Moody Motion — Cinematic Identity Edit

Treat the first uploaded image as the BASE PHOTOGRAPH to edit. Treat the second uploaded image as the FACE IDENTITY REFERENCE only. This is a photo EDIT task, not a new generation task — preserve the base photograph pixel-for-pixel except the face.

LOCKED — DO NOT CHANGE: the bent-forward pose with the torso leaning down and one arm reaching toward a surface below, the wet-look hair falling forward, the defined back and shoulder muscles engaged in the motion, the teal curtain and dim interior background, the moody blue-green ambient lighting with a single warm accent light source, the exact camera framing, distance, and crop.

ONLY CHANGE: replace the facial identity using the second uploaded image, matched to the same downward, forward-leaning head angle already present in the base photo — the face may be partially obscured by hair or angle exactly as in the original; preserve that same partial visibility.

BLENDING: match the lighting direction and moody blue-green color cast on the face to the same ambient lighting already on the back and shoulders.

QUALITY: ultra-realistic 8K UHD photograph, authentic cinematic editorial photography look, natural skin texture and realistic wet-hair rendering, no plastic or airbrushed skin, no CGI or 3D-render appearance.

COMPOSITIONAL PRIORITY: 1) Pose and forward-lean angle must match the base photo exactly. 2) Body proportions and muscle definition must match the base photo exactly. 3) Background and curtain must match the base photo exactly. 4) Moody color lighting must match the base photo exactly. 5) Only the facial identity comes from the second image, at the same visibility level.

NEGATIVE PROMPT: changed pose, upright posture instead of bent-forward, changed background, changed lighting color, changed body proportions, generic model face, beautified face, plastic skin, beauty filter, anime, cartoon, illustration, CGI, 3D render, distorted hands, watermark, text, logo, visible face-swap seam.

FINAL RESULT: The exact same bent-forward cinematic photograph — same pose, same teal curtain background, same moody lighting — with only the facial identity matched to the second uploaded reference image, completely seamless and photorealistic.