JPH05130207A - Loudspeaking simultaneous talking equipment - Google Patents

Abstract

PURPOSE:To provide a loudspeaking simultaneous talking device ensuring excellent talking by accurately discriminating howling generated by rapid fluctuation in the surrounding environment and a double-talking state. CONSTITUTION:A level comparator 108 discriminates whether or not the equipment is in double-talk state and outputs a discrimination signal based on the output signal of a band reject filter 107 and the output signal of a band pass filter 109 in a complementary relation to the band reject filter 107. When the equipment is not in the double talk state, a CPU 103 replies a discrimination signal to implement an adaptive processing, thereby preventing the occurrence of howling and the deterioration in sound quality or the like.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a loudspeaker simultaneous call device, and more particularly to a loudspeaker simultaneous call device having a howling prevention function.

[0002]

2. Description of the Related Art Conventionally, as a kind of speech communication apparatus, there has been used a loudspeaker simultaneous speech communication apparatus having a howling prevention function such as an echo canceller and capable of loudspeaker simultaneous speech. Figure 4
[FIG. 1] is a block diagram of a conventional public address simultaneous voice communication device. In FIG. 4, the speaker 101 is connected to the communication network via the downlink transmission line L ₁ . Further, the downlink transmission line L ₁ is one signal input terminal of a central processing unit (CPU) 103 via an analog / digital (A / D) converter 102 that converts an analog transmission signal from a communication network into a digital signal. It is connected to the. The microphone 105 is a CPU via an A / D converter 104 that converts an analog transmission signal into a digital signal.
It is connected to the other signal input terminal of 103. CPU
An output terminal of 103 is connected to a communication network via a digital / analog (D / A) converter 106 and an upstream transmission line L ₂ . The transmission signal from the downlink transmission line L ₁ and the transmission signal to the uplink transmission line L ₂ are connected to one and the other input terminals of the level comparator 108 for detecting the double talk state, respectively. There is. The output terminal of the level comparator 108 is connected to the control input terminal of the CPU 103.

The CPU 103 performs known arithmetic processing so as to function as an echo canceller (for example, 1
See pages 184 to 193 of "Voice processing and DSP (application example of digital signal processor)" issued by Keigaku Publishing Co., Ltd. on May 31, 989). The operation of the conventional loudspeaker simultaneous communication device configured as described above will be described.

In FIG. 4, the transmission line L ₁ from the communication network
The analog transmission signal received via the
Is output as voice and converted into a digital signal by the A / D converter 102. A / D converter 102
The digital output signal of is subjected to arithmetic processing such as superposition calculation by the CPU 103 to generate a pseudo echo signal.
The pseudo echo signal is transmitted from the speaker 101 to the microphone 10
It is a signal obtained by predicting a signal returning to 5.

On the other hand, the output signal of the microphone 105 (the sum of the signal returned from the speaker 101 to the microphone 105 through the air and the signal generated by the microphone 105 by the voice of the speaker) is output by the A / D converter 104. After being converted into a digital signal, it is output to the CPU 103. On the other hand, the level comparator 108 compares the signals input to its two input terminals and outputs a determination signal indicating that the signal is not in the double talk state when it is not in the double talk state. When the difference between the level of the signal from the downlink transmission line L ₁ and the level of the signal sent to the uplink transmission line L ₂ is a certain value or more, it is determined that the double talk state is not established, and the level comparator 108 outputs the high level signal. A determination signal which is a level signal is output to the CPU 103. When the difference is equal to or less than the predetermined value, it is determined that the state is the double talk state, the level comparator 108 does not output the determination signal, and only outputs the low level signal to the CPU 103.

When not in the double talk state, that is, when the discrimination signal is output from the level comparator 108, C
In response to this discrimination signal, the PU 103 calculates a new parameter for superposition calculation (performs an adaptive process) and also performs superposition calculation to generate a pseudo echo signal. After that, the CPU 103 subtracts the pseudo echo signal from the output signal of the A / D converter 104 and outputs it to the D / A converter 106. As a result, the speaker 101 to the microphone 10
Howling does not occur even if the state of the acoustic path to 5 changes.

In the double talk state, the CPU 10
3 performs superposition calculation to generate a pseudo echo signal, subtracts the pseudo echo signal from the output signal of the A / D converter 104, and outputs the subtracted echo signal to the D / A converter 106. However, if a new parameter is calculated in the double talk state, and the processing such as calculation of the pseudo echo signal and subtraction based on the new parameter is performed, the sound quality deteriorates, so the adaptive processing is not performed.

