JP3505722B2 - Signal level control device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention
The present invention relates to a signal level control device for controlling a level of a signal. 2. Description of the Related Art Conventionally, magnetic recording has been performed with a tape recorder or the like.
When recording audio signals on tape,
By controlling the level of the audio signal
Optimize the recording signal level for the magnetic tape. This
In the case of, the input signal is generally
(Troll) Configuration to supply to recording system via circuit
However, this configuration is the most suitable for the magnetic head of the recording system.
Perform level control so that the recording current flows
I have to. That is, the input audio as described above
An AGC circuit that performs AGC on a signal usually has an input
Audio signal level is the optimal level as described above.
For example, if the signal level is too high
Lower the level if
Try to raise the level. [0004] By the way, the above conventional art
The AGC circuit is used for input input, for example, between music pieces.
Even if there is silence in the audio signal,
Optimum level just like the part (normal music part)
Level adjustment. Therefore, on
In the conventional AGC circuit, the above-mentioned silent portion, that is,
If there is a very low signal level, immediately
Suddenly trying to raise the signal level to the above optimum level
It works in a direction that raises the level drastically. However, as described above, for example, between music pieces
If you suddenly raise the level with, the beginning of the next song
It will be a ruover. That is, this conventional AGC
If the circuit is applied to, for example, a tape recorder, the AGC
Level over at the beginning of the next song in the recording system at the end of the circuit
And distortion or the like occurs. Accordingly, the present invention has been made in view of the above-described circumstances.
Suggestions, like for example between music
The level rises sharply in the silent part (no signal part)
To prevent level over at the beginning of the next song between songs.
To provide a signal level control device that can be prevented.
It is the purpose. [0007] A signal level control according to the present invention.
The device has been proposed to achieve the above objectives
Yes, the signal level for adjusting the signal level of the input signal
Output means from the signal level adjustment means
A predetermined signal processing is performed on the output signal to measure the level of the signal.
Signal processing means for level measurement and level measurement
Based on the level measurement signal output from the signal processing means
Detecting means for detecting a silent part of a signal inputted by the
And the level measurement output from the level measurement signal processing means.
Signal level adjustment of the signal level adjusting means based on the regular signal
Control the amount of adjustment and the
If a silent part of the signal is detected, the signal level
Signal level control means for stopping or delaying the control
You. [0008] In other words, the present invention provides, for example,
Signal level to adjust the signal level of the audio signal
Adjusting means and an output signal output from the signal level adjusting means.
Predetermined signal processing on output signal to measure signal level
Signal processing means for performing level measurement,
Based on the level measurement signal output from the signal processing means.
Silence detecting means for detecting a silence portion of the input signal;
For level measurement output from signal processing means for level measurement
The signal level adjustment amount of the signal level adjusting means based on the signal
And the silence detection means
When silence is detected, control the signal level adjustment amount.
Signal level control means for stopping or delaying,
Automatically adjust the signal level of the above input audio signal
Automatic gain control circuit
When a silent part is detected in the step, the signal level control means
For example, by switching the loop gain, time constant, etc.
It slows down or holds the response speed of the operation.
is there. In the present invention, the signal level control
The means has received a no-signal detection signal from the no-signal detection means.
If there is a reference level input from the signal level adjustment means,
Control the signal level adjusting means in the same way as
It is also possible to use In this case, the level control
Rule circuit has its own time constant (ie shortened by loop gain)
The time adjustment is not done)
So, even in the silent part, the level rises sharply
There is no such thing. According to the signal level control device of the present invention, the
If the silence detection means detects the non-signal part of the signal
If the signal level control means
Slow down or stop the level adjustment
Level rises rapidly even in non-signal areas.
Never. An embodiment of the present invention will be described below with reference to the drawings.
explain. A level control circuit according to an embodiment of the present invention.
Is supplied through an input terminal 91 as shown in FIG.
Variable gain to adjust the signal level of the input audio signal
As a means, an analog signal level is adjusted.
Log attenuator 1 and the analog attenuator 1
Converts analog output to analog / digital (A / D)
Digitally communicates the output of the A / D converter 2
Digital attenuator that adjusts the signal level
And a digital multiplier 3. Further, the level control circuit of the present embodiment
Is the signal based on the output level of the variable gain means.
Control value setting means for calculating a control value for level adjustment;
Then, the limiter unit 11, the normal unit 12,
Has feedback unit consisting of basic unit 13
And silence to detect non-signal portions of the supplied signal.
It has a detection circuit 10. The level control circuit of this embodiment
Wherein the feedback unit detects the silence.
Receiving a no-signal detection signal (silence flag) from the circuit 10
To reduce the speed of signal level adjustment by the variable gain means.
Alternatively, it is stopped (held). That is, the level control of this embodiment
In the circuit shown in FIG.
As shown in A of FIG.
Comparison with a certain silence detection level (for example, -40 dB)
This comparison allows, for example,
If the audio signal level is below the silence detection level (or
Is lower than the silence detection level).
If detected, the silence detection circuit 10 returns to FIG.
A silence flag (silence detection flag) as shown is output.
The silence detection circuit 10 accurately detects silence.
For example, if the average input audio signal level is
A silence detection below a certain reference level (for example, 0 dB)
The above silence flag is output only when the output level falls below
It is also possible. The silence flag from the silence detection circuit 10
Is sent to the feedback unit of the control value setting means.
