JP4882818B2 - Dynamics control device - Google Patents

Description

  The present invention relates to a dynamics control apparatus suitable for use in audio signal dynamics control.

  Noise gates and duckers are known as dynamic control of audio signals (Non-Patent Document 1). Here, the noise gate is used for cutting the noise of the audio signal, and an example of the gain characteristic is shown in FIG. In FIG. 6A, if the input signal level is equal to or higher than the threshold value Thn, the gain is “0 [dB]” and the audio signal is not attenuated. However, if the input signal level is lower than the threshold value Thn, the gain is “− Gn [dB] "and the audio signal is attenuated. Thus, for example, during a period in which a signal having a certain level, such as a musical sound signal or a human voice, is input, the voice signal is output without being attenuated, and the output signal is attenuated in other periods. Therefore, noise generated in the latter period is attenuated.

  Dhaka is a signal that increases or decreases the gain of a certain audio signal (here, called a controlled channel) based on the level of another audio signal (also called a control channel). As shown in FIG. In FIG. 6B, when the level of the input signal of the control channel is less than the threshold value Thd, the gain of the controlled channel is “0 dB” and the audio signal is not attenuated, but the level of the input signal of the control channel is the threshold value. When it is equal to or greater than Thd, the gain of the controlled channel becomes “−Gd [dB]”, and the audio signal of the controlled channel is attenuated. Here, for example, if the controlled channel is BGM (background music) sound and the control channel is an announcement sound, the volume of the BGM sound is automatically reduced when the announcement is started.

  By the way, when the noise gate processing is executed according to the characteristic of FIG. 6A, for example, after the sound once rises, the audio signal is immediately attenuated when the input signal level becomes less than the threshold value Thn even in the period when the reverberation remains. There was a problem that the reverberation part was interrupted. In order to solve this, Patent Document 1 discloses that when the input signal level rises above the threshold value Thn and further falls below the threshold value Thn, a predetermined hold is obtained from the peak timing detected at the input signal level exceeding the threshold value Thn. A technique for holding the gain at 0 [dB] only for a time is disclosed.

Japanese Patent No. 3293240 "PM5D Instruction Manual" Yamaha Corporation, 2004

However, the noise gate function disclosed in Patent Document 1 has a problem in that the hold time is too long or too short because the hold time is constant regardless of the state of the audio signal.
The present invention has been made in view of the above-described circumstances, and an object thereof is to provide a dynamics control device capable of providing an appropriate hold time according to the state of an audio signal.

In order to solve the above problems, the present invention is characterized by having the following configuration. The parentheses are examples.
2. The dynamics control device according to claim 1, wherein a controlled audio signal input unit (2) for inputting a controlled audio signal (first channel) and a controlled audio signal for inputting a control audio signal (first channel). The rectification processing unit (34) that rectifies the control audio signal input from the input unit (2, 32, switch 32 is the gate side) and the control audio signal input unit (2, 32 ) and outputs the rectified signal (S34). And an envelope follower processing section (38) for following the rise of the rectified signal (S34) without delay and outputting an envelope signal (S38) that gradually attenuates at the fall of the rectified signal (S34). , the gain control unit for controlling the gain of said controlled audio signals in accordance with the supplied gain control signal (S46) and (48), and the envelope signal (S38) becomes equal to or higher than a predetermined threshold value (TD) Gay The control unit (48) sets the gain given to the controlled audio signal to a high level, and sets the gain to a low level when the envelope signal (S38) becomes less than the threshold value (TD). And gain control signal generating means (40 to 46) for outputting the gain control signal (S46), and the higher the peak value of the control audio signal , the longer the period for setting the gain to the high level. It is characterized by doing.
Further, in the dynamics control device according to claim 2, the controlled voice signal input unit (2) for inputting the controlled voice signal (first channel) and the control for inputting the control voice signal (first channel). An audio signal input unit (2, 32, switch 32 is on the ducker side) and a rectification processing unit that rectifies the control audio signal input from the control audio signal input unit (2, 32) and outputs a rectified signal (S34) (34) and an envelope follower processing unit (38) that follows the rise of the rectified signal (S34) without delay and outputs an envelope signal (S38) that gradually attenuates at the fall of the rectified signal (S34). ), A gain control unit (48) for controlling the gain of the controlled audio signal in accordance with the supplied gain control signal (S46), and the envelope signal (S38) exceeding a predetermined threshold (TD). in front A gain controller (48) sets the gain given to the controlled audio signal to a low level, and sets the gain to a high level when the envelope signal (S38) becomes less than the threshold value (TD). As described above, a gain control signal generating means (40 to 46) for outputting the gain control signal (S46) is provided, and a period during which the gain is set to the low level as the peak value of the control audio signal increases. Characterized by lengthening .

