Reduction of spatial audio parameters
There is inter alia disclosed an apparatus for spatial audio encoding comprising: means for analysing a plurality of spatial audio parameter sets associated with a frame of one or more audio signals, wherein the plurality of spatial audio parameter sets are associated with a plurality of subframes, a plurality of frequency sub bands and a plurality of sound source directions for the frame of the one or more audio signals; and means for determining from the analysis of the plurality of spatial audio parameter sets at least one spatial audio parameter set for subframes of the frame of the one or more audio signals.
1 . An apparatus for spatial audio encoding comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least:
obtain a plurality of spatial audio parameter sets for a time-frequency tile of a frame of one or more audio signals, wherein a plurality of sound source directions are associated with the time frequency tile, and wherein each spatial audio parameter set is associated with a different sound source direction;
determine whether the plurality of sound source directions for the time frequency tile can be reduced by checking whether a direct-to-total energy ratio value of a spatial audio parameter set associated with a sound source direction is zero or close to zero; and
responsive to determining that the direct-to-total energy ratio value associated with the sound source direction is zero or close to zero, remove the spatial audio parameter set associated with the sound source direction from the plurality of audio parameters sets associated with the time frequency tile.
2 . The apparatus as claimed in claim 1 , wherein parameters of respective ones of the plurality of spatial audio parameter sets further comprise at least one of:
a direction vector comprising an azimuth value and an elevation value;
a surrounding coherence value; and
a spread coherence.
3 . The apparatus as claimed in claim 1 , wherein the initial number of spatial audio parameter sets for frequency sub bands of the frame is dependent on a target coding rate.
4 . A method for spatial audio encoding comprising:
obtaining a plurality of spatial audio parameter sets for a time-frequency tile of a frame of one or more audio signals, wherein a plurality of sound source directions are associated with the time frequency tile, and wherein each spatial audio parameter set is associated with a different sound source direction;
determining whether the plurality of sound source directions for the time frequency tile can be reduced by checking whether a direct-to-total energy ratio value of a spatial audio parameter set associated with a sound source direction is zero or close to zero; and
responsive to determining that the direct-to-total energy ratio value associated with the sound source direction is zero or close to zero, removing the spatial audio parameter set associated with the sound source direction from the plurality of audio parameters sets associated with the time frequency tile.
5 . The method as claimed in claim 4 , wherein parameters of respective ones of the plurality of spatial audio parameter sets further comprise at least one of:
a direction vector comprising an azimuth value and an elevation value;
a surrounding coherence value; and
a spread coherence.