Photo Trends

Verified against ChatGPT · 2026-09-10

1986 couple studio portrait

The framed portrait that sat on the shelf — mottled canvas, a hard key, and the anchoring that stops two faces merging into one.

ChatGPT (GPT Image)Nano Banana (Gemini 3.1 Flash Image)Gemini app

The prompt

Ready to copy — highlighted parts are example details you can swap.

A couple in coordinated jewel-tone 1986 formalwear photographed as a studio portrait against a mottled blue canvas backdrop
Use my uploaded photo, which has two people in it. Anchor each face to its counterpart by position: the person on the left stays the person on the left, the person on the right stays the person on the right. Do not merge, average, swap or blend the two identities, do not let features from one face drift onto the other, and do not make the two people look more alike than they are. Keep each face exactly as it is — same bone structure, nose, eyes, lips, skin tone, age and natural asymmetry, no slimming, lightening or smoothing on either. Change only hair, clothes, jewellery, pose and background. Make this a 1986 Indian portrait-studio photograph of the two of us: coordinated but not identical formalwear in deep jewel tones, padded shoulders on both, a sapphire blouse on one and a wide-lapel burgundy blazer over an open-collar shirt on the other, statement earrings on one and a simple watch on the other. We are seated close on a bench, shoulders angled inward toward each other, heads turned back to the lens and tilted marginally together, with soft closed-mouth smiles, one of us slightly behind and higher so our heads sit at different levels. Behind us a hand-painted mottled canvas backdrop in blue and grey-blue, lighter behind our heads. Hard key high and forty-five degrees left with a defined shadow under each nose, weak fill on the right, a hair light rimming both crowns, catchlights in all four eyes. Medium-format film on glossy paper, 1986: fine grain, warm skin tones, mild vignetting. Vertical 9:16, mid-chest up on both.

Why this works

Two faces fail differently from one. The specific failure is identity bleed — the model averages the two faces toward a shared middle and both people come out looking like siblings. Anchoring each output face to its source counterpart by position gives the model a correspondence to hold. The offset head height matters too: real studio photographers never placed two heads at the same level, so equal heights read as an AI composite even when both likenesses are perfect.

What you get back

A well-modelled two-up studio portrait with offset head heights and both likenesses intact. Check both faces before sharing — bleed is the failure to watch for.

Verified against

ChatGPT GPT Image · 2026-09-10

Nano Banana / Gemini 3.1 Flash Image Gemini app · 2026-09-10

Changelog

  • 2026-09-09 Written for the September 2026 80s photo trend.
  • 2026-09-10 Rewritten as a single copy-and-go paragraph with no fill-ins, and paired with an example image.

Need this built into your business?

If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.

EXPLORE WHAT SCULT BUILDS
All Photo Trends prompts

Check your AI visibility

One URL in, a 0–100 score and the exact fixes out.

RUN THE CHECK

Browse all the tools

15 tools across six categories
13 of them never send your data anywhere

Free · No signup · No trial clock

SEE THE DIRECTORY