Photogpt
A cinematic portrait of the same person from the reference image, preserving exact facial identity, bone structure, proportions, and natural skin texture with high fidelity. Make sure to add focus only till the character after that slowly blur the behind part tot give more depth in the frame but make sure apart from the main character of the refrence no body else is in the frame Scene: an underground subway tunnel platform with curved tiled walls and strong perspective depth. A bench is clearly visible and integrated into the scene. The subject is sitting on the bench, fully grounded and interacting naturally with it (hips resting on bench, legs relaxed forward or slightly apart, hands resting on lap or beside body). Lighting is practical and realistic: overhead fluorescent tube lights running along the tunnel ceiling, emitting a cool teal/green color cast. The light reflects softly across the tiles and floor, creating a continuous ambient glow. No artificial or cinematic lighting. The subject is lit from above with soft shadows under the eyes and jaw, maintaining a moody low-light exposure. The face remains visible but not overly bright. Color grading: dominant teal/green tones with subtle red accents in the far background. Slightly desaturated, cinematic contrast, no neon glow or oversaturation. Composition: subject placed on one side of the frame, strong leading lines from tunnel and tracks, clear vanishing point. Camera: 35mm or 50mm lens, shallow to medium depth of field, sharp focus on subject, background softly blurred. Atmosphere: subtle ambient haze enhancing depth without reducing clarity. Strict requirements: - subject must be sitting (not leaning or standing) - full body weight visibly supported by bench - natural posture, relaxed and grounded - identity must remain unchanged from reference Avoid: standing poses, floating posture, cyberpunk glow, artificial lighting, overexposed highlights, distorted identity.

PhotoGPT 的AI 图片

Gemini Nano Banana 2.1
提示词

A cinematic portrait of the same person from the reference image, preserving exact facial identity, bone structure, proportions, and natural skin texture with high fidelity. Make sure to add focus only till the character after that slowly blur the behind part tot give more depth in the frame but make sure apart from the main character of the refrence no body else is in the frame Scene: an underground subway tunnel…

尺寸1856px x 2304px
参考图片
源图片 1

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 3.1Seedance 2