Nova Patents
EP0446037B1

Hybrid perceptual audio coding

Abstract

This record has no abstract on file.

EP0446037B1, drawing sheet 1
Sheet 1 of 14

Term

Term ended

Expired 6 March 2011, 15.6 years ago.

  1. Priority
  2. Filed
  3. Granted
  4. Expired
  5. Today

8 claims: 1 independent, 7 dependent

  1. 1
    A method of processing an ordered time sequence of audio signals partitioned into a set of ordered blocks, each of the blocks having a discrete frequency spectrum including a first set of frequency coefficients, the method including, for each of the blocks, the steps of:filtering the first set of frequency coefficients into a plurality of frequency groups;determining a tonality metric for each frequency group, the tonality metric corresponding to the degree to which said time sequence of audio signals of the frequency group includes a tone-like quality;determining a noise masking threshold for the frequency coefficients in each frequency group based on the tonality metric, the noise masking threshold indicating the amount of quantization noise that can be introduced into a frequency group without affecting its tone-like quality;and quantizing at least one frequency coefficient in at least one of the frequency groups, the quantizing being based upon the corresponding noise masking threshold, CHARACTERISED IN THAT the filtering step includes filtering the first set of frequency coefficients into a plurality of frequency bands with a bank of filters (201, 202, 203) of a first type, each of the frequency bands representing a contiguous range of frequencies;and filtering at least one of the frequency bands into a plurality of frequency groups with at least one filter (210, 211, 212, 213) of a second type, each of the frequency groups including at least one frequency coefficient, wherein the filtering step filters said frequency coefficients with differing time and frequency resolutions at different frequencies to substantially approximate human auditory response.