THE PROCESS

I generate the backdrop first. The prompts include lighting, style, mood, decor, color. Many, many generations later, I land on one that I think will work. I take into consideration the lighting so that it makes sense when I start putting it together with the people.

Next, I start generating each individual person. I use prompts that describe each person - age, hair color, clothing choices, hair style, shoes, stance, lighting on the face, etc. until I get one that closely resembles the actual person. Then I take a photo of the person and swap it into the AI-generated version of themselves.

Then it’s a matter of photoshopping it all together - matching the lighting, the color, the scale of each person, creating balance in the overall image, and tweaking it to look just right. This is a really condensed explanation of the process so for reference, it’s taken approximately 20 hours to create each one.