KR101981487B1 - Dynamic range compression device for multi-band and control method thereof - Google Patents

Abstract

The present invention relates to a dynamic range processing device for each frequency band and an operating method of the dynamic range processing device for each frequency band. In applying dynamic range amplification (compression processing) to voice signals for each frequency band, So that the effect of the dynamic range amplification processing on the voice signal can be maximized.

Description

TECHNICAL FIELD [0001] The present invention relates to a dynamic range range processing apparatus and a dynamic range range processing apparatus for each frequency band,

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to an apparatus and method for processing dynamic range by frequency band, and more particularly to a dynamic range compression (DRC) And a method for operating the dynamic range processing apparatus for each frequency band.

Generally, users who use a sound output device increase the playback sound volume of the sound output device to create a satisfactory volume magnitude. The dynamic range of the audio signal to be reproduced when the volume is increased is increased. When the reproduction sound volume of the sound output apparatus is increased to a point exceeding the limit related to the dynamic range of the sound signal, distortion such as clipping occurs.

In order to increase the volume while reducing such distortion, a dynamic range compression apparatus (hereinafter, referred to as " dynamic range compression apparatus ") that performs dynamic range compression (DRC), which reduces the volume magnitude difference between a large- DRC device) is used.

Such a DRC apparatus performs dynamic range compression processing on a speech signal in a speech signal preprocessing step before a speech signal is processed by a codec or another speech filter so that a speech signal having a large size and a small size Reduce volume size difference.

However, since the conventional DRC device of this type is a dynamic range compression processing technique for a general audio signal such as a hearing aid or music, the original sound signal itself may be distorted and the tone may be changed and input to the microphone of the mobile terminal. It is inevitable to improve the sound quality in the communication environment of the mobile communication in which the voice signal is transmitted to the other mobile terminal through the network.

For example, in a communication environment of a mobile communication, when the caller makes a call through the mobile terminal, the distance between the caller and the mobile terminal may not be maintained constant, and the original voice signal may be distorted and may not be constant. In particular, This phenomenon can be more prominent when using a call.

Accordingly, in the present invention, a method of maximizing the effect of the dynamic range compression processing on the voice signal in consideration of the specificity of the mobile communication call environment in applying the dynamic range compression processing to the voice signal is proposed.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a dynamic range compression (DRC) A dynamic range processing unit for each frequency band and an operation method of a dynamic range processing unit for each frequency band so that the effect of the dynamic range compression processing on the signal can be maximized.

According to a first aspect of the present invention, there is provided an apparatus for processing a dynamic range according to a frequency band, comprising: a signal divider for dividing an input audio signal into frequency bands; An amplification processing control unit for determining a dynamic range amplification value for each of the frequency bands based on the signal strength of the divided voice signal for each frequency band; An amplification processing unit for amplifying the voice signal for each frequency band based on the dynamic range amplification value for each frequency band determined by the amplification processing control unit; And a signal output unit for combining the voice signals of each frequency band amplified by the frequency band processing unit and outputting one voice signal processed in a dynamic range.

Preferably, the amplification processing control unit includes: a section determination unit that determines whether or not a voice signal of a specific section of the voice signal is a voice section; A type determination unit for determining a type of a voice signal according to a signal strength statistic value calculated on the basis of the signal strength of the voice signal for each frequency band of the specific section for each divided voice signal for each frequency band, And a frequency band amplification value determination unit for determining the amplification range of the dynamic range for each frequency band according to the determination result of the type determination unit.

Preferably, the type determination unit may determine at least one of a maximum value, a minimum value, an average value, and a standard deviation of the signal intensity of the speech signal for each frequency band of the specific section based on the signal strength of the speech signal for each frequency band of the specific section And the voice signal type according to the calculated signal intensity statistical value for the voice signal of each frequency band of the specific section can be determined based on the voice signal type for each predetermined signal intensity statistical value.

Preferably, the frequency band amplification value determination unit determines that the audio signal of the specific period is a voice interval as a result of the determination by the interval determination unit, the plurality of dynamic range input / output An input / output curve selector for selecting a dynamic range input / output curve for each of the frequency bands according to a type of a voice signal for a voice signal for each frequency band of the specific section determined by the type determining unit from a curve, A maximum absolute value determining unit for determining a maximum absolute value of absolute values of signal strengths of respective voice signal samples and determining the determined maximum absolute value as a maximum absolute value of the specific period, The determined maximum absolute value is used as the signal intensity of the input signal, And an amplification value determination unit for determining the amplification range of the dynamic range for each of the plurality of frequency bands.

Preferably, the maximum absolute value determining unit may determine that the maximum absolute value of the specific period and the maximum absolute value of the maximum absolute values of the voice signal that have been confirmed in the predetermined period preceding the specific period are finally determined as the maximum absolute value of the specific period You can decide.

Preferably, the amplification value determining unit determines amplification value of the specific range of the amplified value of the dynamic range for each frequency band determined in relation to the audio signal for each frequency band in the specific section, The amplification value of the dynamic range of the specific frequency band may be reduced by a specific value when there is an amplification value of the dynamic range of the specific frequency band in which the degree of increase of the amplification value is greater than or equal to the predefined specific increase degree as a result of comparison with the amplification value of the dynamic range for each frequency band .

According to a second aspect of the present invention, there is provided a method of operating an apparatus for processing a dynamic range according to a frequency band, comprising: dividing an input speech signal into frequency bands; An amplification process control step of determining a dynamic range amplification value for each of the frequency bands based on the signal strengths of the divided voice signals for each frequency band; Amplifying the voice signal for each frequency band based on the determined dynamic range amplification value for each frequency band; And a signal output step of outputting one voice signal processed in a dynamic range by combining the amplified voice signals for each frequency band.

