Create a **6-second vertical selfie-style emotional video** using the **uploaded picture as the identity reference for the person**.
Preserve the uploaded person’s **exact facial identity, gender traits, hairstyle, skin tone, face shape, eyes, eyebrows, nose, lips, and overall recognizable appearance**. Keep the person photorealistic and consistent throughout.
Recreate the reference video’s main emotional progression: the person begins cheerful and completely normal while walking outside, then their expression suddenly collapses into sadness, their eyes become watery, and visible tears begin streaming down their cheeks as the camera moves closer.
### 0.0–1.0s
Begin with a **wide handheld selfie shot** outdoors on a city sidewalk.
The person walks forward naturally while holding the camera at arm’s length.
They look directly into the lens with a relaxed, cheerful smile.
Show part of the extended arm in the foreground to reinforce the authentic phone-selfie perspective.
Keep the background realistic:
* sidewalk
* parked cars
* building façades
* pedestrians far in the background
* bright natural daylight
Add subtle walking bounce and natural handheld movement.
### 1.0–2.0s
The smile begins fading very quickly.
The person’s eyebrows draw slightly inward and upward.
Their eyes widen subtly, and the mouth opens a little as if they have suddenly remembered or realized something upsetting.
The camera naturally moves closer toward the face, changing from a wider selfie shot into a tighter upper-body framing.
Their walking continues, but the emotional shift becomes the main focus.
### 2.0–3.0s
The expression turns clearly distressed.
The eyebrows become strongly furrowed and raised in the center.
The person’s eyes become glossy and visibly wet.
Their lips part slightly, and the lower lip begins to tense or tremble.
The camera continues pushing closer until the face dominates most of the frame.
Keep the transition sudden but believable, not exaggerated or comedic.
### 3.0–4.0s
Move into a **tight facial close-up**.
Small tears form along the lower eyelids.
The person looks directly into the camera with a shocked, vulnerable, overwhelmed expression.
One tear begins escaping from an eye and rolling naturally down the cheek.
Their breathing appears slightly heavier through subtle mouth and nostril movement.
The city background becomes strongly blurred from the close focal distance.
### 4.0–5.0s
The crying becomes fully visible.
Tears now stream down both cheeks in thin, realistic trails.
The person’s eyes are redder and glassy, with wet reflections along the lower eyelids.
Their eyebrows remain tightly pulled upward toward the center.
The mouth stays slightly open, as if they are trying not to sob.
Keep the face centered and extremely close to the camera.
### 5.0–6.0s
Finish with an **extreme close-up of the crying face**.
Multiple tears continue rolling down the cheeks while fresh moisture remains pooled around the eyes.
The person stares toward the camera with a deeply emotional, stunned expression.
Add only tiny natural head movement and handheld shake.
End while the tears are still falling. Do not suddenly recover or smile again.
### Camera and visual requirements
Match the reference vibe closely:
* vertical 9:16 format
* casual smartphone selfie perspective
* person walking outdoors
* extended arm visible initially
* cheerful opening expression
* sudden emotional collapse
* smooth progression from smile to distress to visible crying
* camera gradually moving from wide selfie to extreme facial close-up
* subtle handheld walking motion
* natural daylight
* realistic urban sidewalk background
* strong background blur toward the end
* photorealistic tears and wet eyes
The tears must behave like **real liquid**, forming at the lower eyelids before rolling down the natural curves of the cheeks. Add realistic specular highlights and small tear trails without making the face look drenched or artificial.
Preserve the uploaded person’s identity throughout. The crying transformation should change only the **expression and tear state**, not their facial structure.
No dialogue, no text, no screaming, no dramatic body collapse, no makeup melting effect, no cartoon tears, no scene cuts, no face distortion, and no identity drift.