Generator Video AI dari Mobil Mainan ke Mobil Asli | Media.io

Use this prompt with a model that accepts both image and video references. Prompt: “Use Image 1 as the vehicle reference and Video 1 as the guide for scene order, camera movement, and transition timing. Replace the reference vehicle with the car from Image 1 in both miniature and full-size form. Preserve its silhouette, paint color, wheels, windows, and lights. Begin inside a moving passenger train with the miniature resting motionless on the interior windowsill while scenery passes outside. Follow the reference into the railway tunnel. Let the taillights illuminate before darkness conceals the window frame and ledge. Continue into the road-tunnel scene with the same vehicle at full size, then follow its exit onto the road. Preserve the reference framing and duration within the model's limits. Use the photo for appearance, not driving direction: the car drives forward with its front leading, the camera follows behind, and the tires stay on the road. Keep one car throughout, with no extra shots, people, hands, distorted panels, extra wheels, paint changes, or added text.”

Start here to turn your car photo into a miniature-car opening image. Image prompt: “Using Image 1 as the car reference, create a photorealistic vertical 9:16 image taken inside a passenger train. A palm-sized die-cast model of this exact car rests on the interior windowsill, viewed from a low rear three-quarter angle with its front pointing along the ledge. Preserve the reference vehicle's body shape, paint color, wheel design, and light shapes. Show a realistic train window frame and enough of the ledge to establish the toy's small scale. Outside the glass, softly blurred countryside suggests the train is moving toward a railway tunnel. Warm late-afternoon light falls across the miniature. Keep the car sharp and the background secondary. Exactly one miniature car; no full-size vehicle outside, no people, no hands, no extra wheels, no floating objects, no added lettering.”

Animate your prepared image with this toy car to real car video prompt. Video prompt: “Animate the supplied train-windowsill image into a cinematic vertical car reveal. During the opening, the miniature car and all four wheels remain still on the ledge while countryside slides past outside. The passenger train enters a railway tunnel and daylight fades. Hold briefly in darkness as the model's red taillights illuminate. Move the camera gently forward, letting the window frame and ledge disappear into shadow. Use this concealed transition to reveal the same car at full size, now driving through a road tunnel. Follow it from a low rear three-quarter angle as it exits onto a wet night highway beneath streetlights. Maintain the same vehicle shape, paint, wheels, and lights. Use continuous-looking camera motion and natural reflections. The car drives forward with its front leading; let the camera follow behind rather than copying the reference photo's orientation. Keep the tires in contact with the road. Exactly one car; no visible melting or stretching, abrupt jump cuts, extra vehicles, people, or added text.”

Keep the train-window setup and give your car a warm daylight finale. Video prompt: “Begin from the supplied image of a miniature car on a passenger train's interior windowsill. Keep the model stationary as the landscape moves outside. The train enters a dark railway tunnel; the miniature's taillights glow and the camera slowly advances. Conceal the window frame, ledge, and scale change in darkness, then reveal the same vehicle at full size driving inside a road tunnel. As it approaches the exit, introduce warm sunlight gradually. Follow the car onto a quiet forest road at golden hour, with long shadows, natural paint reflections, and a smooth low rear tracking shot. Keep a single continuous-looking sequence, with no extra locations or montage. Preserve the car's silhouette, color, wheels, and lights. The car drives forward with its front leading, the camera follows behind, and the tires stay on the road. One car only. No visible body morphing, duplicate vehicles, skidding, sudden exposure jumps, or added text.”