聊天GPT

女性肖像 - 電影 - 移除任何重新定位

為聊天GPT 編寫。不同型號的結果和參考照片支援可能有所不同。

使用提示詞前

列出的工具:聊天GPT。完整說明:8,060 個字元。 1 個提供的範例影像。

This prompt mentions reference images. Read its instructions for the required number of photos and identity constraints. We have not independently tested this prompt. Tool labels and supplied images are not a guarantee of compatibility or output quality.

完整提示詞

Create an ultra-realistic, high-end cinematic dark-fantasy character scene using the provided reference image as the ABSOLUTE STRICT VISUAL, COMPOSITION, POSE, CHARACTER, AND STYLE REFERENCE.

REFERENCE IMAGE PRESERVATION — EXTREMELY IMPORTANT:
Recreate the reference image as closely and accurately as technically possible.

The reference image is the ONLY source for:

exact overall composition

exact number of characters

exact character placement and positions

character scale and proportions

body structure and physique

body orientation

exact poses and gestures

arm and hand positions

head angles

camera angle

perspective

lens perspective

framing and crop

foreground, middle-ground and background arrangement

clothing and armor design

costume construction

fabric and material details

hairstyle styling and silhouette

accessories

environment

background

atmospheric elements

lighting direction

light intensity

shadows

highlights

color palette

contrast

depth

cinematic mood

visual style

overall realism


DO NOT redesign, rearrange, reinterpret, simplify, add, remove, or reposition any character or major visual element.

CHARACTER IDENTITY REPLACEMENT RULE — ABSOLUTE:
The uploaded male face reference is the ONLY source of facial identity for EVERY MALE CHARACTER visible in the reference image.

If there are multiple male characters, apply the SAME uploaded male facial identity to each male character while naturally adapting the face to each character's head angle, perspective, scale, lighting, and physical position.

The uploaded female face reference is the ONLY source of facial identity for EVERY FEMALE CHARACTER visible in the reference image.

If multiple female characters are present, apply the SAME uploaded female facial identity to each female character while naturally adapting the face to each character's head angle, perspective, scale, lighting, and physical position.

NEVER invent a different face.
NEVER use a random face.
NEVER mix facial identities.
NEVER combine male and female facial features.
NEVER transfer identity from the reference image itself.
The reference image provides the CHARACTER'S EXPRESSION and visual design, NOT the facial identity.

FACIAL IDENTITY:
Preserve the uploaded face reference's recognizable identity as accurately as possible:

face shape

forehead

eyebrows

eyes

eye spacing

nose structure

nostrils

cheek structure

lips

jawline

chin

facial proportions

natural skin characteristics

beard/moustache pattern where applicable

distinctive facial features


Do not beautify, redesign, slim, widen, age, de-age, masculinize, feminize, or otherwise alter the identity.

EXPRESSION & EMOTION — CRITICAL:
The FACIAL EXPRESSION of every character must come from the corresponding character in the reference image.

Do NOT copy the reference character's identity.

Instead:
IDENTITY = uploaded face reference
EXPRESSION = reference image
HEAD ANGLE = reference image
GAZE DIRECTION = reference image
POSE = reference image
BODY LANGUAGE = reference image

Preserve the exact emotional feeling visible in the reference:
seriousness, intensity, confidence, menace, calmness, aggression, mystery, determination, or any other emotion shown by that specific character.

The uploaded face must naturally reproduce that expression without losing its recognizable identity.

CHARACTER-SPECIFIC PRESERVATION:
Every character must remain visually faithful to their original role in the reference image.

Do not make all characters look like the same person beyond their facial identity.
Keep their individual:

body proportions

armor/clothing

pose

expression

head position

scale

silhouette

visual presence


EXACT SCENE COMPOSITION:
Maintain the same foreground central character placement and the same supporting characters positioned exactly where they appear in the reference.

Preserve the dramatic low-angle cinematic perspective and the imposing scale relationship between foreground and background characters.

Do not crop out important body parts.
Do not move characters closer together or farther apart.
Do not change the camera height.
Do not change the perspective.

COSTUME & ARMOR:
Reproduce every character's clothing, armor, cloak, hood, belts, straps, metallic components, ornaments, textures, patterns and accessories exactly as seen in the reference.

Preserve the same materials, construction, wear, scratches, dirt, reflections and realistic surface detail.

