เลิกสั่ง AI มั่วๆ! แจก “PROMPT” สร้างวิดีโอ Cinematic ใน Google Flow 

ภาพถ่ายครึ่งตัวที่สมจริงมากของ [นางแบบสาวผิวขาวเนียน รูปร่างสมส่วน] ใน [ฉากภายนอกของปราสาทเอลฟ์โบราณอันงดงามสไตล์ลอร์ดออฟเดอะริงส์ กำลังฉลองปีใหม่]
สวม [ชุดไทยแบบราชสำนัก] พร้อม [กำไลทองและสร้อยคอจี้ทอง]
[อายุ 18 ปี / เชื้อชาติ: ไทย-จีนผสม / ลักษณะใบหน้าแบบเอเชียสมัยใหม่: ดวงตาสวยคม จมูกโด่ง] ผิวดูเป็นธรรมชาติ มีริ้วรอยน้อย
หันหน้าเข้ากล้องด้วยสีหน้ามุ่งมั่น
แสงแบบภาพยนตร์ แสงหลักชัดเจน ขอบภาพคมชัด
ถ่ายด้วย [เลนส์ซูม] ระยะชัดตื้น พื้นหลังเบลอเล็กน้อย
การปรับสีแบบภาพยนตร์ ช่วงไดนามิกสูง เกรนน้อย
สมจริงสมบูรณ์แบบ สัดส่วนถูกต้อง ไม่มีบิดเบี้ยว

Role: You are an Award-Winning Cinematographer, Storyboard Artist, and Film Director. Your mission is to transform a single reference image into a cohesive 10-20 second cinematic sequence, providing keyframes optimized for Video AI generation (like Runway, Pika, or Luma).

Phase 1: Scene Analysis
- Subject ID: Identify all key subjects (A/B/C...). Describe visual traits (clothing/materials/forms), relative positions, and facing directions.
- Environment & Lighting: Interior/Exterior, spatial layout, background elements. Define lighting quality (Hard/Soft, Key/Fill/Rim) and implied Time of Day.
- Visual Anchors: Identify 4-6 features that MUST remain constant (Color palette, specific props, light sources, atmospheric textures like mist/grain).

Phase 2: Narrative & Theme
- Theme: A single sentence defining the core message.
- Logline: A cinematic trailer-style hook.
- Emotional Arc: 4 beats (The Setup → The Build → The Pivot → The Resolution).

Phase 3: Cinematography Strategy
- Sequencing Logic: How the framing evolves to tell the story (e.g., Wide to Close-up).
- Camera Movement Plan: Specific techniques (Push/Pull/Pan/Dolly/Orbit) and the psychological reason behind them.
- Lens & Depth: Focal lengths (e.g., 24mm for wide, 85mm for intimate), Depth of Field (DoF) transitions, and shutter feel.
- Color & Texture: Contrast levels, primary hues, and film grain specifications consistent with the reference.

Phase 4: Keyframes for Video AI (The Output)
Generate 9-12 sequential keyframes. Each must follow this strict format:
- [KF# | Duration | Shot Type (ELS/LS/MS/CU/ECU/etc.)]
- Composition: Subject placement, Fore/Mid/Background, leading lines.
- Action/Beat: Clear, visual movement or change in expression.
- Camera: Movement details (e.g., Slow 5% zoom-in, subtle handheld shake).
- Lens/DoF: mm focal length and focus point.
- Lighting/Grade: Consistency check for highlights/shadows.
- Ambiance: (Optional) Key sound design cues.

Phase 5: Master Storyboard Grid (Visual Output)
- Generate one single consolidated image representing the "Master Contact Sheet".
- The grid should be 3x3 or 4x3 depending on the number of frames.
- Each cell must contain: The frame visual + a clean label (KF#, Shot Type, Duration) in the margin.
- Strict Continuity: Same subject, same outfit, same environment, and consistent cinematic grading across all panels. Realistic DoF (shallow for close-ups, deep for wides).

[ใส่เนื้อเรื่องย่อ หรือ Story Synopsis ของคุณที่นี่]

STORY-TO-STORYBOARD META-PROMPT
IMPORTANT: Do not create the image; create the detailed prompt for the image. The image prompt must make reference to the story provided at the very top of this prompt. The prompt must follow exactly the details of the image prompt structure.

Analyze the story synopsis above and identify:
- The main subject(s) (person, pair, group, creature, vehicle, object)
- Their appearance and defining traits
- The environment and atmosphere
- The emotional or narrative beat
- Lighting/mood implied by the story

Create a full 3×3 cinematic storyboard grid with 9 distinct shots of the same subject(s) in the same environment, using consistent wardrobe, lighting, and atmosphere.
Output a single cohesive AI image prompt that includes all 9 frames (labeled 1–9), using the following structure:

OUTPUT FORMAT: Cinematic 3×3 Storyboard Prompt
Story Synopsis (interpreted): <one-sentence interpretation of the user's synopsis>
Create a professional 3×3 cinematic storyboard grid featuring the same subject(s) from the synopsis in the same environment. Maintain total consistency in appearance, clothing, lighting, mood, and environmental details. Each panel represents a distinct camera shot following cinematic conventions.

Row 1 — Establishing Context
- Extreme Long Shot (ELS): Full environment revealed, subject(s) small in frame. Match the story's setting, lighting, and mood.
- Long Shot (LS): Entire subject(s) visible head-to-toe (or full object/vehicle), standing naturally within the environment.
- Medium Long Shot (MLS / 3-4 / American Shot): Subject(s) framed from knees up (or 3/4 angle for objects), showing stance, posture, and core focus.

Row 2 — Core Coverage
- Medium Shot (MS): Waist-up framing. Capture the key action, attitude, or emotional beat implied by the story.
- Medium Close-Up (MCU): Chest-up. Focus on expression, micro-interaction, or dramatic focus.
- Close-Up (CU): Tight shot of the face (or front detail of an object). Cinematic depth of field, clear emotional state.

Row 3 — Details & Angles
- Extreme Close-Up (ECU): Macro detail: eyes, hands, symbolic object, texture, or a key story element.
- Low Angle Shot (Worm's Eye): Camera looking up at the subject(s) from below. Imposing, grand, or heroic based on the story's tone.
- High Angle Shot (Bird's Eye): Camera looking down from above. Spatial context, scale, or overview of the action.

Global Requirements:
- Same subject from image prompt(s) in all 9 frames
- Same clothing, hairstyle, props, gear, or accessories
- Same lighting conditions and color grading
- Consistent environment and weather
- Correct realism and cinematic depth of field per shot
- Photorealistic textures
- Cinematic camera behavior and focal-length accuracy

Final Line (technical): Extract the still x.y (Where x.y refers to the grid coordinate or still number the user later requests.)

Example of How This Works with a Synopsis:
If the top input was: "A lone desert scout tracks a signal across an ancient canyon at sunset."
Output (shortened example): Story Synopsis: A lone desert scout navigates a vast canyon wasteland at sunset while tracking a mysterious signal.
- ELS: small figure against vast ancient canyon
- LS: scout silhouetted in sunset light
- MLS: knees-up shot with gear visible
- MS: waist-up, scanning horizon
- MCU: chest-up, focused expression
- CU: dust-covered face, tracking device glow
- ECU: close-up of the device display
- Low Angle: heroic stance on canyon ridge
- High Angle: scout from above, canyon below
- Cinematic rule of thirds composition, asymmetrical framing, subject is not centered to allow the environment to tell the story