Within the evolution of perceptual audio coding, there is a long history of exploiting techniques for joint coding of several audio channels of an audio program which are presented simultaneously. The paper describes how such techniques have progressed over time into the recent concept of spatial audio coding, as it is under standardization currently within the ISO/MPEG group. As a significant improvement over conventional techniques, this approach allows the representation of high quality multi-channel audio at bitrates of only 64kbit/s and below.