Can be In the feedback unit,
When the silence flag is received, the input
Audio signal level is very low
Does not suddenly increase the AGC loop gain
In addition, the AGC operation is delayed as shown in FIG.
Or hold, the sudden
We try to prevent a sharp rise in level. Further, in this embodiment, the feed
The back unit is a silence filter from the silence detection circuit 10.
If a lag is received, the reference level
The variable gain means is controlled in the same way as when
It is also possible to control. In this case, this implementation
The level control circuit in the example has its own time constant (ie,
Level adjustment with time constant not shortened by loop gain)
Therefore, even if there is no sound,
There is no such thing as raising the level. FIG. 2D shows this embodiment for comparison.
AGC operation may be delayed depending on silence detection as in
Variable gain means in the case of the conventional example which does not hold or hold
Shows a rapid rise in the level due to. Hereinafter, the configuration of this embodiment shown in FIG. 1 will be described in detail.
Will be described. It should be noted that the configuration of FIG.
Switchable between manual gain control and
First, the configuration for AGC will be described.
You. The configuration of this embodiment shown in FIG.
For example, gain co-processors used in digital tape recorders
This is applied to a control circuit. Note that FIG.
In AGC mode and manual gain control
To switch between the manual and AGC modes.
Of the switches 41 and 42 that are switched according to the
Set the terminal to the switched terminal side of AGC mode side or manual
Switch to selected terminal side in gain control mode
It is realized by doing. In FIG. 1, for example, a microphone input
Input or line input analog audio signal
And supplied to the analog attenuator 1
It is. The analog attenuator 1 is connected to the terminal 91
These analog audio signals can optionally be attenuated (ma
To the next stage A / D converter 2
You. The analog attenuator 1 is in 2dB steps
The attenuation value is set. In addition,
This attenuator not only has its intended purpose, but also
It is also intended to be wider. The A / D converter 2 includes the analog
Analog pre-attenuated by attenuator 1
The audio signal is converted to a digital signal. The data
The digital signal is sent to the next digital multiplier 3. The digital multiplier 3 has an analog
Adjust fine gain that could not be adjusted with tenator 1
You. That is, the digital multiplier 3 uses the A / D
To the audio signal digitally converted by the converter 2
Multiply by a factor of 1 or less to reduce the value
G), the 2 dB of the analog attenuator 1
To perform fine level adjustment by interpolating between steps
What is used. The digital multiplier 3 is approximately
The resolution is set to 0.25 dB step. Ma
For example, the manual gain control mode described later
Mode, such as fade-out
In the case of causing the digital multiplier 3 to attenuate,
As a result, the noise including the noise of the A / D converter 2 is reduced.
Can be attenuated. The next stage DSP of the digital multiplier 3
(Digital signal processor) 4
The digital signal converted by the
A signal to be recorded on a recording medium such as a magnetic tape is formed. this
The output of the DSP 4 is output from the output terminal 92,
Recording on magnetic tape, etc. by the configuration of the recording system at the later stage omitted
Is done. Note that the audio signal is
Pass only through the path of DSP4 from the server. Other structures in FIG.
The components are all used to form the control value of the attenuator.
The configuration (other than the input side of the peak detection circuit 5 described later)
The other component performs processing once every 1/50 second, for example.
U. In this embodiment, the audio signal system is hardware-based.
Software, and everything else is software
Each part is software or hardware
Can also be configured. Here, in order to perform AGC, the above-mentioned Athens
The audio signal whose gain has been adjusted by the
Through the path 5, logarithmic compression circuit 6, and monaural circuit 7.
Is the measurement level. That is, first, the digital power
The output of the calculator 3 is sent to the peak detection circuit 5. The above-described peak detection circuit 5 is modified in the configuration of the preceding stage.
The converted digital data is kept for a certain time (for example, 20 ms).
ec), the maximum volume in that section is detected, and
This digital audio recording for AGC
The representative of Bell. The output of the peak detection circuit 5 is logarithmic
The data is sent to the compression circuit 6. The logarithmic compression circuit 6 performs the peak detection
The value (peak value) detected in Road 5 is matched to the human auditory characteristics.
Logarithmic conversion. In this embodiment,
Floating format (exponent)
Since the part is logarithmic, this is the concatenation of the mantissa part and the MSB
Is non-transmission). Here, the logarithmic compression circuit
6, the floating line for approximating the broken line is used.
The mat is used as it is. Therefore, the exponent is
When it reaches zero, the bits below it are linear,
It is very different from logarithm. Therefore, the exponent part is zero
If the exponent part is compressed in the negative direction like a broken line approximation
To ensure that the error does not become abnormally large.
In the broken line approximation, the part corresponding to the exponent is minus.
, But a negative value in true exponential notation.
The output of the logarithmic compression circuit 6 is sent to a monaural circuit 7.
You. The output of the logarithmic compression circuit 6 is a level meter
Also used for The monaural circuit 7 performs the logarithmic compression process.
Stereo L channel and R channel supplied from road 6
Outputs the larger value of the flannel values. That is,
In AGC, the gains of the left and right channels of L and R are set separately.
If the control is performed in the right direction, the localization will be lost.
