AI art "Small sisters" featuring #Azur Lane
Neues Kunstwerk erzeugen

Small sisters

#Azur Lane
Mehr zu sehen

Kommentare

Two girls from Azur Lane, Taiyuan and Changchun, sit together in shallow water by the seaside at sunset. Taiyuan has short black hair with a purple bow tie and a metal chain belt, wearing a decorated white military uniform with star medals and leg belts. Changchun has long white hair with a fur coat and black fur clothes, wearing pilot goggles around her neck and a white tiger tail. Both have detailed purple eyes, with sweat and wet skin glistening under the soft golden light. The cinematic composition features backlighting, depth of field, and vibrant colors, capturing their relaxed yet elegant poses as they look away and up at the sky.
Parameter anzeigen
Ursprünglicher Prompt
perfecthand,goodhand,goodanatomy, (beautiful eyes, detailed eyes:12),seaside,sit in watter,1girl, solo, wet_skin, sweat, sweatdrop,looking away, looking up, sunset, sit in watter, metal hair pin "102",,cinematic lighting , cinematic angle, soft light , color detail, backlighting, Depth of field, long exposure, light steps, intricate colors, vibrant colors, 2girls,2 girls with taiyuan and changchun,//TaiYuan(azur lane),black hair,purple eye,short hair,number hair pin,bow tie,metal chain,belt,star medal,white military uniform,decorated uniform,boot,leg belt,decorated stripes,//Changchun(azur lane),white hair,purple eyes,knee pad, long sleeve, full white pantyhouse,boots,fur coat,number hair pin,black fur clothes,arm band,binoculars hang around neck,pilot googles,fake cat ear,white tiger tail,
Qualitäts-Tag
high quality, year_2025, year_2024, ﹝Prompts﹞
Größe
768 x 1280
Sampling-Schritte
22
Sampling-Methode
Euler
CFG Scale
4.5
Negativ
bad quality, low quality, worst quality, lowres, nsfw, jpeg artifacts, scan artifacts, mosaic, cropped, error, fewer digits, bad reflection, bad composition, bad anatomy, bad hands, bad fingers, missing fingers, extra hands, extra legs, excess fingers, light particles, artist name, text, watermark, username, copyright name

Verwendetes Modell und LoRA