
A highly specific JSON prompt designed to generate a four-frame, 2x2 seamless photobooth collage featuring two subjects (Sydney Sweeney and Ana de Armas). It strictly requires reference images to lock the subjects' identities and outfits, focusing on an ultra-wide, flash-lit, amateur iPhone photobooth aesthetic with detailed poses for each frame.
{
"task": "Create an iPhone-realistic photobooth collage (4 photos in a 2x2 grid)",
"subjects": [
{
"id": "Subject 1",
"description": "EXACTLY from the Reference Image (same face, hairstyle, skin tone, body proportions, overall look).",
"outfit": "EXACTLY the same as in the Reference Image. Do not change clothing, glasses, or accessories."
},
{
"id": "Subject 2",
"description": "EXACTLY from the Reference Image (same face, hairstyle, skin tone, body proportions, overall look).",
"outfit": "EXACTLY the same as in the Reference Image. Do not change clothing, glasses, or accessories."
}
],
"scene": {
"location": "Inside a small BOX photo booth",
"background": "Matte blue walls and blue floor, tight enclosed corner, clean minimal environment",
"camera_look": "Ultra wide / slight fisheye lens look with subtle perspective distortion typical of photobox cameras. iPhone 17 Pro, photobooth wide-angle.",
"lighting": "Bright neutral-white flash lighting, evenly illuminating faces, soft shadow falloff on blue walls, crisp detail, realistic skin texture. Direct neutral white flash (no yellow / no warm tone).",
"aest


















































