دليل الذكاء الاصطناعيتواصل معنا

VR World-Swap Spot — Seedance 2.0 + GPT Image 2 Tutorial | Meta Quest 3

\

Meta Quest 3 — "Enter Your Worlds"

VR Spec Spot · Seedance 2.0 Production Bible

A 4 × 15s continuity chain (S1 → S2 → S3 → S4). A boy puts on a Meta Quest 3, drops into Zombie Wars, swaps to Kung Fu Master 2026, fights a sumo — who turns out to be his mother calling him for dinner.


0 · Reference asset protocol

Upload these once and reuse the exact @handles across all four scenes so identity, room and product never drift.

Handle

Role (atomic element)

Source

Rule

@boy_ref

Subject identity — 8-yr-old boy, exact face, proportions, curly black hair

reference_1 (the uploaded frame)

Never regenerate the face; always cite this handle

@room_ref

Scene environment — cozy child bedroom, wooden bunk bed, toy shelves, framed pictures, warm window daylight

reference_2

Unchanged until each scene's transition

@quest3_ref

Product identity — Meta Quest 3 headset + box

reference_3

Must stay accurate and visible

@ui_cards_ref

UI design only — the three game cards (Zombie Wars / Football Stars / Kung Fu Master 2026): layout, colors, typography, framing

reference_4

Design only. Never import its background or scene

@prev_scene

Continuation — the previous rendered clip's tail

reference_5 (S2–S4)

Carries lighting, wardrobe, headset state forward

Seedance mapping logic (from the skill's atomic-element table): identity, product and UI = asset references (high info density, must be images). Camera language, VFX and pacing = text (standard cinematic vocabulary is clearer as text). Environment = hybrid (asset base + text modifiers for the world-shift).


1 · Global grade block (paste at the head of every scene)

STYLE — Hollywood live-action blockbuster, shot on 65mm IMAX. Ultra-realistic, physically-based photoreal film — zero CGI feel, real-world physics, natural motivated lighting. Filmic HDR grade: warm-amber practicals against cool teal digital worlds, deep filmic blacks, gentle highlight roll-off, fine 35mm grain, shallow depth of field, subtle anamorphic bloom on light sources. Handheld energy with stabilized precision. No cartoon look, no plasticky skin, no uncanny artifacts, noidentity drift. Duration 15s, 16:9, fast-paced with clean cinematic cuts.

2 · Enhanced Seedance 2.0 prompts

SCENE 1 — The Unboxing → Zombie Wars

sql
STYLE — Hollywood live-action blockbuster, 65mm IMAX, ultra-realistic photoreal film, Unreal-Engine-5-grade fidelity with zero CGI feel, real-world physics, natural motivated lighting, filmic HDR grade, fine 35mm grain, shallow depth of field, subtle anamorphic bloom. Duration 15s, 16:9, fast-paced with clean cinematic cuts.

SUBJECT — the same 8-year-old boy, exact face, proportions and curly black hairstyle preserved. Use @boy_ref as the subject identity reference.
ENVIRONMENT — cozy detailed child bedroom, wooden bunk bed, toys, framed pictures and maps on the walls, soft daylight from the window. Use @room_ref as the environment; it stays unchanged until the transition. Use @quest3_ref for the Meta Quest 3 product. Use @ui_cards_ref for the game-card VISUAL DESIGN ONLY — colors, typography, framing — never its background or scene.

[0:00–0:02] Handheld medium shot, slight front three-quarter angle, natural micro-shake. The boy notices the Meta Quest 3 box on the desk and instantly reaches toward it with excitement; warm window light rakes across his face.
[0:02–0:04] Quick cut, close shot. He already has the headset in his hands and slides it smoothly onto his head in one clean, controlled motion.
[0:04–0:06] Quick cut, medium shot. He grabs both controllers naturally and raises them into position as the camera begins a slow, deliberate push-in.
[0:06–0:08] The push-in continues; a soft digital activation glow blooms around him while the real room stays fully visible and physically lit — no environment change yet.
[0:08–0:10] Three floating holographic UI cards materialize in front of him, slightly curved in space: "Zombie Wars", "Football Stars", "Kung Fu Master 2026". Fully transparent holographic projections integrated into the room with realistic light spill; card design matches @ui_cards_ref only.
[0:10–0:11] He reaches forward and selects "Zombie Wars"; a soft activation pulse ripples from the card, clean UI response, no flicker.
[0:11–0:15] The room begins a smooth digital transformation — walls dissolve softly from the edges into a dark zombie world while lighting transitions realistically. The boy stays centered; his controllers morph seamlessly into a realistic pump-action shotgun that he grips naturally. The camera performs a slow cinematic orbit as the environment fully resolves into the zombie world.

