JPS647688B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an automatic audio level control device that automatically controls the level of an audio program signal handled by, for example, a broadcasting station using a microcomputer.

As is well known, most audio program signals at broadcasting stations are currently processed as analog signals. AGC (automatic gain controller), which controls the audio level, is no exception.The method uses a rectifier and an integrating circuit to detect the audio level, and operates the level control circuit using the feedback of a direct current (DC) control signal. ing. This method is a simple feedback control that always follows a constant level and moves the level more than necessary, which destroys the artistry of music programs, and is insufficient for energy-dense sources during commercials. It is difficult to follow and feels "noisy".

This invention was made based on the above-mentioned circumstances, and was developed by investigating the operations of mixers at broadcasting stations, and by using a microcomputer to process human judgments and movements that approximate human operations. It is something that gives you a feeling, especially
By not performing level control when the difference between the current audio level and the target level is small, unlike conventional AGC, the system always tries to converge the audio level to the target level, resulting in an unnatural hearing sensation. The purpose of this invention is to provide an automatic audio level control device that can suppress such problems.

An embodiment of the present invention will be described below with reference to the drawings.

First, the operation will be schematically explained using FIG. The audio signal input from the input terminal 11 is supplied to an attenuator 12 . This lotuser 1
2 performs substantial level control, and can perform attenuation control of ±16 dB depending on the control amount. The audio signal output from the losser 12 is led to an output terminal 13 and is also supplied to a level detection circuit 14. This level detection circuit 14 digitally detects the average value of the audio signal, and this digital signal is supplied to a microcomputer (μ-CPU) 15.
Further, the audio signal input from the input terminal 11 is supplied to a timing detection circuit 16. This detection circuit 16 processes the waveform of the audio signal using a predetermined envelope, and analyzes whether the envelope is trending upward or downward. This analysis information is supplied to the μ-CPU 15 as level correction start timing information when performing level control.
In this μ-CPU 15, the level detection circuit 14
Discrimination between a human voice and music (discrimination between a pulsed signal and a continuous signal) is performed from the output signal, and the detected level is corrected according to the result of this discrimination. When this corrected signal exceeds a predetermined dead area, a control variable consisting of digital bit data is generated. This signal is supplied to the level control interface section 17 at the control timing of the timing detection circuit 16, converted into an analog DC signal, and supplied to the rotor 12.
The losser 12 is then controlled to bring the audio signal to the target level.

Such processing by the μ-CPU 15 is performed by software, which is configured as follows.

The output signal of the level detection circuit 14 is supplied to a human voice/music discrimination section 18. When the average value of the audio level is constant, it has been confirmed that when comparing the human voice and music, the human voice sounds audibly louder. Therefore, in the discrimination section 18, if the voice is a human voice, the average voice level previously determined in the level detection circuit 14 is intentionally increased by a predetermined amount, and the amount of attenuation in the subsequent rotor 12 is increased. Processing is performed to substantially lower the level. Furthermore, if the level control is performed each time according to the corrected average value obtained by the discrimination section 18, a rather unnatural sound will arise. For this reason, level control is not performed within a predetermined level fluctuation. The dead area detection section 19 performs this. This detection section 19 outputs a prohibition signal that does not perform level control when the target level is within ±5 dB, and
If it exceeds 5dB, an up-count signal and a down-count signal are output respectively. The up-count signal and down-count signal are supplied to an up-down counter 22 via gate circuits 20 and 21, respectively, and the inhibit signal is supplied directly to the counter 22. the gate circuit 20,
21 are controlled on and off by timing signals supplied from the timing detection circuit 16, and in response to this, the counter 22 counts up and down counts signals. Further, 23 is a sequence control section. The control section 23 periodically controls the operation of the level detection circuit 14, and also controls the level detection circuit 14 and the counter 22 in accordance with the output audio level of the losser 12 obtained by the comparison section 24. .

Next, each part will be explained in more detail.

First, the level detection circuit 14 will be explained.

As mentioned above, the level detection circuit in conventional AGC rectifies and smoothes the audio signal, and evaluates the original value as the volume. However, since the rectifying and smoothing circuit has a time constant, it is difficult to accurately detect the volume of an audio signal having a complex waveform. It was confirmed that this volume can be detected by determining the area of the waveform.
The level detection circuit 14 determines the area of the audio signal based on the principle shown in FIG. 2, and calculates the average value of the audio level from this. That is, the audio signal i shown in FIG. 2 is divided by threshold levels T ₁ to T ₇ , the time when the signal exists for each threshold level is detected, and the sum is calculated to determine the audio signal. The area (=volume) of is detected.

