JP2008109491A - Air check device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an air check device capable of automatically extracting and recording a piece of music from various sounds broadcast from a broadcasting station. <P>SOLUTION: Since all sound signals of FM broadcasting are received and stored, a plurality of the same pieces of music are generally included in the sound signals. Since an MC is different in each case even though a piece of music is the same in each case, a plurality of the same pieces of music in which sound signals continuously coincide for a predetermined period or longer are searched out, and the longest piece of music is stored among the same pieces of music. To search a piece of music, starts of a plurality of pieces of music are first specified by detecting vocal components, and in addition, level change rates of sound signals which continuously coincide for the predetermined period or longer are specified as a piece of music. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an air check device, and more particularly, to an air check device that can automatically extract and record music from various sounds broadcast from a broadcasting station.

  Air check is generally used by recording music broadcast on radio or television on a storage medium, and playing the recorded music later for enjoyment. Has been.

  However, the viewer does not know when the desired music will be broadcast. For example, in FM radio broadcasting, it may be known in advance that a desired song is broadcasted in a program, but since CM and DJ are included in the program, when the desired song is broadcast in the program I do not know until. Therefore, in order to record the desired music, the viewer must listen carefully to the program from the beginning.

  Japanese Patent Application Laid-Open No. 11-177511 (Patent Document 1) discloses a digital broadcast receiving and reproducing apparatus for reproducing a song being broadcast or broadcast from a performance start point in a digital broadcast music program. This apparatus uses the control data multiplexed in the program data stream and transmitted at the start and end of the performance of each broadcast song, and the recording start point for recording the audio data as a recorded song for each song. And the recording end point. However, this apparatus can be used only for digital broadcasting in which control data is multiplexed.

  Japanese Patent Laid-Open No. 11-73730 (Patent Document 2) discloses a broadcast recording apparatus that can easily record music when performing an air check, and can easily reproduce desired music from the recorded music. is doing. This device sets a recording start time with a timer, records the received audio data when the set recording start time is reached, stores on-air information including the song name and broadcast start time information of the song, After the air check is finished, the playback start position on the recording medium is set from the broadcast start time information and the recording start time, and a playback operation from the playback start position is instructed. In addition, playback operation can be performed from the list display of song names. However, this apparatus can also be used only for text multiplex broadcasting in which on-air information is multiplexed.

In addition, this applicant refers to the cue sheet in which the title and broadcast time of the music are described in Japanese Patent Application No. 2005-285583 (Patent Document 4) and Japanese Patent Application No. 2005-340069 (Patent Document 5). An air check system that can automatically record music is disclosed. However, this system cannot automatically record desired music without a cue sheet.
JP-A-11-177511 JP-A-11-73730 JP 2001-76461 A Japanese Patent Application No. 2005-285583 (unpublished prior application) Japanese Patent Application No. 2005-340069 (unpublished prior application)

  An object of the present invention is to provide an air check device capable of automatically extracting and recording music from various sounds broadcast from a broadcasting station without the above-described control data, on-air information, cue sheets, and the like. Is to provide.

Means for Solving the Problems and Effects of the Invention

  An air check device according to the present invention includes receiving means, broadcast storage means, music search means, music selection means, and music storage means. The receiving means receives an audio signal broadcast from a broadcasting station. The broadcast storage means stores the audio signal received by the receiving means. The music search means searches the audio signals stored by the broadcast storage means for a plurality of music periods in which the audio signals continuously match each other for a predetermined period or more. The music selection means selects one, preferably the longest, of the plurality of music periods searched by the music search means. The music storage means stores the audio signal of the music period selected by the music selection means.

  According to the present invention, a plurality of music periods in which audio signals coincide with each other continuously for a predetermined period or longer are searched, one music period is selected, and the audio signal of that music period is stored.

  Preferably, the audio signal includes a left channel audio signal and a right channel audio signal. The music searching means includes an inverting means, a synthesizing means, a pulsing means, and a pulse matching means. The inverting means inverts one of the left channel audio signal and the right channel audio signal. The synthesizing unit synthesizes the signal inverted by the inverting unit and the other of the left channel audio signal and the right channel audio signal to generate a synthesized audio signal. The pulsing unit compares the level of the synthesized speech signal generated by the synthesizer with a predetermined level, and outputs a pulse signal having a first logic level when the level of the synthesized speech signal is greater than the predetermined level. When the level is lower than the predetermined level, a pulse signal of the second logic level is output. The pulse matching means identifies a period in which the logic level continuously matches for a predetermined period or more among the pulse signals output from the pulsing means as the music period.