Preferably, the amplification processing control step includes a voice signal section determination step of determining whether a voice signal of a specific section of the voice signal is a voice section, a step of determining whether the voice signal of the specific section Determining a voice signal type according to a signal strength statistic value calculated on the basis of a signal strength of a voice signal for each frequency band of the specific section for each voice signal divided into frequency bands for each frequency band, And an amplification value determination step of determining a dynamic range amplification value for each frequency band according to the determination result of the voice signal interval determination step and the determination result of the voice signal type determination step for each frequency band.

Preferably, the step of determining a voice signal type according to the frequency band includes a step of determining a maximum value, a minimum value, an average value, and a standard value of a signal strength for a voice signal for each frequency band of the specific section, And a deviation of a speech signal according to the calculated signal intensity statistic value with respect to the speech signal of each frequency band of the specific section based on the speech signal type per predetermined signal intensity statistical value, Can be determined.

Preferably, in the case where the voice signal of the specific section is a voice section, the amplification value determination step determines the amplification value from the plurality of dynamic range input / output curves, Output curve for each of the frequency bands according to a type of a voice signal for each of the frequency bands of each of the frequency bands; selecting a dynamic range input / output curve for each frequency band by a frequency band, Determining a maximum absolute value among the absolute values of the input signals, and determining the determined maximum absolute value as a maximum absolute value of the specific period, based on the dynamic range input / output curve selected for each frequency band, The amplitudes of the dynamic range amplification values Frequency determining yeokbyeol may include an amplification value decision step.

Therefore, the method of operating the dynamic range processing device and the dynamic range processing device for each frequency band according to the present invention is characterized in that when the dynamic range amplification processing (compression processing) is applied to the voice signals for each frequency band, So that the effect of the dynamic range amplification processing on the voice signal can be maximized.

1 illustrates a communication system for explaining inter-terminal communication in a communication environment of a general mobile communication.
2 is a control block diagram illustrating an apparatus for processing a dynamic range according to a preferred embodiment of the present invention.
FIG. 3 is a control block diagram specifically illustrating some functional units in a frequency range-specific dynamic range processing apparatus according to a preferred embodiment of the present invention.
4 is a flowchart illustrating an operation method of an apparatus for processing a dynamic range according to a preferred embodiment of the present invention.
5 is an exemplary diagram illustrating a plurality of dynamic range input / output curves used in an apparatus for processing a dynamic range according to a preferred embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

1, in a mobile communication call environment, a call path between a calling terminal 100 and a called terminal 200 is connected through a network 1 including a mobile communication network, Are mutually transmitted and received, thereby realizing a call between the parties.

Of course, in the above description, the mobile communication environment is referred to. However, in the wired communication environment and the Internet communication environment, the communication is realized through voice signal transmission / reception between the calling terminal 100 and the called terminal 200.

For example, both the calling terminal 100 and the called terminal 200 will have the voice of the caller and the surrounding background noise input together through the microphone of the terminal, and the voice signal (voice + background noise) I.e., the calling terminal 100 or the called terminal 200 through the mobile communication network through the coding process, and the other terminal outputs the received voice signal through the speaker through a series of decoding process .

Furthermore, in a mobile communication call environment, when a caller makes a call through a terminal, that is, the calling terminal 100 or the called terminal 200. The distance between the caller and the terminal may not be maintained constant, and the original voice signal, in particular, the voice size may be distorted and may not be constant. In particular, this phenomenon may become more noticeable when talking using handsfree or Bluetooth.

In view of the specificity of the mobile communication environment, the present invention proposes a dynamic range processing device for each frequency band that maximizes the effect of the dynamic range compression processing on voice signals micro-input by the terminal by the caller .

Hereinafter, an apparatus 300 for processing a dynamic range according to a preferred embodiment of the present invention will be described in detail with reference to FIG.

Before describing the present invention, the apparatus 300 may be located within the calling terminal 100 or the called terminal 200 of FIG. 1 and may refer to the terminal 100 for convenience of description. I will.

The voice signal input unit 110 of the terminal 100 receives voice from the outside and outputs the voice signal to the dynamic range processing unit 300 according to the present invention.

For example, the voice signal input unit 110 may output a voice signal according to voice input through a microphone (not shown) receiving voice from the outside to the dynamic range processor 300 according to the present invention.

Accordingly, the apparatus 300 according to the present invention receives the voice signal input through the voice signal input unit 110 and generates a voice signal for each frequency band according to the present invention, Performs dynamic range amplification processing (compression processing), and outputs it to the processing unit 120.

The processing unit 120 transmits a voice signal output from the terminal 100 to the mobile communication network 300. The voice signal output from the dynamic range processing apparatus 300 for each frequency band, Lt; / RTI >

Hereinafter, the apparatus 300 will be described in detail. The apparatus 300 includes a signal divider 310 for dividing an input speech signal into frequency bands, An amplification processing control unit 350 for determining a dynamic range amplification value for each of the frequency bands based on the signal strength of the divided voice signals for each frequency band, An amplification processing unit 330 for amplifying the voice signal for each frequency band based on the frequency band and a voice signal for each frequency band, And a signal output unit 340 for outputting the signal.

Furthermore, the apparatus 300 may further include an input block 320.

The signal divider 310 divides the input audio signal into frequency bands.

For example, the signal divider 310 may include a plurality of band pass filters for passing only the voice signal of a specific frequency band and attenuating or removing the voice signal of the other frequency band, It can be divided by each frequency band.

