北風と太陽
使用したAI
ChatGPT
薄着にする方向の指示は拒否されるChatGPTへの指示は、薄着になる状況設定をすればよいのではないか?
----
この画像を元にZ imageに生成させるための詳細な英語のpromptを、人間が後で修正しやすいように整理して生成して
とZ Image用のpromptを生成させてみると
----
Create a wide horizontal cinematic image in a 16:9 aspect ratio.
SCENE OVERVIEW:
A realistic live-streaming accident moment inside a very small, hot, soundproof recording booth. A mature woman around 35 years old is streaming as a VTuber, but she has accidentally shown her real face on the live stream. She has just noticed the mistake and is shocked, panicked, and desperately trying to stop the stream.
MAIN SUBJECT:
A mature East Asian woman, around 35 years old, realistic appearance, no makeup, slightly tired but attractive in a natural way. Her face is flushed from heat and embarrassment. Her skin is shiny with sweat, especially on her forehead, cheeks, and neck. Her hair is damp and messy from the heat, with strands sticking to her forehead. She is wearing a loose, casual light gray T-shirt and wired earphones.
POSE AND EXPRESSION:
She is seated in a desk chair inside the booth, leaning forward toward the camera. Her eyes are wide open, eyebrows raised, mouth open in panic, as if she has just realized she is visible on the stream. One arm is stretched toward the camera or monitor, fingers spread, trying to stop the broadcast or cover the camera. Her body language should clearly communicate surprise, embarrassment, and urgent panic.
ENVIRONMENT:
The room is a cramped soundproof booth with dark acoustic foam panels on the walls. The space feels hot, humid, and poorly ventilated. Include streaming equipment such as:
- a large condenser microphone
- a pop filter
- a boom arm
- a ring light
- headphones or earphones
- a computer monitor near the camera
- a small desk fan
- cables and small desk clutter
- a digital temperature and humidity display showing a warm, uncomfortable room
LIGHTING AND MOOD:
Use dramatic, cinematic lighting. The ring light creates a bright circular glow on one side, while the rest of the room is dim and warm. The image should feel like a realistic live-stream screenshot captured at the exact moment of discovery. The atmosphere should be tense, awkward, sweaty, and claustrophobic, but not horror-like.
LIVE STREAM UI OVERLAY:
Add a generic live-stream interface overlay, not an exact copy of any real platform. Include:
- a red “LIVE” indicator in the top-left corner
- a running stream timer
- viewer count and reaction icons in the top-right corner
- a vertical chat column on the left side with small Japanese comments reacting to the accident
- a translucent comment input bar at the bottom
- simple generic control icons at the bottom-right
Do not use any real company logos, real YouTube branding, or copyrighted UI elements.
BOTTOM-RIGHT VTUBER OVERLAY:
In the bottom-right corner, add a small superimposed VTuber avatar window, like a streaming software overlay. The avatar should be a cute anime-style female character that represents the streamer’s idealized VTuber persona. The VTuber avatar must mirror the real woman’s exact pose and expression:
- same wide shocked eyes
- same open panicked mouth
- same reaching hand gesture
- same urgent body language
The VTuber character should have long brown hair, large expressive eyes, a cute outfit with soft pink accents, and a polished anime look. The avatar window should have a clean border and should clearly look like an overlay on top of the live stream.
COMPOSITION:
The real woman should be the main focus, positioned near the center. The outstretched hand can be closer to the camera and slightly motion-blurred for urgency. The microphone and ring light should frame the scene. The VTuber overlay should be clearly visible in the lower-right corner without covering the woman’s face. The overall composition should immediately tell the story: “a VTuber streamer accidentally revealed her real face and panicked.”
STYLE:
Photorealistic main scene with a cinematic documentary feel. High detail, realistic skin texture, realistic sweat, realistic room clutter, realistic lighting. The VTuber overlay should be anime-style, colorful, and clean, creating a contrast between the real streamer and her virtual persona.
CAMERA:
Wide-angle lens, close to the desk or monitor, slightly distorted perspective from the webcam or streaming camera. The outstretched hand should appear closer to the viewer. Shallow depth of field, with the woman’s face sharp and the foreground hand slightly blurred.
NEGATIVE PROMPT:
No nudity, no sexual content, no minors, no real YouTube logo, no real brand logos, no gore, no horror, no extreme distortion, no extra fingers, no deformed hands, no duplicated faces, no unreadable messy UI dominating the image, no overly cartoonish main woman, no glamour makeup, no perfect studio room, no calm expression.
----
この画像を元にZ imageに生成させるための詳細な英語のpromptを、人間が後で修正しやすいように整理して生成して
とZ Image用のpromptを生成させてみると
----
Create a wide horizontal cinematic image in a 16:9 aspect ratio.
