Can you guarantee the adult will look the same every time?

No. The eleven-scene synthetic bench was strong, but it used one invented identity and cannot establish performance for every real adult, angle, or reference set. Every result needs an identity review.

What is the biggest warning sign in a reference?

A hidden or ambiguous face. In the illustrated bench, the softest identity results came from the one source where the face was not clearly available. For adults, also avoid filters, sunglasses, motion blur, and group crops.

Did changing outfits change the synthetic model's identity?

Not in this discovery pass. Five outfit-change cases remained usable, including a blazer, sweater, coat, button-up, and trail clothing. That is encouraging evidence from one identity, not a promise for every garment or pose.

What should I do with a scene that is almost right?

If the identity is only almost right, reject it. Re-run from the original references with the relevant angle added or the request simplified. A plausible near-match is still the wrong subject.

Limitations: Identity Drift in Person and Character Consistency

Generative editing preserves what a reference shows clearly and invents what is hidden, tiny, or ambiguous. That rule explains most strong and weak outputs in both lanes. Clear front-facing identity, visible hair, prominent clothing, and simple distinctive features anchor well. New profiles, distant full-body shots, obscured faces, and fine detail demand more scrutiny.

Person lane: 11 of 11 usable on one synthetic adult

We created one fully synthetic adult reference, then tested eleven scene changes one at a time. The cases covered a home workspace, rooftop, podcast studio, cafe, fashion lookbook, business stage, generic product demonstration, wooded trail, side profile, full-body lobby walk, and founder desk. Five changed the outfit; five explicitly locked the original teal overshirt and white shirt. Every output kept the face, short dark curls, skin tone, age, and average-athletic build recognizable enough to use. The locked outfit also held in every locked case.

Mara · synthetic adult — reference portrait, same person — Reference portrait

Mara · synthetic adult — home workspace, same person — Reference portrait

What that result does not prove

It is one synthetic identity, not a representative sample of real adults, ages, hair types, skin tones, cameras, or reference quality.
A usable identity at normal viewing size can still soften subtle marks such as freckles, small scars, or fine hairline detail.
Profile and distant full-body scenes give the reviewer less facial information, so a plausible lookalike can be harder to catch.
The bench did not test crowded groups, extreme poses, heavy occlusion, severe lighting changes, or long chains of derived outputs.

Small sample, strong signal

The 100% discovery result supports a careful person-consistency promise, but not a deterministic guarantee. We publish the sample size and its synthetic provenance so the number cannot be mistaken for broad population validation.

Illustrated lane: 15 of 18 usable across seven characters

The public-domain illustrated bench covered watercolor, engraving, color plate, and line art across cycling, reading, baking, sailing, gardening, and other scene changes. Fifteen scenes held identity, costume, and medium well enough to use. Three were minor flaws rather than total failures. The clearest pattern was weak input: a face-hidden, ambiguous reference led to an invented face and looser style. One otherwise recognizable scarecrow scene shifted from flat line art toward denser cross-hatching.

What to inspect in every person output

Eye spacing, nose and mouth shape, jaw, hairline, curl pattern, visible marks, age, and skin tone.
Body proportions in seated, walking, and full-body shots; do not accept silent reshaping.
Any clothing requested to stay, including color, collar, closures, sleeves, and silhouette.
Hands and object interactions, especially in product demonstrations.
Whether the setting implies an event, endorsement, workplace, credential, or relationship that is not true.

What to inspect in every illustrated output

Face and body proportions, costume silhouette, signature colors, props, and distinctive marks.
Small logos, lettering, badges, and costume filigree, which can soften or turn into texture.
Art medium and rendering density, especially for flat or graphic line styles.
Hidden angles the source never supplied; invention risk rises when the reference turns away or crops the subject.

The operating rule

Do

Use two or three relevant references for new angles and full-body scenes.
Compare identity before judging whether the image is attractive.
Re-run misses from the original references.
Keep permission and truthful context in the review checklist.

Avoid

Treat a polished lookalike as the same person.
Rely on a face-hidden or group reference to establish identity.
Chain a long series from an output that already drifted.
Publish generated context as documentary evidence.

The right expectation is useful, reviewable consistency. Bring strong references, keep the request bounded, and be willing to reject a beautiful miss. The reference guide shows how to choose inputs that leave less for the model to invent.

What can drift when the subject must stay the same