At this time, the signal divider 310 can divide the speech signal into the respective frequency bands by adopting various existing types of band-pass filters. For example, the FIR (Finite Impulse Response) filter or the linear phase FIR filter is adopted Or may employ a Polyphase Quadrature Filterbank (PQF).

The input block unit 320 divides the voice signal of each frequency band into a plurality of sub-bands.

For the sake of convenience of description, referring to a voice signal of one specific frequency band among the voice signals of each frequency band, the input block 320 buffers the input voice signal of a specific frequency band at predetermined time intervals, It is possible to configure the audio signal of a specific frequency band to be input in a unit of a predetermined time interval, that is, a frame unit of a predetermined time interval.

At this time, since the speech signal is inputted in units of samples, one frame will include a plurality of speech signal samples buffered for a predetermined time.

Accordingly, the input block unit 320 can configure each of the voice signals divided into the input frequency bands by the signal divider 310 as a frame unit, as described above.

Hereinafter, for convenience of description, a description will be given of a frame of a voice signal for each frequency band constituted by the input block 320 at the current time and output in units of frames, as a specific frame (e.g., kth frame). In other words, in the input block unit 320, a first frequency band audio signal of a specific frame (e.g., kth frame), a second frequency band audio signal of a specific frame (e.g., kth frame) (N) -th frequency band audio signal of a specific frame (e.g., k-th frame) to the amplification processing unit 330. [

The amplification processing control unit 350 determines the dynamic range amplification value for each frequency band based on the signal strength of the voice signal for each frequency band divided by the signal divider 310. [

That is, the amplification processing control unit 350 determines, for each frequency band, based on the signal strength of each audio signal in each frequency band, which is divided by the signal divider 310 and output from the input block unit 320, The dynamic range amplification value can be variably determined.

Accordingly, in determining the dynamic range amplification value for each frequency band, the amplification processing controller 350 does not always apply the same dynamic range amplification value to one and the same frequency band, but changes the strength of the voice signal of the corresponding frequency band Therefore, the optimum dynamic range amplification value can be adaptively determined and applied. A detailed description related to this will be described later with reference to FIG.

The amplification processing unit 330 amplifies the voice signal for each frequency band based on the dynamic range amplification value for each frequency band determined by the amplification processing control unit 350. [

For example, the amplification processing unit 330 receives the first frequency band audio signal of a specific frame (e.g., kth frame) from the input block unit 320 at the present time, the second frequency band of a specific frame (e.g., kth frame) A third frequency band speech signal of a specific frame (e.g., kth frame), and a (n) frequency band speech signal of a specific frame (e.g., kth frame).

When the amplification processing unit 330 receives the dynamic range amplification value for each frequency band determined by the amplification processing control unit 350 with respect to a specific frame (for example, the kth frame), the amplification processing unit 330 outputs the dynamic range amplification value It is possible to amplify a voice signal for each frequency band (first, second, third, ..., n) of the specific frame (for example, kth frame)

The signal output unit 340 combines the voice signals of each frequency band amplified by the frequency band processing unit 330 and outputs one voice signal processed in the dynamic range.

That is, the signal output unit 340 combines the voice signals for each frequency band amplified by the frequency band processor 330. Then, the signal output unit 340 can change the audio signal of a specific frame (for example, kth frame) combined into a single audio signal sample unit and output it.

Hereinafter, with reference to FIG. 3, the above-described amplification processing controller 350 in the apparatus 300 for processing a dynamic range according to a frequency band according to the present invention will be described in detail.

3, the amplification processing control unit 350 of the apparatus 300 according to an embodiment of the present invention includes a section determination unit 350 that determines whether a voice signal of a specific section of the voice signal is a voice section, (354) for each of the frequency bands, and for each of the frequency bands divided by the frequency bands of the speech signal of the specific interval among the audio signals of the respective frequency bands, A type determination unit 356 that determines a type of a voice signal in accordance with a signal strength statistic value of the frequency band, and a determination unit 356 that determines, based on the determination result of the type determination unit 356, And a frequency band-specific amplification value determination unit 360 for determining the amplification value.

Furthermore, the amplification processing control unit 350 may further include an input block unit 352. The input block unit 352 corresponds to the input block unit 320 described above with reference to FIG.

1, the input block unit 352 receives a voice signal input to the signal divider 310 and receives a voice signal for each frequency band divided by the signal divider 310. [

Accordingly, the input block unit 352 divides the input audio signal, that is, the audio signal before being divided into the respective frequency bands, by intervals of a predetermined time interval.

In other words, the input block unit 352 buffers the input voice signal at a predetermined time interval to form one section, that is, one frame, so that the input voice signal is divided into a plurality of sections of a predetermined time interval, It can be configured in units of frames.

In addition, the input block unit 352 divides the audio signals of each frequency band, which is divided by the signal divider 310, into units of a predetermined time interval, like the input block unit 320 described above with reference to FIG.

In other words, the speech signal of one specific frequency band among the speech signals of the respective frequency bands will be described. The input block 320 buffers the input speech signal of a specific frequency band at predetermined time intervals, It is possible to configure the audio signal of a specific frequency band to be input in a unit of a predetermined time interval, that is, a frame unit of a predetermined time interval.

As described above, the input block unit 352 forms an input speech signal before being divided into the respective frequency bands on a frame-by-frame basis. The input block unit 352 divides each of the audio signals, which are divided and inputted by the signal dividing unit 310, .

