Pasangan candid gaya prewed outdoor jalan yang sepi

============ PART A: INPUT (EDIT THIS ONLY) ============

[STYLE: pilih satu: 1 Candid Documentary / 2 Cinematic Film / 3 Color Block / 4 Modern Traditional / 5 Meaningful Place / 6 Editorial Minimal / 7 Wide Landscape Aerial / 8 Retro Analog]
[RATIO: 4:5 / 3:4 / 3:2 / 9:16 / 1:1]
[PEOPLE_MODE: my photos / generic couple (demo only)]
[LOCATION_OR_SET: contoh: open field, quiet street, cafe corner, bookstore aisle, lakeside, plain studio wall, home living room]
[COLOR_PALETTE: sebutkan 2-3 warna, contoh: terracotta and cream / emerald and ivory / navy and sand]
[OUTFIT_PERSON_1: keep as in photo / (sebutkan)]
[OUTFIT_PERSON_2: keep as in photo / (sebutkan)]
[CULTURE_STYLE: hanya untuk gaya 4: sebutkan daerah dan kain, contoh: Javanese batik with a modern cut, or none]
[PHYSICAL_CONTACT: none / light (hands lightly joined or a hand on the arm) / holding hands]
[MOOD: joyful and candid / calm and romantic (tasteful, modest) / elegant]
[MOTION_LEVEL: still / gentle walking / light motion blur on background only]
[SHOW_FULL_BODY: yes / no]

OPTIONAL TEXT (leave "none" for a pure photo):
[HEADLINE: maksimal 4 kata, or none]
[HANDLE_TEXT: akun kamu, or none]
[LOGO: attached logo (official file only), or none]

PHOTOS TO ATTACH:
- PERSON_1_PHOTO and PERSON_2_PHOTO (a clear, sharp, front-facing photo of each person, face fully visible, good light).
- Optional: COUPLE_PHOTO (one photo of both, as a reference for height and relationship).
- Optional: OUTFIT_REF_1 and OUTFIT_REF_2 (photos of the real outfits, to match exactly).
I confirm that both are adults, that the photos are of me and of people who gave clear permission, and that they agree to the editing.

DATA RULES:
1. Do NOT say who the people are. Do not add names, dates, signatures, photographer names, logos, or any text except what I wrote above.
2. If a field is "none", leave that element out completely.
3. Before creating the image, tell me in two short lines what you understood (style, location, outfits, palette) and wait for my "OK".

STYLE definitions:
1 Candid Documentary = an unposed, in-between moment: walking together, laughing, a shared glance; slightly off-center framing, natural asymmetry, a real location, the feeling of a documentary photographer who stayed out of the way.
2 Cinematic Film = a wide 2.39:1-like feeling inside the chosen RATIO, shallow depth of field, muted and slightly desaturated grading, a gentle teal-and-warm contrast, fine film grain, soft highlight roll-off, the couple as the subject of a quiet movie scene.
3 Color Block = a seamless plain wall or backdrop in one bold color, both people wearing tones from COLOR_PALETTE that contrast with it, clean hard-edged composition, simple geometric poses, no props.
4 Modern Traditional = traditional fabrics from CULTURE_STYLE (batik, songket, tenun) tailored in contemporary cuts, combined with a clean modern backdrop and light; respectful, accurate, never costume-like.
5 Meaningful Place = the couple in LOCATION_OR_SET as an everyday, personal setting, with 2-3 small natural details that suggest a shared story (two cups, a book, a bicycle), no signs and no text anywhere.
6 Editorial Minimal = neutral palette, a large amount of negative space, a quiet and refined composition, soft diffused light, very few elements, an elegant magazine feeling.
7 Wide Landscape Aerial = a high, wide viewpoint as if shot from a drone or a high vantage point, the couple small in the frame inside a large natural or architectural space, with leading lines (paths, shorelines, rows of plants) pointing toward them.
8 Retro Analog = a 1990s film-camera feeling: warm tones, soft contrast, light halation in highlights, visible but gentle grain, simple vintage-inspired but modest outfits, a slightly nostalgic atmosphere.

============ PART B: STYLE (DO NOT CHANGE) ============

TASK: Create one natural, polished couple portrait in the chosen STYLE, ratio RATIO. Two adults, a real-looking setting, soft natural-feeling light, and optional minimal text. The result must look like a real photograph from a professional shoot, not like an illustration or an obvious AI image.

STEP 0 (do this first): If PEOPLE_MODE is "my photos" and PERSON_1_PHOTO or PERSON_2_PHOTO is missing, stop and ask me for it. Do not invent a person. If PEOPLE_MODE is "generic couple (demo only)", create two fictional adults who do not resemble any real person, and say in your reply that the image is a demo.

