Determination of spatial audio parameter encoding and associated decoding

A directional parameter comprising azimuth and elevation values are received for sub-bands of an audio frame. The range of the directional parameter is partitioned, and a primary partition is then identified depending on the value of the directional parameter. The primary partition section is further sub-divided to yield a secondary partitioning of the range of the directional parameter, and one of the secondary partitions is identified. An embedded index value associated with the directional parameter is then obtained by combining the indices associated with the primary and secondary partitions. The primary partition may subdivide the azimuthal and elevational ranges into 8 quarter-hemispheres which may be enumerated using three bits, which represent spherical pseudo-triangles, and which may be sub-divided using the midpoints of the pseudo-triangle's edges (fig. 8). This sub-division may be done recursively. The secondary partition may be selected to minimise distance between its centre and the directional parameter's value.