This processing is performed in order to communicate. Therefore, the measured value
Is monaural (larger for the left and right because of emphasis on high volume)
You. The output of the monaural circuit 7 is the inverse backlash
Sent to Road 8. The reverse backlash circuit 8 is an AGC
The battery inserted in the feedback loop system described later
By the crash element (backlash circuit 14)
Linear as if the system did not vibrate unnecessarily
To compensate for the backlash that pretends to be
You. That is, the last time the attenuator was set (some
Is sometimes the last time before or after)
Set including the difference only for the backlash in the circuit 14
is there. Therefore, the measured peak level is large.
Backlash because of small or small value
Peak when adding amount and not performing backlash processing
Calculate the level. This allows feedback
There seems to be no dead zone element in the loop
Unnecessary vibrations that come and go can be suppressed.
The output of the reverse backlash circuit 8 is a feedback
Limiter unit of the feedback unit of the loop system
Unit 11, normal unit 12, and basic unit 13
Sent. The output of the monaural circuit 7 is
It is also sent to the silence detection circuit 10. The silence detection circuit 10
Detects whether the input is silence as described above.
Circuit for detecting silence when the silence flag is detected.
To change. The silence flag is
Used by the backup units 11, 12, and 13. For example,
From the current attenuator value and the peak detection circuit 5 value,
Calculate absolute level of currently input signal ｛absolute level
= Peak level (log) + attenuation (log)｝
This is, for example, below a certain value (below the silence detection level)
Then, for each of the feedback units 11, 12, and 13,
The input of an internal LPF, which will be described later, is set to zero. to this
No need for AGC volume between music songs, for example
To prevent it from being raised. The output of the silence detection circuit 10 is also
To be sent to the feedback units 11, 12, 13
Has become. Next, the limiter unit 11 and the normal
Feedback of the unit 12 and the basic unit 13
Unit calculations are performed, and these limiters, normal,
Add the output of each basic unit to AGC
Output as an attenuator control value. This file
The detailed configuration of the feedback unit will be described later.
You. Here, the limiter unit 11 is
The fastest (for example, 0.1 seconds or less)
Bottom) A feedback unit that reacts. The normal unit 12 is used for transmitting music and the like.
Medium speed (eg 1 second)
10 to 10 seconds)
You. In addition, the characteristics between line input and microphone input
Switches. Also, the above line input and the above
The characteristics that switch according to the microphone input are described below.
Upper and lower operating levels, upper and lower feedback loop
IN, attack time constant, recovery time constant, and the like. The above-mentioned basic unit 13 is different in the sound source and the like.
The lowest speed (for example, 10 seconds to
Feedback unit that responds in about 100 seconds)
You. This unit also has a
The characteristic switches between the input and the input. Note that, in FIG.
The block corresponding to the force is indicated by L in the figure, and
Corresponding blocks are indicated by M in the figure. Each of the above feedback units 11, 1
The outputs of 2 and 13 are added by an adder 44,
It is sent to the rush circuit 14. The backlash circuit
14 is each feedback unit 11,12,13
A dead band is added to the calculated output to prevent small fluctuations in the input.
The attenuator at the subsequent stage operates to produce unnatural sound.
Is what you want to do. The backlash times
Details of the operation on the road 14 will be described later. The backlash circuit 14 and the reverse back
The delay device 15 provided between the rush circuits 8
When the audio signal level falls, 1
Sample time delay occurs, only when the level falls
This is for delaying one sample time. Sandals
The peak level detected by the peak detection circuit 5
Indicates that the audio signal is increasing one time before, while decreasing
Observe the value of the attenuator value two times before. this
Backlash, which is compensated by the reverse backlash circuit 8
When the audio signal increases, the backlash also becomes
Use the previous backlash value when decreasing.
I have to. Through the delay unit 15, this processing
Be executed. The backlash caused by the delay unit 15
The details of the queue delay will be described later. The level drop is detected by detecting the level drop.
/ Fall determination circuit 9. Level rise / fall judgment times
The path 9 is used to switch the delay in the delay
Measure the audio signal level to determine the next reverse backlash.
For this purpose, a flag corresponding to the rise / fall is set. The output of the backlash circuit 14
Is connected via a switch 41 whose switched terminal on the AGC side is selected.
Then, it is sent to the offset adding circuit 16. The offset
The adder circuit 16 has the above-mentioned fee for a certain reference input.
Reduced in case the calculated value of the debug unit becomes zero.
The amount of decay is arbitrarily determined in advance, and
To the calculated value from the feedback unit via road 14
On the other hand, the predetermined attenuation is added.
You. In other words, offset addition means that the reference input value is
Attenuator value (control value) at the time of input.
Is a value determined in
You. In this embodiment, the value is increased by the offset addition.
The gain obtained is, for example, 10 dB. The office
The output of the bit adder circuit 16 is the attenuator operation range limit circuit.
It is sent to the road 17. The attenuator operation range limiting circuit 17
Is within the operating range of the attenuator volume.
It limits the value. That is, attenuator operation
In the range limiting circuit 17, by setting the loop gain or the like,
Calculates a control value that exceeds the operating range of the attenuator
So limit this. Subsequent stage of the attenuator operation limiting circuit 17
Rounding circuit 18 of each feedback unit
The number of digits of the calculated value does not match the attenuator resolution.