  In this case, the vocal component included in the music period is canceled by the inverting means and the synthesizing means, and the synthesized speech signal becomes almost zero. At this time, the pulse signal output by the pulsing means becomes the second logic level when the synthesized speech signal is substantially zero, and becomes the first logic level otherwise. Since a period in which the logic levels of such pulse signals continuously coincide with each other for a predetermined period or more is specified as a music period, an audio signal in the music period is easily extracted.

  Preferably, the music searching means includes sampling means, change rate calculating means, and level matching means. The sampling means samples the level of the audio signal stored by the broadcast storage means every predetermined period. The change rate calculation means calculates the level change rate by dividing the level sampled by the sampling means by the level sampled before that. The level matching means specifies a period in which the level change rate continuously matches a predetermined period or more from the level change rates calculated by the change rate calculation means as the music period.

  In this case, since the period in which the level change rate of the audio signal continuously matches for a predetermined period or more is specified as the music period, the audio signal of the music period is accurately extracted.

  Preferably, the music search means further includes a first midrange component, a second midrange component higher than the first midrange component, and a second midrange component from the audio signal stored by the broadcast storage means. Extraction means for extracting a third mid-range component higher than the mid-range component and providing it to the sampling means is included.

  In this case, since the level change rates are compared at least at three types of frequencies, the sound signal of the music period is extracted more accurately.

  Preferably, the air check device further includes non-music storage means for storing an audio signal other than the music period selected by the music selection means.

  In this case, separately from the music, an MC or the like including information related to the music is also recorded.

  Preferably, the audio signal includes a left channel audio signal and a right channel audio signal. The music searching means includes an inverting means, a synthesizing means, a pulsing means, a pulse matching means, a synchronizing means, a sampling means, a change rate calculating means, and a level matching means. The inverting means inverts one of the left channel audio signal and the right channel audio signal. The synthesizing unit synthesizes the signal inverted by the inverting unit and the other of the left channel audio signal and the right channel audio signal to generate a synthesized audio signal. The pulsing unit compares the level of the synthesized speech signal generated by the synthesizer with a predetermined level, and outputs a pulse signal having a first logic level when the level of the synthesized speech signal is greater than the predetermined level. When the level is lower than the predetermined level, a pulse signal of the second logic level is output. The pulse matching means specifies a period in which the logic levels of the pulse signals output from the pulsing means coincide continuously for a predetermined period or more as a pulse coincidence period. The synchronizing means synchronizes the audio signal so that the beginning of the pulse matching period specified by the pulse matching means is aligned. The sampling means samples the level of the audio signal stored by the broadcast storage means every predetermined period. The change rate calculation means calculates the level change rate by dividing the level sampled by the sampling means by the level sampled before that. The level matching means compares the level change rates of the audio signals synchronized by the synchronization means at the same timing, so that the level change rate is equal to or greater than a predetermined period from the level change rates calculated by the change rate calculation means. A period that coincides continuously is specified as a music period.

  In this case, the period in which the logic level of the pulse signal continuously matches for a predetermined period or more in the first stage is roughly specified as the music period, and the level change rate of the audio signal continuously matches for a predetermined period or more in the second stage. Since the period is specified as the music period, the audio signal of the music period is easily and accurately extracted.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

  Referring to FIG. 1, an air check device according to an embodiment of the present invention includes an FM (Frequency Modulation) tuner 10, an AD (Analog to Digital) converter (ADC) 11, and an MP3 (MPEG Audio Layer-3) encoder. 12 and an MP3 data file 13.

  The FM tuner 10 receives an analog audio signal broadcast from an FM broadcast station. The analog audio signal is stereo and includes a right channel audio signal and a left channel audio signal. The AD converter 11 converts the analog audio signal received by the FM tuner 10 into a digital audio signal. The MP3 encoder 12 encodes the digital audio signal output from the AD converter 11 into MP3 data. The MP3 data file 13 stores MP3 data output from the MP3 encoder 12.

