Generating an audio signal from multiple microphones based on uncorrelated noise detection

A non-transitory computer readable storage medium in an audio capture system having multiple microphones, the non-transitory computer-readable storage medium storing instructions for generating an output audio signal that when executed by a processor cause the processor to perform steps including:

• receiving at least a first audio signal and a second audio signal from the plurality of microphones;
• generating a first plurality of frequency sub-band signals from the first audio signal corresponding to a plurality of frequency sub-bands and generating a second plurality of frequency sub-band signals from the second audio signal corresponding to the plurality of frequency sub-bands;
• for each of the first and second pluralities of frequency sub-band signals, applying a frequency band-dependent offset to generate a first plurality of offset frequency sub-band signals from the first plurality of frequency sub-band signals and a second plurality of offset frequency sub-band signals from the second plurality of frequency sub-band signals;
• determining an overall correlation metric between the first plurality of offset frequency sub-band signals and the second plurality of offset frequency sub-band signals;
• responsive to the overall correlation metric exceeding a first predefined threshold, processing the audio signals according to a correlated audio signal processing algorithm to generate an output audio signal;
• and
• responsive to the overall correlation metric not exceeding the first predefined threshold, processing the audio signals according to an uncorrelated audio signal processing algorithm to generate the output audio signal.

An audio capture device comprising:

• a plurality of microphones;
• a processor;
• and
• a non-transitory computer readable storage medium storing instructions for generating an output audio signal, the instructions when executed by the processor cause the processor to perform steps including:
• receiving at least a first audio signal and a second audio signal from the plurality of microphones;
• generating a first plurality of frequency sub-band signals from the first audio signal corresponding to a plurality of frequency sub-bands and generating a second plurality of frequency sub-band signals from the second audio signal corresponding to the plurality of frequency sub-bands;
• for each of the first and second pluralities of frequency sub-band signals, applying a frequency band-dependent offset to generate a first plurality of offset frequency sub-band signals from the first plurality of frequency sub-band signals and a second plurality of offset frequency sub-band signals from the second plurality of frequency sub-band signals;
• determining an overall correlation metric between the first plurality of offset frequency sub-band signals and the second plurality of offset frequency sub-band signals;
• responsive to the overall correlation metric exceeding a first predefined threshold, processing the audio signals according to a correlated audio signal processing algorithm to generate an output audio signal;
• and
• responsive to the overall correlation metric not exceeding the first predefined threshold, processing the audio signals according to an uncorrelated audio signal processing algorithm to generate the output audio signal.