The specific configuration of the level detection circuit 14 is shown in FIG.

The audio signal i is supplied to comparators 31a, 31b to 31l. These comparators 31a, 31b to 31l
For example, +8dB, +6dB, ..., 0dB, ...-12dB,
Threshold levels are set at -14dB and 2dB intervals. These comparators 31a, 31b to 31
The output signals of l are sent to AND circuits 32a and 32, respectively.
It is supplied to one input end of b to 32l. Count pulse signals are supplied to the other input terminals of the AND circuits 32a, 32b to 32l, respectively.
This count pulse signal is supplied from a square wave oscillator 33. This oscillator 33 outputs a count pulse signal of a predetermined frequency in response to a start signal supplied from the μ-CPU 15, and this signal is counted by a counter 34 and is also used by the AND circuits 32a, 32b to 32l. supplied to The counter 34 outputs a predetermined number of count pulse signals (count value corresponding to time T shown in FIG. 2).
This count value is μ−CPU1
5 (sequence control unit 23, which will be described later). Then, when the count value reaches a predetermined value, μ-
A stop signal is supplied from the CPU 15 to the square wave oscillator 33, and oscillation is stopped. The output signals of the AND circuits 32a, 32b-32l are supplied to counters 35a, 35b-35l, respectively. The counters 35a, 35b to 35l each count the count pulse signals. That is, the AND circuits 32a, 32b to 32l are controlled to be turned on or off depending on the time of the audio signal output from the comparators 31a, 31b to 31l.
Counters 35a and 35b correspond to this on time.
~35l, the count pulse signal is counted.
The counting outputs of the counters 35a, 35b to 35l are supplied to the average value calculating section 36. This calculation section 3
6, the total count output (area) is calculated, and this value is divided by the count value of the counter 34 to calculate the average value. This average value is the human voice,
The signal is supplied to the music discrimination section 18.

Next, the human voice and music discrimination section 18 will be explained.

In the waveforms shown in FIG. 4, a is a general music signal and b is a human voice. When comparing the auditory sensations of such signals whose waveforms have the same area (=volume) at a certain time t, the human voice shown in FIG. In other words, pulse-like signals are felt to be louder. The human voice/music discrimination unit 18 calculates the height distribution of the waveform within a certain area, and determines the "loudness" of the pulsed sound.
Corrections are being made.

FIG. 5 shows the configuration of the determining section 18.
The average value of the signals obtained by the level detection circuit 14 is supplied to a pulse signal discrimination section 51. The pulse signal discriminator 51 calculates, for example, what percentage of the total signal includes a signal having the average value +8 dB. That is, in FIG. 3, when the average value is 0 dB, that is, the threshold level of the comparator 31e, the count value of the counter 35a corresponding to the comparator 31a that is +8 dB higher than this is taken out. This value is divided by the total count value to determine the percentage. The read address of the read-only memory (ROM) 53 is set in the address specifying section 52 according to this ratio. As shown in FIG. 6, this ROM 53 stores correction levels corresponding to the proportion of high-level signals included in the audio signal.
This correction level is read out in accordance with the designation of the address designation section 52. This read correction level is supplied to the output circuit 54, the average value is increased according to this level, and is supplied to the dead area detection section 19 as a corrected average value.

Next, the dead area detection section 19 will be explained.

As mentioned above, conventional AGC always tries to keep the level constant regardless of the level difference. However, if the level difference between the target volume and the target volume is small, if you try to always converge the volume to the target level, the volume will change frequently, resulting in an unnatural hearing sensation. If the difference between the target level and the current level is small, the dead area detection section 19 reduces the number of adjustment circuits so as not to give an unnatural hearing sensation. The dead area detection section 19 is composed of a comparison section, a bit data generation section, etc., and as shown in FIG. is output. Also, if the corrected average value is +5 dB or more, a down count signal according to (target level) - (corrected average value) is output, and the corrected average value is -
If it is 5dB or less, an up count signal corresponding to (target level) - (corrected average value) is output.
This up count signal is supplied to the counter 22 at the rising edge of the audio signal by the gate circuit 20 controlled by the timing detection circuit 16.
The down count signal is supplied to the counter 22 by a gate circuit 21 which is also controlled by the timing detection circuit 16 when the audio signal falls. According to the count value of the counter 22, the rotor 12 is controlled via the level control interface section 17 as described above.

Next, the timing detection circuit 16 will be explained.

