Enzo Fernandez Celebration Prompt

This is the long version. It walks the model through the whole celebration in order, with a time for each move. Paste it as it is, attach one photo, and set the output to vertical if the tool asks. If your tool has a separate "negative prompt" box, move the last line ("No text, no scoreboard...") there.

Vertical 9:16 video, 8 seconds, realistic, handheld sports broadcast feel. The person in the photo is a footballer in a packed night-time World Cup stadium, floodlights on, a huge crowd in the stands behind. Keep the person's face, hair, skin tone and the clothes from the photo exactly as they are.

0-1.5 s: the person walks fast along the touchline just past the goal, shouting toward the stands, one hand rising toward the ear.
1.5-2.5 s: the person stops, turns to face the camera; the camera drops low into a chest-up shot while the second hand comes up.
2.5-4.5 s: the taunt. Both hands cupped behind the ears, elbows pointing out, a big exaggerated pout, a small side-to-side head sway, eyes on the crowd.
4.5-5.5 s: hands open flat beside the head, the person looks up at the stands with the mouth open, as if saying "louder".
5.5-7 s: the person cups both ears again and holds the pout.
7-8 s: both arms go up wide to the crowd; a teammate in a plain dark navy shirt runs in from behind.

Loud crowd roar and whistles the whole way, rising at the end. No text, no scoreboard, no logos, no badges on the shirts, no captions. One continuous shot, no cuts.

Why it is written like this: video models are good at following a sequence of physical actions and bad at recognizing a named moment. So every line describes hands, elbows, mouth and camera, not who did it first. The phrase "the person in the photo" ties the action to your upload instead of a stock footballer.

The most common complaint with any photo-to-video prompt is that the person stops looking like themselves after a second or two. Most of that comes from the photo, not the prompt. These are the things that make the biggest difference, in order.

1. Use a sharp photo taken from the front, with the face filling at least a quarter of the frame. Group shots and far-away shots lose the face first.
2. Nothing over the face: no sunglasses, no hands, no hair across the eyes, no mask. A cap is fine if the eyes are clear.
3. Even light. Strong shadow on one side of the face often turns into a different jaw or nose in the video.
4. Keep the line "keep the face, hair, skin tone and clothes from the photo exactly as they are" near the start of the prompt, not at the end.
5. Keep the camera close. The chest-up shot in beat two is there on purpose. Wide drone shots of the stadium give the model room to invent a new face.
6. One person per photo. If there are two faces, many tools blend them or pick the wrong one.
7. If the tool offers a "motion strength" or similar slider, start in the middle. Very high motion breaks faces more often.

A dog photo works too, with one change: write "the dog in the photo sits up on its back legs and lifts both front paws beside its ears". See the dog football celebration video page for what that looks like.

Whatever tool you paste these prompts into, only use photos of yourself, of adults who have said yes, or of your pet. Do not use photos of footballers, celebrities or anyone who would not want to be in the clip. The prompts on this page never name a real person, and we suggest you keep it that way.

On our site, photos must be of adults or pets, you confirm permission before you pay, and photos are deleted after 30 days. The stadium, crowd and teammate are generated; there is no real player face, no real kit badge and no broadcast footage in the clip. We are an independent site with no link to the player, his club, his national team, FIFA or the owners of the Topo Gigio character. More on about.