NOISE-REDUCING DIRECTIONAL MICROPHONE ARRAY

• US 20130010982A1
• Filed: 08/28/2012
• Published: 01/10/2013
• Est. Priority Date: 02/05/2002
• Status: Active Grant

• Current Assignee: mh acoustics LLC
• Original Assignee: mh acoustics LLC
• Inventors: Elko, Gary W., Gaensler, Tomas Fritz, Meyer, Jens M.
• Granted Patent: US 9,301,049 B2
• Time in Patent Office: Days
• US Class Current: 381/94.2
• CPC Class Codes: G10L 2021/02166 Microphone arrays; Beamforming
• CPC Class Codes: G10L 21/0216 characterised by the method used for estimating noise
• CPC Class Codes: G10L 21/0264 characterised by the type of parameter measurement, e.g. correlation techniques, zero crossing techniques or predictive techniques
• CPC Class Codes: H04R 1/326 for microphones H04R1/34 and H04R1/40 take precedence
• CPC Class Codes: H04R 2410/01 Noise reduction using microphones having different directional characteristics
• CPC Class Codes: H04R 2410/07 Mechanical or electrical reduction of wind noise generated by wind passing a microphone
• CPC Class Codes: H04R 2430/20 Processing of the output signals of the acoustic transducers of an array for obtaining a desired directivity characteristic H04R2203/12 takes precedence
• CPC Class Codes: H04R 2430/21 Direction finding using differential microphone array [DMA]
• CPC Class Codes: H04R 2430/23 Direction finding using a sum-delay beam-former
• CPC Class Codes: H04R 25/407 Circuits for combining signals of a plurality of transducers
• CPC Class Codes: H04R 3/005 for combining the signals of two or more microphones specially adapted for hearing aids H04R25/407
• CPC Class Codes: H04R 3/04 for correcting frequency response

Abstract

In one embodiment, a directional microphone array having (at least) two microphones generates forward and backward cardioid signals from two (e.g., omnidirectional) microphone signals. An adaptation factor is applied to the backward cardioid signal, and the resulting adjusted backward cardioid signal is subtracted from the forward cardioid signal to generate a (first-order) output audio signal corresponding to a beampattern having no nulls for negative values of the adaptation factor. After low-pass filtering, spatial noise suppression can be applied to the output audio signal. Microphone arrays having one (or more) additional microphones can be designed to generate second- (or higher-) order output audio signals.

Claims

1. A method for processing audio signals, comprising:

(a) generating first and second cardioid signals (e.g., 1509(1) and 1509(2)) from first and second microphone signals (e.g., 1503(1) and 1503(2));

(b) determining whether any of wind noise, thermal noise, and circuit noise are present in the first and second microphone signals;

(c) generating a first adaptation factor (e.g., β

), wherein;

the first adaptation factor is constrained to a first range of values if the determination of step (b) is that wind noise, thermal noise, and circuit noise are not present in the first and second microphone signals; and

the first adaptation factor is constrained to a second range of values, different from the first range of values, if the determination of step (b) is that at least one of wind noise, thermal noise, and circuit noise is present in the first and second microphone signals;

(d) applying the first adaptation factor (e.g., β

) to adjust amplitude of the second cardioid signal (e.g., 1509(2)) to generate an adapted second cardioid signal (e.g., 1513); and

(e) combining the first cardioid signal (e.g., 1509(1)) and the adapted second cardioid signal (e.g., 1513) to generate a first output audio signal (e.g., 1515), wherein, when the first adaptation factor is constrained to the second range of values, the first output audio signal corresponds to a first beampattern having no nulls.

2. The method of claim 1, wherein:

the first cardioid signal (e.g., 1509(1)) is a forward cardioid signal;

the second cardioid signal (e.g., 1509(2)) is a backward cardioid signal;

the adapted second cardioid signal (e.g., 1513) is subtracted from the forward cardioid signal (e.g., 1509(1)) to generate the first output audio signal (e.g., 1515); and

the first beampattern has no nulls for negative values of the first adaptation factor (e.g., β

).

3. The method of claim 1, wherein the first adaptation factor (e.g., β

) is generated based on the second cardioid signal (e.g., 1509(2)) and the first output audio signal (e.g., 1515).

4. The method of claim 3, further comprising the steps of:

determining whether a nearfield source is present by comparing output levels from the first and second cardioid signals; and

decreasing an update step-size used in generating adjustments for the first adaptation factor (e.g., β

) to reduce adaptation speed for generating the first output audio signal (e.g., 1515), if the nearfield source is determined to be present.

5. The method of claim 1, wherein:

steps (b), (c), (d), and (e) are implemented in a subband domain such that the first adaptation factor for a first subband is constrained to the first range of values while the first adaptation factor for a second subband is concurrently constrained to the second range of values.

6. The method of claim 1, further comprising:

(f) applying noise suppression processing to the first output audio signal (e.g., 1515) to generate a noise-suppressed output audio signal, wherein the noise suppression processing is controlled based on the first adaptation factor (e.g., β

) and step (f) comprises;

(1) generating a difference-signal power based on the first and second microphone signals (e.g., 1503(1) and 1503(2));

(2) generating a sum-signal power based on first and second microphone signals (e.g., 1503(1) and 1503(2));

(3) generating a power ratio based on the difference-signal power and the sum-signal power;

(4) generating a suppression value based on the power ratio; and

(5) applying the noise suppression processing to the first output audio signal (e.g., 1515) based on the suppression value to generate the noise-suppressed output audio signal.

