mozgi ← All prompts
Photorealistic livestream frame: an anime girl in school uniform stands in class, with a real schoolgirl in front
GPT Image 2.0

A photorealistic livestream frame: an anime girl in a school uniform stands in a classroom, with a real schoolgirl in the foreground

Prompt

Use the character in the provided reference image as the base design.

Depict “a single frame from just after the livestream begins: an anime character caught on camera in a classroom after school, dragged into a classmate’s ‘踊ってみた’ LIVE stream.”

Accurately preserve the character’s anime-illustration feel from the reference image.

Preserve the face, hairstyle, hair color, eyes, ears, body proportions, silhouette, and the overall impression of the expression.

Absolutely do not compromise the character’s identity.

Do not turn the character into a real person.

Maintain a high-quality anime-illustration style for the character.

The setting is an ordinary Japanese high school classroom.

After school.

The desks and chairs have not been put away.

Bags, textbooks, worksheets, water bottles, pencil cases, and other items are naturally left around.

Do not make the space look like a photography studio.

Depict a lived-in classroom.

The composition is a vertical smartphone LIVE-stream screen.

The viewpoint is that of the streaming camera itself.

This is a screenshot from the LIVE stream as seen by viewers.

Do not show any filming equipment in the frame, such as a smartphone, camera, tripod, or selfie stick.

In the foreground is a real high school girl.

A Japanese high school girl.

Depict her as a person filmed with the front-facing camera of an actual smartphone.

Photographic style.

Photorealistic.

Only the classmate is a live-action person.

Do not turn her into an anime character.

Do not use a cel-animation style.

Do not use anime-style shading.

The foreground figure is a real-world high school girl.

The foreground classmate naturally appears large because she is close to the smartphone’s front-facing camera.

However, she is not the central figure in the composition.

It is fine if about half of her face is cropped out of the frame.

Her shoulder, arm, or hair may also be heavily cut off by the edge of the frame.

The classmate is positioned at the far left or right edge of the screen.

She is only moving closer to the camera while getting ready to stream and should not become the focal point.

She has just finished pressing the button to start the stream.

She is about to return to her mark to start dancing.

A moment before she takes a step back.

Bright and enthusiastic.

Looking forward to the stream.

The dance has not started yet.

Place the character from the reference image in the background.

Wearing a school uniform.

Framing from just above the knees to full-body.

Standing near the middle of the classroom.

Occupying about 30–50% of the screen height.

Although positioned in the background, the character must not be tiny like an extra.

Give the character a strong, clearly visible presence.

Depict the character large enough to naturally draw the viewer’s gaze.

Place the visual center of gravity of the composition on the side of the character from the reference image.

The character from the reference image must remain an anime character.

Do not make the character photorealistic.

The character has been dragged into keeping the classmate company.

Not particularly keen on dancing.

Not the main focus of the stream.

An expressionless look.

Slightly bothered.

Mildly confused.

An awkward smile.

Looking uncomfortable.

Any one of these.

An “Are we starting already?” kind of vibe.

It may also feel as though the character was stopped on the way home.

Just after the LIVE stream begins.

The very first frame of the video.

Nothing has started yet.

Only the classmate is full of enthusiasm.

The character has been swept up in the situation.

Important:

The foreground classmate is a live-action person.

The background character from the reference image is an anime character.

They exist in the same space while retaining their distinct visual styles.

Depict a strange everyday scene in which a real high school girl and an anime character naturally coexist.

The AI must not unify them into the same art style.

Only the foreground is live-action.

Only the background is anime.

Compose the scene so viewers first notice the classmate, then immediately shift their gaze to the anime character in the background.

Display a modern LIVE-stream interface on screen.

LIVE indicator.

Viewer count.

Comment section.

Hearts.

Gifts.

Share button.

Stream title.

Arrange the interface naturally to look like a stream that has just begun.

The comments are in natural Japanese.

Viewers are reacting to the classmate’s stream.

Do not portray the character from the reference image as a celebrity.

Do not portray the character as famous.

Image quality typical of a smartphone livestream.

Light HDR processing.

Subtle sensor noise.

Slight JPEG compression artifacts.

Auto exposure.

High-ISO noise.

Slight chromatic aberration.

The texture of smartphone video.

No cinematic effects.

No dramatic lighting.

No excessive depth-of-field effects.

Avoid the glossy rendering typical of AI.

Prohibited:

Do not turn the foreground figure into an anime character.

Do not turn the foreground figure into a two-dimensional illustration.

Do not unify all the characters under the same art style.

Do not make both the foreground and background anime-style.

Do not make both the foreground and background live-action.

Important:

The main subject on screen is the classmate’s LIVE stream.

Viewers also naturally find their attention drawn to the anime character in the background.

The classmate is large and partially cropped at the edge of the screen.

The character from the reference image naturally commands attention near the center of the screen.

Prioritize the strange everyday feeling of:

“An anime character dragged into a friend’s ‘踊ってみた’ dance stream.”
  • Needs your photo: attach an image to the request — the prompt works with it.
  • The prompt is in Japanese: paste it as is — the model understands it.
Open in catalog
The rights to this prompt belong to its author @Eris_Create_Lab. You are responsible for any commercial use — promoting AI tools, aggregators or Telegram bots: mozgi.io accepts no liability. Details →
Model
GPT Image 2.0
Format
9/16
Author
@Eris_Create_Lab
Original
x.com ↗

Similar prompts

All prompts for GPT Image 2.0

Copied