If this is not the case, round off
It is. Specifically, in the rounding circuit 18,
Calculates the error due to the quantization of the control value of the
Reduce by half. This value is calculated by the above-mentioned silence detection circuit 10.
And is also used here for silence detection. Rounding off
The output of the circuit 18 is the digital / analog distribution circuit 1
9 The digital / analog distribution circuit 19
Is the analog attenuator 1 and the digital multiplier
(Digital attenuator)
This is to distribute the amount of attenuate. That is,
In the digital / analog distribution 19, the rounding circuit is used.
The output of 18 is truncated to analog resolution.
The control value of log attenuator 1
Control to digital multiplier (digital attenuator) 3
Distribute as values. Specifically, in this embodiment,
The resolution of the analog attenuator 1 is 2 dB,
Since the resolution of the multiplier 3 is 0.25 dB,
2 dB steps for analog attenuator 1.
Control value for the digital multiplier 3 is 2 dB
A system that fills the steps with 0.25dB steps
Distribute your value. Although illustration is omitted, digital
Side, there is a time delay due to the A / D converter 2
Therefore, the change time matches the analog attenuator 1.
There is an appropriate delay. The digital / analog distribution circuit 19 and
An antilogarithmic conversion circuit 20 is provided between the digital multiplier 3 and the digital multiplier 3.
Inserted and connected. The antilogarithmic conversion circuit 20 has a logarithmic
Convert the compressed control value to a linear value,
To make the operation of the attenuator in the arithmetic unit 3
Converter. That is, the antilogarithmic conversion circuit 20
Is output from the digital / analog distribution circuit 19.
Digital attenuator (digital multiplier 3)
Since the control value is expressed in logarithm,
The linear ratio of the
Has been converted to an amount. In this embodiment, the above-mentioned feed
The control value from the back unit is stored in the backup RAM 24
Is always held. The backup
The RAM 24 stores a power supply as an initial value for the next AGC operation.
This is a memory for holding a value even after being turned off. Ie
For example, when the set is turned on,
From the backup RAM 24 to the manual adjustment register 22
Data is transferred. This backup RAM 24
The memory location switches between input and microphone input
(Blocks L and M in FIG. 1). Further, the backup RAM 24
The stored control value includes the output of the limiter unit 11.
Is to be subtracted. That is, Limitayu
The knit 11 corresponds to instantaneous large input, and
Since the bell is not expressed, the initial value of the next AGC is
Control from the limiter unit 11 based on the current control amount
Use the value after subtracting the value. Specifically, above
The backup RAM 24 is updated by a subtractor 45.
From the original value of the control value of the tenator, the limiter unit 11
The value obtained by subtracting the control value is further backed up by the rounding circuit 21.
The bit width of the memory of the backup RAM 24 (8 in the embodiment)
Bit) is rounded down and stored
Swelling. The limiter unit 11 operates normally.
The current control value is
(The output of the backlash circuit 14)
Is more present. The output of the backup RAM 24 is
This is sent to the backup limit circuit 23.
The backup limiting circuit 23 operates once when the power is turned on.
Only from the backup RAM 24 to the subsequent manual adjustment register
The upper and lower limits are separately set for the value transferred to 22
It is. Also, in this backup restriction circuit 23,
The limit value can be set differently for the line and microphone.
(Blocks L and M in FIG. 1). Accordingly
The microphone input by the backup limiting circuit 23
And line input will have their ranges limited separately.
You. In addition, it does not affect the current microphone / line input mode.
The microphone input and the line input are both
You. Here, the configuration of the present embodiment is a
In-control is also possible. That is, this
In the case of manual gain control, the analog
The attenuator 1 is not shown in FIG.
Manual adjustment level according to the amount of operation from the recording volume
Based on the manual control value output from the
For example, the analog audio signal is attenuated. That is, the operation from the recording volume
The amount is, for example, the rotation angle data of the dial or the remote control
Operation data output from the volume increase / decrease button.
The operation data is sent to the manual adjustment register 22.
It is. The manual adjustment register 22 has the variable
Control value for level adjustment by the
From the stored control values.
The control value that goes up / down according to the operation amount of the
It is supposed to be. In the present embodiment,
As described above, the analog attenuator 1 is
An attenuator is configured with the arithmetic unit 3, and
Therefore, these attenuators are not
From the manual adjustment register 22 corresponding to the operation amount of
Is controlled by the dynamic control value.
Controls are being implemented. Note that this
The attenuator has a wide input operating range in addition to its original purpose.
It is also intended to be. The output from the manual adjustment register 22
The manual control value that is set depends on the value supplied to the attenuator.
If the sound level adjusted by the attenuator is
It changes in 2dB steps (that is, it changes step by step)
). This manual adjustment register 22
The manual control value is sent to the low-pass filter 26. Here, the low-pass filter 26 is, for example,
It has a time constant of about 1 second, for example.
The manual control value that changes step by step from the
Converting to movement. For example, the above manual adjustment register 2
2 based on the manual control value from
The bell adjustment is increased in the above 2dB step, for example.
, The output of the low-pass filter 26 is
FIG. 3A shows this in association with a change in volume level.
It changes smoothly. The output of the low-pass filter 26 is
It is distributed by a digital / analog distribution circuit 19. This
The digital / analog distribution circuit 19 is used for the manual adjustment.
