JP5673256B2 - Maska sound measuring device and sound masking device - Google Patents

Description

  The present invention relates to a sound masking system that generates a masker sound for making it difficult for a third party to hear a specific speaker's voice, and more particularly to measurement of a masker sound.

  Conventionally, various sound masking techniques have been devised that make it difficult for a third party to hear the voice of a specific speaker. As a similar technique of the sound masking system, Patent Document 1 discloses a technique for adjusting frequency characteristics of reproduced sound of an audio device so that a listener can hear necessary ambient sounds.

  In the case of a sound masking system, a masker sound is emitted to make it difficult for a third party to hear a specific speaker voice. If this is applied to a similar technology, the masker sound is based on the specific speaker voice. Will be adjusted.

JP 2008-228198 A

  However, in the conventional sound masking system, it cannot be determined whether or not a masker sound can be emitted so that a third party cannot really hear a specific speaker's voice. It is also impossible to determine whether the total volume level and frequency characteristics of the masker sound are appropriate.

  Accordingly, an object of the present invention is to realize a masker sound measuring apparatus capable of measuring masker sounds.

  The present invention relates to a masker sound measuring apparatus for measuring a masker sound for making it difficult to hear a specific speaker's voice (speaker voice) in an area to be masked. The masker sound measuring device includes an extraction unit and a voice analysis unit. The extraction unit extracts the speaker position sound picked up at the speaker position, the masker sound emitted to mask the speaker sound, and the masking position sound picked up in the area where the speaker sound should be masked. Are used to extract a determination speaker voice that is a speaker voice in a region to be masked and a determination masker sound that is a masker sound in a region to be masked. The voice analysis unit compares the determination speaker voice with the determination masker sound and analyzes whether masking is performed.

  In this configuration, the speaker voice and masker sound in the region to be masked where the speaker voice and masker sound can be heard are extracted individually. That is, the influence of the external environment is removed, and the component of only the speaker sound that can be heard in the area to be masked and the component of only the masker sound that can be heard in the area to be masked are individually extracted. Accordingly, it is possible to accurately compare the volume level of the masker sound and the volume level of the speaker voice in the area to be masked. Thereby, in the area to be masked, it can be accurately determined whether the volume level of the masker sound is higher than the volume level of the speaker voice, that is, whether the speaker voice is masked.

  Further, the voice analysis unit of the masker sound measuring device according to the present invention compares the determination speaker voice and the determination masker sound for each predetermined frequency component, and the determination masker sound is used for determination in all frequency components. If it is above a predetermined volume level for the speaker's voice, it is determined that it is masked.

  This configuration shows the criteria for determining the masking effect. If the volume level of the masker sound (judgment masker sound) is higher than the volume level of the speaker voice (judgment speaker voice) in the area to be masked, Since it is difficult to hear the person's voice, it is determined that there is a masking effect.

  Further, the voice analysis unit of the masker sound measuring apparatus according to the present invention is configured such that, for all frequency components, the determination masker sound is equal to or higher than a predetermined volume level and within a predetermined volume level range with respect to the determination speaker voice. Judge that it is properly masked.

  This configuration shows a criterion for determining an appropriate masking effect. The volume level of the masker sound (determination masker sound) is greater than or equal to the volume level of the speaker voice (determination speaker voice) and a predetermined level. When it is within the volume level range, it is determined that appropriate masking is performed.

  Moreover, the extraction part of the masker sound measuring apparatus of this invention is provided with the speaker voice extraction part for determination, and the masker sound extraction part for determination. The determination speaker voice extraction unit subtracts the first adaptive filter to which the speaker position voice is input, and the speaker position voice filtered by the first adaptive filter from the masking position voice. And a first post processor that feeds back the signal to the first adaptive filter. The determination speaker voice extraction unit outputs the filtered speaker position voice as the determination speaker voice. The determination masker sound extraction unit subtracts the second adaptive filter to which the masker sound is input, the masker sound filtered by the second adaptive filter from the masking position sound, and the second adaptive signal is the second adaptive signal. And a second post processor that feeds back to the mold filter. The determination masker sound extraction unit outputs the filtered masker sound as the determination masker sound.

  This configuration shows a specific implementation example of the extraction unit. As described above, by using the adaptive filter, it is possible to accurately obtain the transfer function until the speaker voice reaches the area to be masked and the transfer function until the masker sound reaches the area to be masked. By using these transfer functions, the speaker voice (determination speaker voice) and masker sound (determination masker sound) in the area to be masked can be accurately extracted.

  In addition, the masker sound measuring apparatus of the present invention includes a speaker position microphone that picks up the speaker position sound, a masking position microphone that picks up the masking position sound, and a masker sound generating unit that generates a masker sound. . This configuration shows a configuration example in the case where a masker sound measuring device is specifically installed.

  In addition, the masker sound measuring apparatus of the present invention includes a notification unit that notifies the analysis result by the voice analysis unit. In this configuration, it is possible to notify the speaker of the analysis result, that is, whether the masking effect is valid or invalid.

  Moreover, this invention relates to a sound masking apparatus provided with the masker sound measuring apparatus in any one of the above-mentioned. The sound masking device, together with the masker sound measurement device, adjusts the volume level of the masker sound according to the sound emission parameters set so that the masker sound of the appropriate volume level is emitted from the analysis result by the voice analysis unit. And a speaker that emits the adjusted masker sound.

