Short answer
Use one sharp, front-facing photo from the chest or waist up, evenly lit, with a calm, neutral expression. The clip frames you from mid-thigh up for nine seconds and finishes close on your face, so the photo is read twice: once on the turn at 1.5 seconds and once at the end, when there is nowhere to hide.
Try it with one photoStadium Calm Down photo guide
The camera ends right on your face, so this is a portrait job — with one expression rule.
Updated 2026-11-05
The short checklist
- Chest up or waist up. Feet are never in frame, so a full-body photo wastes the pixels.
- Face straight to the lens, head and hair fully in frame.
- Even light from the front — the clip ends close, and a half-dark face stays half-dark.
- A calm, neutral expression. The joke is that none of this concerns you.
- One person only. Two faces merge into one calm one.
Why the expression matters more here
Most trends want a performance. This one wants the absence of one: a deadpan face with the mouth closed while a stadium screams behind it. If your photo is a big grin, the render tends to keep the grin, and the palms down turn into a greeting instead of a verdict.
A photo where you look mildly unimpressed is the best source material there is. The prompt keeps the mouth closed and forbids a wide smile, but a neutral starting face is what makes that hold.
Glasses, beards and hair
Glasses, beards and moustaches all carry over from your photo, and so does the shape of your hair. The original man’s long curly hair and full beard are explicitly discarded, which matters: without that line the model paints your face onto his hair.
If you want the suit look to read properly, a photo with a visible collar or neckline helps — the render has to build a shirt and tie onto whatever you are wearing, and a bare neck gives it less to work with.
What actually breaks
| Photo | What happens |
|---|---|
| Sunglasses | The eyes are invented and the face stops matching at the close-up |
| A big smile | The deadpan collapses and the palms read as a wave |
| Two people in one frame | One calm face is assembled out of two |
| Backlit or dim photo | The face goes muddy exactly when the camera gets close |
| A cropped-at-the-chin photo | The render has to invent the jaw and the hairline |
| A phone screenshot of a photo | Soft edges survive and the close-up looks blurred |
If the render misses
- 1Check 1.5 seconds first — that is the turn, and the clearest look at your face before the walk.
- 2If the face drifts, swap in a sharper photo; the close-up at the end is the harshest test on the site.
- 3If your hands look wrong during the presses, that is the model, not the photo: run it again.
- 4If the suit came out wrong, your photo may be too dark for the collar and tie to be rebuilt.