For example, a 10-bit control value is supplied from the register 22.
For example, the upper 3 bits in the
Distributed to the attenuator 1 side, for example, the lower 7 bits
It is distributed to the digital multiplier 3. This allows
8 levels for analog attenuator 1
The coarse adjustment of the signal is performed by the digital multiplier 3.
And the output level data of the A / D converter 2 as an example.
For example, 1/2 ⁷ Fine adjustment by the level difference of. In the digital / analog distribution circuit 19,
Volume adjusted by the above attenuator for distributed control value
Table 1 shows the relationship with the level change. [Table 1] According to Table 1, the analog athens
The volume level adjusted by data 1 is in 2 dB steps.
The volume level adjusted by the digital multiplier 3 is
This is a 0.25 dB step. That is, the above digital
From the analog distribution circuit 19
The control value sent to data 1 corresponds to the above-mentioned change in volume level.
As shown in FIG. 3B, the volume is increased in 2 dB steps.
It is a value that changes the level.
The signal is sent from the analog distribution circuit 19 to the digital multiplier 3.
Control values associated with the changes in volume level
As shown in C of FIG. 3, the volume level is set in 0.25 dB steps.
A value that changes the bell. Thus, in this embodiment, the manual
When performing level control of the
Level control that does not change
Troll can be realized. In the circuit of this embodiment, the manual adjustment
The control value output from the register 22 and the analog
By adjusting the resolution of the A / D
Avoid damaging the dynamic range of converter 3
are doing. That is, held in the manual adjustment register 22
The set control value is changed to the analog
By limiting the value to a value at which only the tenator 1 operates, the A
Do not impair the dynamic range of / D converter 2
I'm trying. As described above, according to this embodiment, the manual
When performing the operation, the low-pass filter 26
By smoothing the control value from the adjustment register 22,
Thus, variable volume noise can be removed.
Also, in this embodiment, a hand in accordance with the operation amount of the recording volume is set.
The control amount of the dynamic adjustment register 22 is
The dynamic range is matched with the dynamic range.
There is no loss of the sensor. Further, the manual adjustment register 2 of the present embodiment
2, manually adjust the recording level as described above
Volume value (recording volume operation amount)
Output the same manual control value, and, for example,
When going back and forth between the manual gain control modes
To prevent the discontinuity of the gain,
Current calculated by rounding 21 above via switch 42
The adjustment amount (volume value) is always put in
I have. Note that the manual adjustment register 22 of this embodiment is
Independent manual control values corresponding to the microphone input and the microphone input.
I have That is, as shown in FIG.
Al and AGC modes can be switched
For example, switching from AGC to manual operation
Or switch from manual to AGC operation
The output of the feedback unit (limit
Total of the output of each unit
As the initial value of the manual or transfer to AGC
When performing the operation, the manual control value is
It is conceivable to set it to the initial value. In this case, the above AGC
Configuration prevents unnecessary movement of gain
The above backlash element is put in the loop
However, the output of the feedback unit (limit
Total of the output of each unit
If the initial value of the manual is
The difference may result in a gain jump. Also,
Move from AGC to manual while limiter unit is operating
When executed, the instantaneous value is sampled, and the true value and
There is a possibility that the gain may be far apart. Furthermore,
For example, immediately after fade out by manual operation, A
When the operation shifts to the GC operation, the initial gain of the AGC becomes zero.
It takes a long time to reach the appropriate range
There is a possibility that it will happen. In addition, even during manual operation
Only the limiter unit operates to cope with large input
Sometimes it is necessary to do this, but at this time a new backlash
May cause a gain jump. From the above, the configuration of FIG.
For example, in the case of
In this case, the value of backlash processing during AGC operation
Subtracter 45 subtracts output value of limiter unit 11 from
And put it in the manual adjustment register 22
You. Specifically, in this embodiment, a manual
When recording starts and the gain control mode
When the system moves from AGC to manual,
The value of the adjustment register 22 is stored in the low-pass filter
Load into the Vista. In this embodiment, the manual
Since the adjustment register 22 is 1 byte, the low-pass filter
Load the same unit in the register of
Is cleared. This allows a low-pass filter
26 is already converged, so a low-pass
The gain is not changed by the filter 26.
You. After that, the manual gain control
The operation becomes executable. Here, the value after backlash processing is used
What we do is that this is the current attenuator value,
This is because there is no change at the time of manual transfer. Also, the above
Subtractor 45 subtracts the output of limiter unit 11
As described above, the limiter unit 11
If this is sampled when the input is supported,
Because the average control value is not expressed,
The sum of the output of each unit and the basic unit represents the long-term average value.
Because it is. In addition, the manual adjustment register 22
Generally, there is only a resolution of the adjustment step unit, so the control value
Quantize by rounding to the nearest digit of resolution rather than truncating
In this embodiment, this error can be halved.
In consideration of the above, the rounding circuit 21 is inserted. Further, when changing from manual to AGC
In the case of
An unusual value such as a writeout is stored in the manual adjustment register 22.
If it is, the range is limited by the range limiting circuit 25.
To the basic unit of the AGC feedback unit.
By initializing to limit 13 (limiter and normal are zero
To avoid insane settings at the beginning of AGC
And to be able to. In other words, the above range limit times
Road 25 is, for example, at the start of REC or
Changed the gain control mode to AGC operation
Sometimes, the initial value of AGC operation is changed to a backup value or manual.
