
原文
GPT Image 2 · en
Scene:
A square photorealistic studio photograph of a Chinese calligraphy practice copybook opened flat on a bamboo desk. The page is printed on thin ivory practice paper with faint red grid lines, gray model characters, black student brush strokes, ink feathering, and a small ceramic brush rest. The camera is top-down with gentle daylight and crisp page texture.
Subject:
Show a fictional calligraphy copybook page. The exact readable text should include: 书法临摹, 永, 和, 山, 水, 月, 心, SPECIMEN. Arrange the characters in repeated practice boxes: pale gray model characters in the first column, darker traced attempts beside them, and one red rectangular specimen stamp in a corner. No school name, real teacher name, grading score, student identity, or publication imprint.
Important details:
The test is text rendering plus paper physics. Grid lines must align consistently, characters must sit inside boxes, brush strokes should show pressure, dry-brush texture, ink pooling, and slightly uneven human practice. The printed model characters should be cleaner than the practice strokes. Keep the page believable as a cultural stationery product and not as a certificate. The specified characters should remain readable at thumbnail scale.
Use case:
Pseudo-Real Documents single study 4/8, testing copybook grid precision, Chinese character readability, brush texture, and thin practice paper realism.
Constraints:
No readable text except the specified copybook strings, no watermark, no logos, preserve exact 1:1 aspect ratio, no real school name, no personal name, no grading marks, no publisher imprint, no random extra characters, no unreadable ink blot covering the models.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。