When a skilled mixer adjusts the level of the audio signal being broadcast, he or she operates the fader to raise the level when the audio signal increases.
When lowering the level, the fader operation is performed when the audio signal decreases, so that the level adjustment is not perceptible. In order to electrically reproduce the above operations performed by the mixer, this timing detection circuit 16 detects the envelope of the audio signal as shown in FIG. U, I am looking for the level down timing D.

FIG. 9 shows the circuit configuration of the timing detection circuit 16. The input audio signal is sent to the rectifier 9
1 and amplified by an operational amplifier 92. This amplified output signal is transmitted to the first and second time constant circuits 93,
94. The time constant circuits 93 and 94 obtain the envelope of the audio signal, and the first time constant circuit 93 includes a resistor R ₁ whose one end is connected to the output terminal of the operational amplifier 92, and the other resistor R ₁ . A capacitor _C1 with one end connected to the other end and grounded at the other end, and a resistor _R2 with one end connected to the other end of the resistor _R1 .
The charging time constant is set by R ₁ C ₁ and the discharging time constant is set by R ₂ C ₁ . Further, the second time constant circuit 94 includes a resistor R ₃ whose one end is connected to the output terminal of the operational amplifier 92, a capacitor C 2 whose one end is connected to the other end of the resistor R ₃ and whose other end is grounded, and the resistor R ₃ . It consists of a resistor R ₄ whose one end is connected to the other end of R ₃ , and the charging time constant is R ₃ C ₂ and the discharging time constant is
Set by R ₄ C ₂ . The other end of the resistor _R2 is connected to the negative (-) input terminal of the operational amplifier 95, and is also connected to the output terminal of the operational amplifier 95 via the resistor _R5 . Further, the other end of the resistor _R4 is connected to the positive (+) input terminal of the operational amplifier 95 and grounded via the resistor _R6 . The output terminal of the operational amplifier 95 is connected to the positive input terminal and negative input terminal of operational amplifiers 96 and 97, respectively. A negative input terminal and a positive input terminal of the operational amplifiers 96 and 97 are grounded through a series circuit of a resistor R ₇ and a capacitor C ₃ and a series circuit of a resistor R ₈ and a capacitor C ₄ , respectively. Variable resistors VR _{1 and} VR ₂ are connected to the connection between the resistor R 7 and the capacitor C ₃ and the connection between the resistor R ₈ and the capacitor C ₄ , respectively. This variable resistance
+E and -E power supplies are supplied to both ends of VR ₁ and VR _2, respectively, and the operational amplifiers 96 and 97 are set to a predetermined bias. The output terminals of the operational amplifiers 96 and 97 are connected to the gate circuit 20 shown in FIG. 1 through an interface circuit (not shown).
21 respectively.

The operation of the above configuration will be explained using FIG. 10. Note that the first and second time constant circuits 93 and 94
It is assumed that the relationship between charging time constants is R ₁ C ₁ > R ₃ C ₂ and the relationship between discharging time constants is R ₂ C ₁ > R ₄ C ₂ . Also,
In FIG. 10, the dotted line indicates the first time constant circuit 93.
The solid line shows the characteristics of the second time constant circuit 94. Furthermore, the gate circuit 2
It is assumed that 0 and 21 operate in positive logic.

For example, if the input audio signal is on an upward trend, the terminal voltages of capacitors C ₁ and C ₂ will rise faster on the C ₂ side, which has a smaller time constant. For this reason,
The output voltage of the operational amplifier 95 becomes positive, and a positive timing signal is output from the operational amplifier 96 to which this signal is supplied. Therefore, the gate circuit 20 shown in FIG. 1 is turned on, and the up count signal is supplied to the counter 22.

Furthermore, when the input audio signal tends to fall, the terminal voltages of capacitors C ₁ and C ₂ decrease faster on the C ₂ side, which has a smaller time constant. Then, when the terminal voltages of the capacitors C ₁ and C ₂ are inverted, the output voltage of the operational amplifier 95 becomes negative, and the operational amplifier 97 to which this signal is supplied outputs a positive timing signal. Therefore, the gate circuit 21 shown in FIG. 1 is turned on, and a down count signal is supplied to the counter 22.

In this way, level control is performed in the ascending trend of the envelope when increasing the volume, and in the descending trend of the envelope when decreasing the volume, so that it is possible to lower a pitch that is becoming high or to control a pitch that is becoming low. There is no unnatural feeling of raising the level.

