Patent Grant
8208656
1. Field of the Invention
The invention relates to an array microphone system, and more particularly to an array microphone system including two omni-directional microphones to receive sound in a cone-shaped beam.
2. Description of the Related Art
A microphone array is capable of clearly receiving sound from a particular direction while excluding surrounding noise, and is often applied in high-quality audio recorders or communications devices.
The microphone array 70 functions very well. However, the uni-directional microphone 710 included in the microphone array 70 has the problems of being difficult to manufacture because of its design and high costs.
The microphone array 80 has no uni-directional microphones and thus, does not have the accompanying problems of uni-directional microphones. However, sounds coming from the rear of the microphone array 80 can not be excluded due to the pie-shaped beam 830. Thus, limiting actual application of the microphone array 80 to less than that of the microphone array 70.
The invention provides an array microphone system including two omni-directional microphones to receive sound in a cone-shaped beam, thus avoiding the described problems. The array microphone system in accordance with an exemplary embodiment of the invention includes a first omni-directional microphone, a second omni-directional microphone, a gain control, and a beam former. The first omni-directional microphone faces a first direction. The second omni-directional microphone faces a second direction opposing the first direction. When receiving sound, the first omni-directional microphone and the second omni-directional microphone respectively generate a first signal and a second signal. The gain control amplifies the second signal to transform into a third signal, wherein strength of the third signal is equal to that of the first signal when the sound comes from the first direction. The beam former separates an in-beam sound signal and an out-beam sound signal from the first signal and the third signal.
In another exemplary embodiment, the array microphone system further includes a first voice activity detector and a second voice activity detector controlling an operation of the beam former based on the first signal and the third signal.
In yet another exemplary embodiment, the operation of the first voice activity detector and the second voice activity detector is mutually exclusive.
The invention also provides a method for determining a gain of a gain control. The method in accordance with an exemplary embodiment of the invention includes the steps of: first, setting a first omni-directional microphone and a second omni-directional microphone in different positions; second, generating a first sound from a first direction; third, obtaining a first ratio of a first signal from the first omni-directional microphone to a second signal from the second omni-directional microphone; fourth, generating a second sound from a second direction opposing the first direction when the first ratio of the first signal to the second signal exceeds a first predetermined value; fifth, obtaining a second ratio of a third signal from the second omni-directional microphone to a fourth signal from the first omni-directional microphone; and sixth, setting the first ratio as the gain when the second ratio of the third signal to the fourth signal exceeds a second predetermined value.
In another exemplary embodiment, the method for determining a gain of a gain control further includes the step of resetting the first omni-directional microphone and the second omni-directional microphone in different positions when the first ratio of the first signal to the second signal does not exceed the first predetermined value.
In yet another exemplary embodiment, the method for determining a gain of a gain control further includes the step of resetting the first omni-directional microphone and the second omni-directional microphone in different positions when the second ratio of the third signal to the fourth signal does not exceed the second predetermined value.
The invention also provides a method for determining a gain of a gain control. The method in accordance with an exemplary embodiment of the invention includes the steps of: first, setting a first omni-directional microphone and a second omni-directional microphone in different positions; second, generating a first sound from a first direction; third, obtaining a first ratio of a first signal from the first omni-directional microphone to a second signal from the second omni-directional microphone; fourth, generating a second sound from a second direction opposing the first direction when the first ratio of the first signal to the second signal exceeds a first predetermined value; fifth, obtaining a second ratio of a third signal from the second omni-directional microphone to a fourth signal from the first omni-directional microphone; sixth, generating third sound from a third direction perpendicular to the first direction and the second direction when the second ratio of the third signal to the fourth signal exceeds the second predetermined value; seventh, obtaining a third ratio of a fifth signal from the second omni-directional microphone to a sixth signal from the first omni-directional microphone; and eighth, setting the first ratio as the gain when the third ratio of the fifth signal to the sixth signal exceeds a third predetermined value.
In another exemplary embodiment, the method for determining a gain of a gain control further includes the step of resetting the first omni-directional microphone and the second omni-directional microphone in different positions when the first ratio of the first signal to the second signal does not exceed the first predetermined value.
In yet another exemplary embodiment, the method for determining a gain of a gain control further includes the step of resetting the first omni-directional microphone and the second omni-directional microphone in different positions when the second ratio of the third signal to the fourth signal does not exceed the second predetermined value.
A detailed description is given in the following embodiments with reference to the accompanying drawings.
