Want to learn how to create images like this one?
Check out our crash course in prompt engineering & AI art generation!
superten0910
posted 3 months ago
17 views
0 comments
### Positive Prompt（正向提示）
a gothic female silhouette in monochrome,
black-and-white,
grayscale,
wearing an avant-garde opaque black short-sleeve top with sculptural asymmetric seams,
high neckline,
fully covered chest,
no cleavage,
subtle white glow at the sternum,
wide-brim felt hat partly shading the eyes,
high-contrast rim light outlining shoulders and neck,
soft volumetric fog,
slow-shutter motion-blur trails,
cinematic portrait,
ultra-noir,
photorealistic,
4K,
35 mm,
f/1.
4,
masterpiece,
highly detailed
### Negative Prompt（反向提示）
nipples,
nipple,
bare chest,
bare torso,
exposed breast,
underboob,
sideboob,
cleavage,
deep cleavage,
see-through,
sheer,
transparent fabric,
mesh,
lingerie,
corset,
swimsuit,
color,
chromatic,
vivid,
saturated,
red,
blue,
green,
yellow,
pink,
bulky coat,
overcoat,
trench coat,
cape,
hoodie,
(lowres:1.2)
,
(bad anatomy:1.1)
,
(extra limbs)
,
(multiple faces)
,
(deformed)
,
(blurry)
,
(text)
,
(watermark)
,
(logo)
,
(nsfw)
,
(cartoon)
,
(jpeg artifacts)
,
(poorly drawn hands)
,
(bad proportions)
### Sampling & Model Settings
• Model : Stable Diffusion XL 1.
0
• Resolution : 1080 × 1920（9∶16 直式）
• Sampler : DPM++ 2M Karras
• Steps : 30
• CFG Scale : 7.
5
• Seed : -1（先隨機，挑到喜歡的 Seed 再鎖定）
• Clip Skip : 2
• VAE : 官方 SDXL VAE
• Hires Fix : Off
### 可選強化
• LoRA : `gothic_fashion_v1` 權重 0.
6（增加服裝藝術感）
• ControlNet : OpenPose / Canny（若想鎖定姿勢或剪影）
Generation parameters
Model used
Prompt category
More by superten0910
