JPH02283131A - Echo suppressing method - Google Patents

Abstract

PURPOSE:To suppress an echo correctly by providing a comb-line filter being a transmission path for a transmission signal and a reception signal with a complementary frequency characteristic when both the transmission signal and the reception signal are simultaneously detected. CONSTITUTION:When a power PA at a receiver side is higher than a threshold level ThA for detecting a reception signal, the presence of a reception signal is decided and when a power PB at a sender side is higher than a threshold level ThB for detecting a transmission signal, the 2-way communication state is decided switches SW1, SW2 controlled to allow the signal at the sender side to pass a comb-line filter 2 and the signal at the receiver side to pass a comb- line filter 3 thereby suppressing the echo component. A power PC at an output point C of the filter 2 is compared with a threshold level ThC in a level adjustment program 4, and when the power is higher, since it signifies the input of the transmission signal to discriminate the 2-way communication surely and a counter is counted up by one to increase the sensitivity of the threshold level ThB. When the count of the counter is 3 or below, let the attenuation of the threshold level ThB be 6dB, then the sensitivity is increased by 0.5dB resulting in 6.5dB, and the operation above is repeated.

Description

[Detailed Description of the Invention] [Summary] Regarding an echo suppression method for suppressing acoustic echo in a two-way communication device, it is possible to maintain a two-way communication state even if the attenuation or gain of the system from the speaker to the microphone changes. The purpose of the present invention is to provide an echo suppression method that can make accurate judgments and correctly suppress echoes. When only one signal is detected, only the transmission path of the one signal is formed and the transmission path of the other signal is blocked. When both the transmission signal and the reception signal are detected simultaneously, the transmission signal is transmitted with mutually complementary frequency characteristics. and a comb filter forming a transmission path for the received signal, the audio detection control section is provided with a level adjustment means, and the output level of the comb filter on the transmitting side is detected; When it is determined that both the transmitted signal and the received signal are detected simultaneously, if the output level is higher than a predetermined threshold, the detection sensitivity for detecting the transmitted signal is increased, and if it is lower than the prescribed threshold, the transmitted signal is detected. The echo suppressing device is configured to detect a transmitted signal by controlling the detection sensitivity to lower the detection sensitivity, and operate the echo suppressing device based on this.

(Field of Industrial Application) The present invention relates to an improvement in an echo suppression method for suppressing echoes caused by acoustic coupling between speakers and microphones of a TV conference system of a two-way communication device, a loudspeaker telephone, or the like.

In two-way communication devices such as TV conference systems and loudspeaker telephones, echoes and howling caused by acoustic coupling between speakers and microphones pose a problem in speech quality, and echo canceller devices and the like are used to suppress the echoes.

In a two-way communication device that uses a speaker or microphone, during actual use the person may move or the position of the microphone may change.

Although the amount of attenuation or gain between microphones varies, even in this case, it is desired to provide an echo suppression method that correctly detects transmitted signals and received signals in a two-way communication state and suppresses echoes.

[Conventional technology]

FIG. 6 is a block diagram of the main parts of a conventional two-way communication device.

When starting up the two-way communication device, in order to adjust the operation of the echo canceller device 20 to the optimum level, switch SW3 is set to the solid line side, white noise is transmitted from the white noise generator 12, and white noise is transmitted from the speaker 11 to the microphone 9. The gain of the amplifier or attenuator 7 is adjusted so that the level drop that has occurred in The amount of attenuation or amplification is adjusted to the same value as the gain or attenuation, and the rear switch SW3 is used as the dotted line side.

In addition, in this two-way communication device, in order to cancel the echo from the speaker 11 to the microphone 9, the received signal is input to a pseudo-echo generation section 14 composed of a transversal filter to create a signal as close as possible to the actual echo signal. The subtracted signal is added to the subtracter 15, and the subtracted signal is fed back to the 1-pseudo echo generator 14, and the coefficient values of the transpanal filter are updated and controlled to make the echo signal as small as possible.