  In this configuration, the masker sound to be emitted can be adjusted so that effective masking can be performed. That is, the masker sound can be adaptively adjusted and emitted according to the analysis result of the masking effect.

  According to the present invention, it can be determined whether or not the speaker voice is masked by the masker sound in the region to be masked. Thereby, the speaker can grasp | ascertain correctly whether their conversation is masked so that it may be difficult to hear to a third party.

It is a schematic installation block diagram of the sound masking apparatus 1 which concerns on 1st Embodiment. It is a block diagram which shows the structure of the sound masking apparatus 1 which concerns on 1st Embodiment. It is a flowchart which shows the processing flow of the determination of the presence or absence of a masking effect, and the parameter setting and adjustment of a masker sound based on the said determination result. It is a figure which shows the relationship between the frequency characteristic of the volume level of a masker sound, and the volume level of a speaker voice (when appropriate). It is a figure which shows the relationship between the frequency characteristic of the volume level of a masker sound, and the volume level of a speaker voice (when a masker sound is too high). It is a figure which shows the relationship between the frequency characteristic of the volume level of a masker sound, and the volume level of a speaker voice (when a masker sound is too low). It is a figure which shows the relationship of the frequency characteristic of the volume level of a masker sound, and the volume level of a speaker voice (when a masker sound is too low at a specific frequency). It is a schematic installation block diagram of the sound masking apparatus 1A which concerns on 2nd Embodiment. It is a block diagram which shows the structure of 1 A of sound masking apparatuses which concern on 2nd Embodiment.

  A sound masking device including a masker sound measurement unit (masker sound measurement device) according to a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a schematic installation configuration diagram of a sound masking apparatus 1 according to the present embodiment. FIG. 2 is a block diagram showing the configuration of the sound masking apparatus 1 according to this embodiment.

  The sound masking device 1 of this embodiment is used in a situation where a third party can hear the voice of a specific speaker. For example, in FIG. 1, specific speakers 910 and 920 are present in a booth partitioned by a counter 900. A third party 930 is present at a position away from the counter 900 by a predetermined distance. In such a situation, the masker sound Smo is emitted so that the conversation of the specific speakers 910 and 920 is difficult for the third party 930 to hear.

  The sound masking device 1 includes a main body 2, a speaker position microphone MICb, a masking position microphone MICo, a speaker SP, and a display unit 40.

  The speaker position microphone MICb is installed on the table 901 of the counter 900 and is connected to the main body 2. The speaker position microphone MICb collects the conversations of the specific speakers 910 and 920 and outputs the speaker position voice signal Sb, which is a collected signal, to the main body 2.

  The speaker SP is installed in the vicinity of the counter 900 and is connected to the main body 2. The speaker SP emits a masker sound signal Smo obtained by decoding the contents data of the masker sound and adjusting it by a method described later (no adjustment unless necessary). Hereinafter, the masker sound signal Smo emitted from the speaker SP will be referred to as an adjusted masker sound signal Smo regardless of whether or not the adjustment is performed. At this time, the speaker SP emits the adjusted masker sound signal Smo toward the area where masking is desired, including the area where the third party 930 is present. The speaker SP may be integrated with the main body 2. In this case, the main body 2 with the built-in speaker SP is installed in the vicinity of the counter 900.

  The masking position microphone MICo is installed in an area where masking is desired, including the area where the third party 930 is present, and is connected to the main body 2. The masking position microphone MICo picks up sound in the installation area and outputs a masking position sound signal So, which is a sound pickup signal, to the main body 2.

  The display unit 40 includes an indicator lamp, a liquid crystal display, and the like, and is connected to the main body 2. The display unit 40 performs lighting or specific display according to a masking effect determined by a method described later. The display unit 40 may be integrated with the speaker position microphone MICb.

  Next, the configuration of the main body 2 that determines the masking effect will be described. As shown in FIG. 2, the main body 2 includes a masker sound measuring unit 10, a masker sound generating unit 20, a masker sound storage unit 21, and a masker sound adjusting unit 30 corresponding to the “masker sound measuring device” of the present invention.

  The masker sound storage unit 21 combines various masker sounds, for example, disturbing sounds such as hustle sounds created from human voices, and background sounds such as river flowing sounds, ocean wave sounds, and bird chirps. The masker sound content data is stored. The masker sound generation unit 20 appropriately reads out masker sound content data according to an operation input or the like by an operation unit (not shown), and generates a masker sound (original masker sound) signal Sm. The original masker sound signal Sm is output to the masker sound measurement unit 10 and the masker sound adjustment unit 30.

  The masker sound measurement unit 10 includes adaptive filters 111 and 112, post processors 121 and 122, and a voice analysis unit 130.

  The adaptive filter 111 corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb.

  The post processor 121 corresponds to the “first post processor” of the present invention, and receives the masking position sound signal So. The post processor 121 subtracts the speaker position voice signal Sb filtered by the adaptive filter 111 from the masking position voice signal So, and outputs a first subtraction signal Sn1. The first subtraction signal Sn1 is input to the post processor 122 and fed back to the adaptive filter 111.

  The adaptive filter 111 is a filter that adapts filter coefficients so that the speaker position voice signal Sb is not included in the first subtraction signal Sn1. That is, it is a filter that adapts the filter coefficient so as to suppress only the speaker voice from the voice in the area to be masked in which the masker sound and the speaker voice are mixed.