7. The method of claim 6, wherein:

the suppression processing is based on both the power ratio and the first adaptation factor (e.g., β

); and

step (c) comprises generating the first adaptation factor (e.g., β

) based on the power ratio.

8. The method of claim 7, wherein:

step (f) is implemented in a subband domain to generate a suppression level for each subband; and

steps (b), (c), (d), and (e) are implemented in the subband domain such that the first adaptation factor for a first subband is constrained to the first range of values while the first adaptation factor for a second subband is concurrently constrained to the second range of values.

9. The method of claim 1, wherein step (a) comprises filtering at least one of the first and second microphone signals (e.g., 1503(1) and 1503(2)) based on a first weight factor prior to generating the first and second cardioid signals (e.g., 1509(1) and 1509(2)).

10. The method of claim 9, wherein the first weight factor is generated by:

(1) selecting one microphone signal (e.g., 1503(1) and 1503(2)) as a reference signal and another microphone signal (e.g., 1503(1) and 1503(2)) as a calibrated signal;

(2) determining an envelope level for each of the first and second microphone signals (e.g., 1503(1) and 1503(2));

(3) applying a calibration weight factor to the envelope level of the calibrated signal to generate an adjusted calibration-signal envelope level;

(4) updating the calibration weight factor to decrease a difference between the envelope level of the reference signal and the adjusted calibration-signal envelope level; and

(5) applying the updated calibration weight factor to a first low-pass filter (e.g., 606, 616, 716, 1516) to generate the first weight factor for the filtering of step (a).

11. The method of claim 10, further comprising:

(6) determining whether a nearfield source is present, wherein updating of the first weight factor based on the updated calibration weight factor is suspended if any of the wind noise, the thermal noise, and the circuit noise are determined to be present or if the nearfield source is determined to be present.

12. The method of claim 1, wherein:

the first output audio signal (e.g., 1515) is a first-order signal; and

further comprising;

(f) generating third and fourth cardioid signals (e.g., C_F2 and C_B2) from one of the first and second microphone signals (e.g., p₂) and a third microphone signal (e.g., p₃);

(g) generating a second adaptation factor (e.g., β

₁);

(h) applying the second adaptation factor (e.g., β

₁) to the fourth cardioid signal (e.g., C_B2) to generate an adapted fourth cardioid signal;

(i) combining the third cardioid signal (e.g., C_F2) and the adapted fourth cardioid signal to generate a second, first-order output audio signal corresponding to a second beampattern having no nulls for at least one value of the second adaptation factor (e.g., β

₁); and

(j) combining the first output audio signal (e.g., 1515) and the second output audio signal to form a second-order output audio signal corresponding to a third beampattern having no nulls for at least one value of the first adaptation factor and at least one value of the second adaptation factor (e.g., β

₁).

13. The method of claim 12, wherein step (j) comprises:

(1) generating first and second second-order cardioid signals from the first and second first-order output audio signals;

(2) generating a third adaptation factor (e.g., β

₂);

(3) applying the third adaptation factor (e.g., β

₂) to the first second-order cardioid signal to generate an adapted first second-order cardioid signal;

(4) combining the second second-order cardioid signal and the adapted first second-order cardioid signal to generate the second-order output audio signal.

14. The method of claim 1, wherein:

if the wind noise, the thermal noise, and the circuit noise are determined not to be present, then the first adaptation factor is set equal to a specified value in the first range of values; and

if any of the wind noise, the thermal noise, and the circuit noise are determined to be present, then the first adaptation factor is adaptively generated based on the second cardioid signal (e.g., 1509(2)) and the first output audio signal (e.g., 1515) to be in the second range of values.

15. An audio system (e.g., 1500, 1600, 1700, 1800) for processing audio signals, comprising:

(a) means (e.g., 1508) for generating first and second cardioid signals (e.g., 1509(1) and 1509(2)) from first and second microphone signals (e.g., 1503(1) and 1503(2));

(b) means for determining whether any of wind noise, thermal noise, and circuit noise are present in the first and second microphone signals;

(c) an adaptation block (e.g., 1510) adapted to generate a first adaptation factor (e.g., β

), wherein;

the first adaptation factor is constrained to a first range of values if the determination of means (b) is that wind noise, thermal noise, and circuit noise are not present in the first and second microphone signals; and

the first adaptation factor is constrained to a second range of values, different from the first range of values, if the determination of means (b) is that at least one of wind noise, thermal noise, and circuit noise is present in the first and second microphone signals;

(d) a multiplication node (e.g., 1512) adapted to apply the first adaptation factor (e.g., β

) to adjust amplitude of the second cardioid signal (e.g., 1509(2)) to generate an adapted second cardioid signal (e.g., 1513); and

(e) a combiner (e.g., 1514) adapted to combine the first cardioid signal (e.g., 1509(1)) and the adapted second cardioid signal (e.g., 1513) to generate a first output audio signal (e.g., 1515), wherein, when the first adaptation factor is constrained to the second range of values, the first output audio signal corresponds to a first beampattern having no nulls.