主权项 |
1. A method of perceived loudness processing of frequency domain audio data, the method comprising:
accepting blocks of frequency domain audio data comprising transform coefficients that result from applying a lapped transform on overlapping blocks of time samples of audio data, wherein one or more of the accepted blocks have a longest block size, wherein the accepted blocks not having the longest block size have at least one respective short block size shorter than the longest block size, and wherein the longest block size is a respective multiple of each of the one or more short block sizes, such that each of the accepted blocks has the longest block size or one of the one or more short block sizes; for a plurality of the accepted blocks having a particular short block size of the one or more short block sizes, combining the plurality of the accepted blocks having the particular short block size to form a block of frequency domain audio data having the longest block size; and carrying out perceived loudness processing of the accepted blocks, the perceived loudness processing being at the longest block size, including:
determining or accepting one or more sets of perceived loudness parameters of the accepted blocks of frequency domain audio data, or of delayed versions of the accepted blocks of frequency domain audio data, wherein each set of perceived loudness parameters comprises a respective set of perceived loudness parameter values determined at a set of critical bands, wherein each of the one or more sets of perceived loudness parameters of the accepted blocks having the particular short block size are determined at the longest block size, and wherein the one or more determined sets of perceived loudness parameters include at least one of:
a set of determined values of the critical band power spectrum of the accepted blocks or of the delayed versions of the accepted blocks of frequency domain audio data determined at the set of critical bands, anda set of determined values of the specific loudness of the accepted blocks or of the delayed versions of the accepted blocks of frequency domain audio data determined at the set of critical bands. |