2x2 网格,16:9,AI 选取不同的主体:INPUT: $ SUBJECT, 图片 1
1/1

PROMPT RECORD · 公开作品详情

2x2 网格,16:9,AI 选取不同的主体:INPUT: $ SUBJECT

公开 AI 图像创作记录,包含原图、提示词、使用模型、作者、标签与来源。

使用模型
Nano Banana
创作者
@Gdgtify
发布时间
Sat Sep 26 21:37:00 +0000 2026
图片数量
1 张

完整提示词

源提示词按原始换行与标点完整保留
2x2 grid, 16:9, AI picks different subjects: INPUT: $ SUBJECT

// ── SEMANTIC INFERENCE ──────────────────────────────────────────
ARCHETYPE     = infer_subject_archetype($ SUBJECT)         // woman, elder, child, craftsman, musician
MOOD          = infer_mood_key($ SUBJECT)                  // serene, defiant, nostalgic, mischievous
ENVIRONMENT   = infer_setting($ SUBJECT)                   // vintage bathroom, garden shed, taxicab, library
ERA           = infer_era_aesthetic($ SUBJECT)             // 1920s, 1970s, contemporary, near-future
STYLING       = infer_wardrobe($ SUBJECT, ERA)             // fabric, cut, silhouette, color
HAIR          = infer_hair_style($ SUBJECT, ERA)
MAKEUP        = infer_makeup_signature($ SUBJECT, ERA)
JEWELRY       = infer_jewelry_set($ SUBJECT, count=2..3)   // nose ring, earring, chain, watch
BODY_ART      = infer_tattoo($ SUBJECT)                    // placement, style, iconography
POSE          = infer_pose($ SUBJECT)                      // profile, three-quarter, back-turned
EXPRESSION    = infer_expression($ SUBJECT)                // eyes closed, half-lidded, direct gaze
PROPS         = infer_counter_props($ SUBJECT, count=4..7) // bottles, towels, tools, books
WALL_ART      = infer_wall_decor($ SUBJECT, count=2..4)    // mirror, icons, prints, calendar
LIGHTING      = infer_light_recipe($ SUBJECT, ERA)         // overhead + window, lamps, neon, candle
COLOR_GRADE   = infer_color_grade($ SUBJECT, ERA)          // teal-warm, sepia, cold-noir, pastel
PALETTE       = infer_palette($ SUBJECT, ERA)

// ── COMPOSITION FIELD ──────────────────────────────────────────
P(x,y) = subject_gravity(ARCHETYPE) + environment_pull(ENVIRONMENT)
F = ∇P                // directional flow: gaze axis, body orientation
Curl = ∇×P            // organic curves: hair, ribbon, fabric folds, smoke
Div  = ∇·F            // density: texture of skin, cloth, wallpaper, clutter

// ── SUBJECT CONSTRUCTION ───────────────────────────────────────
SUBJECT_LAYER = photoreal_human(
  archetype  := ARCHETYPE,
  pose       := POSE,           // e.g., left profile, chin lifted, eyes closed
  expression := EXPRESSION,     // serene confidence, guarded, bemused
  skin       := 'pores visible, natural texture, subtle blush, no plastic smoothing'
)

STYLING_LAYER = {
  hair   := HAIR,               // shape, weight, fall, dampness or shine
  makeup := MAKEUP,             // lips, liner, contour, freckles
  jewelry:= JEWELRY,            // small silver nose ring, dangling gold earring, chain
  garment:= render_fabric(
      type  := STYLING.material,      // satin, linen, silk, wool
      cut   := STYLING.cut,           // cap sleeves, open back, high collar
      color := STYLING.color,         // deep burgundy, ochre, ivory
      detail:= STYLING.detail         // ribbon ties, seams, embroidery
  )
}

BODY_ART_LAYER = {
  placement := BODY_ART.location,    // left arm, shoulder, collarbone
  style     := BODY_ART.style,       // traditional, fine-line, blackwork
  motifs    := BODY_ART.motifs,      // roses, leaves, sparrows, anchors
  rendering := 'saturated ink, subtle sheen from skin oil'
}