The D / A converter 106 converts the digital input signal into an analog signal and outputs it to the communication network via the upstream transmission line L ₂ . By the above operation, howling can be prevented even if the acoustic state near the communication device changes.

[0009]

However, when the surrounding environment changes suddenly, such as when a loud voice is suddenly input,
Since the adaptive processing is not in time, the pseudo echo signal becomes inaccurate. Therefore, howling occurs and the acoustic feedback of the speaker 101 is included in the output signal of the microphone 105, and the level of the output signal of the D / A converter 106 increases. Therefore, it is difficult to determine whether the double talk state or howling is occurring.

By the way, during the howling, the sound quality deteriorates unless the adaptive processing is continued, but during the double talk, the sound quality deteriorates unless the adaptive processing is stopped. Therefore, since it is difficult to determine whether it is double talk or howling, adaptive processing may be performed during double talk, and there is a problem that sound quality is deteriorated during double talk and sound volume is reduced.

The present invention has been made in view of the above-mentioned problems, and it is a loud voice simultaneous call that can ensure a good call by accurately performing howling caused by a sudden change of the surrounding environment and double talk. The purpose is to provide a device.

[0012]

According to the loudspeaker simultaneous speech communication system of the present invention, a speaker means outputs a voice transmission signal received from a communication network through a downlink transmission path and outputs a discrimination signal from a double talk discrimination means. In the loudspeaker simultaneous communication device which responds to the adaptive processing and outputs the signal output from the microphone means to prevent howling and outputs it to the communication network through the up transmission line, An output signal of the first filter means for passing a signal in a predetermined band of the transmission signal received through the transmission line and a signal for a predetermined band in the transmission signal output to the upstream transmission line are passed. It is characterized in that the discrimination signal is output based on an output signal of the second filter means having a complementary relationship with the first filter means.

[0013]

The double talk discrimination means outputs the discrimination signal based on the output signal of the first filter means and the output signal of the second filter means. The arithmetic means performs arithmetic processing such as adaptive processing based on the discrimination signal and sends it to the upstream transmission path.

[0014]

1 is a block diagram showing an embodiment of the present invention, in which the same parts as those in FIG. 4 are designated by the same reference numerals. In FIG. 1, a speaker 101 as a speaker means is
It is connected to a communication network via a band reject filter (BRF) 107 as a first filter means and a downlink transmission line L ₁ . The band reject filter 107 is
It has the transmission characteristics shown in FIG. 2, and only a signal in a specific frequency band is attenuated in order to delete a part of the band to the extent that sound quality is not affected. Band reject filter 10
The output terminal 7 is provided with a central processing unit (C) as an arithmetic means via an analog / digital (A / D) converter 102 that converts an analog transmission signal from a communication network into a digital signal.
PU) 103 is connected to one signal input terminal.
The microphone 105 as a microphone means is the A / D converter 104.
Is connected to the other signal input terminal of the CPU 103 via. An output terminal of the CPU 103 is connected to a communication network via a digital / analog (D / A) converter 106 for converting a digital signal after arithmetic processing into an analog signal and an upstream transmission line L ₂ . Further, the level comparator 10 as the double talk detecting means for detecting the double talk.
One and the other input terminals of 8 are connected to the output terminal of the band reject filter 107 and the output terminal of the band pass filter 109 as the second filter means, respectively. The bandpass filter 109 has the transmission characteristics shown in FIG. 3, and passes only signals in a specific frequency band. The band reject filter 107 and the band pass filter 109 have a complementary relationship, and the frequency band attenuated by the band reject filter 107 matches the frequency band passed by the band pass filter 109. The input terminal of the bandpass filter 109 is connected to the communication network via the upstream transmission line L ₂ . Incidentally, C
The PU 103 and the level comparator 108 perform the same operations as those shown in FIG.

The operation of this embodiment configured as described above will be described with reference to FIGS. 1 to 3. 1 to 3, the analog transmission signal from the communication network is output as voice from the speaker 101, while the band reject filter 1 is used.
After the signal of the predetermined band is removed by 07, it is input to the A / D converter 102 and converted into a digital signal. A / D
The digital output signal of the converter 102 is subjected to arithmetic processing such as superposition calculation by the CPU 103 as in the case of FIG. 4, and a pseudo echo signal is generated. The pseudo echo signal is a signal obtained by predicting an acoustic feedback signal returned from the speaker 101 to the microphone 105 via the air.

