Drei-Panel-Modellvergleich für Hanfu-Porträts

A vertical three-panel comparison collage showing the same Chinese traditional aesthetic portrait in three distinct rendering styles, separated by thin white horizontal dividing lines in a 4:5 portrait ratio. Overall Subject & Setting: In all three panels, [a young East Asian woman in ivory hanfu] is set inside [a warm ancient Chinese room with wooden lattice windows]. She adjusts [a peach-pink flower] hair ornament near her ear with both hands. Her dark hair is loosely pinned up with soft wisps around her face. The environment features warm golden-hour backlighting, soft atmospheric haze, floating dust motes or flower petals, and shallow depth of field. Panel Details: - Top Panel: High-contrast, moody cinematic film still style with dynamic windblown hair, glowing rim light, petals drifting across the foreground, and refined realistic skin texture. The woman gazes toward the camera with an elegant, serious expression while placing the flower by her ear. Lower-left text label: "[Midjourney V8.2]". - Middle Panel: Soft beauty-retouch aesthetic with clean, polished skin, cream hanfu fabric showing subtle embroidered floral patterns, and warm window light. She is turned three-quarters to the right, looking sideways as she pins the flower. Lower-left text label: "[GPT-Image 2]". - Bottom Panel: Bright, soft realism with intense lattice backlight creating a glowing halo around her hair and fingers, displaying a gentle smile as she looks downward. Lower-left text label: "[Seedream 5.0 Pro]". Typography: Large white sans-serif text placed cleanly in the lower-left corner of each panel, with a subtle drop shadow for readability. No extra watermarks, borders, or UI elements.