  The FM tuner 10, the AD converter 11 and the MP3 encoder 12 operate in principle, and record all broadcast contents from the FM broadcast station in the MP3 data file 13. Therefore, in the MP3 data file 13, not only music but also MC (Master of Ceremony) such as DJ is recorded together. The MP3 data is saved as one file for each program, for example. As described above, since all FM stereo broadcast programs are stored in the MP3 data file 13 as MP3 data, there is a possibility that the same music is repeatedly recorded several times. The embodiment of the present invention pays attention to the fact that the music is the same every time while the MC is different every time. By searching for the same data in the MP3 data file 13, only the music is automatically found. is there.

  The air check device further includes an MP3 decoder 14 and a DA (Digital to Analog) converter (DAC) 15. The MP3 decoder 14 decodes the MP3 data stored in the MP3 data file 13 into a digital audio signal. The DA converter 15 converts the digital audio signal output from the MP3 decoder 14 into an analog audio signal.

  The air check device further includes bandpass filters (BPF) 16 and 17, an inverter 18, a synthesizer 19, a pulse generator 20, a pulse data generation module 21, A data file 22.

  The band pass filter 16 extracts a mid range (for example, 800 to 3 kHz) from the left channel audio signal output from the DA converter 15. The band pass filter 17 extracts a mid range (for example, 800 to 3 kHz) from the right channel audio signal output from the DA converter 15. The inverter 18 inverts the mid-range left channel audio signal output from the bandpass filter 16. The synthesizer 19 synthesizes (adds) the mid-range left channel audio signal inverted by the inverter 18 and the mid-range right channel audio signal output from the bandpass filter 17 to generate a synthesized audio signal. .

  An example of the synthesized speech signal is shown in FIG. The synthesized audio signals C1 to C4 shown here are generated from audio signals of different programs. The music includes a period in which only the vocal component VO is included. In such a period, the audio signals of both channels are canceled and the synthesized audio signals C1 to C4 become zero. The timing of the MC component MC is random, while the timing of the vocal component VO is the same even if the program is different. In the example shown in FIG. 2, in any of the synthesized speech signals C1 to C4, the vocal component VO of 500 milliseconds appears twice with an interval of 1 second. Therefore, there is a high possibility that the same music is recorded during this period.

  The pulsator 20 compares the level of the synthesized speech signals C1 to C4 generated by the synthesizer 19 with a predetermined level. When the level of the synthesized speech signals C1 to C4 is greater than the predetermined level, the pulsator 20 is at the H (logic high) level. Pulse signals P1 to P4 are output, and L (logic low) level pulse signals P1 to P4 are output when the levels of the synthesized speech signals C1 to C4 are lower than a predetermined level. In the example shown in FIG. 2, the pulse signals P1 to P4 are at the L level during a period in which the vocal component VO is included, and are at the H level in other periods. The predetermined level is set to 0 V in an ideal case where the left and right vocal components VO are completely canceled, but is actually set to, for example, 500 mV in consideration of a margin.

The pulse data generation module 21 is a computer program, and generates pulse data PD1 to PD4 as shown in the following Table 1 based on the pulse signals P1 to P4 output from the pulse generator 20.

  In the pulse data PD1 to PD4, when the pulse signals P1 to P4 are at the H level, “1” indicating the H level is associated with a period (milliseconds) in which the pulse signals P1 to P4 are at the H level. When P1 to P4 are at the L level, “0” indicating the L level is associated with a period in which the pulse signals P1 to P4 are at the L level. The pulse data file 22 stores the pulse data PD1 to PD4 generated by the pulse data generation module 21.

  By matching the pulse data PD1 to PD4 with each other, it is possible to specify the period during which the same music is recorded. However, there is a problem that music and MC cannot be accurately separated only by such a pulse matching method. Hereinafter, this problem will be described.

  For example, when the music is faded in from the middle of the MC, as shown in FIG. 3A, the audio signal A1 includes the MC component MC and the music component BGM in the same period. Since the MC component MC cancels out, it does not appear in the synthesized speech signal C1, but the synthesized speech signal C1 does not become zero because of the music component BGM. When the music is started after the end of MC, as shown in FIG. 3B, the audio signal A2 has a period in which only the MC component MC is included, and then only the music component BGM is included. There is a period. In the period in which only the MC component is included, the MC component MC cancels out, so the synthesized speech signal C2 becomes zero. Further, when MC enters in the middle of the music, as shown in FIG. 3C, the audio signal A3 includes the MC component MC and the music component BGM in the same period. Since the MC component MC cancels out, it does not appear in the synthesized audio signal C3, but the synthesized audio signal C3 does not become zero because of the music component BGM.

