
原文
GPT Image 2 · en
Scene:
Create a square cinematic still for Aurora Glacier Silent Theatre. Use production-grade photoreal lighting, disciplined lens behavior, and a composition that feels like a finished key frame from an original film rather than a casual snapshot. An Arctic glacier plain opens under a sweeping aurora curtain, with deep night sky, luminous green and faint violet ribbons overhead, fractured ice catching reflected color, and a breath-frozen atmosphere that feels almost soundless. Pull back enough to explain the broader geography or architectural envelope while keeping the square crop intentional and balanced.
Subject:
Build a coherent original scene whose emotional center comes from atmosphere, light physics, and spatial clarity. Avoid relying on celebrity likeness, franchise iconography, or stylized illustration shortcuts. A solitary human figure in a red parka should appear tiny against the glacier geometry, while crevasses, wind-scoured snow, and ice walls provide a strong sense of scale without cluttering the frame. The wider frame should reveal how the main light or weather phenomenon organizes the surrounding world and not just the immediate subject.
Important details:
Preserve a believable physical world with strong depth layering, careful color separation, and premium image discipline. Keep the aurora soft and physically plausible rather than graphic, with believable ice reflectance, low-angle moonless ambient light, and careful contrast so the red parka remains a scale marker rather than a fashion focal point. Make scale legible through spacing, horizon control, and foreground-to-background separation rather than through extra props.
Use case:
Aurora Glacier single study 8/9.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no tents, no readable expedition logos, no fantasy animals, no saturated nightclub colors, no fireworks effect.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。