Create a 15-second Japanese action-comedy with synchronized sound. B interrupts spectacular sword co

Prompt

Create a 15-second Japanese action-comedy with synchronized sound. B interrupts spectacular sword combat by capturing Shibuya's giant monster with an enormous arcade claw. Its dangling legs kick helplessly. End on A's familiar silent, stunned face. Use the selected aspect ratio. REFERENCES A = black-haired heroine from @Image 1; voice = @Audio 1. B = white-haired heroine from @Image 2; voice = @Audio 2. Monster = @Image 3, approximately 100 meters tall. Preserve exact faces, hair, costumes, accessories, proportions and reference art styles. Identity references only; no sheet layouts. One A, one B, one monster. Allied heroines, equal standing height excluding ornaments. Heroines remain clean-lined, cel-shaded 2D anime. Monster retains its reference style. Shibuya, claw, cable, glass, concrete, dust and reflections are photorealistic. Match scene lighting without changing illustrated faces. VOICE ONE spoken line. A uses @Audio 1 at 0.3–1.1s and says: 「いくよ!」 Natural Japanese pronunciation and accurate lip sync. References define voices, not scripts; never replay recordings. B is silent; @Audio 2 is unused. After this line: no speech, narration, screams, grunts, gasps, laughter or audible character breaths. Music and effects continue. LOCATION AND PROPS Rain-wet nighttime Shibuya crossing: zebra stripes, towers, neon reflections. Screens show abstract colors, no text. No civilians, traffic or extra monsters. A fights on the near crossing; B stands on the nearby sidewalk. Monster occupies the far side between buildings. Preserve this geography and all damage. A has one right-hand katana and one waist scabbard. Her blade never changes shape. B stands behind a plain waist-high arcade pedestal, present from the opening. ONE joystick, ONE red button, no screen or labels. Left hand moves joystick; right index finger presses button. Claw: one giant silver three-pronged arcade grabber, central housing and ONE taut steel cable extending upward beyond the frame. Rounded gripping ends, consistent mechanical design. Open span exceeds the monster's shoulder width. CAPTURE RULES Joystick aligns the claw. ONE button press triggers descent, closure and lifting. Show the claw approaching OPEN, surrounding the upper torso, making contact, closing beneath the shoulders and THEN lifting. Three prongs form a supporting cage around the torso without piercing, crushing or passing through it. Head, arms and legs remain visible. The monster retains its size and design; only the dangling posture resembles a prize. No plush transformation. Feet leave the road only after the closed claw is pulled upward by its cable. No levitation before gripping. Once captured, it stays held and continues rising. No release or escape. Heroines remain outside the debris path and are never captured. 0.0–4.0s — EXPLOSIVE COMBAT Open immediately: the monster's huge beam cuts across the far road, rupturing asphalt in a traveling line of impacts. A sprints diagonally clear of the beam, trailing teal light and water spray. She kicks off a low facade and launches into ONE airborne sword slash. A speaks the opening line in the specified window. Her katana releases a broad teal-white crescent that visibly strikes the monster's torso. Brief impact slowdown, sparks and an air ring, then explosive follow-through. Monster staggers but stays standing. Its beam stops. A lands at the near crossing and raises her blade for another attack. Briefly establish B beside her pedestal, calmly watching. Fast tracking; readable silhouettes. 4.0–6.0s — THE BUTTON Medium shot includes B's face, hands, joystick and button. B moves the joystick once with her left hand while looking upward, aligning the claw above the monster. Her right index finger visibly presses the red button ONCE. Ordinary arcade click; neutral expression. Cut upward: clouds part around the enormous OPEN claw beginning its vertical descent, cable visible above it. Button press precedes descent. 6.0–8.5s — THE CLAW DESCENDS Low three-quarter wide view includes monster, claw and buildings. The open claw drops between the towers. Air pressure rolls the clouds outward and drives dust along the street. An outer prong clips an empty building corner, shearing off glass and concrete. Heavy fragments fall away from the heroines. The claw slows near the monster. Preserve clear body and prong silhouettes. Open prongs surround the torso, contact its sides and close beneath the shoulders in one readable movement. A deep mechanical clunk marks the completed grip. Show closure before lifting, without a cut hiding capture. 8.5–11.0s — AN ENORMOUS PRIZE Cable retracts, lifting the closed claw. Show both monster feet leaving the street and loose rubble falling beneath them. The monster hangs upright, supported beneath the shoulders, torso slightly slumped. Both legs make short, frantic alternating kicks. Arms flail without changing anatomy. Hold a wide view long enough to read the helpless dangling posture. As it rises, one swinging foot strikes an upper facade and breaks glass and concrete outward. Show foot contact before destruction. Claw stays closed, cable taut; no shrinking or toy transformation. Tilt upward with the ascent, preserving the claw and monster outlines against the clouds. 11.0–12.5s — B WINS Medium two-shot near the crossing: A left, B right beside the pedestal. Captured monster continues rising in the distant background, legs still kicking. B releases the controls and gives one small satisfied right-handed fist pump. A holds her katana ready with nothing left to fight. She looks up after the monster, then toward B. Her determined expression fades. Keep her blade clear of B and the pedestal. 12.5–15.0s — THE FAMILIAR FACE Frontal close-up of A, exact reference face and hairstyle. Wide eyes, small pupils, raised brows; lips part ONCE and hold still. Her familiar stunned, uncomprehending expression. A looks toward camera; no blush, crying or scream face. Short controlled push-in; hold the final full second with tiny natural movement. Sword outside the crop. Damaged Shibuya remains blurred behind her. End on A, no extra gag or cut back. CAMERA AND SOUND Directional tracking, brief impact slowdown, wide capture framing. Blur backgrounds and debris, preserve silhouettes. Instrumental percussion, strings and brass; no choir or vocal samples. Synchronize combat impacts, button, winch, claw and structural damage. Reduce dramatic music as the claw closes. Stop it when the monster's feet leave the road, revealing machinery, frantic foot impacts and falling rubble. Final face: faint wind, distant winch and settling debris. No monster vocals, canned laughter, comedy sting or victory announcement. EXCLUDE Text, subtitles, logos, watermarks, HUD, score displays, extra characters, clones, mixed identities, costume changes, extra limbs or fingers, fused bodies, morphing weapons, gore, injured heroines, changing claw geometry or monster release.

Reference Images

@tokyo_Valentine7

You may also like