Simply Ray Seedance

Seedance

AI動画プロンプト集: 雨の東京デートとクレーンゲームの景品

二人が小さな店で食事を分け、傘で雨をしのぎ、ゲーム店でぬいぐるみを獲得します。Seedanceの動画プロンプトで、手持ちPOV、自然な笑い声、服装や人物の一貫性を保つ構成を学べます。

元のプロンプト

プロンプトは原文のまま掲載しています。コピーしても翻訳や動画生成は行われません。

Create a 30-second, 16:9 ultra-realistic live-action POV vlog using the uploaded image as the exact character reference. Preserve the woman’s facial identity, features, skin tone, hairstyle, body proportions, and appearance. Keep the same boyfriend, identity, outfits, and proportions throughout. Their relationship feels natural, playful, affectionate, and unscripted. STYLE: Handheld smartphone footage with an early-2000s MiniDV/home-video feel—natural shake, imperfect framing, autofocus breathing, motion blur, mild noise, exposure shifts, soft evening light. No cinematic stabilization, commercial look, or influencer posing. WARDROBE: Woman: cream knit top, olive-gray trousers, black-and-white sneakers, burgundy shoulder bag, silver earrings, low ponytail. Boyfriend: light-gray jacket, muted blue T-shirt, black trousers, white sneakers, canvas tote. Keep identical throughout. 00:00–00:05: Shaky selfie outside a Tokyo apartment. Boyfriend appears and suggests going out. She smiles and follows him. 00:05–00:12: They discover a tiny local restaurant, eat together, share bites, laugh naturally, and playfully fight over the last piece of food. 00:12–00:17: They leave and discover light rain. He opens an umbrella; they walk closely through wet streets, laughing and avoiding puddles. 00:17–00:24: They find a small old arcade. She films him trying a claw machine. He fails, then wins a tiny plush toy and proudly shows it to the camera. 00:24–00:30: Warm evening storefronts glow. They walk together, joke in a shop-window reflection, then at a crossing he holds up the plush like a trophy. She laughs and the camera swings toward the glowing Tokyo street before cutting naturally. AUDIO: Authentic Tokyo ambience only—street noise, restaurant sounds, laughter, rain, footsteps, arcade sounds, traffic, bicycle bells, and crossing signals. No music, narration, voiceover, subtitles, captions, logos, watermarks, or random text. REALISM: Consistent identities, realistic Korean facial features, Japanese environment, food, hands, rain, reflections, physics, and natural couple interactions. No distorted faces, extra fingers, duplicated people, outfit changes, exaggerated acting, CGI, or tourist-landmark montage.