Multi-channel echo compensation system
View Patent ↗A system reduces echoes in an audio system by de-correlating audio signals. Adaptive filters provide compensation based on the de-correlated signals. A controller controls the de-correlation based upon the adaptation state of the adaptive filters.
1. A method for multi-channel echo compensation, comprising:
receiving a first audio input signal from a first channel, and receiving a second audio input signal from a second channel;
de-correlating the first and second audio input signals to generate a first de-correlated audio signal and a second de-correlated audio signal, respectively;
providing the first de-correlated audio signal to a first loudspeaker to generate a first loudspeaker signal and providing the second de-correlated audio signal to a second loudspeaker to generate a second loudspeaker signal;
detecting the first and second loudspeaker signals and a desired signal, by a microphone to generate a microphone output signal; and
adaptively filtering the microphone output signal to generate an echo compensated signal using an echo compensation filter, wherein de-correlating of the first and the second audio signals includes controlling a value of a time-varying parameter based on an adaptation state of the echo compensation filter and generating the first and second decorrelated audio signals by applying a time-varying all-pass filter followed by adaptive non-linear processing to a particular one of the input audio signals, the time-varying all-pass filter using the time-varying parameter.
2. The method according to claim 1 , further comprising controlling the value of a degree of non-linear processing on the bases of the adaptation state of the adaptive echo compensation filter.
3. The method according to claim 2 , further comprising determining the correlation of the first and the second audio signals and where the de-correlating is controlled by the determined correlation between the first and second audio input signals.
4. The method according to claim 3 , where the correlation is determined by calculating a short time correlation value or a short time coherence value of the first and second audio input signals.
5. The method according to claim 1 , wherein the adaptation state of the echo compensation filter is determined on the basis of the echo compensation signal.
6. The method according to claim 1 , wherein the adaptation state of the echo compensation filter is determed on the basis of the filter coefficients of the echo compensation filter.
7. The method according to claim 1 , where the time-varying filtering comprises all-pass filtering according to the equation:
x ′( n )=−β( n ) x ( n )+ x ( n− 1)+β( n ) x ′( n− 1)
where β is a time-varying parameter, n is a discrete time index, x is an audio signal of one channel and x′ is a filtered audio signal.
8. The method according claim 1 wherein determining the adaptation state of the adaptive echo compensation filter on the basis of the adapted filter coefficients of the echo compensation filter is performed by calculating a system distance D(n) according to the formula:
D
(
n
)
=
10
log
10
(
N
N
T
∑
i
=
0
N
T
-
1
(
h
^
1
,
i
2
(
n
)
+
h
^
2
,
i
2
(
n
)
)
)
where n is a discrete time index, N is a length of an echo compensation filter, N T is a number of sampling times and ĥ 1 (n) and ĥ 2 (n) are filter coefficients of the echo compensation filter for the respective first and second channels.
9. The method according to claim 1 , comprising detecting the first and second loudspeaker signals using a plurality of microphones arranged in a microphone array.
10. The method according to claim 9 , comprising beam-forming the microphone signal generated by the microphone array to generate a beam-formed microphone signal.
11. The method according to claim 1 , further comprising:
controlling the value of a degree of non-linearity of the adaptive non-linear processing on the basis of the adaptation state of the adaptive echo compensation filter.
12. The method according to claim 11 , further comprising determining correlation of the first and second audio signals and wherein de-correlating the first and the second audio signals is controlled by the determined correlation of the first and second audio signals.
13. A non-transitory computer-readable storage medium having processor executable
instructions to compensate for echoes in a multi-channel system, the processor executable instructions comprising:
processor code for receiving a first audio input signal from a first channel, and receiving a second audio input signal from a second channel;
processor code for de-correlating the first and second audio input signals to generate a first de-correlated audio signal and a second de-correlated audio signal, respectively;
processor code for providing the first de-correlated audio signal to a first loudspeaker to generate a first loudspeaker signal and providing the second de-correlated audio signal to a second loudspeaker to generate a second loudspeaker signal;
processor code for detecting the first and second loudspeaker signals and a desired signal, by at least one microphone to generate a microphone output signal; and
processor code for adaptively filtering the microphone output signal to generate an echo compensated signal using an echo compensation filter, wherein the processor code for de-correlating of the first and the second audio signals includes controlling a value of a time-varying parameter based on an adaptation state of the echo compensation filter and generating the first and second decorrelated audio signals by applying a time-varying all-pass filter followed by adaptive non-linear processing to a particular one of the input audio signals, the time-varying all-pass filter using the time-varying parameter.
