Determination of the significance of spatial audio parameters and associated encoding

A method implemented by an audio encoder for spatial audio encoding comprising:

• receiving or determining for one or more audio signals, spatial audio parameters on a sub band basis for providing spatial audio reproduction, wherein the spatial audio parameters comprise a surround coherence value for each sub band of a plurality of sub-bands of a frame;
• receiving an energy ratio value for each of the plurality of sub bands;
• determining a first significance measure for the coherence values of the plurality of sub bands of the frame by (i) determining a proportion of coherent non-directional energy for each of the plurality of sub bands, wherein the proportion of coherent non-directional energy for a sub band is determined as a non-directional energy ratio for the sub band multiplied by the surround coherence value for the sub band, wherein the non-directional energy ratio is related to a ratio of energy in the sub band remaining after the energy ratio value for the sub band has been removed, and (ii) summing, for all the plurality of sub bands, the determined proportion of coherent non-directional energy;
• determining a second significance measure for the plurality of sub bands of the frame by being caused to (i) multiply, for each sub band, the proportion of coherent non-directional energy by the non-directional energy ratio, (ii) sum, for all the plurality of sub bands, the multiple of the proportion of coherent non-directional energy and the non-directional energy ratio for each sub band, and (iii) normalize the summed multiple of the proportion of coherent non-directional energy and the non-directional energy ratio for each sub band by a summation of the non-directional energy ratio for each sub band;
• determining a significance measure for the plurality of sub bands of the frame as the maximum of the first significance measure and the second significance measure for the coherence values of the plurality of sub bands of the frame;
• determining whether to encode the coherence values of the plurality of sub bands of the frame based on a comparison of the significance measure to a threshold value associated with the frame;
• and
• encoding the surround coherence values of the plurality of sub bands of the frame that were determined to be encoded.

A computer program product embodied on a non-transitory computer readable medium, comprising computer program code configured to, when executed on at least one processor, cause an apparatus of an audio encoder to:

• receive or determine for one or more audio signals, spatial audio parameters on a sub band basis for providing spatial audio reproduction, wherein the spatial audio parameters comprise a surround coherence value for each sub band of a plurality of sub-bands of a frame;
• receive an energy ratio value for each of the plurality of sub bands;
• determine a first significance measure for the surround coherence values of the plurality of sub bands of the frame by (i) determining a proportion of coherent non-directional energy for each of the plurality of sub bands, wherein the proportion of coherent non-directional energy for a sub band is determined as a non-directional energy ratio for the sub band multiplied by the surround coherence value for the sub band, wherein the non-directional energy ratio is related to a ratio of energy in the sub band remaining after the energy ratio value for the sub band has been removed, and (ii) summing, for all the plurality of sub bands, the determined proportion of coherent non-directional energy;
• determine a second significance measure for the plurality of sub bands of the frame by being caused to (i) multiply, for each sub band, the proportion of coherent non-directional energy by the non-directional energy ratio, (ii) sum, for all the plurality of sub bands, the multiple of the proportion of coherent non-directional energy and the non-directional energy ratio for each sub band, and (iii) normalize the summed multiple of the proportion of coherent non-directional energy and the non-directional energy ratio for each sub band by a summation of the non-directional energy ratio for each sub band;
• determine a significance measure for the plurality of sub bands of the frame as the maximum of the first significance measure and the second significance measure for the coherence values of the plurality of sub bands of the frame;
• determine whether to encode the surround coherence values of the plurality of sub bands of the frame based on a comparison of the significance measure to a threshold value associated with the frame;
• and
• encode the surround coherence values of the plurality of sub bands of the frame that were determined to be encoded.