belong strictly inside [EXECUTION].
  • Camera Declaration Style:
    • Express camera optics directly in natural cinematographic prose (e.g., Medium close-up at eye level, fixed side profile. ..., Centered medium wide shot in ..., Establishing wide framing of ..., POV handheld medium shot through ...).

5. The 6 Essential Canvas Dimensions

A production-ready [OPENING] establishes the following 6 physical coordinates:

  1. Optical Perspective & Shot Size (Cinematographer):
    • Initial shot size (WS, MS, MCU, CU, POV), camera height (eye-level, low-angle, high-angle), and lens perspective relative to the 180° line of action.
  2. In-Frame Subject Geography (Visual Staging):
    • Explicit horizontal coordinates (frame-left, frame-center, frame-right) and depth layers (foreground, midground, background) for all visible subjects.
  3. Physical Posture, Eyelines & Costume Drape (Costume & Performance):
    • Body posture (seated, standing, leaning), facing direction, gaze targets, and physical contact or proximity between visible subjects, including how garments hang or crease in that pose.
  4. Assembled Light & Shadow Interaction (Lighting / Gaffer):
    • How the scene’s illumination actively strikes the visible subjects and surfaces in this frame (e.g., amber desk lamp light skimming the edge of Robert’s jaw, fluorescent backlight silhouetting Samira against steamed glass, diagonal rain-streaked window light cutting across midground). Never restate abstract lookbook specs; describe the active light falloff in this composition.
  5. Pre-Loaded Hand Allocations & Prop Grips (Props Master):
    • If a subject interacts with an object in early execution beats, their hand-grip state must be pre-loaded at \(t = 0\) (e.g., “holding a heavy bowling ball at chest height”, “right hand resting on the carved armrest”).
    • Why: Unspecified hands default to resting positions, causing props to spontaneously teleport into grip when motion starts.
  6. Environment & Background Anchors (Production Design):
    • Permanent physical set pieces, architectural textures, background depth elements, and practical light fixtures framing the shot (e.g., “spinning dryer drums behind them”, “rain-streaked sash windows on frame-right”, “glowing amber desk lamp at center”).

6. The “No Mid-Scene Environment Hallucination” Rule

The video generation model anchors its spatial reality to the environment established in \(S_0\):

[!IMPORTANT] Ground the Background Early: You cannot introduce new static environment set pieces in the middle of [EXECUTION] out of thin air unless the camera explicitly moves or changes angle. Always establish the background architecture, surfaces, and practical light sources visible in that frame within [OPENING].

7. The First Macro-State Handoff Mandate (\(S_0 \to S_1\))

Line 1 of [EXECUTION] (\(S_1\)) is strictly bound to the ground-truth canvas established in [OPENING] (\(S_0\)):

[!IMPORTANT] The Zero-Jump Law & Initial Tag Declaration: 1. Declare Matching [CAM] on Line 1: Because [OPENING] contains no bracketed tags, Line 1 of [EXECUTION] must declare the initial Sub-State [CAM] (or named anchor [CAM 01]) matching the exact optical framing and perspective established in [OPENING] (e.g., [CAM 01] MCU, eye-level, fixed side profile -> [ACT] ...). 2. Zero Optical or Spatial Jump: Line 1 must not jump-cut to a new angle, change camera distance, or teleport characters to new frame coordinates. 3. Prop Grip Continuity: Held props established in \(S_0\) must be the physical origin point for any kinetic interaction on Line 1.

8. Prohibited Anti-Patterns & What to Avoid

  • ❌ No Square-Bracket Tags: Never place [CAM], [ACT], or [BLOCK] tags inside [OPENING]. Tags belong exclusively inside [EXECUTION].
  • ❌ No Unfolding or Narrative Action:
    • Never describe motion that takes time (e.g., avoid “John walks into the room, greets Robert, and sits down”).
    • Describe the frozen seated state at \(t = 0\) (e.g., “John sits slouched frame-left opposite Robert frame-right”).
  • ❌ No Off-Screen Entity Injections:
    • Never describe where off-screen characters are standing if they are not in this camera view.
  • ❌ No Audio Cues or Foley:
    • [OPENING] is a silent visual photograph (\(S_0\)). All sound effects, foley impacts, and spoken dialogue belong in [EXECUTION] (or static room tone in [AESTHETIC]).
  • ❌ No Lookbook Redundancy:
    • Do not re-list full wardrobe color specs, film grain types, or lighting palettes from [AESTHETIC] unless directly needed to describe the immediate physical composition.
  • ❌ No Initial Angle Disconnect:
    • Line 1 of [EXECUTION] must inherit the exact camera angle, subject positions, and prop grips established in [OPENING].

9. System Mechanism & Downstream Relations

\[\text{[OPENING]} \equiv S_0 \implies S_1 = f(S_0 \mid \text{STAGING})\]

  • State System Role: Acts as the Initial State Vector (\(S_0\)) at \(t = 0\), establishing the frozen spatial ground truth before time progression begins.
  • Upstream Input: Materializes the static lookbook parameters from [AESTHETIC] within the spatial boundaries and 180° axis locked in [LOGIC].
  • Downstream Interface:
    • To [EXECUTION]: Forms the direct starting baseline for Line 1 (\(S_1 = f(S_0 \mid \text{STAGING})\)). Execution Line 1 mutates coordinates directly from \(S_0\) without spatial teleportation, angle resets, or hand-grip discrepancies.
Back to top