Abstract: Content to be encoded by an encoder may be defined by a sequence of values chosen from a set of possible values. Individual values of the set of possible values may have a probability of occurrence in the sequence of values. The content may form a sequence of frames, such as a video sequence. The encoder may allocate sets of buffers for storage of reference frames, frame contexts (including probability tables), and encoded frames. Individual context buffers (including individual probability tables) may be associated with individual reference frames. Encoding of a particular frame may be based on selecting a particular frame context from a context buffer, the particular probability table therein, and the content of the particular frame.