// ── ENVIRONMENT ────────────────────────────────────────────────
WALLS = wallpaper_or_paint(
    color   := PALETTE.wall,
    texture := 'hand-applied, subtly uneven, faint scuffs'
)
WALL_DECOR = for each item in WALL_ART:
    render_frame(item.kind, item .art, item.position)
    // gold-frame mirror, icon panel, rosary, cross, sconce
COUNTER = surface_dressing(
    props   := PROPS,               // perfume bottles, tubes, towel, lipstick
    clutter := 'light, arranged, not chaotic — each prop readable'
)

// ── LIGHTING RECIPE ────────────────────────────────────────────
KEY        : warm overhead lamp, diffused, ~2900K, casting soft downward light
FILL       : soft window glow from opposite side, ~5200K daylight diffused
RIM        : faint atmospheric rim from window on hair and shoulder edges
SHADOWS    : soft, present, not crushed
SPECULAR   : subtle on skin, satin highlight on fabric, glass glint on bottles

// ── COLOR GRADE ────────────────────────────────────────────────
Grade   : cool shadows, warm midtones, slightly desaturated reds
Look    : dreamy retro, teal-shifted shadows, filmic rolloff
Palette : wall = pale mint; garment = deep burgundy; skin = warm ivory;
          tattoo = saturated floral; props = amber glass, cream, blush

// ── CAMERA ─────────────────────────────────────────────────────
Lens       : 50mm prime, f/1.8–f/2.0
Distance   : medium-close portrait, upper body in frame
Focus      : tack-sharp on nearest eye / cheekbone / tattoo highlight
Bokeh      : creamy, subtle, golden-amber highlights from props and sconce
Resolution : 8K
Framing    : rule of thirds, subject slightly off-center, gaze space preserved

// ── FINISH ─────────────────────────────────────────────────────
No plastic smoothing, no heavy retouch, no drawn edges.
Skin retains pores, fine hairs, lip texture, subtle creases.
Fabric retains weave, sheen, and weight.
Background props rendered in same photoreal fidelity, softened by DOF.
No text, no watermark, no logos.

Output: one 8K photoreal cinematic portrait, fully inferred from $ SUBJECT.

提示词译文

译文与源提示词分开显示,不替代原文
2x2 网格,16:9,AI 选取不同的主体:INPUT: $ SUBJECT

// ── 语义推理 ──────────────────────────────────────────
ARCHETYPE     = infer_subject_archetype($ SUBJECT)         // 女性、长者、儿童、工匠、音乐人
MOOD          = infer_mood_key($ SUBJECT)                  // 宁静、桀骜、怀旧、狡黠
ENVIRONMENT   = infer_setting($ SUBJECT)                   // 复古浴室、花园棚屋、出租车、图书馆
ERA           = infer_era_aesthetic($ SUBJECT)             // 1920年代、1970年代、当代、近未来
STYLING       = infer_wardrobe($ SUBJECT, ERA)             // 面料、剪裁、廓形、颜色
HAIR          = infer_hair_style($ SUBJECT, ERA)
MAKEUP        = infer_makeup_signature($ SUBJECT, ERA)
JEWELRY       = infer_jewelry_set($ SUBJECT, count=2..3)   // 鼻环、耳环、项链、手表
BODY_ART      = infer_tattoo($ SUBJECT)                    // 位置、风格、图像学
POSE          = infer_pose($ SUBJECT)                      // 侧面、四分之三、背对
EXPRESSION    = infer_expression($ SUBJECT)                // 闭眼、半眯、直视
PROPS         = infer_counter_props($ SUBJECT, count=4..7) // 瓶子、毛巾、工具、书本
WALL_ART      = infer_wall_decor($ SUBJECT, count=2..4)    // 镜子、圣像、画作、日历
LIGHTING      = infer_light_recipe($ SUBJECT, ERA)         // 顶光+窗光、灯具、霓虹、烛光
COLOR_GRADE   = infer_color_grade($ SUBJECT, ERA)          // 蓝绿暖调、棕褐、冷硬黑色、柔和粉彩
PALETTE       = infer_palette($ SUBJECT, ERA)