14. The non-transitory computer-readable storage medium according to claim 13 , further comprising:
processor code for controlling the value of a degree of non-linear processing on the bases of the adaptation state of the adaptive echo compensation filter.
15. The non-transitory computer-readable storage medium according to claim 14 , further comprising:
processor code for determining the correlation of the first and the second audio signals and where the de-correlating is controlled by the determined correlation between the first and second audio input signals.
16. The non-transitory computer-readable storage medium according to claim 15 , where the correlation is determined by calculating a short time correlation value or a short time coherence value of the first and second audio input signals.
17. The non-transitory computer-readable storage medium according to claim 13 , wherein the adaptation state of the echo compensation filter is determined on the basis of the echo compensation signal.
18. The non-transitory computer-readable storage medium according to claim 13 , wherein the adaptation state of the echo compensation filter is determined on the basis of the filter coefficients of the echo compensation filter.
19. The non-transitory computer-readable storage medium according to claim 13 , where the time-varying filtering of the processor code for de-correlation comprises all-pass filtering according to the equation:
x ′( n )=−β( n ) x ( n )+ x ( n− 1)+β( n ) x ′( n− 1)
where β is a time-varying parameter, n is a discrete time index, x is an audio signal of one channel and x′ is a filtered audio signal.
20. The non-transitory computer-readable storage medium according to claim 13 , wherein the adaptation state of the adaptive echo compensation filter on the basis of the adapted filter coefficients of the echo compensation filter is determined by calculating a system distance D(n) according to the formula:
D
(
n
)
=
10
log
10
(
N
N
T
∑
i
=
0
N
T
-
1
(
h
^
1
,
i
2
(
n
)
+
h
^
2
,
i
2
(
n
)
)
)
where n is a discrete time index, N is a length of an echo compensation filter, N T is a number of sampling times and ĥ 1 (n) and ĥ 2 (n) are filter coefficients of the echo compensation filter for the respective first and second channels.
21. A system for reducing echoes in an audio system, comprising:
a de-correlation circuit configured to de-correlate a plurality of audio signals and to generate respective de-correlated audio signals;
an adaptive echo compensation filter configured to echo compensate microphone signals, the echo compensation based on the de-correlated audio signals; and
a controller configured to control the de-correlation circuit based upon an adaptation state of the adaptive echo compensation filter;
wherein the de-correlating circuit is configured to generate the de-correlated audio signals by applying a time-varying all-pass filter followed by adaptive non-linear processing to a particular one of the audio signals and wherein the time-varying all-pass filter uses a time-varying parameter and the controller is configured to control the time-varying parameter based on an adaptation state of the adaptive echo compensation filter.
22. The system according to claim 21 , further comprising a multi-channel source configured to provide the plurality of audio signals.
23. The system according to claim 22 , where the controller controls the decorrelation circuit based on the determined correlation between the the audio signals.
24. The system according to claim 23 , where the controller determines the correlation state based on a short time correlation value or a short time coherence value of the audio signals.
25. The system according to claim 22 , further comprising:
a plurality of loudspeakers configured to receive the respective de-correlated audio signals and to generate loudspeaker output signals based on the respective de-correlated audio output signals, where each loudspeaker generates the respective loudspeaker output signal based on one audio input signal provided by a multi-channel source; and
one microphone configured to receive the loudspeaker output signals and a desired signal, and to generate a microphone output signal;
the echo compensation filter configured to echo compensate the microphone output signal and generate an echo compensated acoustic signal; and
the controller configured to receive the desired signal and to control the de-correlation circuit based on the desired signal.
26. The system according to claim 21 , where the de-correlation circuit includes a non-linear processor.
27. The system according to claim 26 , where the non-linear processor includes a half-wave rectifier.
28. The system according to claim 21 , where the controller is configured to calculate an adaptation state of the adaptive echo compensation filter based on system distance measure according to the equation:
D
(
n
)
=
10
log
10
(
N
N
T
∑
i
=
0
N
T
-
1
(
h
^
1
,
i
2
(
n
)
+
h
^
2
,
i
2
(
n
)
)
)
where n is a discrete time index, N is a length of an echo compensation filter, N T is a number of sampling times and ĥ 1 (n) and ĥ 2 (n) are filter coefficients of the echo compensation filter for the respective first and second channels.
29. The system according to claim 21 , further comprising at least one microphone and/or a microphone array having at least one directional microphone.
30. The system according to claim 29 , further comprising a beam-forming circuit connected to the at least one microphone and/or microphone array, and configured to generate a beam-formed microphone signal.