Vivaldi Dance AI Video Generator from Image | Media.io

Take the main dancer’s place while following your selected choreography. Prompt: “Use Image 1 as the sole identity and outfit reference. Use Video 1 as the Vivaldi dance motion and scene reference. Replace only the main dancer with the person from Image 1. Preserve their facial features, hairstyle, skin tone, clothing, and body proportions. Follow the reference footwork, arm gestures, turns, pauses, and weight shifts in the same order and timing. Preserve the background, lighting, camera movement, framing, and shot sequence from Video 1. Keep planted feet stable and maintain natural limb articulation. Match the reference aspect ratio and duration within the model’s supported limits. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No original dancer’s face, extra limbs, foot sliding, outfit flicker, added choreography, text, or watermark.”

Keep your clothing and accessories consistent through the routine. Prompt: “Create a Vivaldi dance remix with the person in Image 1 replacing only the main dancer in Video 1. Preserve the person’s face, hairstyle, body proportions, and complete outfit from Image 1. Keep garment colors, fabric patterns, sleeve length, hemlines, shoes, and accessories unchanged. Follow Video 1 for choreography, body orientation, footwork, gestures, and timing. Let fabric and accessories move naturally with the body without changing their design. Preserve the reference setting, lighting, framing, camera movement, and shot order. Maintain stable foot contact and coherent hand-to-body occlusion. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No copied outfit from the reference dancer, changing shoes, accessory swaps, clothing flicker, distorted hands, extra limbs, text, or watermark.”

Animate an original humanoid character without losing its illustrated look. Prompt: “Use the single full-body humanoid character in Image 1 as the appearance and visual-style reference. Use Video 1 only for the Vivaldi dance choreography and timing. Animate the character in the setting from Image 1 with a fixed full-body camera. Preserve its face, silhouette, colors, outfit, limb count, and illustrated rendering style. Adapt the reference poses, arm gestures, footwork, and weight shifts to the character’s proportions while keeping joints coherent and feet grounded. Leave room around the hands and feet throughout the motion. SAME CHARACTER. NO DUPLICATE. NO NEW SCENES. No photorealistic face, changing body proportions, extra limbs, broken joints, foot sliding, background changes, added props, text, or watermark.”

Create a vertical dance excerpt with the full routine visible inside the frame. Prompt: “Use Image 1 as the sole identity and outfit reference. Use the trimmed 9:16 Video 1 as the Vivaldi dance motion, scene, and framing reference. Replace only the main dancer with the person from Image 1. Preserve their face, hairstyle, skin tone, outfit, and proportions. Follow the opening pose, choreography, gestures, footwork, and final pose in the original order and timing. Keep the reference background, lighting, camera movement, and vertical composition. Maintain visible space around the head, hands, and feet. Generate in 9:16 for the supported reference duration. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No added introduction, extra cuts, new zoom, automatic loop, cropped feet, face morphing, foot sliding, clothing flicker, text, or watermark.”