動画 画像
ログイン

One Door Five Worlds

回視聴1,801
いいね203
コメント23
シェア78
ブックマーク14
引用0
Create a 28-second ultra-cinematic surreal fashion adventure starring my female AI model. Preserve her EXACT facial identity, facial structure, hairstyle, skin tone, body proportions and recognizable appearance throughout the entire video.

The film must feel like ONE continuous impossible journey, not separate AI-generated clips. Photorealistic, mysterious, premium and visually spectacular.

0–4 SEC — THE MYSTERIOUS HOTEL

Begin with an extreme cinematic close-up of my model's eyes.

Camera smoothly pulls backward, revealing her standing inside an impossibly long, luxurious old-European hotel corridor at night.

Warm golden wall lamps illuminate hundreds of identical doors stretching into the distance.

She wears an elegant flowing black evening dress with subtle jewellery.

She looks toward one particular door and confidently approaches it.

CAMERA: Slow dolly backward → smooth over-the-shoulder orbit as her hand reaches toward the brass handle.

She opens the door.

Intense white light floods through it and completely fills the lens.

SEAMLESS TRANSITION THROUGH THE LIGHT.

4–9 SEC — STORMY ICELAND

The camera passes through the doorway directly behind her.

Without stopping her walking motion, she emerges onto a spectacular Icelandic black-sand beach.

Towering dark cliffs, crashing ocean waves, mist and light rain surround her.

Strong wind naturally moves her hair and black dress.

CAMERA: Begin low beside her feet → rapidly rise into a cinematic tracking orbit around her.

Standing completely alone on the beach is another mysterious hotel door.

She reaches it and immediately opens it.

Camera follows directly through.

9–14 SEC — FUTURISTIC TOKYO

She steps through the same doorway directly into a futuristic Tokyo street at midnight.

Her walking movement continues seamlessly.

Wet streets reflect enormous neon signs, futurist

✨ 生成 元の投稿 Hitboard で開く

このプロンプトの使い方

プロンプトをコピーして Seedance 2.5 に貼り付けます。ショット構成・カメラワーク・タイミングは原文のまま残してください。それがこの結果を生んだ要素です。被写体・舞台・スタイルの語だけを差し替え、プロンプトが参照画像に触れていれば添付します。最初の生成がずれたら、指示を増やすよりプロンプトを短くしましょう。