“Create a hyper-realistic cinematic portrait of the same person from my uploaded photo (keep face identity, skin tone, hairstyle as in my photo). He is a young Indian man, fit and muscular, bare-chested, wearing a white cotton dhoti with a golden border and rudraksha mala beads on neck and wrists, with a red tilak on forehead.
Scene: inside an ancient stone Hindu temple (dark carved pillars, symmetrical aisle), shallow depth of field, smoky incense haze. Dramatic volumetric god rays coming from above, warm golden light, high contrast, moody cinematic color grade.
Action: he is performing Shiva Lingam abhishekam, holding a brass lota in his right hand and pouring a continuous stream of white milk onto a black Shiva lingam in the foreground. Milk is flowing down the lingam, wet reflective stone, small water splashes in the circular yoni base, subtle glowing sacred tripundra marking on the lingam.
Camera: medium close-up, slightly low angle, 85mm lens look, f/1.8 bokeh, ultra-detailed skin texture, realistic hands, sharp focus on face + milk stream + lingam, 8k, HDR, cinematic.”
Negative prompt (paste in “negative”)
“cartoon, anime, CGI, illustration, lowres, blurry, over-smoothed skin, bad anatomy, extra fingers, deformed hands, duplicated face, text, watermark, logo, signature, oversaturated, harsh neon, plastic skin”
Quick settings (if your tool asks)
Aspect ratio: 2:3 (vertical like your sample)
Style: photoreal / cinematic
Strength (if using img2img): 0.35–0.55 to preserve your face
Tell me which tool you’re using (Midjourney / SDXL / Leonardo / Bing / etc.) and I’ll rewrite the prompt in that exact format (with the best “image weight / denoise / CFG” values).