Hereinafter, for the sake of convenience of explanation, a frame of a speech signal constituted and output by the input block unit 320 at the current time point, and a frame of a speech signal of each frequency band divided by each frequency band in the speech signal, (E.g., the k-th frame).

In other words, the input block 320 outputs the audio signal of a specific frame (for example, the k-th frame) configured at the present time to the interval determination unit 354 and the frequency band amplification value determination unit 360, (E.g., kth frame), a second frequency band audio signal of a specific frame (e.g., kth frame), a third frequency band of a specific frame (e.g., kth frame) Audio signal ... The (n) frequency band audio signal of a specific frame (e.g., kth frame) is output to the type determination unit 356. [

The section determination unit 354 determines whether or not a voice signal of a specific section of the voice signal is a voice section. Here, the specific period means a specific frame (e.g., kth frame) corresponding to the current time point described above.

That is, the section determination unit 354 receives the audio signal composed of the frame units from the input block unit 320 as described above.

Accordingly, the interval determining unit 354 determines whether a voice signal of a specific frame (e.g., kth frame) received at the current point of time corresponds to a voice interval or a non-voice interval.

At this time, the section determination unit 354 determines whether the voice signal of a specific frame (e.g., kth frame) is a voice section or a non-voice section, and adopts one of various existing voice and non- can do.

For example, the section determination unit 354 determines whether or not a speech feature element designated in advance is included in the speech signal of the input specific frame (e.g., kth frame). If the speech feature element is included as a speech section It can be judged as a non-voice section.

More specifically, the section determination unit 354 calculates a characteristic value or a quantitative value for a speech signal after a noise attenuation for a speech signal of a specific frame (for example, kth frame) input in accordance with a predetermined calculation method , And it is possible to confirm whether or not the voice characteristic element designated in the voice signal is included by comparing the calculated specific value or the quantitative value with the voice determination threshold.

The type determination unit 356 determines a type of a specific frame (for example, kth frame) for a specific frame (e.g., kth frame) of the received voice signal for each frequency band, For example, the kth frame) of the audio signal according to the signal strength statistical value calculated on the basis of the signal strength of the audio signal for each frequency band.

That is, the type determination unit 356 will receive the audio signals for each frequency band, which is configured in units of frames, from the input block unit 320 as described above.

The type determination unit 356 determines the type of the audio signal of each frequency band of a specific frame (for example, kth frame) for the audio signal of each frequency band of a specific frame (e.g., kth frame) And determines the type of the voice signal according to the signal strength statistic value calculated based on the strength.

More specifically, the type determination unit 356 determines the type of a specific frame (e.g., k-th frame) based on the signal strength of a voice signal for each frequency band of a specific section, A signal intensity statistic value including at least one of a maximum value, a minimum value, an average value, and a standard deviation of a signal intensity with respect to a voice signal per frequency band is calculated, ) Can be determined for each of the frequency bands of each of the frequency bands.

Here, the types of speech signals according to the pre-designated signal intensity statistical values can be determined by a combination of a maximum value, a minimum value, an average value, and a standard deviation of a signal strength of a speech signal. May be the information that is designated through < RTI ID = Furthermore, it is preferable that the types of the voice signals according to the signal strength statistical values are designated differently for each frequency band.

Here, the voice signal type will be described with reference to the voice of the caller. For example, it is possible to describe how the size of the voice to be pronounced varies, for example, whether the size of the voice is constant, whether the voice is large or small, How many times the voice grows smaller or becomes smaller, how often it repeats itself, whether the pronunciation of the voice corresponds to a consonant, whether it is a vowel, whether it is a consonant, etc. Lt; / RTI >

To be specific, the type determination unit 356 first verifies the signal strength of a speech signal for each frequency band of a specific frame (e.g., kth frame). This can be realized by using an existing signal strength verification algorithm, so a detailed description thereof will be omitted.

The type determination unit 356 determines the maximum value, the minimum value, the average value, and the standard value of the signal intensity for each of the voice signals in each frequency band based on the signal strength of the voice signal for each frequency band of a specific frame The deviation can be calculated.

As described above, the type determination unit 356 calculates the maximum value, the minimum value, the average value, and the standard deviation calculated for each of the speech signals for each frequency band of a specific frame (e.g., kth frame) It is possible to determine the type of the voice signal according to the signal strength statistical value calculated for each of the voice signals for each frequency band of a specific frame (e.g., k-th frame) based on the voice signal type by value.

The band-by-band amplification value determination unit 360 determines the amplification range of the dynamic range for each frequency band according to the determination result of the section determination unit 354 and the determination result of the type determination unit 356

The frequency band-specific amplification value determination unit 360 includes an input / output curve selection unit 362, a maximum absolute value determination unit 364, and an amplification value determination unit 366 ).

The input / output curve selection unit 362 selects an input / output curve curve based on the input signal strength increase versus the dynamic range amplification value as a result of the determination by the interval determination unit 354 when the voice signal of the specific period (e.g., kth frame) Range input / output curves for each frequency band according to the type of voice signal for each frequency band of a specific frame (e.g., k-th frame) determined by the type determination unit 356 from a plurality of dynamic range input / Select.

That is, the input / output curve selection unit 362 selects the input / output curve selection unit 362 in the case where it is determined in the type determination unit 356 that the audio signal of a specific frame (e.g., kth frame) k-th frame) according to each type of voice signal determined for each of the frequency bands of the frequency bands of the frequency bands of the first to k-th frames.

The input / output curve selection unit 362 selects the input / output curve of the first frequency band and the dynamic range input / output curve of the second frequency band among the plurality of dynamic range input / output curves in relation to the specific frame (e.g., kth frame) , The dynamic range input / output curve of the third frequency band ... the dynamic range input / output curve of the (n) frequency band can be selected.

