Stable diffusion教程多，但感觉目前教程对于prompt 的讲解不直观，不方便收藏复制，很多好的作品也过分繁琐。因此，想化繁为简，整理那些精准的prompt。

说在前面：

演示用案例：女仆

演示用基础模型：

(资料图片)

Model：cetusMix_Version35.safetensors

LORA：bronyaZaychikSilverwingNEX_v09.safetensors

Prompt：

(masterpiece:1.3),(best quality:1.3), ((from side)),1girl, (full body:1.2), standing, beautiful detailed girl, perfect figure, shiny skin, (detailed hair:1.2), (extremely detailed skin), light on face,best shadow, living room,maid, pleated skirt, capelet, robe, gothic,<lora:bronyaZaychikSilverwingNEX_v09:0.8>Negative prompt: EasyNegative, duplicate, ((2girls)), bad-hands-5

Steps: 30, Sampler: DPM++ SDE Karras, CFG scale: 8, Seed: 457314261, Size: 512x512, Model hash: a611cf9c19, Model: cetusMix_Version35, Denoising strength: 0.7, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires upscaler: R-ESRGAN 4x+ Anime6B

from below：从下

from side：从侧面

from down：貌似是正面

from behind

from above