cathyno1nu

發布於 Jul 2, 2026

cathyno1nu

1girl, masterpiece, best quality:1.2, highres, ultra-detailed, cinematic lighting, dramatic angle, action scene, retro style, cat's eye, hitomi kisugi, blue bodysuit, yellow sash, lying, cinematic shot, eye-level, alley, streetlight, kicking, motion blur, wet pavement, shadow, wall, wet asphalt, amber light, blue neon, crime thriller, arri alexa, volumetric fog, rim light, deep shadows, film grain, 35mm lens, shallow dof, photorealistic

顯示參數

原始提示詞: , (masterpiece, best quality:1.2), highres, ultra-detailed, cinematic lighting, dramatic angle, action scene,，retro style（cat's eye），uterine prolapse，, guro，，hitomi kisugi, , blue bodysuit,yellow_sash, ,lying，，，Cinematic shot, , Cinematic shot, eye-level from outside the alley, A woman's lower body and legs are still visible in the weak streetlight, just outside the corner — her upper body has already been dragged around the corner，upper body is out of frame ，upper body hidden behind the wall edge. legs are kicking wildly, feet scrambling against the wet pavement, motion blur on the limbs. The corner wall cuts the frame diagonally, shadow swallowing where her torso should be. Wet asphalt reflecting amber streetlight and cold blue neon from the main road. Gritty crime thriller mood, Arri Alexa texture, volumetric fog, rim light on her calves, deep shadows, film grain, 35mm lens, shallow DoF. masterpiece, best quality, photorealistic.
尺寸: 1280 x 768
取樣步數: 24
取樣方式: Euler a
CFG Scale: 6.2
負面: nsfw, worst quality, bad quality, low quality, lowres, anatomical nonsense, artistic error, bad anatomy, interlocked fingers, extra fingers, text, artist name, signature, bad feet, extra toes, ugly, poorly drawn, censor, blurry, watermark, simple background, transparent background, old, oldest, glitch, deformed, mutated, disfigured, long body, bad hands, missing fingers, extra digit, fewer digits, cropped, very displeasing, sketch, jpeg artifacts, username, censored, bar_censor, mosaic_censor, conjoined, bad ai-generated, nsfw, long neck, skin blemishes, skin spots, acne, the wrong limb, error, black line, excess hands，manga style，bad anatomy, bad hands, missing limbs, extra limbs, floating limbs, detached body parts, severed, dismembered, broken bones protruding, exposed muscle tissue, organs visible，legs out of frame，foot out of frame，这是一组经过优化的提示词，专门针对“倒挂”、“绳索束缚”和“布料物理（披风）”进行了强化。正面提示词 (Positive Prompt) (masterpiece, best quality:1.2), ultra-detailed, cinematic lighting, dramatic angle, Cassandra Cain, Batgirl (DC Comics), Asian female, young adult, wearing torn black and yellow tactical suit, blood stains on face, disheveled hair, defeated, unconscious, eyes closed, (hanging upside down:1.4), (rope tied tightly around one ankle:1.3), suspended in mid-air, body swinging gently, motion blur on extremities, black cape flowing downward, heavy fabric physics, gravity affected cloth, dynamic folds, weighty drape, dark atmosphere, abandoned warehouse background, dim lighting, dust particles, low angle shot, worm's eye view 负面提示词 (Negative Prompt) (worst quality, low quality:1.4), bad anatomy, bad hands, missing fingers, extra digits, fewer digits, cropped, jpeg artifacts, signature, watermark, blurry, text, error, mutated, deformed, ugly, floating, levitating, standing, cape clipping through body, stiff fabric, flat cloth, plastic skin, clean costume, smiling, happy 关键修改解析（为什么这么写） 1. 关于“倒挂”与“摇晃”： * "(hanging upside down:1.4)"：这是核心动作，高权重确保构图正确。 * "body swinging gently" + "motion blur on extremities"：AI 很难凭空想象出“摇晃”的感觉，加入运动模糊能让静止的图片产生“正在晃动”的视觉残留感。 * "worm's eye view"（虫视角）：倒挂题材的最佳视角，能极大增强压迫感和空间感。 2. 关于“披风动态”（这是最难的部分）： * "heavy fabric physics, gravity affected cloth"：这是告诉 AI 披风是有重量的，不是纸片。 * "weighty drape"：防止披风像超人那样飘在身后，而是因为重力自然下垂包裹身体。 * 避坑：负面词里加了 "cape clipping through body"（披风穿模）和 "stiff fabric"（硬布），这是 AI 画倒挂最常见的两个错误。 3. 关于“绳索”： * "(rope tied tightly around one ankle:1.3)"：精确到脚踝，避免 AI 把绳子画在膝盖或大腿这种奇怪的位置。进阶参数建议（PixAI 设置） * 采样器 (Sampler)：推荐使用 DPM++ 2M Karras 或 Euler a。Euler a 对动态模糊的表现力更强。 * CFG Scale：设置在 7 - 9 之间。太低会乱画绳子，太高会让姿势僵硬。 * ControlNet（如果有）：强烈建议使用 OpenPose 或 Depth 控制，手动摆一个倒挂的骨架，否则 AI 经常会把头画得太大或身体比例失调。风格微调选项 * 如果想要更绝望/黑暗：在正面词加 "single spotlight, chiaroscuro lighting, deep shadows"。 * 如果想要更像电影截图：在正面词加 "35mm photograph, film grain, depth of field, bokeh"。需要我帮你把敌人（比如一只抓着绳子的手或者阴影）也加入到画面描述中，增加构图的层次感吗？