What kind of text prompts work best with Seedance 2.0 for cinematic videos?

Effective Seedance 2.0 prompts describe the scene, subject, motion, lighting, and camera movement together. For example: "A woman walking through a sunlit forest, slow-motion, golden hour lighting, camera tracking forward, cinematic depth of field" produces much better results than a vague prompt like "woman in forest." Including motion cues (slow-motion, time-lapse, dolly shot), lighting descriptions (golden hour, studio lighting), and environment context (rainy city street, ocean waves) significantly improves Seedance output quality.