  When these synthesized speech signals C1 to C3 are pulsed by the pulse generator 20, pulse signals P1 to P3 can be obtained. In this example, the period in which the pulse signals P1 to P3 coincide with each other is at least 3 seconds. Therefore, if MP3 data corresponding to this period is extracted, it is possible to cut out only a piece of music, but there are some in which only the music component BGM is included in the first second as shown in FIG. As shown in FIGS. 3A and 3C, there are some which include not only the music component BGM but also the MC component MC. As described above, since the music and MC cannot be accurately separated only by the pulse matching method, it is preferable to use the level matching method described later together.

  In order to realize the level matching method, the air check device further includes bandpass filters (BPF) 23 to 25, AD converters 26 to 28, level change rate calculation modules 29 to 31, and level data generation modules 32 to 34. And a level data file 35. The band pass filter 23 extracts a mid-frequency component of 500 Hz from the audio signal output from the DA converter 15. The bandpass filter 24 extracts a mid-frequency component of 1 kHz from the audio signal output from the DA converter 15. The band pass filter 25 extracts a midrange component of 2 kHz from the audio signal output from the DA converter 15.

  The AD converter 26 samples the level of the 500 Hz audio signal output from the bandpass filter 23 every predetermined period (for example, 1 second), and AD-converts it into level data. The AD converter 27 samples the level of the 1 kHz audio signal output from the bandpass filter 24 every predetermined period (for example, 1 second), and AD-converts it into level data. The AD converter 28 samples the level of the 2 kHz audio signal output from the bandpass filter 25 every predetermined period (for example, 1 second), and AD-converts it into level data.

  The level change rate calculation module 29 calculates the level change rate for each predetermined period by dividing the level data sampled by the AD converter 26 by the level data sampled before the predetermined period. The level change rate calculation module 30 calculates the level change rate for each predetermined period by dividing the level data sampled by the AD converter 27 by the level data sampled before the predetermined period. The level change rate calculation module 31 divides the level data sampled by the AD converter 28 by the level data sampled before a predetermined period before calculating the level change rate for each predetermined period.

The level data generation module 32 arranges the level data output from the AD converter 26 every predetermined time in order, and arranges the level change rates calculated by the level change rate calculation module 29 in order in association with the level data. The level data generation module 33 arranges the level data output from the AD converter 27 every predetermined time in order, and arranges the level change rates calculated by the level change rate calculation module 30 in order in association with the level data. The level data generation module 34 arranges the level data output from the AD converter 28 every predetermined time in order, and arranges the level change rates calculated by the level change rate calculation module 31 in order in association with the level data. The level data file 35 stores level data generated by the level data generation modules 32 to 34. An example of the level data is shown in Table 2 below.

  Level data of 500 Hz output from the AD converter 26 every predetermined time is sequentially arranged in the upper stage, level data of 1 kHz output from the AD converter 27 every predetermined time is sequentially arranged in the middle stage, and AD data is arranged in the lower stage. Level data of 2 kHz output from the converter 28 every predetermined time is arranged in order.

  The level change rates calculated by the level change rate calculation module 29 are arranged in order under the 500 Hz level data, and the level change rates calculated by the level change rate calculation module 30 are listed under the 1 kHz level data. The level change rates calculated by the level change rate calculation module 31 are arranged in order under the 2 kHz level data. For example, among the 500 Hz level data, a change rate of 75% (= 150/200 × 100) is recorded between the 200 mV level data and the 150 mV level data one second later.

  Level data LD1 is obtained from the audio signal A1 shown in FIG. 3 (a), level data LD2 is obtained from the audio signal A2 shown in FIG. 3 (b), and from the audio signal A3 shown in FIG. 3 (c). Level data LD3 is obtained. When the level change rates of the level data LD1 to LD3 are compared with each other, the level change rates surrounded by a thick frame in Table 2 are almost the same. Therefore, when MC is included immediately after the start of music as shown in FIG. 3A, the level change rate does not match during that period, so the music is not extracted. Also, as shown in FIG. 3C, when MC is included in the middle of the music, the level change rate does not match during that period, so the music is not extracted.

  According to such a level matching method, music and MC can be separated more accurately than the pulse matching method described above. Further, although it is possible to separate music and MC using only the level matching method, the amount of computation becomes very large only by the level matching method. Therefore, it is preferable to use these methods in combination, first specify the music roughly by the pulse matching method, and then specify the music in detail by the level matching method.

