把声音分成对白、环境、动作和音乐层。先让对白清楚，再用环境建立空间，用动作声强调剪辑点，最后以音乐控制情绪；每层都保留独立音量控制。 Separate dialogue, ambience, action, and music into functional layers. Make dialogue intelligible first, use ambience for space, action sounds for edit points, and music for emotion. Keep independent level control for every layer.