gpt-imageIMAGE

Backlit silhouette prompt uses partial reveal to draw viewers in

A minimal backlit silhouette formula relies on foreground translucency and ambiguous poses so viewers fill in the image themselves, with any subject producing a new story.

DeepBlue深藍

@DeepBlueX0Archived 2026.09.22Originally posted 2026.09.21

Backlit silhouette prompt uses partial reveal to draw viewers in (image 1 of 4)

Backlit silhouette prompt uses partial reveal to draw viewers in (image 2 of 4)

Backlit silhouette prompt uses partial reveal to draw viewers in (image 3 of 4)

Backlit silhouette prompt uses partial reveal to draw viewers in (image 4 of 4)

DeepBlue深藍 @DeepBlueX0Originally posted 2026.09.21

Open original on X

为什么“若隐若现”总比直接展示更有吸引力？因为看不完整，反而会让人主动补全画面。最近在玩一套视觉玩法通用🔵 GPT image 2.5 Prompt：极简逆光剪影 × 暧昧克制 × 前景透映 × 【主体】逆光留下轮廓，前景制造朦胧，暧昧藏在距离和姿态里。最好玩的地方，是主体可以天马行空地自由组合。词很短，玩法却很多。神话人物可以，古代人物可以，影视角色可以，现实中的职业组合也可以，甚至两个完全不搭的主体放在一起也行。同一套视觉规则，换一个主体，就是一个新故事。以下4图主体看图应该就能猜出来吧？答案放评论区～

This post quotes another post: x.com/deepbluex0/status/2101610754424361206

x.com/deepbluex0/status/2101856621471396270Archived 2026.09.22 10:11