AUDIO — soft movement, subtle breathing, clean UI hum, soft selection click, rising low cinematic tension.

CONSTRAINTS — use ONLY the UI card visuals from @ui_cards_ref; import NO background or environment from it; UI stays stable with no flicker; no face change, no identity drift; no cartoon effects; realistic light interaction between holograms and room; smooth believable transformation physics.

SCENE 2 — First-Person Zombie Wars → Kung Fu

sql
STYLE — Hollywood live-action blockbuster, 65mm IMAX, ultra-realistic photoreal film, zero CGI feel, real-world physics, natural motivated lighting, filmic HDR grade, fine 35mm grain, shallow depth of field, anamorphic bloom. Duration 15s, 16:9, fast-paced with clean cinematic cuts.

CONTINUATION — direct continuation of @prev_scene.
SUBJECT — the same 8-year-old boy, exact face, proportions and hair preserved. Use @boy_ref for identity, @room_ref for the base room, @quest3_ref for the product, and @ui_cards_ref for the card design ONLY (no background transfer).

[0:00–0:01] Hard cut into first-person view from the boy's eyes; subtle headset vignette, clean realistic optics.
[0:01–0:04] The room is now semi-transformed. Windows shake violently as zombies break through and climb inside — realistic glass fragments, drifting dust, natural physics. The boy raises the pump shotgun into frame from first person.
[0:04–0:08] First-person shooting sequence, controlled aim: shot 1 — a clean hit, zombie falls backward; shot 2 — a side hit with realistic impact reaction; shot 3 — a third zombie gets close and is dropped at short range. No exaggerated gore, grounded physical reactions, natural recoil and slight handheld shake.
[0:08–0:09] The action pauses subtly; the holographic UI reappears in first-person space — "Zombie Wars", "Football Stars", "Kung Fu Master 2026", consistent with @ui_cards_ref design only.
[0:09–0:11] The boy's hand reaches forward into frame and selects "Kung Fu Master 2026"; a soft activation pulse, clean UI response.
[0:11–0:15] The environment dissolves smoothly from the zombie scene into a traditional kung-fu dojo — wooden floors, warm light, balanced shadows. A large sumo wrestler appears directly ahead in first person, grounded stance, realistic body weight and presence. The weapon vanishes naturally from the hands into empty hands, ready for combat; subtle breathing motion and a slight stance shift as the fight is about to begin.

AUDIO — distant zombie sounds fading out, shotgun blasts and shell sounds, clean UI hum, soft selection click, ambience shifting into a calm dojo, low tension build.

CONSTRAINTS — first-person continuity from @prev_scene; use ONLY @ui_cards_ref card design; no background transfer from it; stable UI, no flicker; no identity drift; no cartoon effects; realistic optics and physics.

SCENE 3 — Kung Fu Master · The Fight

Visual Basic
STYLE — Hollywood live-action blockbuster, 65mm IMAX, ultra-realistic photoreal film, zero CGI feel, real-world physics, natural motivated lighting, filmic HDR grade, fine 35mm grain, shallow depth of field, anamorphic bloom. Duration 15s, 16:9, fast-paced with clean cinematic cuts.

CONTINUATION — direct continuation of @prev_scene.
SUBJECT — the same 8-year-old boy, exact face, proportions and hair preserved. Use @boy_ref for identity, @quest3_ref for the headset (must stay visible and accurate on his head).

[0:00–0:01] Hard cut to a third-person cinematic view from outside. The boy — clearly wearing the Meta Quest 3 — and the sumo wrestler are both framed in the dojo. Camera low and dynamic.
[0:01–0:04] The boy sprints forward confidently as the camera tracks laterally; he drops low and slides smoothly between the sumo's legs with realistic friction and body control.
[0:04–0:06] He emerges behind the sumo in one continuous motion and, without stopping, delivers a sharp kick to the back of the knees; the knees buckle naturally and the sumo drops into a kneeling position with realistic weight shift.
[0:06–0:09] The boy moves around to the front with stylish, controlled footwork, jumps and performs a clean spinning kick in mid-air; the kick connects to the upper body, knocking the sumo fully off balance.
[0:09–0:11] The sumo falls heavily to the ground — realistic impact, slight floor vibration, dust settling.
[0:11–0:13] The sumo recovers quickly, stands, and grabs the boy — still wearing the headset — pulling him slightly closer.
[0:13–0:15] Tone shift. The sumo looks at the boy and speaks in a calm female voice. Dialogue (clear, natural English): "Dinner is ready." The camera holds a medium shot, contrasting the action with the humor.