  By performing such processing, the speaker position voice signal Sb after the filtering process becomes a voice signal that depends on the transfer characteristics of the speaker voice from the speaker position to the area to be masked. The speaker position voice signal Sb is equivalent to the speaker voice signal Sbc indicating the speaker voice that has reached the area to be masked.

  The adaptive filter 111 outputs the speaker voice signal Sbc thus calculated (corresponding to the “speaker voice for determination” of the present invention) to the post processor 121 as described above, and also the voice analysis unit. To 130.

  In this way, by using the adaptive filter 111 and the post processor 121 having the above-described configuration, it is possible to extract the speaker voice signal Sbc that has reached the area to be masked. The adaptive filter 111 and the post processor 121 correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 112 corresponds to the “second adaptive filter” of the present invention, and receives the original masker sound signal Sm.

  The post processor 122 corresponds to the “second post processor” of the present invention, and receives the first subtraction signal Sn1. The post processor 122 subtracts the original masker sound signal Sm filtered by the adaptive filter 112 from the first subtraction signal Sn1, and outputs a second subtraction signal Sn2. The second subtraction signal Sn2 is fed back to the adaptive filter 112.

  The adaptive filter 112 is a filter that adapts filter coefficients so that the second subtracted signal Sn2 does not include the original masker sound signal Sm. That is, the filter coefficient is adapted to further suppress only the masker sound while suppressing the speaker sound from the sound in the region to be masked where the masker sound and the speaker sound are mixed.

  By performing such processing, the original masker sound signal Sm after the filter processing becomes an audio signal depending on the transfer characteristic of the masker sound from the installation position of the speaker Sp to the area to be masked. The original masker sound signal Sm is equivalent to the masker sound signal Smc that has reached the area to be masked.

  The adaptive filter 112 outputs the masker sound signal Smc thus calculated (corresponding to the “determination masker sound” of the present invention) to the post processor 122 as described above, and also to the voice analysis unit 130. Output.

  Thus, by using the adaptive filter 112 and the post processor 122 having the above-described configuration, it is possible to extract the masker sound signal Smc that has reached the area to be masked. The adaptive filter 112 and the post processor 122 correspond to the “determination masker sound extraction unit” of the present invention.

  The voice analysis unit 130 compares the speaker voice signal Sbc for determination with the masker sound signal Smc for determination, and determines whether there is a masking effect based on the level of the volume level. Specifically, the presence / absence of the masking effect is determined by the flow from S104 of the following flow onward.

  FIG. 3 is a flowchart showing a process flow for determining whether or not there is a masking effect, and for setting and adjusting masker sound parameters based on the determination result. 4, 5, 6, and 7 are diagrams showing the relationship between the frequency characteristics of the volume level of the masker sound and the volume level of the speaker voice. FIG. 4 shows the case where the volume level of the masker sound is appropriate, the solid line indicates the masker sound signal Smc for determination, and the broken line indicates the speaker voice signal Sbc for determination.

  FIG. 5 shows a case where the volume level of the masker sound is too high in the entire frequency band, the two-dot chain line indicates the unadjusted determination masker sound signal Smc, the broken line indicates the determination speaker voice signal Sbc, and the solid line Indicates the masker sound signal Smo after adjustment. FIG. 5A shows the frequency characteristics before adjustment, and FIG. 5B shows the frequency characteristics after adjustment and the adjustment concept.

  FIG. 6 shows the case where the volume level of the masker sound is low in the entire frequency band, the two-dot chain line indicates the unadjusted determination masker sound signal Smc, the broken line indicates the determination speaker voice signal Sbc, and the solid line indicates The adjusted masker sound signal Smo is shown. 6A shows the frequency characteristics before adjustment, and FIG. 6B shows the frequency characteristics after adjustment and the adjustment concept.

  FIG. 7 shows the case where the volume level of the masker sound is partially low in the frequency domain, the two-dot chain line shows the unadjusted determination masker sound signal Smc, the broken line shows the determination speaker voice signal Sbc, The solid line indicates the masker sound signal Smo after adjustment. FIG. 7A shows frequency characteristics before adjustment, and FIG. 7B shows frequency characteristics after adjustment and an adjustment concept.

  First, when a masker sound is emitted from the speaker SP (S101), the speaker position sound signal Sb is acquired by the speaker position microphone MICb and the masking position sound signal So is acquired by the masking position microphone MICo as described above. (S102). The masker sound emission may be started in response to, for example, power-on of the sound masking device 1 or triggered by a predetermined operation input or detection of a predetermined volume level by the speaker position microphone MICb. Also good.

  Next, as described above, the speaker voice signal Sbc for determination and the masker sound signal Smc for determination are extracted (S103).

  Next, the voice analysis unit 130 performs frequency analysis on the determination speaker voice signal Sbc and the determination masker sound signal Smc by FFT processing or the like, and acquires each frequency component for each signal. The voice analysis unit 130 compares the volume level for each frequency component with respect to the speaker voice signal Sbc for determination and the masker sound signal Smc for determination (S104).

