Multi-band noise gate

1. A method for processing a plurality of audio signals comprising:

receiving the plurality of audio signals in a frequency domain, each audio signal having been captured by a respective audio recording device and transformed from a time domain to the frequency domain, wherein the frequency domain is divided into a plurality of frequency bins that collectively span the frequency domain and each audio signal is represented by a plurality of magnitudes respectively corresponding to the plurality of frequency bins of the frequency domain, wherein each magnitude represents a magnitude of the audio signal at the corresponding frequency bin;

determining an overall attenuation multiplier based on the plurality of audio signals and an overall lookup table that relates decibel values to different overall attenuation multipliers;

determining an attenuation vector comprising a plurality of bin-specific attenuation multipliers, each bin-specific attenuation multiplier respectively corresponding to a different frequency bin of the plurality of frequency bins, wherein determining the attenuation vector includes:

separating the plurality of audio signals into a plurality of frequency bands, wherein the plurality of frequency bands collectively span the frequency domain, and each respective frequency band is spanned by two or more frequency bins of the plurality of the frequency bins; and

for each respective frequency band:

obtaining a band-specific lookup table from a plurality of band-specific lookup tables, wherein the band-specific lookup table corresponds to the respective frequency band and relates decibel levels to bin-specific attenuation values; and

for each respective frequency bin within the frequency band:

 determining an approximate decibel level at the respective frequency bin based on the plurality of audio signals;

 determining a bin-specific attenuation multiplier corresponding to the respective frequency bin based on an average decibel level at the respective frequency bin and the respective band-specific lookup table; and

 inputting the bin-specific attenuation multiplier to the attenuation vector;

scaling each bin-specific attenuation value in the attenuation vector with the overall attenuation multiplier; and

editing each of the audio signals based on the scaled bin-specific attenuation values in the attenuation vector to produce an attenuated audio signal.

2. The method of claim 1 , wherein determining the approximate decibel level at the respective frequency bin comprises:

for each audio signal, determining a complex magnitude at the respective frequency bin,

calculating a root mean square value based on the complex magnitudes at the respective frequency bin; and

calculating a logarithm of the root mean square value to obtain the approximate decibel level at the respective frequency bin.

3. The method of claim 1 , wherein each bin-specific attenuation multiplier is between zero and one.

4. The method of claim 1 , wherein the method is executed by a processor of an image capture device and each band-specific lookup table is configured for a make and model of the image capture device.

5. The method of claim 1 , wherein determining the overall attenuation multiplier comprises:

determining an approximate decibel value of all of the audio signals of the plurality of audio signals; and

determining the overall attenuation multiplier from the overall lookup table using the approximate decibel value of all of the audio signals.

6. The method of claim 5 , wherein the overall attenuation multiplier is a value between zero and one.

7. The method of claim 6 , wherein scaling each of the bin-specific attenuation values based on the overall attenuation value includes multiplying the bin-specific attenuation values by the overall attenuation value.

8. The method of claim 5 , further comprising:

smoothing the approximate decibel value of all of the audio signals of the plurality of audio signals based on a previously determined approximate decibel level.

9. The method of claim 1 , wherein editing each of the audio signals comprises:

for each respective audio signal and for each respective frequency bin, multiplying a magnitude of the respective audio signal at the respective frequency bin by the scaled bin-specific attenuation value.

10. The method of claim 1 , further comprising:

smoothing the scaled bin-specific attenuation values in the attenuation vector.

11. The method of claim 1 , further comprising:

performing an inverse transform of the attenuated audio signal to the time domain; and

storing the inverse transformed attenuated audio signal.

12. A method for processing a plurality of audio signals comprising:

receiving the plurality of audio signals in a frequency domain, each audio signal having been captured by a respective audio recording device and transformed from a time domain to the frequency domain, wherein the frequency domain is divided into a plurality of frequency bins that collectively span the frequency domain and each audio signal is represented by a plurality of magnitudes respectively corresponding to the plurality of frequency bins of the frequency domain, wherein each magnitude represents a magnitude of the audio signal at the corresponding frequency bin;

determining an attenuation vector comprising a plurality of bin-specific attenuation multipliers, each bin-specific attenuation multiplier respectively corresponding to a different frequency bin of the plurality of frequency bins, wherein determining the attenuation vector includes:

separating the plurality of audio signals into a plurality of frequency bands, wherein the plurality of frequency bands collectively span the frequency domain, and each respective frequency band is spanned by two or more frequency bins of the plurality of the frequency bins; and

for each respective frequency band:

obtaining a band-specific lookup table from a plurality of band-specific lookup tables, wherein the band-specific lookup table corresponds to the respective frequency band and relates decibel levels to bin-specific attenuation values; and

for each respective frequency bin within the frequency band:

determining an approximate decibel level at the respective frequency bin based on the plurality of audio signals;

determining a bin-specific attenuation multiplier corresponding to the respective frequency bin based on an average decibel level at the respective frequency bin and the respective band-specific lookup table; and

inputting the bin-specific attenuation multiplier to the attenuation vector; and

editing each of the audio signals based on the bin-specific attenuation values in the attenuation vector to produce an attenuated audio signal.

13. The method of claim 12 , wherein determining the approximate decibel level at the respective frequency bin comprises:

for each audio signal, determining a complex magnitude at the respective frequency bin;

calculating a root mean square value based on the complex magnitudes at the respective frequency bin; and

calculating a logarithm of the root mean square value to obtain the approximate decibel level at the respective frequency bin.

14. The method of claim 12 , wherein each bin-specific attenuation multiplier is between zero and one.

15. The method of claim 12 , wherein the method is executed by a processor of an image capture device and each band-specific lookup table is configured for a make and model of the image capture device.

16. The method of claim 12 , wherein each band-specific lookup table is based on a root mean square level of a noise floor, averaged over time.

17. The method of claim 12 , further comprising:

performing an inverse transform of the attenuated audio signal to the time domain; and

storing the inverse transformed attenuated audio signal.

18. The method of claim 17 , wherein the inverse transform is an inverse Fast Fourier Transform (IFFT) or an inverse Laplace transform.