A method for composing an intermediate representation of a video sequence from at least two representations of the video sequence of different quality, spatially partitioned into a plurality of tiles, each tile being associated with compressed data in the representation. The tiles can be decoded independently of each other and, at a given time, the different representations are identically partitioned. When the sequence is being played on a playback device, the method includes: obtaining a context data item representing the activity of viewers located near the playback device; selecting, on the basis of the context data item obtained and at least one maximum rate criterion for the intermediate representation resulting from the composition, a set of tiles forming a partition of the sequence from the at least two representations; and composing an intermediate representation of the video sequence from the compressed data associated with the selected set of tiles.