  The voice analysis unit 130 determines whether or not the volume level of the determination masker sound signal Smc is higher than the volume level of the determination speaker voice signal Sbc in all frequency components. At this time, the voice analysis unit 130 determines whether or not the volume level of the determination masker sound signal Smc is equal to or higher than the volume level obtained by adding the margin Ms to the volume level of the speaker voice signal Sbc for determination. The margin Ms is a value experimentally calculated in advance, and is set to about 6 [dB], for example. The margin Ms can be adjusted as appropriate according to the situation.

  As shown in FIGS. 4 and 5, the voice analysis unit 130 has a volume level of the determination masker sound signal Smc equal to or higher than the volume level obtained by adding the margin Ms to the speaker voice signal Sbc for determination in all frequency components. If there is (S105: Yes), it is determined that the masking effect is effective, and the process proceeds to step S106.

  Specifically, as shown in FIG. 4, a value (Gf1-Gf9) obtained by subtracting the volume level of the determination speaker voice signal Sbc from the volume level of the determination masker sound signal Smc of each frequency (f1-f9). ) Is equal to or greater than the margin Ms, or, as shown in FIG. 5A, the volume level of the speaker voice signal Sbc for determination from the volume level of the masker sound signal Smc for determination of each frequency (f1-f9). When the value obtained by subtracting (Gf11−Gf91) is larger than the margin Ms, it is determined that the masking effect is effective.

  On the other hand, as shown in FIGS. 6 and 7, the voice analysis unit 130 has a margin between the volume level of the determination masker sound signal Smc and the volume level of the speaker voice signal Sbc for determination in at least one frequency component. If it is lower than the volume level obtained by adding Ms (S105: No), it is determined that the masking effect is not effective at present, and the process proceeds to step S107.

  Specifically, as shown in FIG. 6A, a value obtained by subtracting the volume level of the speaker voice signal Sbc for determination from the volume level of the masker sound signal Smc for determination of each frequency (f1-f9) ( When Gf12−Gf92) is a negative value (smaller than the margin Ms), or as shown in FIG. 7A, the volume level of the masker sound signal Smc for determination of some frequencies (f1−f5) When the value obtained by subtracting the volume level of the speaker voice signal Sbc for determination (Gf13−Gf53) becomes a negative value (when smaller than the margin Ms), it is determined that the masking effect is not effective.

  If it is determined that there is a masking effect (S105: Yes), the voice analysis unit 130 determines whether the volume level of the determination masker sound signal Smc is within an appropriate volume level range. More specifically, the voice analysis unit 130 determines that the volume level of the masking sound signal Smc for determination is the volume level of the speaker voice signal Sbc for determination at each frequency (f1, f2,..., F9). And the margin Ms as a lower limit, it is determined whether or not it is within the level range Wm. This level range Wm has also been set to a range where the volume level of the masker sound is felt not to be higher than necessary, that is, a range where it is felt that it is not too loud, by an operation input using an operation unit (not shown). Adjustments can be made later as appropriate.

  If the voice analysis unit 130 determines that the volume level of the determination masker sound signal Smc is within an appropriate range for all frequency components (f1-f9) as shown in FIG. 4 (S106: Yes), The sound emission parameter of the original masker sound signal Sm is not changed, and the sound emission parameter change instruction is not output.

  If the voice analysis unit 130 determines that the volume level of the determination masker sound signal Smc is not within an appropriate range for at least one frequency component (S106: No), the process proceeds to step S108.

  Returning to step S105 and determining that the masking effect is not effective (S105: No), the voice analysis unit 130 reads the currently set sound emission parameter and detects whether the volume level has reached the upper limit. To do. If the current sound emission parameter has reached the upper limit volume level (S107: Yes), the voice analysis unit 130 indicates that the volume level of the masker sound is the upper limit and that the masking effect is not effective as it is. Notification is made via the display unit 40 (S112).

  On the other hand, if the current sound emission parameter is not the upper limit volume level (S107: No), the voice analysis unit 130 proceeds to the process of step S108.

  In step S108, when the voice analysis unit 130 determines that the volume level of the determination masker sound signal Smc is not within an appropriate range in all frequency components (S108: Yes), the volume level is similarly applied to all frequency components. The volume is adjusted to adjust (S109). Specifically, as shown in FIG. 5B, when the volume level of the determination masker sound signal Smc is high for all frequency components, the volume level is decreased for all frequency components. At this time, the volume adjustment is performed so that the difference (Gf11C-Gf91C) in the adjusted volume level is within the above-described level range Wm. As shown in FIG. 6B, when the volume level of the determination masker sound signal Smc is low for all frequency components, the volume level is improved for all frequency components. At this time, the volume adjustment is performed so that the difference (Gf12C−Gf92C) in the adjusted volume level is within the above-described level range Wm.

  If the voice analysis unit 130 determines that all frequency components can be adjusted within the level range Wm (S111: Yes), the voice analysis unit 130 outputs the volume parameter Pv to the masker sound adjustment unit 30 as a sound emission parameter.

  On the other hand, if the voice analysis unit 130 determines that all the frequency components cannot be adjusted within the level range Wm (S111: No), the process proceeds to step S110.

  If the specific frequency component does not fall within the level range Wm (S108: No), the voice analysis unit 130 sets the equalizing parameter Pe so that the volume level of the specific frequency falls within the level range Wm (S110). The equalizing parameter Pe and the volume parameter Pv at this time are output to the masker sound adjustment unit 30 as sound emission parameters.