AUDIO — foot slides, fabric movement, impact hits, controlled breathing, then clean dialogue over soft room ambience.

CONSTRAINTS — direct continuation of @prev_scene; Meta Quest 3 headset stays visible and consistent; no identity drift; no cartoon effects; realistic martial-arts motion and physics; clean readable choreography; natural impact reactions; dialogue clear and natural in English.

SCENE 4 — The Reveal · Back to Reality

sql
STYLE — Hollywood live-action blockbuster, 65mm IMAX, ultra-realistic photoreal film, zero CGI feel, real-world physics, natural motivated lighting, filmic HDR grade, fine 35mm grain, shallow depth of field, anamorphic bloom. Duration 15s, 16:9, fast-paced with clean cinematic cuts.

CONTINUATION — direct continuation of all previous scenes via @prev_scene.
SUBJECT — the same 8-year-old boy, exact face, proportions and hair preserved. Use @boy_ref for identity, @room_ref for the real bedroom, @quest3_ref for the product.

[0:00–0:02] Continuation, medium shot. The boy is still held by the sumo, still wearing the Meta Quest 3, and reacts with slight confusion. Dialogue: "What?"
[0:02–0:05] The sumo reaches up and begins removing the Meta Quest 3 from the boy's head; as the headset lifts, a smooth transformation begins.
[0:05–0:08] Mid-motion transformation — the sumo seamlessly morphs into the boy's mother while removing the headset, and the environment transitions naturally from the dojo back into the real bedroom. No hard cut: a fully blended transformation with continuous, realistic lighting.
[0:08–0:12] Now fully in the real room, the mother stands in front of the boy holding the headset, calm and natural. Dialogue (grounded tone): "Dinner is ready, Johny. That's enough gaming."
[0:12–0:14] The mother turns and walks out of the room naturally; the boy stays seated, slightly frozen, processing the shift.
[0:14–0:15] Fast clean product-shot transition: the Meta Quest 3 box slides into frame and settles on the desk in front of camera — perfect product framing, clean lighting, minimal motion settle. END.

AUDIO — soft dialogue clarity, subtle room ambience, light fabric movement, smooth transition sound design, clean product impact.

CONSTRAINTS — direct continuation of all previous scenes; perfect identity preservation for the boy and a natural mother appearance; no cartoon morphing — the transformation must feel realistic and seamless; smooth, physically believable environment transition; Meta Quest 3 clearly visible and accurate; clean cinematic product ending frame; dialogue natural and clearly audible in English.

3 · GPT Image 2.0 keyframe prompts (for the storyboard frames)

Generate these first as still keyframes, then feed the matching still into Seedance as the scene's opening frame for maximum lock. Written in film-photography language (35mm cinematic film still) rather than "photorealistic" — GPT Image 2.0 renders faces far better when framed as film. In most frames the headset covers the eyes, which further protects likeness. Attach @boy_ref / @quest3_ref as image references where the tool supports it.

KF-01 · S1 — Spots the box

Visual Basic
A cinematic 35mm film still, wide-aperture medium shot: an 8-year-old boy with curly black hair in a color-block bomber jacket (olive, tan, rust panels, navy ribbing) stands in a cozy sunlit child's bedroom — wooden bunk bed, toy shelves, framed animal and car pictures on patterned wallpaper — and reaches with bright excitement toward a Meta Quest 3 box on a wooden desk. Warm morning window light rakes from frame left, soft filmic grain, shallow depth of field, teal-warm blockbuster grade, gentle highlight roll-off. Handheld energy, natural skin, no plastic sheen.

KF-02 · S1 — Holographic game menu

sql
A cinematic 35mm film still, medium shot: the same curly-haired 8-year-old boy wearing a white Meta Quest 3 headset, holding two controllers raised, standing centered in his warm sunlit bedroom. Three softly glowing, semi-transparent holographic game cards float in a gentle arc in front of him, cool teal light spilling onto his face and the room — cards read "ZOMBIE WARS", "FOOTBALL STARS", "KUNG FU MASTER 2026" in clean game-UI typography. The real bedroom stays fully visible behind the holograms. Filmic HDR grade, fine grain, shallow depth of field, realistic light interaction, IMAX blockbuster look.

KF-03 · S1 — World dissolve to zombie

Visual Basic
A cinematic 35mm film still, slow-orbit medium shot: the curly-haired boy in the Meta Quest 3 headset stands centered as the cozy bedroom dissolves from the edges into a dark, moody zombie world — crumbling walls, cold blue moonlight cutting through dust. In his hands the controllers have become a realistic pump-action shotgun, gripped naturally. Warm-to-cold lighting transition mid-morph, deep filmic blacks, subtle anamorphic bloom, fine grain, tense blockbuster atmosphere, no CGI feel.

