A method for decoding a data stream containing audio/video substreams and control substreams comprises buffering nodes having the possibility to buffer multiple data packets in the same buffer. This may be achieved by having separate parameters for the allocated buffer size and any stored packet. Thus, not only multiple packets may be stored in the buffering node, but also such node may exist while its buffer is empty, so that the node may be reused later. This is particularly useful for buffering and selectively accessing multiple audio packets in MPEG-4 audio nodes or sound nodes.