  Specifically, for example, as shown in FIG. 7B, when the volume level of the determination masker sound signal Smc is low in the frequency component (f1-f5), the frequency component (f1-f5) is set. On the other hand, an equalizing process for improving the volume level is performed. At this time, the volume adjustment is performed so that the difference (Gf13C−Gf53C) in the volume level after adjustment falls within the level range Wm.

  The masker sound adjusting unit 30 includes an equalizing unit 301 and a volume adjusting unit 302. The equalizing unit 301 performs an equalizing process for adjusting the volume level of the specific frequency component of the original masker sound signal Sm according to the equalizing parameter Pe of the sound emission parameter. The volume adjustment unit 302 performs a volume adjustment process for adjusting the volume level of all frequency components of the original masker sound signal Sm after the equalizing process, based on the volume parameter Pv of the sound emission parameter. The adjusted masker sound signal Smo is output to the speaker SP and emitted from the speaker SP.

  By performing the processing as described above, it is possible to accurately determine whether or not the speaker voice is masked by the masker sound. Furthermore, based on the determination result, a masker sound that can reliably mask the speaker voice can be emitted. Furthermore, it is possible to automatically adjust so that the masker sound is emitted at an optimum level. As a result, the speaker voice can be reliably masked without causing discomfort to a third party or the like.

  Next, a sound masking apparatus 1A according to a second embodiment will be described with reference to the drawings. Unlike the sound masking device 1 of the first embodiment, the sound masking device 1A of the present embodiment receives a plurality of speaker position sound signals Sb1, Sb2 and a plurality of masking position sound signals So1, So2, So3. FIG. 8 is a schematic installation configuration diagram of the sound masking apparatus 1A according to the present embodiment. FIG. 9 is a block diagram showing the configuration of the sound masking apparatus 1A according to this embodiment.

  Note that the masker sound adjustment unit 30A of the sound masking device 1A shown in the present embodiment includes an equalizing unit and a volume adjustment unit similar to the masker sound adjustment unit 30 shown in the first embodiment, and performs the same processing. Therefore, detailed functional block illustrations and processing descriptions are omitted.

  The sound masking apparatus 1A includes a main body 2A, speaker position microphones MICb1, MICb2, masking position microphones MICo1, MICo2, Mico3, speakers SP1, SP2, SP3, and display units 40A, 40B.

  The speaker position microphone MICb1 is installed on the table 901A of the counter 900, and is connected to the main body 2A. The speaker position microphone MICb1 collects the conversations of the specific speakers 910A and 920A and outputs the speaker position voice signal Sb1 to the main body 2A.

  The speaker position microphone MICb2 is installed on the table 901B of the counter 900 and is connected to the main body 2A. The speaker position microphone MICb2 collects the conversations of the specific speakers 910B and 920B and outputs the speaker position voice signal Sb2 to the main body 2A.

  The speakers SP1, SP2, SP3 are installed in the vicinity of the counter 900 and are connected to the main body 2A. The speakers SP1, SP2, SP3 emit the adjusted masker sound signal Smo. At this time, the speakers SP1, SP2, and SP3 emit the adjusted masker sound signal Smo toward an area where masking is desired including an area where a plurality of third parties 930 are present.

  Masking position microphones MICo1, MICo2, and Mico3 are installed at appropriate intervals in an area where masking is to be performed, including areas where a plurality of third parties 930 are present, and are connected to main body 2A. The masking position microphones MICo1, MICo2, and Mico3 respectively pick up sounds in the installation area and output masking position sound signals So1, So2, and So3 to the main body 2A.

  The display units 40A and 40B are constituted by an indicator lamp, a liquid crystal display, and the like, and are connected to the main body 2A. The display units 40A and 40B individually perform lighting or specific display according to a masking effect determined by a method described later. The display units 40A and 40B may be integrated with the speaker position microphones MICb1 and MICb2.

  Next, the configuration of the main body 2A that determines the masking effect will be described. As shown in FIG. 9, the main body 2A includes a masker sound measuring unit 10A, a masker sound generating unit 20, a masker sound storage unit 21, and a masker sound adjusting unit 30A corresponding to the “masker sound measuring device” of the present invention. Since the masker sound generation unit 20 and the masker sound storage unit 21 are the same as those in the first embodiment, description thereof is omitted. The original masker sound Sm from the masker sound generation unit 20 is output to the adaptive filters 112A, 112B, and 112C.

  The masker sound measurement unit 10A includes adaptive filters 111A, 111B, 111C, 111D, 111E, 111F, 112A, 112B, 112C, post processors 121A, 121B, 121C, 121D, 121E, 121F, 122A, 122B, 122C, and voice analysis. Part 130A.

  The adaptive filter 111A corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb1. The post processor 121A corresponds to the “first post processor” of the present invention, and receives the masking position sound signal So1. The post processor 121A subtracts the speaker position audio signal Sb1 (corresponding to the speaker audio signal Sbc1A) filtered by the adaptive filter 111A from the masking position audio signal So1, and outputs a first subtraction signal Sn11. The first subtraction signal Sn11 is input to the post processor 121B and fed back to the adaptive filter 111A.

  The adaptive filter 111A outputs the speaker voice signal Sbc1A calculated in this way (corresponding to the “speaker voice for determination” of the present invention) to the post processor 121A as described above, and the voice analysis unit Output to 130A.

