e5

e5 asked a precise question: can AI produce catalogue and campaign images that sit next to its real photography without anyone seeing the join? We answered on real Marc O'Polo pieces from its online store, a light-blue knit for men and a pink bouclé for women.

Analysing the real campaign first. Before generating anything, we studied e5's own pictures: the white of the studio set, the light and the shadows, the framing and the poses, the way knit, logo and colour read on screen. Those became the rules every image had to follow.

Aligned, then extended. On the same garments we produced front, back and close-up shots that match the existing catalogue, then went further: the same pieces on different bodies and ages, plus-size and older models included, in a Mediterranean campaign setting, with walk-in videos for every product.

A technical sheet for every garment. Before a single image is generated, each piece gets its own sheet: fabric, weave and stripe, label, buttons, pocket monogram and fit, then the scene, its location, light, shadows and styling. Two examples below, a Dansaert Blue linen shirt and a sage-green suit, each checked against the real product page until texture, details and colour read true.

AI as a complement, not a replacement. The point was never to replace the photo shoot but to complete it: the products that miss the shoot, the extra sizes and ages, the campaign shots there was no budget for. A hybrid creative process, where photography and AI share one visual language.

Side by side, their photography and ours read as one catalogue. To close, the campaign clone: the same look, now in motion.