For example, as shown in FIG. 5, the apparatus 300 for processing a dynamic range according to a frequency band according to an embodiment of the present invention may be configured such that a signal intensity of an input signal is represented by a y- It is possible to have a plurality of dynamic range input / output curves with different rates of increase of the dynamic range amplification value compared to the signal intensity increase.

Accordingly, the input / output curve selection unit 362 selects the input / output curve for each frequency band in accordance with each type of the audio signal determined for each of the audio signals for each frequency band, with respect to a specific frame (e.g., kth frame) (Related to the magnitude of the input signal), the dynamic range input / output curve can be selected from a plurality of dynamic range input / output curves.

On the other hand, the input / output curve selection unit 362 determines whether or not a speech signal of a specific frame (e.g., kth frame) is a non-speech period in the interval determination unit 354, (1) which has the smallest amplification degree for all frequency bands among the above-mentioned plurality of dynamic range input / output curves, regardless of the type of the voice signal determined for each of the voice signals for each of the frequency bands of the It is possible.

Of course, when the interval determiner 354 determines that the speech signal of a specific frame (e.g., kth frame) is a non-speech interval, the apparatus 300 amplifies the speech signal for each frequency band by the frequency band- It would also be possible not to perform.

The maximum absolute value determination unit 364 determines the maximum absolute value of the absolute values of the signal strengths of the respective voice signal samples included in the voice signal of the specific frame (e.g., kth frame) (the voice signal not divided for each frequency band) , And determines the maximum absolute value as a maximum absolute value of a specific frame (e.g., kth frame).

Furthermore, the maximum absolute value determination unit 364 can store the maximum absolute value of the specific frame (e.g., kth frame) confirmed as described above. Therefore, the maximum absolute value determination unit 364 may store the maximum absolute values of the k-1th, k-2th, and k-3th frames preceding the specific frame (e.g., kth frame) of the current time.

The maximum absolute value determining unit 364 then determines whether or not the maximum absolute value of the specific frame (e.g., kth frame) confirmed as described above is greater than the maximum absolute value of the predetermined period (e.g., kth frame) (k-1) th frame, (k-2) th frame, and (k-3) th frame), the largest absolute maximum value among the maximum absolute values determined through the above- have.

This is to prevent a sudden increase in signal strength that may occur in a voice signal of a specific frame (e.g., kth frame).

The amplification value determination unit 366 inputs the maximum absolute value determined by the maximum absolute value determination unit 364 based on the dynamic range input / output curve selected for each frequency band with respect to a specific frame (e.g., kth frame) as described above The scale factor _k (m) is determined for each frequency band of a specific frame (e.g., kth frame) by using the signal strength of the signal. Here, k means the above-mentioned k-th frame, and m means a frequency band (m = 1,2,3 ... n).

For example, with reference to FIG. 5, a dynamic range input / output curve? For the first frequency band, a dynamic range input / output curve? For the second frequency band, a third frequency band? The amplification value determination unit 366 determines that the maximum absolute value of the input signal is equal to or larger than a predetermined value (for example, kth frame) 256, b / 256 as the amplification value of the dynamic range in the second frequency band, and b / 256 as the amplification value of the dynamic range in the third frequency band.

At this time, the amplification value determiner 366 determines the amplification value of the specific frame ( _k ) among the dynamic range amplification values (scalefactor _k (m)) for each frequency band determined in relation to the audio signal for each frequency band of a specific frame (Scalefactor _k-1 (m)) determined for each frequency band in relation to the speech signal of each frequency band of the k-1th frame, that is, the scale factor _k-1 (m) It is preferable to reduce the dynamic range amplification value of this specific frequency band by a specific value when there is the dynamic range amplification value of a specific frequency band which is higher than or equal to the predetermined specified increase rate.

This is also intended to prevent a sudden increase in signal strength that may occur in a specific frame (e.g., kth frame) audio signal.

Thus, as described above, the amplification processing section 330 amplifies the voice signal for each frequency band based on the dynamic range amplification value for each frequency band determined by the amplification processing control section 350. [

That is, when the amplification processing unit 330 receives the dynamic range amplification value (scalefactor _k (m)) for each frequency band determined by the amplification processing control unit 350 with respect to a specific frame (e.g., kth frame) (1, 2, 3, ..., n) of the specific received frame (for example, kth frame) based on the band-specific dynamic range amplification value (scalefactor _k Can be processed.

Then, as described above, the signal output unit 340 combines the voice signals for each frequency band amplified by the frequency-band-specific processing unit 330 and outputs the voice signals of a specific frame (for example, kth frame) It is possible to output the audio signal subjected to the dynamic range amplification processing by changing it in units of the audio signal sample.

As described above, according to the present invention, an apparatus 300 for processing a dynamic range according to a frequency band divides an input speech signal into frequency bands and divides the divided speech signals into frequency bands, The dynamic range amplification value for each frequency band is determined by using the dynamic range input / output curve selected for the frequency range, and the dynamic range amplification processing is performed for the voice signal for each frequency band using the determined dynamic range amplification value for each frequency band, So that the dynamic range amplification can be performed by adaptively applying the optimum dynamic range amplification value instead of always applying the fixed dynamic range amplification value to the voice signal of the same frequency band.

Therefore, in applying the dynamic range amplification processing (compression processing) to the voice signals for each frequency band, the apparatus 300 for frequency-band-specific dynamic range according to the present invention calculates the dynamic range The effect of the range amplification processing can be maximized.

