Audio signal processing

An audio signal processing apparatus comprising:

• a receiver,
• wherein the receiver is arranged to receive a set of decimated subband signals at a decimated sample rate,
• wherein the set of decimated subband signals represents an audio signal at a first sampling frequency,
• wherein the set of decimated subband signals has a maximum frequency not exceeding a threshold frequency,
• wherein the threshold frequency is lower than half the first sampling frequency by a first frequency margin,
• wherein the decimated subband signals is frequency shifted to a frequency interval which is a multiple of a frequency interval from zero to half the decimated sample frequency and subsequently decimated by a decimation factor;
• and
• a synthesizer, wherein the synthesizer is arranged to synthesize an audio output signal at the first sampling frequency from the set of decimated subband signals, the synthesizer comprising:
• an upsampler, wherein the upsampler is arranged to upsample the decimated subband signals to the first sampling frequency so as to generate upsampled subband signals;
• a frequency shifter, wherein the frequency shifter is arranged to apply a reverse frequency shift to the upsampled subband signals, wherein the reverse frequency shift for a first decimated subband signal results in the first decimated subband signal being shifted to a frequency range corresponding to the frequency range of the decimated subband signal in the audio signal prior to frequency shifting and decimation;
• a subband merger, wherein the subband merger is arranged to merge, following the application of the reverse frequency shift, the upsampled subband signals to generate a merged audio signal;
• a set of subband filters, wherein the subband filters arranged to filter the upsampled processed subband signals to attenuate upsample alias spectra for each subband, wherein each filter of the set of filters has a bandwidth not exceeding half the first sampling frequency.

A method of audio signal processing comprising:

• receiving an audio signal sampled at a first sampling frequency, wherein the audio signal has a maximum frequency not exceeding a threshold frequency, wherein the threshold frequency is lower than half the first sampling frequency by a first frequency margin;
• a filter bank,
• wherein the filter bank generates subband signals for the audio signal,
• wherein the filter bank comprises a set of overlapping sub-filters providing a set of subbands;
• applying a frequency shift to at least one subband of the set of subbands;
• decimating the subband signals by a decimation factor resulting in decimated subband signals having a decimated sample frequency,
• wherein the decimated sample frequency is equal to the first sampling frequency divided by the decimation factor,
• wherein the decimated sample frequency being at least twice a bandwidth of each of the overlapping sub-filters;
• and
• wherein the frequency shift for a subband shifts the subband to a frequency interval which is a multiple of a frequency interval from zero to half the decimated sample frequency.