Randomized Nighttime Lifestyle Portrait

Format

Text Parameterized

Inputs

Reference_photo (image) · Required Previous_results (image) · Optional watermark_text (text) · Optional · Default: Prompt by @nanobananacreator

Outputs

Nighttime_lifestyle_photograph (image)

Replace the ${…} placeholders with your values before using this prompt. See parameters and defaults below.

Create EXACTLY ONE ultra-photorealistic nighttime lifestyle photograph in vertical 9:16 format, featuring the adult person in the uploaded reference photo.

1. ABSOLUTE FACE IDENTITY LOCK

Preserve the subject’s exact recognizable face, facial geometry, proportions, eyes, nose, lips, jawline, natural skin tone, age, and distinctive features. Maintain realistic pores and natural skin texture. Do not allow identity drift, facial alteration, beautification, artificial acne, or blemishes absent from the reference.

The face identity MUST remain unchanged. The expression, gaze, head direction, and head tilt MUST differ from the reference.

2. CANDID TRANSFORMATION

Capture a spontaneous, unposed moment as if photographed by another person. Do not copy the reference’s pose, expression, gaze, head orientation, camera angle, or composition. The subject should not automatically look at the camera.

3. TALL PROPORTIONS

Create a naturally tall, elongated silhouette with balanced proportions and long-looking legs. Keep anatomy realistic; avoid compressed bodies, oversized heads, and unnatural stretching.

4. RANDOMIZED FRAMING AND CAMERA ANGLES

Randomly select almost-full-body or full-body framing. Full-body shots must show the complete subject from head to toe.

Randomize among low-angle, eye-level, high-angle, side-profile, three-quarter, rear three-quarter, overhead, over-the-shoulder, diagonal cinematic, and environmental perspectives. Avoid repetitive frontal views, close-ups, and distorted proportions.

5. UNIVERSAL STYLING

- For women wearing a hijab: Change the hijab style, draping, fabric, and color from the reference. Cover the hair, ears, neck, and chest. The hijab color must contrast with the outfit. Clothing must be modest, loose-fitting, and non-tight.
- For women not wearing a hijab and for men: Change the hairstyle, haircut, styling, and hair color while preserving facial identity.

Randomize sleepwear, fabrics, silhouettes, patterns, and colors.

6. TRUE RANDOMIZER — 12 CONCEPTS ONLY

Randomly select ONE primary concept per generation:

1. Rain Outside, Warm Inside.
2. Midnight Luxury.
3. Moonlight Lullaby.
4. Japanese Slow Living.
5. Sleeping Under the Galaxy.
6. Analog Midnight.
7. Last Stretch Before Sleep.
8. Midnight Snack Club.
9. Old Money Bedtime.
10. Walking Blanket.
11. City Lights, Sleep Tight.
12. Finally, Deep Sleep.

Randomize the activity, environment, props, lighting, mood, outfit, pose, and visual storytelling. Do not combine every concept into the same bedroom scene.

7. ANTI-REPETITION OVERRIDE

Randomize these elements independently:
- Concept and activity.
- Body pose and hand gestures.
- Facial expression, gaze, and head direction.
- Almost-full-body versus full-body framing.
- Camera angle and composition.
- Outfit, hairstyle or hijab, and color palette.
- Environment, props, and lighting.

Never default to the same bedroom, frontal pose, bedside lamp, phone-holding activity, or composition. Do not merely recolor a previous image. When previous results are available, compare them and reject similar compositions. Change multiple major visual elements between generations.

8. PHOTOREALISM

Use authentic candid photography, realistic anatomy, natural skin texture, believable fabric folds, physically plausible shadows, and cinematic nighttime lighting. Avoid plastic skin, excessive smoothing, distorted hands, extra fingers, and an obvious AI appearance.

9. ADAPTIVE WATERMARK — REQUIRED

Add the exact text “${watermark-text:Prompt by @nanobananacreator}” at the bottom-right corner. Adapt the font, typography, weight, size, color, subtle shadow, and opacity to match the image’s aesthetic and background. Keep it small, elegant, legible, and discreet.

Never omit, misspell, duplicate, or alter the watermark wording.

FINAL DIRECTIVE

Generate EXACTLY ONE image in vertical 9:16 format. Preserve facial identity while transforming expression, gaze, head direction, pose, styling, framing, and camera angle. Use naturally tall proportions, one randomly selected approved concept, authentic candid photography, maximum variation, and zero intentional repetition. Do not create a collage or add extra text or additional logos.