Thus, according to the configuration according to claim 1, from a longer period to hold the gain higher peak values occurring in the controlled audio signal is high to a high level, a relatively long time finish against strong peaks It can function as a noise gate that can output without attenuation .
Further, according to the configuration according to claim 2, since a longer period to hold the gain higher peak values occurring in the controlled audio signal is high to a low level, when a strong peak occurs, attenuates the controlled audio signal It can function as a ducker that ensures a relatively long period .

1． Next, the configuration of a dynamics control device according to an embodiment of the present invention will be described with reference to FIG.
In the figure, reference numeral 2 denotes a waveform input unit, which converts a plurality of channels of analog input signals into digital signals. Reference numeral 4 denotes a DSP (digital signal processor) that performs processing such as dynamics control on each input signal. Reference numeral 6 denotes a waveform output unit that converts digital output signals output from the DSP 4 into analog signals. Reference numeral 10 denotes a display that displays various information to the user.

  Reference numeral 12 denotes an operation unit, which includes various switches. A CPU 14 controls each unit via the bus 8 according to a control program stored in the ROM 16. In particular, the CPU 14 loads a microprogram for controlling the dynamics to the DSP 4 based on the parameter setting command input via the operation unit 12 and controls the dynamics for the registers in the DSP 4. The following various parameters (SEL, HD, TD, AD, RD, etc.) are set. Reference numeral 18 denotes a RAM, which is used as a work memory for the CPU 14.

  Next, an algorithm implemented by the microprogram and parameters in the DSP 4 will be described with reference to FIG. In the figure, the controlled channel that receives the dynamics control is the first channel (Ch-1). The dynamics control applied to the first channel is specified by a parameter called G / D selection mode SEL. This designates either the noise gate (G) or the ducker (D). In FIG. 2, when performing ducking, a control channel for controlling the dynamics of the controlled channel is defined as a second channel (Ch-2).

  Reference numeral 32 denotes a control channel selection switch. When the G / D selection mode SEL is a noise gate (G), the first channel is selected. When the G / D selection mode SEL is a ducker (D), the second channel is selected. Select as control channel. A rectification processing unit 34 rectifies the audio signal of the selected control channel and outputs it as a rectification signal S34. Reference numeral 36 denotes a logarithmic converter that calculates the logarithm of the rectified signal S34 and outputs the resulting decibel signal as a signal S36.

  A first envelope follower processing unit 38 follows the rising of the signal S36 without delay, and gradually falls linearly with a predetermined slope (hold rate) HD with respect to the falling of the signal S36. Is output. Here, examples of waveforms of the signals S36 and S38 are shown in FIGS. 3 (a) and 3 (b). Note that the signals related to FIGS. 3A to 3E are all decibel scale signals, but for ease of understanding, in FIGS. 3A to 3E, the waveforms are all linear scales. indicate.

  Returning to FIG. 2, reference numeral 40 denotes a threshold processing unit. When the signal S38 is equal to or greater than the predetermined threshold TD, the signal S40 is “S38−TD”, and when the signal S38 is less than the threshold TD, Output. A rectangular wave conversion unit 42 converts the signal S40 into a rectangular wave having a predetermined amplitude RG according to the G / D selection mode SEL. That is, when the G / D selection mode SEL is a noise gate (G), a rectangular wave signal S42 that is high when the signal S40 is positive and low when the signal S40 is "0" is output. . On the other hand, when the G / D selection mode SEL is Ducker (D), a rectangular wave signal S42 that is low when the signal S40 is positive and high when the signal S40 is “0” is output.