  As described above, by using the adaptive filter 111A having the above-described configuration and the post processor 121A, it is possible to extract the speaker voice signal Sbc1A that has reached the masking position microphone MICo1 in the region to be masked. Note that the adaptive filter 111A and the post processor 121A correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 111B corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb2. The post processor 121B corresponds to the “first post processor” of the present invention, and receives the first subtraction signal Sn11. The post processor 121B subtracts the speaker position voice signal Sb2 (corresponding to the speaker voice signal Sbc2A) filtered by the adaptive filter 111B from the first subtraction signal Sn11, and outputs a second subtraction signal Sn12. The second subtraction signal Sn12 is input to the post processor 122A and fed back to the adaptive filter 111B.

  The adaptive filter 111B outputs the speaker voice signal Sbc2A calculated in this way (corresponding to the “speaker voice for determination” of the present invention) to the post processor 121B as described above, and the voice analysis unit Output to 130A.

  As described above, by using the adaptive filter 111B having the above-described configuration and the post processor 121B, the speaker voice signal Sbc2A that has reached the masking position microphone MICo1 in the region to be masked can be extracted. The adaptive filter 111B and the post processor 121B correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 112A corresponds to the “second adaptive filter” of the present invention, and receives the original masker sound signal Sm.

  The post processor 122A corresponds to the “second post processor” of the present invention, and receives the second subtraction signal Sn12 from the post processor 121B. The post processor 122A subtracts the original masker sound signal Sm filtered by the adaptive filter 112A from the second subtraction signal Sn12, and outputs a third subtraction signal Sn13. The third subtraction signal Sn13 is fed back to the adaptive filter 112A.

  The adaptive filter 112A outputs the masker sound signal Smc1 calculated in this way (corresponding to the “determination masker sound” of the present invention) to the voice analysis unit 130A.

  As described above, by using the adaptive filter 112A having the above-described configuration and the post processor 122A, it is possible to extract the masker sound signal Smc1 that has reached the masking position microphone MICo1 in the region to be masked. The adaptive filter 112A and the post processor 122A correspond to the “determination masker sound extraction unit” of the present invention.

  The adaptive filter 111C corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb1. The post processor 121C corresponds to the “first post processor” of the present invention, and receives the masking position sound signal So2. The post processor 121C subtracts the speaker position voice signal Sb1 (corresponding to the speaker voice signal Sbc1B) filtered by the adaptive filter 111C from the masking position voice signal So2, and outputs a fourth subtraction signal Sn21. The fourth subtraction signal Sn21 is input to the post processor 121D and fed back to the adaptive filter 111C.

  The adaptive filter 111C outputs the speaker voice signal Sbc1B calculated in this way (corresponding to the “speaker voice for determination” of the present invention) to the post processor 121C as described above, and the voice analysis unit Output to 130A.

  As described above, by using the adaptive filter 111C having the above-described configuration and the post processor 121C, the speaker voice signal Sbc1B that has reached the masking position microphone MICo2 in the region to be masked can be extracted. The adaptive filter 111C and the post processor 121C correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 111D corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb2. The post processor 121D corresponds to the “first post processor” of the present invention, and receives the fourth subtraction signal Sn21. The post processor 121D subtracts the speaker position voice signal Sb2 (corresponding to the speaker voice signal Sbc2B) filtered by the adaptive filter 111D from the fourth subtraction signal Sn21, and outputs a fifth subtraction signal Sn22. The fifth subtraction signal Sn22 is input to the post processor 122B and fed back to the adaptive filter 111D.

  The adaptive filter 111D outputs the speaker voice signal Sbc2B (corresponding to the “determination speaker voice” of the present invention) calculated in this way to the post processor 121D as described above, and also the voice analysis unit Output to 130A.

  As described above, by using the adaptive filter 111D having the above-described configuration and the post processor 121D, the speaker voice signal Sbc2B that has reached the masking position microphone MICo2 in the region to be masked can be extracted. The adaptive filter 111D and the post processor 121D correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 112B corresponds to the “second adaptive filter” of the present invention, and receives the original masker sound signal Sm.

  The post processor 122B corresponds to the “second post processor” of the present invention, and receives the fifth subtraction signal Sn22 from the post processor 121D. The post processor 122B subtracts the original masker sound signal Sm filtered by the adaptive filter 112B from the fifth subtraction signal Sn22, and outputs a sixth subtraction signal Sn23. The sixth subtraction signal Sn23 is fed back to the adaptive filter 112B.

  The adaptive filter 112B outputs the masker sound signal Smc2 calculated in this way (corresponding to the “determination masker sound” of the present invention) to the voice analysis unit 130A.

  As described above, by using the adaptive filter 112B and the post processor 122B having the above-described configuration, it is possible to extract the masker sound signal Smc2 that has reached the masking position microphone MICo2 in the region to be masked. The adaptive filter 112B and the post processor 122B correspond to the “determination masker sound extraction unit” of the present invention.

  The adaptive filter 111E corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb1. The post processor 121E corresponds to the “first post processor” of the present invention, and receives the masking position sound signal So3. The post processor 121E subtracts the speaker position voice signal Sb1 (corresponding to the speaker voice signal Sbc1C) filtered by the adaptive filter 111E from the masking position voice signal So3, and outputs a seventh subtraction signal Sn31. The seventh subtraction signal Sn31 is input to the post processor 121F and fed back to the adaptive filter 111E.