  The air check device further includes a separation module 36, a music data file 37, and an MC (Master of Ceremony) data file 38. The separation module 36 is a computer program that performs pulse matching processing based on the pulse data stored in the pulse data file 22 and performs level matching processing based on the level data stored in the level data file 35. The MP3 data stored in the MP3 data file 13 is separated into music data and other MC data (non-music data). The music data file 37 stores music data output from the separation module 36. The MC data file 38 stores MC data output from the separation module 36.

  The processing contents of the separation module 36 are shown in FIG. The separation module 36 compares the pulse data with each other to find a period (hereinafter referred to as “pulse coincidence period”) in which the pulses continuously match for a predetermined time (for example, 2 seconds) (S1). In the example shown in FIG. 2 and Table 1, among the four pulse data PD1 to PD4, the L level for 500 milliseconds, the H level for 1 second, and the L level for 500 milliseconds continuously match, and Since the total period is 2 seconds, this period is specified as the pulse matching period PM. The music data starts around the pulse coincidence period PM. Since it is only necessary to find out the start time of the music data, it is not necessary to specify how far the pulses have been matched.

  Since the timing of the pulse coincidence period PM differs for each pulse data, the separation module 36 synchronizes the MP3 data based on the identified pulse coincidence period PM (S2). Specifically, as shown in FIG. 5, the phases of the pulse signals P1 to P4 and the synthesized speech signals C1 to C4 are shifted so that the pulse matching period PM is aligned, and MP3 data corresponding to the pulse matching period PM is specified. Thereby, the starting position of the music data can be generally specified in the MP3 data for each program.

  Next, the separation module 36 compares the level change rates with each other so that the level change rates are continuously matched for a predetermined time (for example, 10 seconds) or longer (hereinafter referred to as “level change rate matching period”). ) Is searched (S3). However, the level change rates do not need to be completely matched, and may be matched within a range of ± 10% as a margin. In the example shown in Table 2, since the third and subsequent level change rates from the left are almost the same between the level data LD1 and LD2, this period is specified as the level change rate matching period LM1. In addition, since the level change rates after the fourth from the left are almost the same between the level data LD2 and LD3, this period is specified as the level change rate matching period LM2. Further, since the level change rates after the fourth from the left are almost the same between the level data LD3 and LD1, this period is specified as the level change rate matching period LM3.

  Even in this level matching process, the level change rate is compared with each other as in the previous pulse matching process. However, since the start position of the music data has already been specified in the pulse matching process, the level change rate is brute force. There is no need to contrast. The level matching process itself is more computationally intensive than the pulse matching process itself, but it is only necessary to compare the level change rate in order from the start position of the already specified music data. Is not so much.

  Next, the separation module 36 selects the longest one from the level change rate matching periods LM1 to LM3, and selects MP3 data corresponding to that period (S4). Next, the separation module 36 divides the selected MP3 data into music data and MC data based on the corresponding level change rate matching period (S5). That is, the part corresponding to the level change rate matching period in the MP3 data is determined as music data, and the other part is determined as MC data. Then, the separation module 36 stores the music data in the music data file 37 and the MC data in the MC data file 38 (S6).

  As described above, according to the embodiment of the present invention, a plurality of music periods in which MP3 data match each other continuously for a predetermined period or longer are searched for, the longest music period is selected, and the MP3 of the music period is selected. Since data is stored as music data, music can be automatically extracted and recorded from various audio broadcasts from broadcast stations without conventional control data, on-air information, cue sheets, etc. be able to.

  In the above embodiment, the beginning of the music period is roughly specified by the pulse matching process in the first stage, and the music period is specified in detail by the level matching process in the second stage. It is also possible to specify the period. In this case, since the level matching process is not performed, the process (23 to 34) for generating the level data is not required in addition to the synchronization process (S2).

  On the other hand, it is also possible to specify the music period only by level matching processing. In this case, since the pulse matching process is not performed, the process (16 to 21) for generating the pulse data is not required in addition to the synchronization process (S2).

  While the embodiments of the present invention have been described above, the above-described embodiments are merely examples for carrying out the present invention. Therefore, the present invention is not limited to the above-described embodiment, and can be implemented by appropriately modifying the above-described embodiment without departing from the spirit thereof.