SCENE OVERVIEW:
A realistic live-streaming accident moment inside a very small, hot, soundproof recording booth. A mature woman around 35 years old is streaming as a VTuber, but she has accidentally shown her real face on the live stream. She has just noticed the mistake and is shocked, panicked, and desperately trying to stop the stream.
MAIN SUBJECT:
A mature East Asian woman, around 35 years old, realistic appearance, no makeup, slightly tired but attractive in a natural way. Her face is flushed from heat and embarrassment. Her skin is shiny with sweat, especially on her forehead, cheeks, and neck. Her hair is damp and messy from the heat, with strands sticking to her forehead. She is wearing a loose, casual light gray T-shirt and wired earphones.
POSE AND EXPRESSION:
She is seated in a desk chair inside the booth, leaning forward toward the camera. Her eyes are wide open, eyebrows raised, mouth open in panic, as if she has just realized she is visible on the stream. One arm is stretched toward the camera or monitor, fingers spread, trying to stop the broadcast or cover the camera. Her body language should clearly communicate surprise, embarrassment, and urgent panic.
ENVIRONMENT:
The room is a cramped soundproof booth with dark acoustic foam panels on the walls. The space feels hot, humid, and poorly ventilated. Include streaming equipment such as:
- a large condenser microphone
- a pop filter
- a boom arm
- a ring light
- headphones or earphones
- a computer monitor near the camera
- a small desk fan
- cables and small desk clutter
- a digital temperature and humidity display showing a warm, uncomfortable room
LIGHTING AND MOOD:
Use dramatic, cinematic lighting. The ring light creates a bright circular glow on one side, while the rest of the room is dim and warm. The image should feel like a realistic live-stream screenshot captured at the exact moment of discovery. The atmosphere should be tense, awkward, sweaty, and claustrophobic, but not horror-like.
LIVE STREAM UI OVERLAY:
Add a generic live-stream interface overlay, not an exact copy of any real platform. Include:
- a red “LIVE” indicator in the top-left corner
- a running stream timer
- viewer count and reaction icons in the top-right corner
- a vertical chat column on the left side with small Japanese comments reacting to the accident
- a translucent comment input bar at the bottom
- simple generic control icons at the bottom-right
Do not use any real company logos, real YouTube branding, or copyrighted UI elements.
BOTTOM-RIGHT VTUBER OVERLAY:
In the bottom-right corner, add a small superimposed VTuber avatar window, like a streaming software overlay. The avatar should be a cute anime-style female character that represents the streamer’s idealized VTuber persona. The VTuber avatar must mirror the real woman’s exact pose and expression:
- same wide shocked eyes
- same open panicked mouth
- same reaching hand gesture
- same urgent body language
The VTuber character should have long brown hair, large expressive eyes, a cute outfit with soft pink accents, and a polished anime look. The avatar window should have a clean border and should clearly look like an overlay on top of the live stream.
COMPOSITION:
The real woman should be the main focus, positioned near the center. The outstretched hand can be closer to the camera and slightly motion-blurred for urgency. The microphone and ring light should frame the scene. The VTuber overlay should be clearly visible in the lower-right corner without covering the woman’s face. The overall composition should immediately tell the story: “a VTuber streamer accidentally revealed her real face and panicked.”
STYLE:
Photorealistic main scene with a cinematic documentary feel. High detail, realistic skin texture, realistic sweat, realistic room clutter, realistic lighting. The VTuber overlay should be anime-style, colorful, and clean, creating a contrast between the real streamer and her virtual persona.
CAMERA:
Wide-angle lens, close to the desk or monitor, slightly distorted perspective from the webcam or streaming camera. The outstretched hand should appear closer to the viewer. Shallow depth of field, with the woman’s face sharp and the foreground hand slightly blurred.
NEGATIVE PROMPT:
No nudity, no sexual content, no minors, no real YouTube logo, no real brand logos, no gore, no horror, no extreme distortion, no extra fingers, no deformed hands, no duplicated faces, no unreadable messy UI dominating the image, no overly cartoonish main woman, no glamour makeup, no perfect studio room, no calm expression.
呪文
呪文を見るにはログイン・会員登録が必須です。
イラストの呪文(プロンプト)
jacket partially removed, heart in eye, burnt clothes, holding fishing rod, kanji, doujin cover, pentagram, tape gag, adjusting headwear, red socks, friends, cloud print, coke-bottle glasses, oral invitation, competition school swimsuit, barbell piercing, gradient legwear, prisoner, blood on breasts, wind chime, carrying over shoulder, tape measure, flaming weapon
イラストの呪文(ネガティブプロンプト)
入力なし