The dynamic range processing apparatus 300 according to the present invention is located in a terminal (called a calling terminal 100 or a called terminal 200, for example, terminal 100 in Fig. 1) as shown in Fig. 2 (Compression process) on the voice signal of each frequency band according to the present invention as described above to receive the voice signal input through the voice signal input unit 110 and to transmit the voice signal from the terminal 100 to the mobile communication network To the processing unit 120, which processes it for transmission.

Alternatively, the apparatus 300 according to the present invention may be located in an equipment (not shown) in a mobile communication network, not in a terminal, decodes a voice signal transmitted / received between terminals in a mobile communication network, It is also possible to perform the dynamic range amplification process (compression process) on the voice signal for each frequency band according to the present invention as described above, and to perform the coding process again and output the original transmission direction.

Hereinafter, an operation method of the dynamic range processor according to the preferred embodiment of the present invention will be described in detail with reference to FIG. For convenience of description, the reference numerals of FIGS. 1 to 3 and 5 will be described.

The operation method of the dynamic range processing device for each frequency band according to the present invention divides an input speech signal into frequency bands (S20) when a speech signal is inputted (S10).

For example, an operation method of a dynamic range processor for each frequency band according to the present invention includes a plurality of band pass filters that pass only a voice signal of a specific frequency band and attenuate or remove a voice signal of a different frequency band, The signal can be divided by each frequency band through a plurality of band-pass filters.

Thereafter, the operation method of the dynamic range processing device for each frequency band according to the present invention performs the steps S30 and S35 described below in parallel.

That is, the operation method of the dynamic range processing device according to the present invention is characterized in that the speech signal inputted in step S10, that is, the speech signal before being divided by each frequency band, and the speech signal of each frequency band divided in step S20, (S30).

In other words, the method of operating the dynamic range processing device for each frequency band according to the present invention comprises the steps of buffering input speech signals at predetermined time intervals to form one section, i.e., one frame, Unit, that is, a frame unit of a predetermined time interval.

In addition, the operation method of the dynamic range processing device for each frequency band according to the present invention will be described with reference to a speech signal in one specific frequency band among the divided speech signals for each frequency band. The audio signal of a specific frequency band to be input can be configured in units of intervals of a predetermined time interval, that is, a frame unit of a predetermined time interval.

Accordingly, the method of operating the dynamic range processing apparatus according to the present invention is characterized in that an input speech signal before being divided for each frequency band is configured for each frame, and each of the divided speech signals for each frequency band is divided into frames Can be configured.

Hereinafter, for convenience of explanation, a frame of a speech signal constituting and outputting frame by frame at the present time and a frame of a speech signal of each frequency band divided by each frequency band in the speech signal are referred to as a specific frame (e.g., kth frame) .

In other words, the operation method of the dynamic range processing apparatus for each frequency band according to the present invention is characterized in that the speech signal of a specific frame (for example, kth frame) configured at the present time and the speech signal of a specific frame A second frequency band audio signal of a specific frame (e.g., kth frame), a third frequency band audio signal of a specific frame (e.g., kth frame) (N) -th frequency band speech signal of FIG.

After step S30, the operation method of the dynamic range processor according to the present invention determines whether a voice signal of a specific frame (e.g., kth frame) corresponds to a voice interval or a non-voice interval (S40) .

In addition, the method of operating the dynamic range processing apparatus according to the present invention may further comprise the step of determining, at step S40, whether or not the speech signal is a speech signal according to a signal strength statistic value calculated on the basis of a signal strength of a speech signal for each frequency band of a specific frame Determine the type.

More specifically, a method of determining a type of a voice signal of a voice signal for each frequency band will now be described. A method of operating a dynamic range processor for each frequency band according to the present invention includes: Calculates a signal strength statistic value including at least one of a maximum value and a minimum value, an average value and a standard deviation of signal strengths for the voice signals for each frequency band of a specific frame (for example, kth frame) And the type of the voice signal according to the calculated signal strength statistic value can be determined for the voice signal for each frequency band of a specific frame (for example, kth frame) based on the voice signal type according to the pre-designated signal strength statistical value.

Here, the types of speech signals according to the pre-designated signal intensity statistical values can be determined by a combination of a maximum value, a minimum value, an average value, and a standard deviation of a signal strength of a speech signal. May be the information that is designated through < RTI ID = Furthermore, it is preferable that the types of the voice signals according to the signal strength statistical values are designated differently for each frequency band.

Here, the voice signal type will be described with reference to the voice of the caller. For example, it is possible to describe how the size of the voice to be pronounced varies, for example, whether the size of the voice is constant, whether the voice is large or small, How many times the voice grows smaller or becomes smaller, how often it repeats itself, whether the pronunciation of the voice corresponds to a consonant, whether it is a vowel, whether it is a consonant, etc. Lt; / RTI >

First, the operation method of the dynamic range processing device according to the present invention checks the signal strength of a voice signal for each frequency band of a specific frame (e.g., k-th frame).

Accordingly, the method of operating the dynamic range processing apparatus for each frequency band according to the present invention is characterized in that, based on the signal strength of a voice signal for each frequency band of a specific frame (for example, kth frame) , The lowest value, the mean value, and the standard deviation.

As described above, the operating method of the dynamic range processing device for each frequency band according to the present invention calculates the maximum value, the minimum value, the average value, and the standard deviation calculated for each of the voice signals for each frequency band of a specific frame (for example, kth frame) The type of the voice signal according to the signal strength statistical value calculated for each of the voice signals according to the frequency bands of a specific frame (for example, the k-th frame) can be determined based on the voice signal types according to the pre-designated signal strength statistical values.

