Living Room Podcast — Interviewing a Cat

Gemini Omni FlashRealisticComedyCatCute PetsAnimalsLip SyncImage to VideoStructured Prompt10s
Three reference images first build out the shot-reverse-shot setup for a single podcast room, then Shoot 1-5 has the person and the cat trading lines one at a time. The complete workflow for a two-camera dialogue short.
PROMPT · Video Prompt
Create with AI
Shoot 1 @Image 1 : The Man Camera: (static camera, no movement) realistic arm movements and subtle micro-movements. Micro-action: [He is reading the news article displayed on his phone to the cat on his right. After finishing reading, he lifts his head, and reacts with a naturally surprised expression. The reaction is subtle and realistic, not exaggerated.] Script: “Okay this is wild — scientists say cats sleep 16 hours a day.” Voice: male voice, mid-30s, bright. Accent: English with a neutral American accent, informal and conversational, like a real casual chat in English. Shoot 2 @Image 2 : The Cat Camera: (static camera, slow zoom in) realistic movements and subtle micro-movements. Micro-action: [The cat nods in agreement with what it just heard and says:] Script: “Yeah, and the other 8 we spend judging you.” Voice: Cat speaks in male voice, mid-30s, deep. Accent: English with a neutral American accent, informal and conversational, like a real casual chat in English. Shoot 3 @Image 1 : The Man Camera: (static camera, no movement) realistic arm movements and subtle micro-movements. Micro-action: [He pauses for a moment as if thinking about his answer, then says:] Script: “Wait — so when you stare at me at 3am, that’s judgment, not love?” Voice: male voice, mid-30s, bright. Accent: English with a neutral American accent, informal and conversational, like a real casual chat in English. Shoot 4 @Image 2 : The Cat Camera: (static camera, slow zoom in) realistic movements and subtle micro-movements. Micro-action: [The cat tilts its head slightly, deadpan, then says:] Script: “Can’t it be both?” Voice: Cat speaks in male voice, mid-30s, deep. Accent: English with a neutral American accent, informal and conversational, like a real casual chat in English. Shoot 5 @Image 3 : The Cat + Man Micro-action: [Both react at the same time — man laughs and shakes his head, cat’s whiskers twitch as if amused — they laugh together.] Camera shake slightly as they laugh.
✍️ Editor’s Notes

这条是对话脚本体,五个Shot按对话轮次切分,每轮标清机位——静止不动还是缓慢推近,还有角色微表情、台词和配音要求(口音、音色、年龄),几乎是给Gemini Omni Flash这种支持语音画面联动的模型写的分镜剧本。语言精炼到每段只留一个动作加一句台词,信息密度靠轮次堆叠而不是单句形容词。

Create with AI
Image Prompt
PROMPT · Full Prompt
Create with AI
Image 1 : An iPhone photo of a the man seated comfortably on a light fabric couch during a podcast recording, holding a smartphone while reading from it, speaking into a professional black podcast microphone mounted on a boom arm. Modern podcast studio with beige curtains in the background, cozy minimalist interior, realistic recording setup, casual creator atmosphere. iPhone photo, eye-level framing, candid dump vibe, direct harsh iPhone flash, slight grain, realistic skin detail, authentic imperfections, natural lighting balance, documentary feel, unposed moment, relatable creator aesthetic. No text, no watermark, no distortion. Image 2: Using this image @Image 1 only as a reference for the setting and style, recreate the SAME podcast in the SAME room — same kind of sofa, the same boom podcast microphone on its arm, the same cozy warm living-room atmosphere and a similar curtain backdrop — but seen from the OTHER side of the podcast, as if filming the second guest sitting across the table. It does NOT need to be an exact copy of the background: the plant, the exact lighting, the cushions and small props can be different and arranged naturally, as long as it clearly reads as the same room. Do NOT make it a mirror image of the reference. Show it from the opposite seat: the microphone now comes in from the LEFT side of the frame and the person faces toward the left / center, as if looking at the guest sitting opposite them. Put a Cat @Image 2 there. Photorealistic, natural skin texture and lighting. Single image. Image 3: @Image 1 @Image 2 combine them
✍️ Editor’s Notes

这条其实是图像生成的前置准备,先用Image1描述一张iPhone随拍质感的播客照片,定下构图和「candid dump vibe」的氛围词,再用Image2要求同房间换机位生成对角座位,最后Image3才把人和猫合成到一张图里。三步走的是搭场景、建对称机位、合成的图像工程套路,是给后续视频生成铺参考图的准备工作,不是视频提示词本身。

Create with AI
Related Works
×