Use the uploaded reference image as the exact identity and appearance reference for the seated person. Create an 11-second ultra-realistic cinematic stunt video. The person from the reference image is sitting calmly on a black folding chair in the middle of a dry, open field with dusty uneven ground. Preserve his exact face, identity, hairstyle, skin tone, body proportions and overall appearance from the reference image. He wears an all-black outfit and black sunglasses. He sits confidently with one leg crossed over the other, looking toward the camera. 0–3 seconds: The person is already seated and remains calm and still. Keep his pose, position, chair and appearance consistent. 3–6 seconds: A RED luxury SUV suddenly approaches from behind him at high speed, kicking up natural dust. 6–8 seconds: The SUV hits the uneven ground and jumps into the air, moving toward the camera at an impressive cinematic angle. Make the jump physically realistic with believable vehicle weight, suspension and tire movement. 8–9 seconds: The SUV lands naturally with visible suspension compression and realistic dust reacting to the impact. 9–11 seconds: The SUV continues past the seated person while dust spreads naturally across the ground. The person remains completely calm and seated in the same position. Keep the uploaded reference image as the primary identity reference for the person throughout the entire video. Photorealistic professional-camera footage, natural sunlight, realistic shadows, realistic dust particles, cinematic depth of field, subtle motion blur, realistic physics, detailed 4K quality. Keep the camera perspective, composition, environment and person's position consistent throughout. No identity drift, no face changes, no body deformation, no pose changes, no extra people, no duplicate person, no changing clothes, no changing chair, no floating vehicle, no unrealistic physics, no scene changes, no text, no watermark, no CGI or cartoon appearance.