  The adaptive filter 111E outputs the speaker voice signal Sbc1C calculated in this way (corresponding to the “determination speaker voice” of the present invention) to the post processor 121E as described above, and the voice analysis unit Output to 130A.

  As described above, by using the adaptive filter 111E having the above-described configuration and the post processor 121E, it is possible to extract the speaker voice signal Sbc1C that has reached the masking position microphone MICo3 in the region to be masked. The adaptive filter 111E and the post processor 121E correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 111F corresponds to the “first adaptive filter” of the present invention, and receives the speaker position voice signal Sb2. The post processor 121F corresponds to the “first post processor” of the present invention, and receives the seventh subtraction signal Sn31. The post processor 121F subtracts the speaker position voice signal Sb2 (corresponding to the speaker voice signal Sbc2C) filtered by the adaptive filter 111F from the seventh subtraction signal Sn31, and outputs an eighth subtraction signal Sn32. The eighth subtraction signal Sn32 is input to the post processor 122C and fed back to the adaptive filter 111F.

  The adaptive filter 111F outputs the speaker voice signal Sbc2C calculated in this way (corresponding to the “speaker voice for determination” of the present invention) to the post processor 121F as described above, and the voice analysis unit Output to 130A.

  As described above, by using the adaptive filter 111F having the above-described configuration and the post processor 121F, the speaker voice signal Sbc2C that has reached the masking position microphone MICo3 in the region to be masked can be extracted. The adaptive filter 111F and the post processor 121F correspond to the “determination speaker voice extraction unit” of the present invention.

  The adaptive filter 112C corresponds to the “second adaptive filter” of the present invention, and receives the original masker sound signal Sm.

  The post processor 122C corresponds to the “second post processor” of the present invention, and receives the eighth subtraction signal Sn32 from the post processor 121F. The post processor 122C subtracts the original masker sound signal Sm filtered by the adaptive filter 112C from the eighth subtraction signal Sn32, and outputs a ninth subtraction signal Sn33. The ninth subtraction signal Sn33 is fed back to the adaptive filter 112C.

  The adaptive filter 112C outputs the masker sound signal Smc3 calculated in this way (corresponding to the “determination masker sound” of the present invention) to the voice analysis unit 130A.

  As described above, by using the adaptive filter 112C having the above-described configuration and the post processor 122C, it is possible to extract the masker sound signal Smc3 that has reached the masking position microphone MICo3 in the region to be masked. The adaptive filter 112C and the post processor 122C correspond to the “determination masker sound extraction unit” of the present invention.

  The voice analysis unit 130A compares the speaker voice signals Sbc1A and Sbc2A for determination with the masker sound signal Smc1 for determination, and determines the presence or absence of the masking effect based on the level of the volume level. Specifically, the determination is made by the same method as in the first embodiment. Thereby, the masking effect at the point of the masking position microphone MICo1 in the region to be masked can be determined.

  The voice analysis unit 130A compares the speaker voice signals Sbc1B and Sbc2B for determination with the masker sound signal Smc2 for determination, and determines the presence or absence of the masking effect based on the level of the volume level. Specifically, the determination is made by the same method as in the first embodiment. Thereby, the masking effect at the point of the masking position microphone MICo2 in the region to be masked can be determined.

  The voice analysis unit 130A compares the speaker voice signals Sbc1C and Sbc2C for determination with the masker sound signal Smc3 for determination, and determines the presence or absence of the masking effect based on the level of the volume level. Specifically, the determination is made by the same method as in the first embodiment. Thereby, the masking effect at the point of the masking position microphone MICo3 in the region to be masked can be determined.

  The voice analysis unit 130A sets sound emission parameters to the masker sound adjustment unit 30A based on the analysis results of these masking effects. Thereby, a masker sound can be emitted so that the masking effect with respect to all the speaker voices becomes effective over the entire region to be masked.

  In the second embodiment, two types of speaker voices and three masking position microphones are installed. However, each of the microphones has another number, and the above-described configuration and processing are applied to perform masking. The masker sound can be automatically adjusted so that the effect judgment and the masking effect become effective.

  In the above description, an example in which a speaker speech post-processor and a masker sound post-processor are individually provided has been described, but these may be unified. Thereby, the number of components can be reduced. Further, the connection relationship between the speaker voice post processor and the masker sound post processor may be reversed.

  In the above description, a configuration is shown in which output signals from the post processors are fed back to the corresponding adaptive filters. However, the configuration may be such that the final output signals of the post processor groups arranged in series are fed back to an adaptive filter that gives signals to these post processor groups. Specifically, for example, in the case of the first embodiment (see FIG. 2), the first subtraction signal Sn1 output from the post processor 121 is output from the post processor 122 instead of being fed back to the adaptive filter 111. A configuration in which the second subtraction signal Sn2 is fed back to the adaptive filters 111 and 112 may be used. Further, for example, in the case of the second embodiment (see FIG. 9), the first subtraction signal Sn11 output from the post processor 121A and the second subtraction signal Sn12 output from the post processor 121B are applied to the adaptive filters 111A and 111B, respectively. The third subtraction signal Sn13 output from the post processor 122A may be fed back to the adaptive filters 111A, 111B, and 122A.

  In the above description, an example in which it is determined whether the volume level of the determination masker sound signal Smc is within an appropriate range for each frequency component. However, the volume level including the entire frequency band is appropriate for masking. It is good also as a judgment standard whether it is in a safe range. For example, the following two methods may be used.