// ── 构图场 ──────────────────────────────────────────
P(x,y) = subject_gravity(ARCHETYPE) + environment_pull(ENVIRONMENT)
F = ∇P                // 方向流:视线轴、身体朝向
Curl = ∇×P            // 有机曲线:头发、丝带、织物褶皱、烟雾
Div  = ∇·F            // 密度:皮肤、织物、墙纸、杂物的质感

// ── 主体构建 ───────────────────────────────────────
SUBJECT_LAYER = photoreal_human(
  archetype  := ARCHETYPE,
  pose       := POSE,           // 例如:左侧面、下巴微抬、闭眼
  expression := EXPRESSION,     // 宁静自信、戒备、玩味
  skin       := '毛孔可见、自然纹理、微妙红晕、无塑料感磨皮'
)

STYLING_LAYER = {
  hair   := HAIR,               // 形状、重量、垂坠、湿润或光泽
  makeup := MAKEUP,             // 唇、眼线、修容、雀斑
  jewelry:= JEWELRY,            // 小银鼻环、垂坠金耳环、项链
  garment:= render_fabric(
      type  := STYLING.material,      // 缎、亚麻、丝、羊毛
      cut   := STYLING.cut,           // 短袖露肩、露背、高领
      color := STYLING.color,         // 深酒红、赭石、象牙白
      detail:= STYLING.detail         // 丝带系带、缝线、刺绣
  )
}

BODY_ART_LAYER = {
  placement := BODY_ART.location,    // 左臂、肩膀、锁骨
  style     := BODY_ART.style,       // 传统、细线、黑灰写实
  motifs    := BODY_ART.motifs,      // 玫瑰、树叶、麻雀、锚
  rendering := '饱和墨色、皮肤油脂带来的微妙光泽'
}

// ── 环境 ────────────────────────────────────────────────
WALLS = wallpaper_or_paint(
    color   := PALETTE.wall,
    texture := '手工涂刷、微微不均、轻微磨损'
)
WALL_DECOR = for each item in WALL_ART:
    render_frame(item.kind, item .art, item.position)
    // 金框镜子、圣像板、念珠、十字、壁灯
COUNTER = surface_dressing(
    props   := PROPS,               // 香水瓶、软管、毛巾、口红
    clutter := '轻盈、有序、不混乱——每个道具都清晰可辨'
)

// ── 布光方案 ────────────────────────────────────────────
KEY        : 温暖的顶灯,漫射,约2900K,投下柔和的向下光
FILL       : 来自对侧的柔和窗光,约5200K 日光漫射
RIM        : 来自窗户的微弱大气边缘光,勾出头发和肩膀轮廓
SHADOWS    : 柔和、存在、不死黑
SPECULAR   : 皮肤上的微妙高光、面料的缎面光泽、瓶子的玻璃反光

// ── 色彩分级 ────────────────────────────────────────────────
Grade   : 冷调阴影、暖调中间调、红色略去饱和
Look    : 梦幻复古、青蓝偏移阴影、电影感曲线下沿
Palette : 墙面 = 淡薄荷;服装 = 深酒红;肤色 = 暖象牙;
          纹身 = 饱和花卉;道具 = 琥珀玻璃、奶油色、腮红粉

// ── 相机 ─────────────────────────────────────────────────────
Lens       : 50mm 定焦,f/1.8–f/2.0
Distance   : 中近景肖像,框入上半身
Focus      : 对最近的眼睛 / 颧骨 / 纹身高光锐利清晰
Bokeh      : 奶油般柔润、微妙,由道具和壁灯产生金琥珀色光斑
Resolution : 8K
Framing    : 三分法,主体略偏中心,保留视线空间

// ── 后期 ─────────────────────────────────────────────────────
无塑料感磨皮,无重度修图,无描边。
皮肤保留毛孔、细毛、唇纹、微妙褶皱。
面料保留织纹、光泽与垂感。
背景道具以同样的写实保真度渲染,由景深柔化。
无文字、无水印、无标志。

输出:一张由 $ SUBJECT 完整推理而成的 8K 写实电影感肖像。
50mm 标准镜头浅景深暖色调中景多分屏轮廓光柔光多面板布局人物肖像电影感摄影文生图