This is fine if only the received signal exists and the near-end speaker does not speak, but if the near-end speaker speaks, this state must be detected and the updating operation of the pseudo echo generator 14 must be stopped. Must be.

The double talk (two-way communication state) detection unit 13 in FIG. 6 is the part that performs this control, and detects the respective levels of the transmitted signal and the received signal, and the ratio thereof is 6 dB with a preset attenuation amount.
If the value is smaller than the value of , it is determined that there is double talk, and the adaptive operation of the echo canceller device 20 is stopped.

In this way, the echo of the two-way communication device is canceled out.

[Problems to be Solved by the Invention] However, in actual use, the amount of attenuation in the system from the speaker 11 to the microphone 9 varies by %JJ due to the movement of the person or the change in the position of the microphone.

For example, if the amount of attenuation changes from 6 dB to 5 dB, it may be mistakenly judged as double talk due to noise even if the near-end speaker is not speaking, and if it changes to 7 dB, echoes due to wrap-around will decrease. If the near-end speaker's voice is low, it will be erroneously determined that the speaker is not speaking, and the two-way conversation state cannot be correctly determined, and the echo canceller device 20 will not operate properly, resulting in a drop in voice quality.

The present invention enables accurate determination of two-way communication status even if the attenuation or gain of the system from the speaker to the microbond changes.
The purpose is to provide an echo suppression method that can suppress echoes correctly.

[Means to solve the problem]

FIG. 1 is a block diagram of the principle of the present invention.

As shown in FIG. 1, a voice detection control section 1 detects a transmission signal and a reception signal, respectively, and when the voice detection control section 1 detects only one of the transmission signal and reception signal, When both the transmitted signal and the received signal are detected at the same time, the transmitted signal and the received signal are formed with mutually complementary frequency characteristics. The voice detection control unit 1 is provided with a level adjustment means 4 to detect the output level of the comb filter 2 on the transmitting side, and detect the output level of the comb filter 2 on the transmitting side. When it is determined that both the output level and the received signal are detected at the same time, if the output level is greater than a predetermined threshold, the detection sensitivity for detecting the transmitted signal is increased, and if it is smaller than the predetermined threshold, the transmitted signal is detected. The transmitted signal is detected by controlling the detection sensitivity to be lowered, and the echo suppressing device 10 is operated based on this.

[For production]

In the present invention, as the echo suppressing device 10, an echo suppressing device for which the present applicant applied for a patent in Japanese Patent Application No. 62-216469 on September 1, 1988 is used.

When the echo suppressing device 10 detects only one of the transmitted signal and the received signal, it forms only the transmission path for the one signal and blocks the transmission path for the other signal, as shown in FIG. As shown in the attenuation characteristics of the comb filters 2 and 3 having complementary frequency characteristics in FIG. It has a comb filter 3 on the receiving side that does not pass the frequency signal that has passed through the comb filter 2 on the transmitting side, and in a two-way communication state, as shown in FIG. This is to suppress the echo.

In the present invention, the audio detection control section 1 is further provided with a level adjustment means 4, which detects the output level of the comb filter 2 on the transmitting side, and when it is determined that both the transmitted signal and the received signal are detected at the same time. If the output level is greater than a predetermined threshold, it is determined that a two-way conversation is indeed occurring since a transmission signal is being input, and the detection sensitivity for detecting the transmission signal is increased.
If it is smaller than a predetermined threshold, it is determined that the transmitted signal is not being transmitted manually and is not in a two-way communication state, and control is performed to lower the detection sensitivity for detecting the transmitted signal.

In this way, for example, the amount of attenuation from the microphone to the speaker will be reduced, and even if it is determined that a two-way conversation is occurring, the level of the transmitted signal will increase compared to the looping echo. 2 is not passed and the output level is below the threshold level, the threshold level for determining that there is a transmission signal is increased and the sensitivity is lowered, so there is no possibility of a two-way communication state erroneously occurring.