After step S40, the method of operating the dynamic range processor according to the present invention determines the dynamic range amplification value for each frequency band according to the determination result of the voice interval at step S40 and the voice signal type determination result for each frequency band.

More specifically, in the method of operating the dynamic range processing device for each frequency band according to the present invention, when the voice section of the specific section, that is, the specific frame (e.g., kth frame) In step S50, a dynamic range input / output curve is selected for each frequency band according to a type of a voice signal for a voice signal for each frequency band of a specific frame (e.g., kth frame) determined in step S40 from the plurality of dynamic range input / output curves.

Accordingly, an operation method of an apparatus for processing a dynamic range according to a frequency band according to the present invention is characterized in that, with respect to a specific frame (e.g., kth frame), a dynamic range input / output curve of a first frequency band, Band dynamic range input / output curve, dynamic range input / output curve of the third frequency band ... The dynamic range input / output curve of the (n) frequency band can be selected.

For example, as shown in FIG. 5, an operation method of an apparatus for processing a dynamic range according to a frequency band according to the present invention is characterized in that a signal strength of an input signal is represented by a y- It is possible to have a plurality of dynamic range input / output curves with different rates of increase of the dynamic range amplification value compared to the signal intensity increase.

Accordingly, the method of operating the dynamic range processor according to the present invention is characterized in that, with respect to a specific frame (for example, the k-th frame), the optimal method for each frequency band The dynamic range input / output curve can be selected from among a number of dynamic range input / output curves for processing with volume magnitude (related to the voice signal signal strength).

After step S50, the operation method of the dynamic range processing apparatus for each frequency band according to the present invention is characterized in that each speech signal sample signal included in a speech signal of a specific frame (for example, kth frame) The maximum absolute value among the absolute values of the intensity is confirmed, and the determined maximum absolute value is determined as the maximum absolute value of the specific frame (e.g., kth frame) (S60).

Furthermore, the operation method of the dynamic range processing apparatus according to the present invention can store the maximum absolute value of a specific frame (e.g., k-th frame) confirmed as described above. The operation method of the dynamic range processing apparatus according to the present invention is characterized in that the maximum absolute value of k-1, k-2, and k-3 frames preceding a specific frame (e.g., kth frame) Will be storing.

The operation method of the dynamic range processor according to the present invention is characterized in that the maximum absolute value of a specific frame (for example, kth frame) confirmed as described above and a maximum absolute value of a specific frame (e.g., kth frame) (K-th frame), which is the largest absolute value among the maximum absolute values obtained through the above-described process in a predetermined section (e.g., k-1, k-2, As the maximum absolute value of the maximum value.

The operation method of the dynamic range processing apparatus according to the present invention after step S60 is the same as that of the dynamic range input / output curve determined in step S60, based on the dynamic range input / output curve selected for each frequency band with respect to a specific frame The dynamic range amplification value (scalefactor _k (m)) is determined for each frequency band of a specific frame (e.g., kth frame) by using the maximum absolute value as the signal strength of the input signal (S70).

The operation method of the dynamic range processing apparatus for each frequency band according to the present invention is characterized in that a dynamic range amplification value (scalefactor _k (m)) for each frequency band determined in relation to a voice signal for each frequency band of a specific frame (for example, (Scalefactor _k-1 (m)) for each frequency band determined in relation to the speech signal for each frequency band of the immediately preceding section, that is, the k-1th frame preceding the specific frame (e.g., kth frame) It is desirable to reduce the dynamic range amplification value of this specific frequency band by a specific value when there is a dynamic range amplification value of a specific frequency band in which the degree of increase of the amplification value is equal to or higher than a predetermined specified increase degree.

For example, with reference to FIG. 5, the dynamic range input / output curve? For the first frequency band, the dynamic range input / output curve? For the second frequency band, the third frequency If the dynamic range input / output curve 3 & cir & is selected for the band and the maximum absolute value determined in relation to a specific frame (e.g., kth frame) is 256, a method of operating the dynamic range- A / 256 as the amplification value of the dynamic range of the first frequency band, b / 256 as the amplification value of the dynamic range of the second frequency band, and b / 256 as the amplification value of the dynamic range of the third frequency band.

Meanwhile, apart from step S30 described above, the method of operating the dynamic range-dependent processing apparatus according to the present invention divides speech signals of each frequency band, which is divided in step S20, into sections of a predetermined time interval at step S35.

For convenience of explanation, the speech signal of one specific frequency band among the speech signals of each frequency band is referred to as an operating method of the dynamic range processing unit for each frequency band according to the present invention, So that the input speech signal of a specific frequency band can be configured in units of intervals of a predetermined time interval, that is, a frame unit of a predetermined time interval. At this time, since the speech signal is inputted in units of samples, one frame will include a plurality of speech signal samples buffered for a predetermined time.

Accordingly, the method of operating the dynamic range processor according to the present invention can configure each of the voice signals for each frequency band, which is divided at step S20, as a frame unit as described above.

Hereinafter, for convenience of explanation, the frame of the audio signal constituting each frequency band constituting and outputting at the present time point will be described as a specific frame (e.g., kth frame) as described above.

In operation S80, the voice signal for each frequency band is amplified based on the dynamic range amplification value for each frequency band determined in step S70, as described above.

That is, the method of operating the dynamic range processor for each frequency band according to the present invention is characterized in that, based on the dynamic range amplification value (scalefactor _k (m)) for each frequency band determined in relation to a specific frame (e.g., kth frame) It is possible to amplify a voice signal for each frequency band (first, second, third, ..., n) of a specific frame (e.g., kth frame).