Note that the level up timing and level down timing are determined by the first and second time constant circuits 93 and 9.
It can be changed as appropriate depending on the set time constant of 4.

Next, the sequence control section 23 will be explained.

The mixer will quickly set the audio level to the target level at the start of the broadcast, and if the audio level suddenly increases during the broadcast or there is a pause period, the audio level will suddenly decrease after these periods. I'm trying to make sure it doesn't get too high. The sequence control unit 23 performs an operation similar to this. This sequence control section 23 varies the detection period of the level detection circuit 14, that is, the generation period of the start signal and stop signal of the square wave oscillator 33 shown in FIG. This period is varied as shown in FIG. In addition, in FIG. 11, the shaded area indicates the level control period, and Ta,
Tb and Tc indicate the level detection period in the level detection circuit 14, and this period is Ta<Tb<Tc
The relationship is as follows.

FIG. 11a shows a case, for example, at the start of broadcasting, where level detection and level control are performed in the order of Ta, Tb, and Tc, and the target level is quickly reached. After reaching the target level, level detection is performed at intervals of Tc.

Figure b shows control when the audio level suddenly increases, for example, during broadcasting. Until then Tc
When the audio level, which has been maintained at a constant level for a period of Th, suddenly increases during a period of Th, the audio level is first forcibly reduced to a predetermined level, and
Level detection and level control are performed in the order of Ta, Tb, and Tc to quickly return to the target level. The operation of forcibly lowering the level is performed as follows. A sudden rise in the audio level is detected by the comparison section 24 shown in FIG. 1, and this information is supplied to the sequence control section 23. A predetermined down count signal is supplied from the control section 23 to the gate circuit 21. This gate circuit 21
is turned on under the control of the timing detection circuit 16 when the audio signal is in a downward trend, and a down count signal is supplied to the counter 22. The counted value of the counter 2 is then supplied to the losser 12 via the level control interface section 17, and the audio level is lowered to a predetermined level.

Further, FIG. 11c shows control when there is a pause period. The audio level, which had been maintained at the target level for a period of Tc up to that point, is placed in a pause state for a period of Tp. Pause period (Tp＜8sec)
In this case, if the level fluctuates due to background noise or the like, the level suddenly becomes high or low after the pause is released, which is inconvenient. Therefore, the level is fixed during the pause period. Also, the pause period
Before and after Tp, Tc ₁ and Tc ₂ have a relationship of Tc ₁ +Tc ₂ =Tc. In such an operation, when a no signal is detected by the comparator 24 shown in FIG. It is done by being taken as.

FIG. 11d shows control when a no-signal state occurs. If the audio level, which had been maintained at the target level for a period of Tc, becomes a no-signal state for a period of T _N (T _N > 8 seconds), first, as in the case where the audio level suddenly increases as described above,
The audio level is set to a predetermined level, and then Ta,
Level detection and level control are performed in the order of Tb and Tc to quickly reach the target level.

According to the above configuration, the level detection circuit 14 detects the level of the audio signal output from the rotator 12 and supplies it to the microcomputer 15 to calculate the control amount of the rotator 12, while the timing detection circuit At step 16, a rising or falling tendency of the input audio signal to the losser 12 is detected, and level control is performed by supplying the control amount to the losser 12 in accordance with the timing of this rise or fall. Therefore, there is no audible unnaturalness, and level control similar to that performed by a mixer can be automatically performed, which is extremely convenient.

Further, the level detection circuit 14 calculates the area of the input audio waveform over a predetermined time period, and calculates the average value of the audio level from this. Therefore,
Since detection is more accurate than average value detection using a rectifier circuit in conventional detection circuits, the actual volume can be reliably detected.

In addition, the human voice/music discrimination unit 18 calculates the ratio between the average voice level and the level exceeding this, since human voices are generally more pulse-like than music. The average value is corrected. Therefore, pulsed sounds such as the human voice are attenuated by a large amount in the Lotusser 12, making the human voice "loud" compared to music.
can be suppressed.

Furthermore, the dead area detecting section 19 does not perform level control if the level difference from the target sound volume is within ±5 dB. Therefore, the conventional
It is possible to suppress the unnatural hearing sensation caused by constantly trying to converge the volume to the target level like AGC.

Further, in the timing detection circuit 16, the first and second
Due to the difference in time constants set in the time constant circuits 93 and 94, the rise and fall tendencies of the audio waveform are detected by the envelope according to the time constants, and the supply timing of the control amount supplied to the losser 12 is determined. It's in control. Therefore, when increasing the volume, the level control by the rotuser 12 is performed in the ascending trend of the envelope, and when decreasing the volume, it is performed in the descending trend of the envelope. There is no unnatural feeling of raising the sound that is becoming.