Also, if the amount of attenuation from the speaker to the microphone increases and the amount of echo wraparound decreases, even if the near-end speaker speaks in a low voice, this voice will pass through the comb-shaped filter 2 on the transmitting side and will not be output. The level is equal to or higher than the threshold level.

Therefore, once the voice detecting side jlH unit 1 determines that a two-way communication state exists, the threshold level for determining that a transmission signal is present is lowered and the sensitivity is increased, so that the two-way communication state can be correctly detected.

That is, even if the attenuation from the speaker to the microbond changes, the echo suppressor 10 will operate correctly and the voice quality will not deteriorate.

[Embodiment] FIG. 2 is a block diagram of main parts of a two-way communication device according to an embodiment of the present invention, FIG. 3 is a flowchart of a control and level adjustment program for a voice detection control unit according to an embodiment of the present invention, and FIG. The figure shows comb filters 2 and 3 with complementary frequency characteristics as shown in Figure 2.
FIG. 5 is a diagram showing the control state of the voice detection control section l of FIG. 2.

In the case of FIG. 2, as in the case of the conventional example shown in FIG. Adjust the gain or attenuation of the amplifier or attenuation R'& 7 so that the relationship between points A and B is 6 dB, and adjust the attenuation or gain of the attenuator or amplifier 8 to the gain or attenuation of the amplifier or attenuator 7. This is the same as setting the switch SW3 to the dotted line side after canceling the .

The echo suppressing device 10 is one for which the above-mentioned applicant filed a patent application in Japanese Patent Application No. 62-216469 on September 1, 1988, and the attenuation amount of the comb filter 2.3 shown in FIG. 4
As shown in Figures (a) and (b), the signal has complementary frequency characteristics, and the frequency signal that has passed through the comb filter 3 on the receiving side shown in (b) is filtered by the comb filter 3 on the transmitting side shown in (a). 2 is not passed through, and the frequency signal that has passed through the transmitting side comb filter 2 shown in (a) is passed through the receiving side comb filter 3 shown in (b).
cannot pass through.

Next, the control of the audio detection control section 1 and the control by the level adjustment program will be explained with reference to FIG.

In step 1, the power PA at point A on the receiving side is calculated, and in step 2, the power PB at point B on the transmitting side is calculated, and the process proceeds to step 3.

In step 3, if the power PA on the receiving side is greater than the threshold ThA for detecting a received signal, it is determined that there is a received signal and the process proceeds to step 4, and if it is smaller than the threshold ThA, it is determined that there is no received signal and the process proceeds to step 5.

In step 5, if the power P on the transmitting side is smaller than the thresholds T and B for detecting the transmitted signal, it is determined that there is no transmitted signal, and the switches SWI and SW2 are controlled, and as shown in FIG. If it is larger than the threshold ThB, it is assumed that there is a transmission signal and the process goes to step 6, where the switch SW
I and SW2 are controlled, and as shown in FIG. 5, the transmitting side is passed through, and the receiving side is set to a cutoff state through an attenuator 6, thereby blocking the echo component.

In step 4 above, if the power P8 on the transmitting side is smaller than the threshold ThB for detecting a transmitted signal, it is determined that there is no transmitted signal;
Proceeding to step 8, the switch swL SW2 is controlled, and as shown in FIG. If it is larger than the detection thresholds T and B, the process proceeds to step 9 as a two-way communication state, and controls the switches SWI and SW2, so that the transmitting side uses the comb filter 2 and the receiving side uses the comb filter 3. The echo component is suppressed as it passes through.

Next, in this state, the process proceeds to step 10, and control is performed as follows using the level adjustment program.

In step 10, the output C of the comb filter 2 on the transmitting side is
The power PC at the point is compared with the threshold 'I't,C, and if it is larger, it means that a transmission signal is being input, so it is determined that a two-way conversation is indeed in progress, and in step 11, in order to increase the sensitivity of the threshold ThB. If the value of the counter incremented by 1 is 3 or less, the process proceeds to step 12, and for example, the threshold T for detecting the transmitted signal is set to 0.5 d if 1 is an attenuation value of 6 dB.
Raise the B sensitivity to 6.5 dB, return to step 1, and repeat the above operation.