Set value (both values in the manual adjustment register 22)
However, in this embodiment, for example, the manual control value is an insane value.
If (for example, the last time you finished with a fade-out)
In addition, by limiting this, the AGC error at the beginning of recording is disabled.
It is trying to ease the situation. Also in this case back
Rush is zero. Specifically, the gain control
Roll mode changes from manual to AGC mode
The value of the manual adjustment register 22 is
5 to limit the range
Register of LPF of knit 13 (low-pass fill of FIG. 4)
Initially set in the register 114) of the The above range limit
The range limitation in the circuit 25 depends on the nature of the sound source.
Decide how much range to affect the initial value of AGC
Things. Also in this case, the manual adjustment register 22
Because it is 1 byte, load to the same unit of LPF
And clear the other bytes. In the configuration of the present embodiment shown in FIG.
The limiter, normal, and basic feedback units
G11,12,13 are from the above manual to AGC
For example, at the time of REC start (REC PA
USE, REC PLAYPAUSE)
It is initialized even when the operation mode is switched.
You. In this case, the limiter unit 11 and the normal unit
The unit 12 is initialized to zero and the basic unit 13 is manually adjusted.
The value of the integer register 22 is initialized by limiting the range. That is, the configuration shown in FIG.
REC start (for example, RECPAUSE, REC
Mode to perform various operations such as PLAY PAUSE
The setting switch 32 and the mode setting switch 32
Mode recognition hand for recognizing operation modes corresponding to various operations
And a controller 31 as a stage.
In response to recognizing the operation mode in
Immediately before the operation mode is switched, the feedback unit
The control value set in the backup RAM 24
And when the above operation mode is switched,
The control value stored in the backup RAM 24 is initialized to the initial value.
To the feedback unit as
The above control supplied as an initial value to the value calculation unit 71
Value for adjusting the signal level in the attenuator 83.
It is used as a control value. Next, the feedback loop unit
Will be described with reference to FIG. Limi above
Tta, normal and basic units 11, 12, 13
Actually consists of a single configuration,
Change for each unit and microphone / line input mode
In this way, five operations are executed. In this FIG.
Here, the block indicated by L in the figure corresponds to the line input
This is a block that holds the parameters.
Block holds the parameters corresponding to the microphone input.
In the figure, M and L are microphones and line inputs.
A block that holds the same parameters by force. This
These parameter holding blocks are basic, normal, and
Provided according to the operation of each unit of the limiter (limiter
Only microphone and line input have the same parameters). That is, in FIG.
Is supplied with the output of the reverse backlash circuit 8.
The input signal is supplied to subtractors 116 and 120.
Each of the upper operating points (upper operating levels) is
Each parameter from the holding unit 115 that holds the parameter is
Are supplied to the subtractor 120.
Holds each parameter of operating point (lower operating level)
Each parameter is supplied from the unit 122
I have. The output of the subtracter 116 is a comparator
117, and the output of the subtractor 120 is
Is sent to the The comparator 123 is supplied
If the obtained signal (difference) is negative (that is, the terminal 121
(When the input signal supplied to the above is smaller than the above lower operating point)
Is supplied with the output of the subtractor 120 of the switch 119.
And the supplied signal (difference) is
In the positive case (ie, the input signal supplied to the terminal 121)
Switch 119 when the signal is above the lower operating point)
A switching control signal for selecting the switching terminal a (zero input) is output.
The comparator 117 determines that the supplied signal (difference) is positive.
(That is, the input signal supplied to the terminal 121)
Is greater than the above upper operating point).
Select the switched terminal a to which the output of the subtractor 116 is supplied.
And if the supplied signal (difference) is negative (ie
The input signal supplied to the terminal 121 is equal to or higher than the above operating point.
(Lower case) the output of the switch 119 of the switch 118
Outputs switching control signal to select switched terminal b to which power is supplied
I do. In other words, the output from the switch 118
As for the output, the input from the terminal 121
Larger case, smaller than lower operating point and intermediate
And zero in the middle case
If it is larger than the point, the difference from the upper operating point is
If it is smaller than the point, the difference from the lower operating point is the next stage.
Sent to the configuration. That is, with these configurations,
AGC does not operate at the interim level without feedback.
A dead zone. Also, set the lower operating point to the minimum value.
Responds only to large input like limiter
You can make a unit that does. In addition, the upper operating point and lower
The same operating point eliminates the dead zone,
You can create the characteristics of this unit. The output terminal of the switch 118 is connected to a switch.
The terminal 105 is connected to the switched terminal b. The switch
The switched terminal a of the switch 105 has a zero input. The
The switch 105 is connected to the silence detection circuit via the terminal 101.
Switched according to the silence flag supplied from the road 10
When the silence flag is supplied (silence detection
Output), the zero input on the switched terminal a side is output,
If there is no flag, switch to the switched terminal b side.
It is what is done. The output of the switch 105 is supplied to the multiplier 106
Sent to The multiplier 106 has an upper gain and a lower gay
Parameters from the parameter holding units 102 and 103
In response to the switching control signal of the comparator 117
As supplied via switch 104 which is switched
It has become. The switch 104 corresponds to the comparator
The switching terminal a when the switching control signal from the
Is selected, and conversely, in the case of a negative value, the switched terminal b is selected. That is, in the multiplier 106, the end
If the input of the child 121 is larger than the above upper operating point,
The gain is set so that the input of the terminal 121 is lower than the lower operating point.