On the other hand, the output signal of the microphone 105 (the sum of the acoustic feedback of the transmission signal from the speaker 101 to the microphone 105 through the air and the transmission signal generated by the microphone 105 by the voice of the speaker) is A. After being converted into a digital signal by the / D converter 104, it is output to the CPU 103. The level comparator 108 compares the output signal of the band reject filter 107 with the output signal of the band pass filter 109 to determine whether or not the double talk state is present. When not in the double talk state, the level comparator 108 is a high level signal. Is output. The level of the output signal of the band reject filter 107 and the band pass filter 109
If the difference from the output signal level of 1 is equal to or more than a certain value, it is determined that the double talk state is not established, and the level comparator 108 outputs a determination signal to the CPU 103. On the other hand, when the difference is equal to or less than the predetermined value, it is determined that the double talk state is generated, and the determination signal is not output from the level comparator 108,
Only a low level signal is output to the CPU 103.

When not in the double talk state, that is, when the discrimination signal is output from the level comparator 108, C
In response to this discrimination signal, the PU 103 calculates a new parameter for superposition calculation (performs an adaptive process) and also performs superposition calculation to generate a pseudo echo signal. After that, the CPU 103 subtracts the pseudo echo signal from the output signal of the A / D converter 104 and outputs it to the D / A converter 106. As a result, the speaker 101 to the microphone 10
Cancel the acoustic return to 5. In the double talk state, the CPU 103 performs superposition calculation to generate a pseudo echo signal, and outputs the pseudo echo signal from the output signal of the A / D converter 104 to the subtraction D / A converter 106. No adaptive processing is performed.

After the above processing is performed, the output signal of the D / A converter 106 is output to the communication network via the upstream transmission line L ₂ . When determining whether the state is the double talk state as described above, since the band reject filter 107 and the band pass filter 109 have a complementary relationship, the output signal of the band pass filter 109 includes the speaker 10
No signal is returned from 1 to the microphone 105 via the air, and therefore, howling and double talk due to sudden changes in the surrounding environment are not erroneously determined.

The above-described embodiment can be modified in various ways. For example, the band reject filter 107 and the band pass filter 109 can be configured by digital filters executed by calculation processing. Further, the characteristics of the band reject filter 107 and the band pass filter 109 are not limited to those in the examples of FIGS. 2 and 3 and may have a complementary relationship such that a signal in a band passed by one filter does not pass by the other. ..

The loudspeaker simultaneous call device of the above embodiment can be used by being incorporated in various devices such as a telephone terminal device, an intercom, a PBX loudspeaker trunk, and a radio. As described above, according to the present embodiment, since it is determined whether or not the double talk is based on the signals from the band reject filter 107 and the band pass filter 109 having complementary characteristics, the band pass filter. Since the output signal of 109 does not include the acoustic return signal from the speaker 101 to the microphone 105, there is no erroneous determination of howling and double talk due to a sudden change in the surrounding environment. Therefore, it is possible to prevent erroneous adaptive processing at the time of double talk, and it is possible to make a good voice call.

[0021]

According to the present invention, it is possible to accurately determine whether howling or double talk even when the surrounding environment changes suddenly, and it is possible to make a good voice call without deterioration of sound quality. it can.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a characteristic diagram of a band reject filter used in the present invention.

FIG. 3 is a characteristic diagram of a bandpass filter used in the present invention.

FIG. 4 is a block diagram of a conventional loudspeaker simultaneous call device.

[Explanation of symbols]

101 ... Speakers as speaker means 102, 104 ... A / D converter 103 ... Central processing unit as arithmetic means 105 ... Microphone as microphone means 106 ... D / A converter 107・ ・ ・ Band reject filter as first filter means 108 ・ ・ ・ Level comparator as double talk discrimination means 109 ・ ・ ・ Band pass filter as second filter means L ₁・ ・ ・ Downward transmission line L ₂ ...・ Upstream transmission path

Claims (1)

[Claims] In order to prevent howling, a transmission signal received from a communication network via a downlink transmission line is output as voice by a speaker means, and adaptive processing is performed in response to a determination signal from a double talk determination means. In the loudspeaker simultaneous communication device that arithmetically processes the output signal of the microphone means and sends the output signal to the communication network through the up transmission line, the double talk determination means is a transmission signal received through the down transmission line. The output signal of the first filter means for passing the signal in the predetermined band and the signal for the predetermined band in the transmission signal output to the upstream transmission line are passed.
A loudspeaker simultaneous call device, wherein the discrimination signal is output based on an output signal of a second filter means having a complementary relationship with the filter means.