Further, when the sensitivity of the threshold value TkB for detecting the transmitted signal is increased to, for example, 7.5 dB as an attenuation value in step 12, the count value becomes 3, and thereafter the process returns from step 11 to step 1 and the above operations are performed sequentially.

In step 10 above, if the power PC at the output point C of the comb filter 2 on the transmitting side is smaller than the thresholds T and C, no transmitting signal is input, so both.

It is determined that there is no incoming call, and the process proceeds to step 13. If the threshold value ThB for detecting a transmitted signal is, for example, an attenuation value of 6 dB, the sensitivity is lowered by 0.5 dB to 5.5 dB, and the process returns to step 1.

By repeating the above, the sensitivity of thresholds T and B decreases one after another by 4.
．． It becomes 5dB.

In this case, in step 4, the power P on the transmitting side becomes smaller than the threshold value T, B for detecting a transmitted signal, so there is no transmitted signal, and the process proceeds to step 8, where switches SWI, S
W2 is controlled, and as shown in FIG. 5, the receiving side is passed through and the transmitting side is passed through an attenuator 5 to be cut off.

That is, even if the attenuation or gain of the system from the speaker 11 to the microphone 9 changes, the two-way communication state can be detected correctly and echoes can be suppressed correctly, so that voice quality will not deteriorate.

(Effects of the Invention) As explained in detail above, according to the present invention, the speaker 11
Even if the attenuation or gain of the system from the microphone 9 to the microphone 9 fluctuates, the two-way communication state can be detected correctly and echoes can be suppressed correctly, so there is an effect that voice quality will not be degraded.

[Brief explanation of drawings]

Fig. 1 is a block diagram of the principle of the present invention, Fig. 2 is a block diagram of main parts of a two-way communication device according to an embodiment of the present invention, and Fig. 3 is a control and level of the voice detection control unit according to an embodiment of the present invention. A flowchart of the adjustment program, FIG. 4 is a characteristic diagram of the attenuation amount versus frequency of the comb filters 2 and 3 having complementary frequency characteristics in the second time, and FIG. 5 shows the control state of the voice detection control unit 10 of FIG. FIG. 6 is a block diagram of the main parts of a conventional two-way communication device. In the figure, 1 is a voice detection control unit, 2.3 is a comb filter, 4 is a level adjustment means, a level adjustment program, and 5.6.
is an attenuator, 7 is an amplifier or attenuator, 8 is an attenuator or amplifier, 9 is a microphone, 10 is an echo suppression device, 11 is a speaker, 12 is a white noise generator, 13 is a double talk detector, 14 is a pseudo echo a generation section; 15 a subtraction section; 20 an echo canceller device; SWI SW2. SW3 indicates a switch.

Claims (1)

[Claims] When suppressing acoustic echo of a two-way communication device, a voice detection control unit (
1), when the voice detection control unit (1) detects only one of the transmitted signal and the received signal, only the transmission path of the one signal is formed and the transmission path of the other signal is cut off, and the transmission comb filters (2, 3) that form transmission paths for the transmitted signal and the received signal, respectively, with mutually complementary frequency characteristics when both the signal and the received signal are detected simultaneously;
The voice detection control section (1) is provided with a level adjustment means (4) to detect the output level of the comb filter (2) on the transmission side, When it is determined that both the output level and the received signal are detected at the same time, if the output level is greater than a predetermined threshold, the detection sensitivity for detecting the transmitted signal is increased, and if it is smaller than the predetermined threshold, the transmitted signal is detected. An echo suppression method characterized in that a transmission signal is detected by controlling the detection sensitivity to be lowered, and the echo suppression device (10) is operated based on the detected transmission signal.