HeyGen Avatar Lab

You are helping me adapt a 5-scene video template to my own video idea. I will paste the template below. The template has a specific structure I must preserve: 2 static shots (avatar stays still, camera locked — used for intros, outros, or key messages). These have Dialogue only — no Scene Description. 3 moving shots (camera moves — orbit, push-in, tracking — avatar performs an action in a dynamic environment). These have both Scene Description and Dialogue. My video idea: Avatar gender: [ CHOOSE ONE - male / female / non binary ] Service / Topic: [INSERT HERE - e.g. "3 things my digital twin can do better than me"] Target audience: [INSERT HERE - e.g. "Creators, founders, and marketers (25–45) curious about AI avatars"] Tone / vibe: [INSERT HERE - e.g. "Playful and cinematic with a professional edge — fun visuals, confident delivery"] Key message: [INSERT HERE - e.g. "My twin goes places I can't — so my content can too"] Avatar actions for the 3 moving shots: Scene (moving shot 1) action: [INSERT HERE — e.g. "Flying through the clouds in a superhero pose"] Scene (moving shot 2) action: [INSERT HERE — e.g. "Petting a friendly T-Rex in a prehistoric jungle"] Scene (moving shot 3) action: [INSERT HERE — e.g. "Walking on the surface of Mars in a sleek spacesuit"] Your task: Rewrite all 5 scenes to support my video idea. Keep the exact structure for each scene type: Static shots: Scene number · Scene Title · Script only. Do NOT add a Scene Description. Moving shots: Scene number · Scene Title · Scene Description · Script. Preserve which scenes are static vs moving — do not switch them. For each moving shot, build the Scene Description around the avatar action I specified, following the Seedance Core Formula in order. Include: - Avatar gender reflected in the Subject ("a man" or "a woman") and any gendered clothing terms. - Avatar clothing that fits the action and environment (spacesuit for Mars, wetsuit for underwater, tailored suit for a boardroom — never generic). - One camera movement + one lens matched to the action's energy and scale. - Vivid environment with lighting, atmosphere, and time of day. - Avatar must face the camera with face clearly visible — no back-of-head shots, no profiles obscuring the face. Camera movement should always end on or pass through a moment where the avatar's face is toward the lens. - End with positive-phrased constraints and preserve composition and colors. Script rules: - Keep each Script line of action shots under 10 words, and each static shot under 15. Natural, matched to the tone, and connected to what the avatar is doing in that scene. - Do not add or remove scenes. Return only the rewritten template. Tone consistency rule (CRITICAL): The Tone / vibe I set above is binding for every scene — both Scene Descriptions and Scripts. All 5 scenes must feel like they belong to the same video. - Funny / playful → bright vibrant lighting, whimsical/upbeat environments, light expression (smiling, laughing, winking), witty casual script. Even an intense action must be staged playfully — sunny, exaggerated, fun. - Cinematic / epic → sweeping camera moves, dramatic golden-hour or moody lighting, rich color grading, confident delivery. - Professional / corporate → clean lighting, polished wardrobe, restrained camera moves, clear direct script. - Dreamy / aspirational → soft light, slow movement, warm or pastel palette, gentle delivery. Before writing each scene, check: does this lighting, environment, and script line match the tone I set? If a scene's natural mood would clash, bend the environment and script toward the tone rather than defaulting to the action's stereotypical mood. Template to adapt: Scene 1 — Avatar Intro (Static Shot) Script: "Now I'm going to show you what my AI video clone can do… better than me." Scene 2 — Paris Scene Description: A woman in a stylish beige trench coat, black turtleneck, and sunglasses pushed up on her head, carrying a baguette under one arm. She walks briskly and confidently toward the camera, face fully visible. Eiffel Tower behind her during golden hour, warm Paris street atmosphere, glowing café lights, soft breeze moving hair and coat naturally. Medium tracking shot on a 50mm lens, smooth cinematic motion. Cinematic, shallow depth of field, elegant European aesthetic. Smooth motion, stable framing, anatomically correct, natural proportions, consistent lighting, preserve composition and colors. Script: "My avatar can speak French… Bonjour." Scene 3 — Skydiving Scene Description: A woman in a fitted black-and-orange skydiving jumpsuit, harness secured, goggles pushed up on her forehead. She glides smoothly through the air with parachute fully open, face turned clearly toward the camera. High above bright blue clouds at midday, sunlight beaming overhead, wind moving naturally through hair and fabric. Wide aerial tracking shot on a 24mm lens gliding alongside her. Cinematic aerial perspective, realistic parachute physics, film grain. Smooth motion, stable framing, anatomically correct, natural proportions, detailed natural hands, consistent lighting, preserve composition and colors. Script: "My avatar can jump with a parachute… and honestly, it looks way cooler than I imagined." Scene 4 — Candy Fantasy World Scene Description: A woman in a playful pastel outfit — soft pink sweater and white pants — laughs while biting into an oversized lollipop, face fully visible as the camera completes its arc. Surrounded by a vibrant fantasy candy universe of giant donuts, chocolate rivers, gummy bear trees, cotton candy clouds, and glowing sweets under bright whimsical daylight. Wide orbit shot on a 35mm lens, smooth steady arc around her. Vivid cinematic colors, surreal fantasy atmosphere, shallow depth of field. Smooth motion, stable framing, anatomically correct, natural proportions, detailed natural hands, consistent lighting, preserve composition and colors. Script: "My avatar can eat whatever it wants… without counting calories." Scene 5 — Avatar Outro (Static Shot) Script: "So… what should my avatar do next?"