  Reference numeral 44 denotes a second envelope follower processing unit. In the case of a noise gate, the time constant of the attack rate AD is applied to the rising time of the rectangular wave signal S42 and the release rate RD is applied to the falling time. The waveform is transformed so as to gradually change logarithmically according to the constant, and the result is output as a gain control signal S44. Here, examples of waveforms of the signal S40, the rectangular wave signal S42, and the gain control signal S44 are shown in FIGS. In the case of performing the ducker process, since the rectangular wave signal S42 has the opposite polarity, the time constant of the attack rate AD with respect to the falling time of the rectangular wave signal S42 and the time constant of the release rate RD with respect to the rising time. Each should be applied.

  The rectangular wave signal S42 shown in FIG. 3 (d) is an example when the noise gate (G) is selected as the G / D selection mode SEL, but if the ducker (D) is selected, A signal obtained by inverting the high level and the low level becomes a rectangular wave signal S42. Comparing FIGS. 3A and 3D, the timing at which the rectangular wave signal S42 falls is longer than the timing at which the signal S36 falls below the threshold TD by the hold time HDT. The hold time HDT becomes longer as the peak value PK of the signal S40 becomes higher.

  Returning to FIG. 2, a linear converter 46 converts the decibel scale gain control signal S44 into a linear scale gain control signal S46. A multiplier 48 multiplies the input signal of the first channel by the gain control signal S46 and outputs the result as an output signal of the first channel.

2． Next, the operation of this embodiment will be described.
Assume that the noise gate (G) is selected as the G / D selection mode SEL of the first channel. Here, when a low peak (slightly exceeding the threshold TD) occurs in the input signal of the first channel, the hold time HDT is shortened and the gain control signal S44 falls quickly. Noise can be removed. Further, when a high peak occurs in the input signal of the first channel, the hold time HDT becomes long, and the gain control signal S44 is held at a high level for a relatively long time. As a result, the lingering sound that continues to be generated can be output without being attenuated.

  Further, it is assumed that Ducker (D) is selected as the G / D selection mode SEL for the first channel. Here, if a low peak (slightly exceeding the threshold value TD) occurs in the input signal of the second channel, the hold time HDT is shortened and the gain control signal S44 falls quickly, so the output signal level of the first channel Will be lowered for a relatively short time. Further, when a high peak occurs in the input signal of the second channel, the hold time HDT becomes long, and the output signal level of the first channel is lowered for a relatively long time.

3． Modifications The present invention is not limited to the above-described embodiments, and various modifications can be made as follows, for example.
(1) In the above embodiment, the dynamics control is performed by the micro program operating on the DSP 4, but only this micro program is stored in a recording medium such as a CD-ROM or a memory card and distributed, or through the transmission line. You may distribute it.

(2) Various modifications can be made to the algorithm (FIG. 2) realized by the DSP 4 in the above embodiment. For example, instead of the first envelope follower processing unit 38 and the subsequent threshold processing unit 40 in FIG. 2, a threshold processing unit 52 and a subsequent first envelope follower processing unit 54 are employed as shown in FIG. Also good. In this case, the signal S52 output from the threshold processing unit 52 is “S36−TD” when the signal S36 is equal to or greater than the predetermined threshold TD, and is “0” when the signal S36 is less than the threshold TD. That is, the signal S52 is a signal obtained by extracting only a portion exceeding the threshold value TD of the waveform shown in FIG. 3A, and the first envelope follower processing unit 54 determines the falling edge of the signal S52 according to the hold rate HD. The inclination is controlled, and a signal S54 similar to the signal S40 shown in FIG.