The invention can be more fully understood by reading the subsequent detailed description and examples with references made to the accompanying drawings, wherein:
The following description is of the best-contemplated mode of carrying out the invention. This description is made for the purpose of illustrating the general principles of the invention and should not be taken in a limiting sense. The scope of the invention is best determined by reference to the appended claims.
Referring to
Referring to
Referring to
The first VAD 151 and the second VAD 152 receives the signals S1 and S3 from the main microphone 111 and the gain control 120, and provides voice detection signals S8 and S9 corresponding to the received sound for controlling the operation of the beam former 140. The operation of the first VAD 151 and that of the second VAD 152 are mutually exclusive. If the first VAD 151 is on, then the second VAD 152 will be off. On the other hand, the first VAD 151 will be off if the second VAD 152 is on. When the in-beam sound 20 is received by the array microphone system 110, the first VAD 151 is on and the second VAD 152 is off. When there is no in-beam sound 20 but out-beam sound 21, 22, or 23, the first VAD 151 is off and the second VAD 152 is on.
The beam former 140 receives the signal S1 from the main microphone 111, the amplified signal S3 from the gain control 120, and the voice detection signals S8 and S9 from the first VAD 151 and the second VAD 152, and separates an in-beam sound signal S7 and an out-beam sound signal S5 from the signal S1 and the amplified signal S3.
The operation of the array microphone system 110 is introduced in detail in the following three cases:
In the first case, there is no out-beam sound 21, 22, and 23, and the array microphone system 110 only receives the in-beam sound 20. The main microphone 111 and the reference microphone 112 respectively generate signals S1 and S2, both of which correspond to the in-beam sound 20. Because the in-beam sound 20 comes from the front and the reference microphone 112 faces the rear, the signal S2 generated by the reference microphone 112 is much weaker than the signal S1 generated by the main microphone 111 (S2<<S1). The gain control 120 amplifies the signal S2 and outputs an amplified signal S3 wherein S3≈S1.
The first VAD 151 and the second VAD 152 receives the signals S1 and S3 from the main microphone 111 and the gain control 120. After calculation, the first VAD 151 is on and the second VAD 152 is off. The first VAD 151 outputs voice detection signals S8 and S9 to the beam former 140.
In the beam former 140, an adaptive filter 141a receives the amplified signal S3 from the gain control 120, the voice detection signal S9 from the first VAD 151, and a feedback signal S5 from a summer 141b, calculates the parameters for the linear correlation between the signals S1 and S3, and provides a filtered signal S4 which is approximately equal to the amplified S3 (i.e. S4≈S3). As described, S3≈S1. Thus, S4≈S1. The signal S4 is then subtracted from the signal S1 by the summer 141b to obtain a signal S5. The signal S5 is very small (≈0) because S1≈S4, which is reasonable because the signal S5, as described, corresponds to the out-beam sound. In the first case, there is no out-beam sound.
Another adaptive filter 142a receives the signal S5 from the summer 141b, the voice detection signal S8 from the first VAD 151, and a feedback signal S7 from a summer 142b, calculates the parameters for the linear correlation between the signals S1 and S5, and provides a signal S6. Because the signal S5 input to the adaptive filter 142a is very small, the signal S6 output from the adaptive filter 142a is very small. That is, S6≈0. The signal S6 is then subtracted from the signal S1 by the summer 142b to obtain a signal S7. The signal S7, corresponding to the in-beam sound 20, is approximately equal to the signal S1 (S7≈S1) because S6≈0.
In the first case, the array microphone system 110 only receives the in-beam sound 20, and separately outputs two signals S7 and S5, wherein the signal S7 corresponds to the in-beam sound 20 and the signal S5 corresponding to the out-beam sound is very small.
In the second case, there is no in-beam sound 20, and the array microphone system 110 receives out-beam sound 21, 22, and/or 23. For simplification, there is only the out-beam sound 21 coming from the rear. The main microphone 111 and the reference microphone 112 respectively generate signals S1 and S2, both of which correspond to the out-beam sound 21. Because the out-beam sound 21 comes from the rear and the main microphone 111 faces the front, the signal S1 generated by the main microphone 111 is much weaker than the signal S2 generated by the reference microphone 112 (S1<<S2). The gain control 120 amplifies the signal S2 and outputs an amplified signal S3 wherein S3>S2. Note that S3>>S1 because S2>>S1.