Furthermore, since the control timing can be varied depending on the set time constants of the first and second time constant circuits 93 and 94, it is possible to easily approximate the operation of a mixer man, and to obtain an extremely good hearing sensation. is possible.

In addition, the sequence control unit 23 also
The detection period of the level detection circuit 14 is varied depending on the audio signal output from the level detection circuit 14. Therefore, when the audio level suddenly increases, when there is a pause state, or when there is no signal, the recovery is quickly carried out and unnatural hearing sensations are suppressed.

As detailed above, according to the present invention, level control is not performed when the difference between the current audio level and the target level is small, so that the audio level is always maintained as in conventional AGC. It is possible to provide an automatic audio level control device that can suppress an unnatural hearing sensation caused by trying to converge the audio level to a target level.

[Brief explanation of drawings]

The drawings show an automatic audio level control device according to the present invention; FIG. 1 is a schematic diagram of the device, FIG. 2 is a diagram for explaining the principle of the level detection device, and FIG. 3 is a level detection circuit. FIG. 4 is a diagram explaining the discrimination principle of the human voice and music discrimination section. FIG. , FIG. 5 is a block diagram showing the human voice and music discrimination section, FIG. 6 is a diagram showing the memory contents of the ROM shown in FIG. 5, and FIG. 7 is shown to explain the operation of the dead area detection section. 8 is a diagram shown to explain the principle of the timing detection circuit, FIG. 9 is a circuit configuration diagram of the timing detection circuit, FIG. 10 is a diagram shown to explain the operation of FIG. Figures a to d are diagrams showing different control operations of the sequence control section, respectively. 12...Lotuser, 14...Level detection circuit, 15
...Microcomputer (μ-CPU), 16...Timing detection circuit, 17...Level control interface section, 18...Human voice, music discrimination section, 19...
Dead area detection unit, 20, 21... gate circuit, 22
... Counter, 23... Sequence control section, 24... Comparison section.

Claims (1)

[Claims] 1. A variable attenuator into which an audio signal is input, and an average audio level detected by finding the waveform area within a unit time of the audio signal output from the variable attenuator and dividing by the unit time. input the average audio level detected by this circuit and compare it with a preset target level, and if the difference between the average audio level and the target level is within a preset control prohibited range, If the control prohibition range is outside the control prohibition range and the average audio level is higher than the target level, an attenuation increase command signal is generated. A command signal generation means for generating an attenuation reduction command signal, a control prohibition command signal, an attenuation increase command signal, and an attenuation reduction command signal outputted from the command signal generation means are input, and the above-mentioned control signals are inputted according to each command signal. 1. An automatic audio level control device comprising attenuation amount control means for controlling an attenuation amount of a variable attenuator. 2. The level detection circuit includes arithmetic means for detecting an average audio level by determining the waveform area within a unit time of the audio signal output from the variable attenuator and dividing it by the unit time; A determination level is determined by adding a preset reference level to the average audio level, and the waveform area within the unit time of the audio signal that is equal to or higher than this determination level is determined, and the waveform area equal to or higher than this determination level and the level are determined. and a correction means for determining a ratio of the area of the waveform above the average audio level determined by the detection circuit, and correcting the average audio level of the level detection circuit based on this ratio, the control command signal generating means: 2. The automatic audio level control device according to claim 1, wherein each of the control command signals is generated based on the average audio level corrected by the correcting means. 3. The attenuation amount control means inputs the audio signal, detects its envelope, detects rising and falling trends of the envelope, and outputs a level up timing signal when there is an upward trend, and a level down timing signal when there is a downward trend. a timing detection means for generating the
a first gate circuit that derives an attenuation increase command signal outputted from the command signal generation means in accordance with the level down timing signal; and a second gate circuit that derives the attenuation amount in response to the timing signal, and adjusts the attenuation amount of the variable attenuator in response to the attenuation amount increase command signal and the attenuation amount decrease command signal derived from the first and second gate circuits. The automatic audio level control device according to claim 1, characterized in that the audio level automatic control device is adapted to control. 4. The attenuation amount control means includes a comparison section that detects that the level of the output of the variable attenuator fluctuates outside a preset range, and controls the attenuation amount of the variable attenuator according to the output of the comparison section. 2. The automatic audio level control device according to claim 1, further comprising a sequence control section.