(3) In the above embodiment, instead of the logarithmic conversion unit 36 and the first envelope follower processing unit 38 in the subsequent stage shown in FIG. 2, a first envelope follower processing unit 62 and a subsequent stage as shown in FIG. The logarithmic conversion unit 64 may be employed. In the example of FIGS. 2 to 4, since the decibel scale signal is input to the first envelope follower processing units 38 and 54, the hold rate HD is a parameter for controlling the linear falling slope of the signal in the decibel scale. However, since the linear scale rectified signal S34 is input to the first envelope follower processing unit 62 in the example of FIG. 5, the hold rate HD has a logarithmically falling time constant of the signal in the linear scale. It is better to set the parameter to be controlled.

  That is, the fall of the signal S62 output from the first envelope follower processing unit 62 is logarithmic, but the signal S64 obtained by converting the signal S62 by the logarithmic conversion unit 64 is shown in FIG. The signal has the same shape as the signal S38 shown. That is, the operation in the example of FIG. 5 is equivalent to the operation according to the configuration shown in FIGS. As described above, the present invention is not limited by the order of each component of the algorithm, or the like, as long as the same result as in the above embodiment is obtained.

(4) In the above embodiment, the amplitude RG of the rectangular wave signal S42 is a predetermined value. However, the amplitude RG need not be a predetermined value, and may be increased or decreased according to the peak value PK. For example, when Ducker (D) is selected as the G / D selection mode SEL for the first channel, the amplitude RG may be decreased as the peak value PK of the signal S40 increases.

(5) Further, the dynamics control device of the above embodiment is not limited to one used as a single device, and may be incorporated in, for example, a mixer.

It is a block diagram of the dynamics control device of one example of the present invention. It is a block diagram of the algorithm performed in DSP4 of one Example. It is a wave form diagram of each part in FIG. It is a block diagram of the algorithm performed in DSP4 of a modification. It is a block diagram of the algorithm performed in DSP4 of another modification. It is a figure which shows an example of the gain characteristic of a noise gate and Dhaka.

Explanation of symbols

  2: Waveform input unit, 4: DSP, 6: Waveform output unit, 8: Bus, 10: Display, 12: Operation unit, 14: CPU, 16: ROM, 18: RAM, 32: Control channel selection switch, 34: rectification processing unit, 36, 64: logarithmic conversion unit, 38, 54, 62: first envelope follower processing unit, 40, 52: threshold processing unit, 44: second envelope follower processing unit, 46: linear conversion unit, 48: Multiplier (gain setting unit).

Claims (2)

A controlled audio signal input unit for inputting the controlled audio signal;
A control audio signal input unit for inputting a control audio signal;
A rectification processing unit that rectifies the control audio signal input from the control audio signal input unit and outputs a rectified signal;
An envelope follower processing unit for following the rise of the rectified signal without delay and outputting an envelope signal that gradually attenuates at the fall of the rectified signal;
A gain control unit for controlling the gain of the controlled audio signal according to the supplied gain control signal;
When the envelope signal is equal to or higher than a predetermined threshold, the gain control unit sets a gain to be given to the controlled audio signal to a high level, and when the envelope signal is less than the threshold, the gain is set to a low level. A gain control signal generating means for outputting the gain control signal so as to set to
The dynamics control device characterized in that as the peak value of the control audio signal becomes higher, the period during which the gain is set to the high level is lengthened. A controlled audio signal input unit for inputting the controlled audio signal;
A control audio signal input unit for inputting a control audio signal;
A rectification processing unit that rectifies the control audio signal input from the control audio signal input unit and outputs a rectified signal;
An envelope follower processing unit for following the rise of the rectified signal without delay and outputting an envelope signal that gradually attenuates at the fall of the rectified signal;
A gain control unit for controlling the gain of the controlled audio signal according to the supplied gain control signal;
When the envelope signal is equal to or higher than a predetermined threshold, the gain control unit sets a gain to be given to the controlled audio signal to a low level, and when the envelope signal is less than the threshold, the gain is set to a high level. as set, comprising a gain control signal generating means for outputting the gain control signal,
As the peak value of the control audio signal increases, a period during which the gain is set to the low level is lengthened .

Images