Photogpt2026年3月15日

REFERENCE IMAGE LOCK:
Use the uploaded image as a strict visual reference.
Preserve the exact same person, facial structure, body proportions, hairstyle, and identity.
Do NOT change who the person is.
No face replacement, no new features, no beautification.

TRANSFORMATION GOAL:
Convert the reference image into a raw, low-fidelity surveillance camera (CCTV) style image.

SCENE & COMPOSITION:
The subject appears centered from mid-torso upward, standing behind a cramped retail checkout counter.
Camera is high-mounted with a steep downward, top-down surveillance angle.
Vertical 4:5 aspect ratio.
Subtle fisheye distortion visible at the edges of the frame.

SUBJECT POSE & EXPRESSION:
The same person from the reference image.
Head tilted back, looking up toward the camera.
Relaxed yet defiant posture.
One hand raised toward the camera with the middle finger extended.
Gesture should be readable but softened and partially ghosted by motion blur.
Facial features partially obscured by blur and noise.
Dark oversized sunglasses reflecting ambient light.
Mouth slightly open, tongue faintly visible.

ENVIRONMENT:
Cramped retail shop interior.
Foreground includes a cluttered checkout counter with loose papers and a white shopping bag with red rope handles.
Background shelves densely packed with small items, souvenirs, and figurines, appearing abstract due to low resolution.

LIGHTING & COLOR:
Harsh, flat fluorescent lighting typical of indoor retail CCTV.
Blown highlights, dull shadows.
Cool green-blue color cast.
Uneven color reproduction with visible banding and reduced dynamic range.

IMAGE QUALITY & ARTIFACTS:
Intentionally degraded, low-fidelity image.
Strong motion blur, digital smearing, compression artifacts.
Heavy noise, scan lines, low perceived sharpness.
Details appear stretched and indistinct despite deep depth of field.

MOOD:
Raw, gritty, confrontational, unpolished, authentic surveillance footage.

NEGATIVE:
Different person, face change, identity change, beauty retouching, cinematic lighting, studio lighting, sharp focus, high resolution, HDR, DSLR look, professional photography, glamour, fashion portrait, clean image, vibrant colors, polished aesthetic.

PhotoGPT 的AI 图片

Photogpt2026年3月15日

提示词

REFERENCE IMAGE LOCK: Use the uploaded image as a strict visual reference. Preserve the exact same person, facial structure, body proportions, hairstyle, and identity. Do NOT change who the person is. No face replacement, no new features, no beautification. TRANSFORMATION GOAL: Convert the reference image into a raw, low-fidelity surveillance camera (CCTV) style image. SCENE & COMPOSITION: The subject appears…

尺寸1856px x 2304px

参考图片

开始探索

加入拥有数千名活跃用户的 Discord 社区，了解最新提示词灵感并探索社区作品

生成图像加入 Discord

Google Veo 3.1 Seedance 2

PhotoGPT 的AI 图片

开始探索

相关作品