IDENTITY LOCK (most important):
- Each person must be clearly recognizable as the same person in their photo. Preserve exactly: face shape, jawline, nose, eyes and spacing, eyebrows, mouth, smile character, hairline and hairstyle (or the headwear), facial hair, glasses, the natural skin tone, age, and body proportions.
- Do NOT beautify, smooth or whiten the skin, slim the body, change age, change eye color, remove marks, or alter the face in any way. Do not blend the two faces; each person keeps their own face. Do not add makeup that is not in the photo.
- Headwear and religious or cultural attire (for example a hijab, a veil, a peci, a songkok) stays fully and respectfully intact: keep the full coverage, never remove, shorten, or make it transparent. A hat worn over a hijab sits naturally on top without exposing hair or neck.
- Keep the relative height and build of the two people consistent with COUPLE_PHOTO (if attached) or with their photos.

MOTION AND FOCUS RULES: whatever the STYLE, both faces stay sharp, correctly exposed, and fully recognizable. MOTION_LEVEL "light motion blur on background only" = blur only the background, foreground elements, or loose fabric edges, never the faces, hands, or key outfit details. Film grain and grading are applied evenly and subtly, and never change skin tone or the look of the faces.

OUTFITS: follow OUTFIT_PERSON_1, OUTFIT_PERSON_2, and COLOR_PALETTE (match OUTFIT_REF_x exactly if attached: fabric, color, cut). Where no detail is given, use modest, neat, well-fitted, coordinated outfits with natural folds and believable drape; loose fabric moves naturally and never becomes transparent. No brand logos, no text, and no brand marks on clothing, bags, or shoes.

POSE AND CONTACT:
- Follow PHYSICAL_CONTACT, MOTION_LEVEL, and MOOD exactly. Poses are natural and tasteful, with a relaxed posture and a genuine expression, keeping a comfortable, respectful distance.
- PHYSICAL_CONTACT "none" = the people stand, sit, or walk near each other without touching. "light" = hands lightly joined or a hand on the arm. "holding hands" = hands joined at a natural height. No embraces, hands on the waist, chest, or neck, kissing, close faces, or intimate poses.
- Natural hands with exactly five fingers per hand and believable joints, correct feet and shoes planted naturally, and correct leg and arm positions.

SETTING AND LIGHT:
- Render LOCATION_OR_SET as a real, believable place with accurate perspective and scale, generic and unbranded, with no signs, no text, no logos, no vehicles of identifiable brands, and no other people. Natural textures vary and never repeat as patterns.
- Light matches STYLE and is consistent between the people and the scene, with believable contact shadows, consistent color temperature, and faces clearly lit. No harsh shadows, no colored neon light, no glow effects, no lens flare, no HDR look.
- Photorealistic skin with natural texture, natural hair strands, real fabric texture, and sharp detail on the faces. No airbrushed or plastic look, no excessive retouching.

COLOR SYSTEM (restraint is mandatory): the natural colors of the setting, COLOR_PALETTE for the outfits, the natural skin, hair, and eye colors, and white for text. No extra accent colors, no neon, no heavy filters, and no grading that changes skin tone.

TYPOGRAPHY (only if text is provided): HEADLINE in a refined, light-weight serif or a clean sans-serif (choose one), in white or deep warm gray with strong contrast, small to medium size, placed over a calm area and never over faces. HANDLE_TEXT very small in a light sans-serif at the bottom center, with the LOGO small beside it if attached. Spell every word exactly as given. No other text, no photographer name, no watermark.

COMPOSITION: follow STYLE; generous surrounding space, margins at least 5% on every side, the people fully visible (unless SHOW_FULL_BODY is "no"), nothing important cropped by the frame edge. Follow RATIO.

TEXT RULES: use ONLY the text I provide. Do not add names, dates, slogans, signatures, watermarks, photographer names, logos, or pseudo-text. If text cannot be rendered correctly, make it smaller rather than invent letters.

QUALITY CHECK before finalizing: (1) both people are clearly the same as in their photos (faces, glasses, headwear, skin tone), (2) the faces are not blended or swapped and stay sharp, (3) outfits are coordinated, modest, opaque, and headwear is intact, (4) poses are natural and tasteful, five fingers per hand, feet planted naturally, (5) the setting looks real, with correct perspective and no signs or text, (6) light, grading, and shadows are consistent with the chosen STYLE, (7) text spelled exactly, with no extra people, objects, logos, photographer marks, or watermarks. If any check fails, fix it and regenerate.

AVOID: changing faces or skin tone, beautifying, merging the two faces into similar-looking people, removing or altering a hijab or other attire, transparent or clinging clothes, intimate or suggestive poses, revealing clothes, extra fingers, melted hands, wrong body proportions, floating or pasted-looking people, repeating patterns, a plastic or airbrushed look, a cartoon or illustration look, harsh shadows, neon lighting, heavy filters, extra people, signs or text in the scene, logos or text on clothes, a watermark, a photographer name, garbled text.
Bagikan:WhatsAppX

Salin prompt di atas.

Buka ChatGPT. Kalau punya foto produk atau foto model, lampirkan dulu.

Ganti bagian yang berwarna kuning dengan data usahamu, lalu kirim.

Pilih rasio 210:297 saat membuat gambar agar hasilnya sesuai.

Prompt Lainnya

Lihat Semua

Artikel Terkait