原文
GPT Image 2 · en
A photorealistic close-up tabletop scene showing a large mound of uncooked long-grain white rice poured onto a warm brown wooden surface, filling almost the entire frame. The camera is slightly elevated and angled toward the pile so the rice forms a broad dome centered in the image, with thousands of individual grains sharply visible and softly lit by warm indoor lighting. Hide a tiny easter egg in the rice: exactly 1 unusual rice grain near the lower-left area is rotated horizontally and has the text "{argument name="engraved text" default="GPT Image 2"}" neatly engraved on its surface in small legible letters. Add a circular magnified callout in the bottom-left corner with a bright red outline, zooming in on that single engraved grain among surrounding rice grains so the text becomes readable. Also add exactly 1 small picture-in-picture inset in the bottom-right corner showing a casual indoor livestream or interview setting with 4 seated people around a round wooden table and an open laptop, warm ambient lighting, and faces slightly indistinct; the inset should feel like a secondary contextual thumbnail, much smaller than the main rice image. Prioritize realistic texture, convincing macro detail, natural shadows, subtle ivory color variation in the rice, and clear but believable engraved text on the highlighted grain.使い方
- パラメータを含む元のプロンプト全体をコピーします。
- 被写体、文字、構図を自分の要件に置き換えます。
- 対応する外部ツールで使用し、利用条件と結果を確認します。
開始前の確認
- 参考画像の要否を確認してください。互換モデルの表示は、そのモデルでの生成実績を意味しません。
- 名前、数値、画像内の文字を確認します。モデルの版で結果は変わります。