If so, a process of multiplying the lower gain is performed. Gay on this
And the lower gain are feedback loop gains.
When the gain is increased, the level is compressed and returns to the reference level.
Get closer. Also, lowering the gain weakens the compression force,
The ability to approach the sub-level is also reduced. The output of the multiplier 106 is supplied to a subtractor 111.
Through the multiplier 112 and the comparator 1
07. The comparator 107 has the power
When the output of the arithmetic unit 106 is zero or more,
A switching control signal for selecting the switching terminal a is output, and the multiplier
When the output of the switch 106 is smaller than zero,
Outputs a switching control signal for selecting the switched terminal b side of
is there. The switched terminal a of the switch 110 has an attack
Parameter holding unit 10 for holding each parameter of the time constant
8 are supplied, and the switched terminal b is
Parameter holding to hold each parameter of the coverage time constant
Each parameter is supplied from the unit 109
I have. The output of the switch 110 is supplied to the multiplier 112.
Sent. The multiplier 112 and the subtractor 111 are
IIR type buffer with adder 113 and register 114
A low pass filter. That is, the above multiplier
The output of the register 112 is supplied to the register 11 via the adder 113.
4 and the output of the register 114 is
It is fed back to the adder 113 and the subtractor 111
The parameter is stored in the multiplier 112.
The parameters (multiplication coefficients) from the holding units 108 and 109 are supplied.
Is supplied to form an IIR low-pass filter.
ing. Specifically, pass through the low-pass filter
To prevent the system from oscillating and slow down the AGC
Thus, the characteristics of each feedback unit are formed.
This low-pass filter has a time constant of attack and
Integral characteristics can be adjusted because it can be set separately for the balance
You. Further, above the IIR type low pass filter,
The output of the register 114 is fed back from the terminal 124.
Taken as the output of the group unit. In addition, above
The register 114 is connected to the
Loaded with the char value. FIG. 5 shows the structure of the backlash circuit 14.
2 shows a flowchart of backlash in the embodiment. Here, the analog attenuator 1
The control value and the value logarithmically compressed by the logarithmic compression circuit 6 are used.
If the unit is different, the backlash circuit 14
By multiplying the output of these backlash amounts by a coefficient,
I have to correct it. Note that, in this embodiment,
The attenuator side is in 2 dB units and the peak detection side is in 6 dB units.
The backlash amount is divided by 3
You. That is, in FIG.
The value input to the rush circuit 14 is
The difference from the output value of each time is taken. Next, in step S2,
The difference in step S1 exceeds the width of the backlash
Check (check up and down by backlash width). In step S2, the backlash width
If it is determined that the value does not exceed the limit, the process proceeds to step S8.
Only. 3 minutes difference in backlash output in step S8
(The difference at this time is the amount of backlash.)
The backlash output is used as the reverse backlash value.
Is done). Then, backlash processing is performed in step S9.
Output remains unchanged (dead zone), backlash
Processing ends. In step S2, the backlash
If it is determined that the width exceeds the width, step S3
Proceed to. In step S3, it is determined whether the difference is positive.
Disconnection is made. If it is determined to be positive in step S3,
Goes to step S5, and the
Output one third of the negative backlash width
Power. Next, in step S7, the backlash processing is executed.
Force is the value obtained by subtracting the backlash width from the current input value.
(Ie, output a value smaller by the backlash width)
). Thereafter, the backlash processing ends. If it is determined in step S3 that the value is negative,
Proceeds to step S4. In step S4, the battery
Output 1/3 of backlash width to crash output
You. Next, in step S6, the output of the backlash processing is now
To the value obtained by adding the backlash width to the input width
That is, a value larger by the backlash width is output). That
Thereafter, the backlash processing ends. Note that the back in the above flowchart is
The rush width is a value on one side, and is 0.75 in the embodiment.
dB. Also, the backlash output is delayed as described above.
The signal is input to the reverse backlash circuit 8 through the extension device 15.
Therefore, it is divided by 3 in advance to match the units. Referring to FIG. 6 and FIG.
The backlash delay will be described. A in FIG.
FIG. 7 shows an example in which the audio signal is increasing, and FIG.
A shows an example where the audio signal is decreasing.
Show. 6B and FIG. 7B show the AGC execution period.
6C and 7C show the amount of backlash one time before.
The data of FIG. 6D and FIG.
FIG. 6 and FIG.
T _C Indicates the timing of the changing point of the attenuator.
T _P Indicates an example of a peak detection point. That is, for example, as shown in FIG.
While the audio signal is increasing, the attenuator control value
When the value is changed, the level in the latter half of the peak detection section is newly controlled.
Value and the signal is increasing,
The maximum value of the section is in the latter half. Therefore, the next peak detection
, A peak level reflecting the new control value is detected. I
Therefore, the audio signal increases as shown in FIG.
When backlash correction is performed
Is, for example, the arrow R in the figure. ₆ The error of the new control value
The data of the backlash amount one time before C in FIG.
use. For example, as shown in FIG.
