【韓国トレンド・AI画像生成】推しと自撮り!? 韓国で流行｢맥북3컷｣をGeminiで再現!!

맥북3컷 風!! 画像作成プロンプト

A comprehensive prompt for generating a vertical smartphone photo collage in 3 panels. Don’t need a photo frame The final image must have the authentic, raw quality of a photo taken with an iPhone at night.

Crucial Subject & Scene Instructions:
The two individuals in the generated image must be precise reproductions of the subjects in the two accompanying reference images (reference_image_A.png and reference_image_B.png).

The AI must accurately render each person’s specific gender, hairstyle, hair length, facial features (eye size and shape, nose height, mouth size, lip thickness), hair color, makeup, and clothing as seen in their respective reference images. The combination of subjects can vary (two females, two males, or one male and one female), and the generation must reflect the provided references exactly.

All three panels must be set in the exact same location within a stylish, dimly lit room at night. The direction of light, the background elements, and the overall atmosphere must remain consistent throughout the entire collage, creating a cohesive sequence of moments in the same space. If the reference image is visible only in part of the top or only in the upper body, create the appropriate tops, pants, or skirts to match the style of the image and complete the costume.

Panel 1 (Top): The two subjects, as defined by the reference images, are positioned close together, looking directly at the camera. Both are performing the popular “looda heart” (cheek heart) pose with their hands. The same pink heart filter effect seen above the heads. This heart filter is made visible on the heads of each of the two characters, and it’s made possible to see this filter according to the shape of each head
The lighting is warm and moody, consistent with the room’s night setting.

Panel 2 (Middle): The same two subjects are taking a mirror selfie, jointly holding a horizontal white iPhone and looking intently at the screen. The background and warm lighting are identical to Panel 1. The camera angle and face sizes are identical to Panel 1 (Top).

Panel 3 (Bottom): The two subjects are
gazing into each other’s eyes and smiling warmly, sharing an intimate moment. They may be doing slight face-framing poses. they sit in the same dimly lit, cozy spot as the previous panels. The camera angle and face sizes are identical to the top and middle panels.