It is a functional block diagram which shows the structure of the air check apparatus by embodiment of this invention. It is a timing diagram which shows the synthetic | combination audio | voice signal output from the synthesizer in the air check apparatus shown in FIG. 1, and the pulse signal output from a pulsator. It is a timing diagram which shows the audio | voice signal output from the band pass filter in the air check apparatus shown in FIG. 1, the synthetic | combination audio | voice signal output from a synthesizer, and the pulse signal output from a pulsator. It is a flowchart which shows the process of the separation module in the air check apparatus shown in FIG. FIG. 5 is a timing chart showing a synchronization process of the separation module shown in FIG. 4.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 FM tuner 13 MP3 data file 16, 17, 23-25 Band pass filter 18 Inverter 19 Synthesizer 20 Pulse generator 21 Pulse data generation module 22 Pulse data file 29-31 Level change rate calculation module 32-34 Level data generation Module 35 Level data file 36 Separation module 37 Music data file 38 Data file

Claims (6)

Receiving means for receiving an audio signal broadcast from a broadcasting station;
Broadcast storage means for storing the audio signal received by the receiving means;
Music search means for searching for a plurality of music periods in which the audio signals continuously match each other from the audio signals stored by the broadcast storage means; and
Music selection means for selecting one of a plurality of music periods searched by the music search means;
An air check apparatus comprising: a music storage unit that stores an audio signal of a music period selected by the music selection unit. The air check device according to claim 1,
The audio signal includes a left channel audio signal and a right channel audio signal,
The music searching means is
Inverting means for inverting one of the left channel audio signal and the right channel audio signal;
Synthesizing means for synthesizing the signal inverted by the inverting means and the other of the left channel audio signal and the right channel audio signal to generate a synthesized audio signal;
The level of the synthesized voice signal generated by the synthesizing means is compared with a predetermined level, and when the level of the synthesized voice signal is larger than the predetermined level, a pulse signal having the first logic level is output, and the level of the synthesized voice signal is predetermined. Pulsing means for outputting a pulse signal of the second logic level when the level is smaller than the level;
An air check apparatus comprising: pulse matching means for specifying a period in which a logic level continuously matches a predetermined period or more among pulse signals output by the pulsing means as the music period. The air check device according to claim 1,
The music searching means is
Sampling means for sampling the level of the audio signal stored by the broadcast storage means every predetermined period;
A rate of change calculation means for calculating a level change rate by dividing the level sampled by the sampling means by a level sampled before that;
An air check apparatus comprising: a level matching unit that identifies a period in which a level change rate continuously matches a predetermined period or more from among the level change rates calculated by the change rate calculating unit. The air check device according to claim 3,
The music search means further includes
Among the audio signals stored by the broadcast storage means, a first mid-range component, a second mid-range component higher than the first mid-range component, and higher than the second mid-range component An air check device comprising: extraction means for extracting a third mid-range component and providing it to the sampling means. The air check device according to claim 1, further comprising:
An air check apparatus comprising non-music storage means for storing an audio signal other than the music period selected by the music selection means. The air check device according to claim 1,
The audio signal includes a left channel audio signal and a right channel audio signal,
The music searching means is
Inverting means for inverting one of the left channel audio signal and the right channel audio signal;
Synthesizing means for synthesizing the signal inverted by the inverting means and the other of the left channel audio signal and the right channel audio signal to generate a synthesized audio signal;
The level of the synthesized voice signal generated by the synthesizing means is compared with a predetermined level, and when the level of the synthesized voice signal is larger than the predetermined level, a pulse signal having the first logic level is output, and the level of the synthesized voice signal is predetermined. Pulsing means for outputting a pulse signal of the second logic level when the level is smaller than the level;
Pulse matching means for specifying a period in which the logic level of pulse signals output by the pulsing means coincides continuously for a predetermined period or more as a pulse matching period;
Synchronizing means for synchronizing the audio signal so that the beginning of the pulse matching period specified by the pulse matching means is aligned;
Sampling means for sampling the level of the audio signal stored by the broadcast storage means every predetermined period;
A rate of change calculation means for calculating a level change rate by dividing the level sampled by the sampling means by a level sampled before that;
By comparing the level change rates of the audio signals synchronized by the synchronization means at the same timing, the level change rates continuously calculated from the level change rates calculated by the change rate calculation means for a predetermined period or more. An air check device comprising: level matching means for specifying a matching period as the music period.

Images