ヒ・そば・屋のコーディングチェック

プロンプト

Wan 3.0 実送信プロンプト（24秒・自然発話優先） · revision 1

Create exactly one 24-second 16:9 video with synchronized Japanese dialogue and sound, 480P. A tense supernatural examiner parody that abruptly ends as a mundane trespassing comedy. Cinematic 3D characters in a realistic small Japanese apartment, preserving both different reference body proportions. Four scenes, one continuous generation. No subtitles or captions.
NATURAL SPEECH TAKES PRIORITY: All shot timestamps below are approximate editorial guides, NOT speech-duration targets or mandatory utterance start/end times. Deliver each Japanese line at a natural conversational pace for its emotion. Never stretch syllables, vowels, final words or pauses to fill a time window; never slow down or rush a line to hit a timestamp. The written 「やでー」 permits only a brief natural Kansai ending, not a sustained vowel. Let each utterance finish naturally, then let the mouth rest. Use any remaining screen time for silent reactions, a glance, a small gesture, room ambience or the departing car. Shift internal cut boundaries as needed to preserve complete natural utterances and the short question-and-answer pause. Keep the scene order and total video duration of 24 seconds. Do not add dialogue to fill time.
References:
Image 1: Yametaro exact identity and purple patterned shirt, compact stylized body, black hair and round glasses with white lenses.
Image 2: Sobaya exact identity: huge muscular body, neutral gray skin on all exposed areas, short black hair, white rigid mask with the exact black openings and red markings. White tight short-sleeve T-shirt, black trousers. Preserve mask for EVERY frame, including arrest and inside the police car. Use no beer mug in this episode; his prop is one ordinary playing card. Do not alter the mask into Hisoka's face. Hisoka-inspired performance, poised wrist gestures, predatory examiner confidence and playful sing-song timing; retain Sobaya's canonical voice identity.
Use character sheets only for design, never reproduce labels, panel borders, white backdrops or multiple views as extra characters.
Audio 1: Yametaro voice-timbre reference only. Audio 2: Sobaya voice-timbre reference only. Do not copy reference words or timing. Generate the following new Japanese dialogue. Symbols heart/spade are acting annotations only, never spoken.
Yametaro identity lock — Image 1 is the exact design authority, not loose inspiration. Preserve its head width-to-height ratio, hair silhouette and hairline, cheek and ear placement, short neck and compact body proportions in every frame. Keep the round black spectacle frames and the sheet's simple white lens/eye areas; do not invent visible pupils, irises, realistic eyeballs, eye sockets, eyelids or eyebrows. Keep the tiny black forehead mark separate from the low, subtle skin-colored nose; no black nose dot, outlined nostrils or projecting nose. Preserve the round cheeks and flat pink blush marks, smooth skin-colored chin, ears and nape, with no black contour bands or facial creases. Do not lengthen the forehead/head, inflate the hair, enlarge ears or add realistic lips. Animate the small inset mouth using the sheet's rest and A/I/U/E/O views as articulation guidance, with visible mouth interior when speaking; do not freeze the smile, protrude the lips or stretch the mouth into a new facial design. Emotion and cinematic lighting must not redesign the face. Wardrobe follows the scene instructions independently of identity. A previous generated take is not an identity reference.

Overall continuity: One Yametaro and one Sobaya. Afternoon. A modest private apartment, laptop on desk, entrance behind desk. Sobaya is an uninvited stranger posing as an examiner, not an authorized employee. The laptop shows defocused code-like shapes, no legible UI or company logo. No magical powers or combat. Hisoka-like tension comes from framing, pauses, delicate card handling and a sweet threatening cadence. The final arrest is dry and absurd.
Shot 1 [0.0-6.0 s] — Medium close-up of Yametaro seated at his laptop at home. Typing stops. With hopeful casual Kansai delivery, after a brief opening beat he says: 「アクシデンチュアのコーディング試験に提出するやでー」. Hold until the final vowel ends. Company pronunciation: あくしでんちゅあ. Animate his mouth, jaw and cheeks to the generated phonemes using Image 1 mouth designs, white lenses unchanged.
Shot 2 [6.0-14.0 s] — Begin this scene with a distinct door-latch click, gacha. The door opens and Sobaya steps inside without permission, the same white mask correctly worn. Once he enters, settle into a chest-up shot, card held below face. He glances toward the laptop, gently tilts his head, then says: 「うーん、君は不合格」 with a sweet heart-like lifted ending. After a brief natural pause, he says: 「君の敗因はメモリの無駄遣い」, calm and pointed with a sharp final fall, at an ordinary conversational pace without drawing out words. Keep the rigid mask unchanged, never fabricate moving lips on its surface; voice and subtle head and body gestures indicate his speech. No exposed face.
Shot 3 [14.0-19.5 s] — Reverse close-up of Yametaro after Sobaya finishes naturally. He leans back and sharply blurts: 「誰やお前！」 in one spontaneous, naturally paced outburst. Make the shout forceful through voice, posture and reference-based speaking mouth; keep the same white lenses, head proportions and hair silhouette, with no bulging eyes or enlarged face. After Yametaro finishes, cut to Sobaya's masked close-up. One relaxed hand points to himself, card now held still. He answers naturally: 「僕？ 僕はヒソバ」. Pronounce ぼく？ ぼくはひそば, with a playful pause after the question and a self-important theatrical introduction. His mask remains rigid and attached, no face replacement. Hold a tiny ominous silence.
Shot 4 [19.5-24.0 s] — Abrupt hard cut to the apartment street, flat ordinary daylight; all sinister tension disappears. The rear door of a stationary black-and-white Japanese patrol car is already open. One uniformed Japanese police officer guides the same masked Sobaya, already standing beside the open door, into the rear seat. A second officer is already in the driver's seat. Show Sobaya duck, sit fully inside, rear door close, then car gently drive away, red beacon flashing. His white mask never shifts or disappears; hands are now empty, card is no longer held. No dialogue from anyone in this scene. End on the departing police car at 24.0 s; the arrest itself is the visual punchline.
Audio: Japanese dialogue exactly once each in order. No narrator, extra words or reference-audio phrases. Preserve canonical voice timbres; Sobaya's manner becomes relaxed, teasing and theatrical, while his articulation stays naturally paced; express confidence through tone and brief gestures rather than prolonged syllables. Yametaro uses natural Kansai delivery. Lip sync only on Yametaro's visible uncovered mouth: generate phonemes and mouth animation together, close mouth between phrases, never cut before final sound. Sobaya speaks from behind his unchanged rigid mask. Keyboard taps, door latch, soft footsteps, car door closing, engine and distant short siren, all generated with the video and quieter than dialogue. No background music. No spoken suit names.
Hard constraints: natural speech with no duration-filling vowel elongation or artificial speech stretching; internal timestamps are flexible editorial guides; exactly 24 seconds total, one task, four sequential scenes; no subtitles, title cards, overlays, speech bubbles or readable screen text; no duplicate characters or disembodied heads; no change of Yametaro design; Sobaya's canonical white mask stays correctly attached in every frame, never lifted, removed, translucent, reshaped or replaced with a human face; no beer mug; no arrest of Yametaro; Sobaya is visibly inside the car before it moves; never animate the rigid mask as a mouth.