While the audio signal is decreasing, change the control value of the attenuator.
The second half of the peak detection period is affected by the new control value.
However, since the maximum value of the section is decreasing,
In the next peak detection, it becomes the peak value before the new control value.
You. For this reason, the peak detection value reflected by the new control value
Next (D in FIG. 7). Therefore, FIG.
If the audio signal is decreasing as
When backlash correction is performed, for example, ₇ To
As shown, the data of the backlash amount two times before D in FIG.
Use data. FIG. 8 shows a specific configuration of another embodiment of the present invention.
Show. The configuration shown in FIG.
In roll mode, the feedback loop is off
The configuration shown in FIG.
Feedback even when in gain control mode
The same limiter as described above is performed by the loop limiter unit 11.
This is a configuration that makes it possible to perform a shutter operation. What
In FIG. 8, the same components as those in FIG.
The reference numerals are attached and the description is omitted. In FIG. 8, the switch shown in FIG.
41, an adder 51 is provided instead of the adder 51.
1 and the low-pass filter 26
Switch 52 which is turned on in the control mode
From the output of the low-pass filter 26
A subtractor 53 for subtracting the value of the menu width is inserted and connected.
The rounding circuit 21 and the backup RAM 24
, Between the normal unit 12 and the adder 44, and
Between the basic unit 13 and the adder 44.
Switches 56, 54 and 55 that are turned on when
I have. That is, a limiter is added to the manual operation.
Backlash output for AGC in Fig. 1
(The output of the limiter unit 11 via the adder 44) and
Manual operation output (output of low-pass filter 26)
Add (however, the output from the low-pass filter 26
Subtract the width of the crash)
Input to road 16. Here, the low pass is performed by the subtractor 53.
Subtracting the backlash width from the output of filter 26
For the following reasons. For example, the limiter unit 11 moves momentarily
After making and returning, the backlash always moves in the forward direction.
With the width of the crash (the limiter operates for a moment)
Large input is well beyond the backlash width)
After the limiter has been operated even once,
It will be a normal value. Therefore, the above backlash width is subtracted.
Correct manual control value that is not affected by the limiter operation.
Output to the attenuator. Before the limiter is activated,
In the case of
Backlash is in the forward direction when shifting to Al mode
Is initialized to have the backlash width. You
That is, the output value of backlash = + backlash
Width, backlash output = + backlash width / 3
You. As described above, according to the configuration of this embodiment,
AGC with backlash, manual and AG
Trouble such as gain jump when going back and forth between C operations
Can be avoided. As described above, the signal level control of the present invention
In the device, the no-signal detection that detects the no-signal portion of the signal
Signal level adjusting means having signal output means
Is the signal level control means for controlling the amount
These non-signal detection signals are received by this signal level adjusting means.
Slow down or stop the signal level adjustment
Control, such as between music tunes,
The sudden rise in level in the sound part (no signal part)
To prevent level over at the beginning of the next song between songs.
Is possible.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block circuit diagram showing a specific configuration of a level control circuit according to an embodiment of the present invention. FIG. 2 is a diagram illustrating a relationship between silence detection and level adjustment according to the present embodiment. FIG. 3 is a diagram illustrating a manual control value and a change in volume corresponding to a specific configuration of the embodiment. FIG. 4 is a block circuit diagram showing a specific configuration of a feedback unit. FIG. 5 is a flowchart of a backlash process. FIG. 6 is a diagram for explaining delay of backlash when an audio signal is increasing. FIG. 7 is a diagram for explaining a delay of backlash when an audio signal is decreasing. FIG. 8 is a block circuit diagram showing another specific configuration of the present embodiment. [Explanation of Signs] 1 ... Analog Attenuator 2 A / D Converter 3 Digital Multiplier 4 ... DSP 5 peak detection circuit 6 logarithmic compression circuit 7 monaural circuit 8 reverse backlash circuit 9 Level rise / fall determination circuit 10 Silence detection circuit 11 Limiter unit 12 Normal unit 13 Basic unit 14 Backlash circuit 15 Delay unit 16 Offset adder circuit 17 Attenuator operation range limitation Circuits 18, 21 ··· Rounding circuit 19 ··· Digital / analog distribution circuit 0: antilog conversion circuit 22: manual adjustment register 23: backup restriction circuit 24: backup RAM 25: ..Range limiting circuit 26 ........ Low-pass filter

Continuation of the front page (56) References Japanese Utility Model Sho 58-5005019 (JP, U) Japanese Patent Publication Sho 50-40653 (JP, B2) Japanese Patent Publication Sho 62-5363 (JP, B1) (58) Fields surveyed (Int .Cl. ⁷ , DB name) G11B 20/00-20/04 G11B 5/09 311 H03G 3/20-3/34

Claims (1)

(57) [Claim 1] Signal level adjusting means for adjusting a signal level of an input signal, and said output for measuring the level of an output signal output from said signal level adjusting means. Level measurement signal processing means for performing predetermined signal processing on the signal; silence detection means for detecting a silence portion of the input signal based on a level measurement signal output from the level measurement signal processing means; Controlling the signal level adjustment amount of the signal level adjustment means based on the level measurement signal output from the level measurement signal processing means, and detecting the silent part of the input signal by the silent detection means; A signal level control unit for stopping or delaying the control of the signal level adjustment amount.