The present invention concerns a method for generating media files from video sequences, the method comprising by a server: —obtaining from the video sequences, video data composed of a plurality of samples; —generating a video track based on the obtained video data, each video track comprises samples of a video sequence, and the video track is associated with descriptive metadata, the descriptive metadata comprises: a spatial information related to one or more samples of the associated video track; and a composition information for organizing generated video tracks to get a full picture when displayed by a client; and —generating media files including the generated video tracks.