The first VAD 151 and the second VAD 152 receives the signals S1 and S3 from the main microphone 111 and the gain control 120. After calculation, the first VAD 151 is off and the second VAD 152 is on. The second VAD 152 outputs the voice detection signals S9 and S8 to the beam former 140.
In the beam former 140, an adaptive filter 141a receives the amplified signal S3 from the gain control 120, the voice detection signal S9 from the second VAD 152, and a feedback signal S5 from the summer 141b, calculates the parameters for the linear correlation between the signals S1 and S3, and provides a filtered signal S4. Note that the signal S4, deriving from the signal S3, is much greater than the signal S1 because S3>>S1. The signal S4 is then subtracted from the signal S1 by the summer 141b to obtain a signal S5 corresponding to the out-beam sound 21.
The adaptive filter 142a receives the signal S5 from the summer 141b and the voice detection signal S8 from the second VAD 152, calculates the parameters for the linear correlation between the signals S1 and S5, and provides a signal S6 which is approximately equal to the signal S1. The signal S6 is then subtracted from the signal S1 by the summer 142b to obtain a signal S7. The signal S7 corresponding to the in-beam sound is very small (S7≈0) because S6≈S1.
In the second case, the array microphone system 110 only receives the out-beam sound, and separately outputs two signals S7 and S5, wherein the signal S7 corresponding to the in-beam sound is very small and the signal S5 corresponds to the out-beam sound.
In the third case, the array microphone system 110 simultaneously receives the in-beam sound 20 from the front and the out-beam sound 21 from the rear. The main microphone 111 and the reference microphone 112 generate signals S1 and S2, respectively. The signal S1 generated by the main microphone 111 contains an in-beam part and an out-beam part. Because the out-beam sound 21 comes from the rear and the main microphone 111 faces the front, the out-beam part of the signal S1 is small. Similarly, the signal S2 generated by the reference microphone 112 contains an in-beam part and an out-beam part. Because the in-beam sound 20 comes from the front and the reference microphone 112 faces the rear, the in-beam part of the signal S2 is small. The signal S2 is then amplified by the gain control 120 into a signal S3 wherein the in-beam part of the amplified signal S3 is approximately equal to that of the signal S1.
The first VAD 151 and the second VAD 152 receive the signals S1 and S3 from the main microphone 111 and the gain control 120. After calculation, the first VAD 151 is on and the second VAD 152 is off. The first VAD 151 outputs the voice detection signals S9 and S8 to the beam former 140.
In the beam former 140, the adaptive filter 141a receives the amplified signal S3 from the gain control 120, the voice detection signal S9 from the first VAD 151, and a feedback signal S5 from the summer 141b, calculates the parameters for the linear correlation between the signals S1 and S3, and provides a filtered signal S4, wherein the in-beam part of the filtered signal S4 is approximately equal to that of the signal S1. The filtered signal S4 is then subtracted from the signal S1 by the summer 141b to cancel out the in-beam part and obtain a sound signal S5 corresponding to the out-beam sound 21.
The adaptive filter 142a receives the signal S5 from the summer 141b, the voice detection signal S8 from the first VAD 151, and the feedback signal S7 from the summer 142b, calculates the parameters for the linear correlation between the signals S1 and S5, and provides a filtered signal S6 which is approximately equal to the out-beam part of the signal S1. The signal S6 is then subtracted from the signal S1 by the summer 142b to cancel out the out-beam part and obtain a sound signal S7 corresponding to the in-beam sound 20.
In the third case, the array microphone system 110 simultaneously receives the in-beam sound 20 and the out-beam sound 21, and separately outputs two signals S7 and S5, wherein the signal S7 corresponds to the in-beam sound 20 and the signal S5 corresponds to the out-beam sound 21.
However, step S470 and step S480 can be omitted. That is, if S2′/S1′>Y (step S460), then the ratio A is set as the gain of the gain control 120 (step S490).
As described, the array microphone system 110 includes two omni-directional microphones 111 and 112 to receive sound in a cone-shaped beam 12, thus avoiding the previously mentioned problems of conventional array microphones.
While the invention has been described by way of example and in terms of preferred embodiment, it is to be understood that the invention is not limited thereto. To the contrary, it is intended to cover various modifications and similar arrangements (as would be apparent to those skilled in the art). Therefore, the scope of the appended claims should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements.
| Number | Name | Date | Kind |
|---|---|---|---|
| 20030053639 | Beaucoup et al. | Mar 2003 | A1 |
| 20060002570 | Vaudrey et al. | Jan 2006 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20100322436 A1 | Dec 2010 | US |
