Documento originale
GPT Image 2 · en
A dramatic, self-portrait-style close-up of a stage scene shows a Japanese idol performer leaning towards the camera from a low angle, her upper body and outstretched arms dominating the frame. She has long, straight to slightly wavy hair, and her hair color is {argument name="hair color" default="dark reddish brown"} Her hair was glossy, and thick, side-swept bangs covered one eye. Her expression was cute and intimate; she placed one hand near her mouth, striking a shy idol pose, as if interacting directly with fans at a concert. She wore a black and dark {argument name="accent color" default="Purple"} The gothic idol stage costume features five clearly visible elements: a shiny, corset-style cropped top with metallic trim, a black belt with silver metal eyelets at the waist, black sheer mesh opera gloves, multiple straps on the upper arms, and a dangling silver chain across the chest and hem. The addition of glossy patent leather, studs, buckles, lace, and layered ruffles creates a highly detailed J-pop or anime concert look. The backdrop is a large indoor live stage with bright blue-white spotlights, smoke, truss installations, and LED screens flashing soft, abstract shapes, creating a strong backlighting and bokeh effect. The cinematic concert photography style employs high contrast, cool lighting with purple-blue highlights, shallow depth of field, slight lens distortion from a close-up arm perspective, dynamic foreground perspective, and vertical hand-held composition. This results in a vibrant, charismatic, and fan-service-oriented visual, much like…{argument name="idol group concept" default="A dark fantasy-style idol live performance"} .Come usare
- Copia tutto il prompt originale, inclusi i parametri.
- Adatta soggetto, testo e composizione alle tue esigenze.
- Usa uno strumento esterno compatibile; verifica accesso e risultato.
Prima di iniziare
- Controlla se serve un’immagine di riferimento. La compatibilità non prova il modello usato per l’esempio.
- Verifica nomi, numeri e testi; i risultati cambiano con le versioni del modello.