Do not replace the costume with generic fantasy clothing.

CENTRAL CHARACTER:
Preserve the central character's exact hooded appearance, dark metallic facial armor/mask design, layered black clothing, ornate chest armor, circular shoulder ornaments, belt, long flowing dark cloak, arm armor, gloves and overall imposing silhouette exactly as shown.

If the reference contains a visible facial area beneath/within the mask, use the uploaded male face identity there while preserving the mask/helmet design and the original expression as much as physically possible. Specially stand like dr. Doom from Marvel.

SUPPORTING CHARACTERS:
Preserve the exact appearance, armor, scale, posture, stance and placement of every supporting character.

Their facial identity must come from the appropriate uploaded face reference, while their expression and visual presentation must remain faithful to the reference image.

ENVIRONMENT:
Recreate the same dark apocalyptic fantasy environment:
dramatic smoky atmosphere,
burning or illuminated landscape,
floating smoke and dust,
dark rocky terrain,
intense atmospheric depth,
and the same overall epic battlefield-like setting visible in the reference.

LIGHTING:
Match the reference lighting extremely closely.

Preserve the powerful warm golden-orange backlight behind the characters, dramatic sunlight/sunset glow, strong rim lighting around silhouettes, deep shadows on the dark armor and clothing, realistic reflected highlights, atmospheric haze and volumetric light.

The faces must receive lighting consistent with their exact position in the scene.

Do not use flat studio lighting.
Do not brighten the scene excessively.
Do not change the color mood.

CAMERA & FRAMING:
Match the reference camera angle, focal perspective, subject distance, vertical framing, crop, depth of field and cinematic composition.

The final image should feel like the same photograph/film frame recreated with different facial identities.

REALISM:
Ultra-photorealistic cinematic quality.
Extremely detailed realistic skin and facial texture.
Realistic pores and subtle imperfections.
Physically accurate armor materials.
Realistic metal reflections.
Realistic cloth and leather textures.
Natural atmospheric haze.
Photorealistic volumetric lighting.
High dynamic range.
Natural depth and dimensionality.
Professional blockbuster cinematography.
No artificial plastic appearance.

IDENTITY CONSISTENCY:
All male characters = uploaded male face identity.
All female characters = uploaded female face identity.

Do not mix identities between characters.

The facial identity must remain recognizable even under masks, helmets, shadows, dramatic lighting, or unusual viewing angles.

FINAL PRIORITY ORDER:

1. EXACT REFERENCE IMAGE COMPOSITION


2. EXACT CHARACTER COUNT AND POSITIONS


3. EXACT POSES AND BODY LANGUAGE


4. EXACT CLOTHING / ARMOR / CHARACTER DESIGN


5. EXACT EXPRESSION, GAZE AND HEAD ANGLE FROM REFERENCE


6. UPLOADED FACE REFERENCES FOR IDENTITY


7. EXACT LIGHTING AND ENVIRONMENT


8. ULTRA-REALISTIC CINEMATIC QUALITY



The final result must look like the original reference image was recreated with the SAME characters, SAME poses, SAME costumes, SAME expressions, SAME environment and SAME cinematic lighting — with ONLY their facial identities replaced by the uploaded male and female face references.

NO EXTRA CHARACTERS.
NO MISSING CHARACTERS.
NO POSITION CHANGES.
NO POSE CHANGES.
NO EXPRESSION CHANGES.
NO COSTUME CHANGES.
NO BACKGROUND CHANGES.
NO CAMERA CHANGES.
NO IDENTITY MIXING.
NO RANDOM FACES.
NO TEXT.
NO LOGOS.
NO WATERMARKS.

複製完整的提示並根據要求使用您自己的參考照片。

此提示提供的範例影像。此提示尚未經過我們網站的測試。

照相館
AI 產生的範例

為您自己的照片提供一點靈感。

1加入您的照片
A 人

保持臉部清晰可見

B 人

保持臉部清晰可見

JPG、PNG 或 WebP · 最大。每個 10 MB

僅當您選擇「產生」時才會發送您的照片。使用您有權使用的照片。

2讓它成為你的
更改詳情
編輯您的提示
開放雙子座

準備好…

登入您的帳戶

繼續使用您的 Google 帳戶。

您無需登入即可瀏覽和複製提示。

我們如何使用您的資訊: 隱私權政策