  (A) The average value of the sound volume levels of the masking sound signals Smc for determination obtained by picking up sound with all the installed speaker position microphones is calculated. When the average value of the sound volume level is within a range appropriate for masking, it is determined that the sound volume is masked.

  (B) When the volume level of the determination masker sound signal Smc obtained by picking up the sound at each speaker position microphone is within a range appropriate for masking, it is determined that the masking is performed.

  (C) The determination masker sound signal Smc having the minimum volume level is acquired from the volume levels of the plurality of determination masker sound signals Smc obtained by collecting the sounds at the respective speaker position microphones. When the volume level of the minimum determination masker sound signal Smc is equal to or higher than the minimum volume level necessary for masking, it is determined that the masking sound signal Smc is masked.

  Even if these methods are used, the masking effect can be reliably determined as described above.

  In each of the above-described methods for determining based on the volume level, the volume level may be adjusted for each method, for example, by the following method.

  (AA) In the case of (A) described above, the sound emission levels of all the arranged speakers are uniformly changed, and the average value of the sound volume levels is adjusted to be within a range suitable for masking.

  (AB) In the case of (B) described above, the sound emission level of the speaker that is close in distance to the microphone that is not in the appropriate range is varied, and the volume level of all the judgment masker sound signals Smc is masked. Adjust within the appropriate range.

  (AC) In the case of (C) described above, the sound emission level of all the arranged speakers is uniformly changed so that the volume level of the minimum judgment masker sound signal Smc falls within the appropriate range for masking. Adjust to. Further, the sound emission level of the speaker that is close in distance to the microphone that is determined to be the minimum determination masker sound signal Smc is increased, and the volume level of the minimum determination masker sound signal Smc is in a range suitable for masking. Adjust so that it is inside. In these cases, it is preferable to adjust so that the volume levels of all the determination masker sound signals Smc are within a range suitable for masking.

1, 1A: Sound masking device, 2, 2A: Main body,
10, 10A: Masker sound measuring unit, 20: Masker sound generating unit, 21: Masker sound storage unit, 30, 30A: Masker sound adjusting unit, 40, 40A, 40B: Display unit,
111, 111A, 111B, 111C, 111D, 111E, 111F, 112, 112A, 112B, 112C: Adaptive filter, 121, 121A, 121B, 121C, 121D, 121E, 121F, 122, 122A, 122B, 122C: Post processor , 130: voice analysis unit, 301: equalizing unit, 302: volume adjusting unit 900: counter, 901: table, 910, 910A, 910B, 920, 920A, 920B: specific speaker, 930: third party MICb, MICb1, MICb2: Speaker position microphone, MICo, MICo1, MICo2, MICo3: Masking position microphone, SP, SP1, SP2, SP3: Speaker

Claims (7)

Speaker position sound that is sound collected at the speaker position, masker sound emitted to mask the speaker sound, and sound picked up in the area where the speaker sound should be masked An extraction unit that extracts a determination speaker voice that is a speaker voice in the area to be masked and a determination masker sound that is a masker sound in the area to be masked using a masking position voice When,
A masker sound measuring apparatus comprising: a voice analysis unit that compares the determination speaker voice and the determination masker sound and analyzes whether masking is performed. The masker sound measuring device according to claim 1,
The voice analysis unit
The determination speaker voice and the determination masker sound are compared for each predetermined frequency component, and the determination masker sound is equal to or higher than a predetermined volume level with respect to the determination speaker voice in all frequency components. Masker sound measuring device that determines that it is masked in some cases. The masker sound measuring device according to claim 2,
The voice analysis unit
A masker that determines that masking is appropriately masked when the determination masker sound is equal to or higher than a predetermined volume level and within a predetermined volume level range with respect to the determination speaker voice in all frequency components. Sound measuring device. A masker sound measuring device according to any one of claims 1 to 3,
The extraction unit includes:
A first adaptive filter to which the speaker position voice is input, a speaker position voice filtered by the first adaptive filter is subtracted from the masking position voice, and a first subtraction signal is subtracted from the first adaptive type. A determination speaker voice extraction unit configured to output the filtered speaker position voice as the determination speaker voice, the first post processor feeding back to the filter;
The second adaptive filter to which the masker sound is input, the masker sound filtered by the second adaptive filter is subtracted from the masking position sound, and the second subtracted signal is fed back to the second adaptive filter. A determination masker sound extraction unit configured by a second post-processor and outputting the filtered masker sound as the determination masker sound;
A masker sound measuring device comprising: A masker sound measuring device according to any one of claims 1 to 4,
A speaker position microphone for picking up the speaker position voice;
A masking position microphone for picking up the masking position sound;
A masker sound measuring apparatus comprising: a masker sound generating unit that generates the masker sound. A masker sound measuring device according to any one of claims 1 to 5,
A masker sound measuring apparatus comprising a notifying unit for notifying an analysis result by the voice analyzing unit. A masker sound measuring device according to any one of claims 1 to 6,
A masker sound adjusting unit that adjusts the volume level of the masker sound according to a sound emission parameter set so that a masker sound of an appropriate volume level is emitted from the analysis result by the voice analysis unit;
A sound masking device comprising: a speaker that emits an adjusted masker sound.

Images