KF-04 · S2 — First-person zombie breach

Visual Basic
A cinematic 35mm film still, first-person POV with subtle VR headset vignette: looking down the barrel of a raised pump-action shotgun held in a child's hands, into a semi-transformed bedroom-turned-battleground. Two windows shatter as silhouetted zombies climb through amid flying glass fragments and drifting dust, lit by cold blue backlight and warm interior spill. Grounded, non-gory, physically real. Filmic grain, shallow depth of field, high dynamic range, gritty IMAX blockbuster grade.

KF-05 · S2 — Dojo forms, sumo appears (POV)

less
A cinematic 35mm film still, first-person POV with soft headset vignette: empty child-sized hands raised in a ready guard, facing a large sumo wrestler standing in a traditional kung-fu dojo — warm wooden floors, paper screens, balanced directional light and soft shadows. The sumo has real body weight and grounded presence. Calm warm palette after chaos, fine film grain, shallow depth of field, photoreal, IMAX blockbuster look.

KF-06 · S3 — Slide between the legs

sql
A cinematic 35mm film still, low dynamic third-person tracking shot: the curly-haired 8-year-old boy wearing a white Meta Quest 3 headset slides low across a polished dojo floor, passing between the wide stance of a large sumo wrestler, motion blur on his trailing limbs, dust kicking up. Warm dojo light, wooden floor reflections, deep filmic blacks, shallow depth of field, crisp readable action choreography, IMAX blockbuster grade, real physics.

KF-07 · S3 — Mid-air spinning kick

less
A cinematic 35mm film still, dynamic medium action shot: the curly-haired boy in a Meta Quest 3 headset caught mid-air in a clean spinning kick, foot connecting with the upper body of a large off-balance sumo wrestler in a warm-lit dojo. Subtle motion blur, dust and light streaks, powerful grounded impact, deep filmic contrast, fine grain, shallow depth of field, blockbuster martial-arts cinematography, no cartoon effects.

KF-08 · S3 — "Dinner is ready" beat

sql
A cinematic 35mm film still, held medium shot: a large sumo wrestler gently holds the curly-haired 8-year-old boy — still wearing a white Meta Quest 3 headset — close, both in a warm-lit dojo. The tone is calm and slightly comedic, a beat of stillness after action. Soft warm key light, gentle rim, natural skin, fine film grain, shallow depth of field, IMAX blockbuster grade.

KF-09 · S4 — Sumo morphs into mother

Visual Basic
A cinematic 35mm film still, blended-transformation medium shot: mid-morph, a large sumo wrestler seamlessly dissolving into a warm, natural-looking mother as she lifts a white Meta Quest 3 headset off the curly-haired boy's head; the environment simultaneously cross-blends from a warm dojo into a cozy sunlit child's bedroom. Continuous realistic lighting through the blend, no hard cut, soft warm palette, fine grain, shallow depth of field, believable and seamless, IMAX blockbuster grade.

KF-10 · S4 — Product hero end frame

Visual Basic
A cinematic 35mm film still, clean locked product shot: a Meta Quest 3 retail box settling on a wooden desk in a softly blurred cozy child's bedroom background, warm window light, crisp key light on the box, subtle reflection, shallow depth of field, premium commercial product framing, fine film grain, minimal motion, polished IMAX blockbuster finish. End card composition.

4 · Production notes

  • Render order matters. Render S1 first, then chain each next scene from the tail frame of the previous render via @prev_scene. This is what keeps wardrobe, lighting, headset state and the boy's identity locked across all 60 seconds.
  • Keyframe-then-animate. For the trickiest beats (world dissolves, the sumo→mother morph), generate the GPT Image 2.0 still first and hand it to Seedance as the opening frame. It dramatically reduces drift on transformations.
  • UI discipline. The three cards come only from @ui_cards_ref — design, not scene. Repeat the "no background transfer" constraint in every scene that shows the menu (S1, S2).
  • Headset visibility. In S3 and S4 the Quest 3 must stay on the boy's head — it's the product and the joke. Call it out in the constraints every time.
  • Two dialogue lines, both clean English: "Dinner is ready." (S3, calm female voice from the sumo) and "Dinner is ready, Johny. That's enough gaming." (S4, mother). Keep "What?" (S4, boy) short and dry.
  • Grade continuity. Warm bedroom → cool teal digital → red-tinged zombie → warm dojo → warm resolution. The storyboard color ribbon tracks this arc; hold the grade steady within each world so only the world changes, not the film stock.