Thereafter, the method of operation of the dynamic range processing device for each frequency band according to the present invention combines the voice signals for each frequency band amplified and processed (S90), and the voice signal of a specific frame (for example, kth frame) The sound signal amplified in the dynamic range can be output by changing the audio signal sample unit (S100) and outputting it (S110).

The method of operating the dynamic range processor according to an embodiment of the present invention may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

According to the operation method of the dynamic range processing device and the dynamic range processing device for each frequency band according to the present invention, in applying the dynamic range amplification process (compression process) to the voice signal for each frequency band, the specificity of the mobile communication call environment is considered In order to maximize the effect of the dynamic range amplification processing on the voice signal, the present invention overcomes the limitations of the prior art, This is an invention that is industrially applicable because it can be practically done.

300: Dynamic range processing unit by frequency band
310: Signal divider 320: Input block
330: amplification processing unit 340: signal output unit
350: Amplification processing control unit

Claims (10)

An amplification processing controller for determining a dynamic range amplification value for each of the frequency bands based on a signal strength of a voice signal for each frequency band obtained by dividing an input voice signal by each frequency band;
An amplification processing unit for amplifying the voice signal for each frequency band based on the dynamic range amplification value for each frequency band determined by the amplification processing control unit; And
And a signal output unit for combining the voice signals of each frequency band amplified by the amplification processing unit and outputting one voice signal processed in a dynamic range;
Wherein the amplification-
Outputting a dynamic range input / output curve for each of the frequency bands according to a type of a voice signal for a voice signal for each frequency band divided for each frequency band from a voice signal of the specific period, ≪ / RTI >
And outputting an output signal of each of the plurality of frequency bands based on a dynamic range input / output curve selected for each of the frequency bands, And the dynamic range amplification value is determined for each frequency band. The method according to claim 1,
Wherein the amplification-
A section determining unit for determining whether a voice signal of a specific section of the voice signal is a voice section,
A signal intensity statistic value calculated on the basis of a signal intensity of a voice signal for each frequency band of the specific section for each frequency band of the voice signal divided for each frequency band from the voice signal of the specific section among the voice signals for each frequency band, A type determination unit for determining a type of a voice signal according to the type of the voice signal,
And a frequency band amplification value determining unit for determining a dynamic range amplification value for each of the frequency bands according to the determination result of the interval determination unit and the determination result of the type determination unit. 3. The method of claim 2,
Wherein,
Calculating a signal strength statistic value including at least one of a maximum value, a minimum value, an average value and a standard deviation of the signal intensity for each voice signal in each frequency band of the specific section on the basis of the signal strength of the voice signal for each frequency band of the specific section And determines a speech signal type according to the calculated signal strength statistic value for a speech signal for each frequency band of the specific section based on a speech signal type according to a predetermined signal intensity statistical value. 3. The method of claim 2,
The frequency band-specific amplification value determination unit determines,
And outputting an output signal of the specific period determined by the type determination unit from a plurality of dynamic range input / output curves whose increase rate of the dynamic range amplification value is different from the signal strength increase of the input signal as a result of the determination of the interval determination unit An input / output curve selector for selecting a dynamic range input / output curve for each of the frequency bands according to a type of a voice signal for a voice signal for each frequency band,
A maximum absolute value determining unit for determining a maximum absolute value of the absolute values of the signal strengths of the respective voice signal samples included in the voice signal of the specific period and determining the determined maximum absolute value as the maximum absolute value of the specific period,
And an amplification value determination unit for determining an amplification value of the dynamic range for each of the frequency bands by using the determined maximum absolute value as the signal strength of the input signal based on the dynamic range input / output curve selected for each of the frequency bands. Range processing device. 5. The method of claim 4,
Wherein the maximum absolute value determining unit comprises:
Determining a maximum absolute value of the specific period and a maximum absolute value of a maximum absolute value of the specific period and a maximum absolute value of the voice signal in a predetermined period preceding the specific period, as a maximum absolute value of the specific period, Processing device. 5. The method of claim 4,
Wherein the amplification value determination unit determines,
And comparing the dynamic range amplification values for the respective frequency bands with respect to the audio signals for each frequency band in the immediately preceding section preceding the specific section among the dynamic range amplification values for the respective frequency bands determined in relation to the audio signal for each frequency band in the specific section Wherein the dynamic range amplification value of the specific frequency band is decreased by a specific value when there is a dynamic range amplification value of a specific frequency band in which the degree of increase of the amplification value is greater than or equal to a predetermined specific increase degree. An amplification process control step of determining a dynamic range amplification value for each frequency band based on a signal strength of a voice signal for each frequency band obtained by dividing an input voice signal by each frequency band;
Amplifying the voice signal for each frequency band based on the determined dynamic range amplification value for each frequency band; And
And a signal output step of outputting one voice signal processed in a dynamic range by combining the amplified voice signals for each frequency band;
Wherein the amplification processing control step comprises:
Outputting a dynamic range input / output curve for each of the frequency bands according to a type of a voice signal for a voice signal for each frequency band divided for each frequency band from a voice signal of the specific period, ≪ / RTI >
And outputting an output signal of each of the plurality of frequency bands based on a dynamic range input / output curve selected for each of the frequency bands, Wherein the dynamic range amplification value is determined for each frequency band. 8. The method of claim 7,
Wherein the amplification processing control step comprises:
A voice signal interval determination step of determining whether a voice signal of a specific section of the voice signal is a voice section,
A signal intensity statistic value calculated on the basis of a signal intensity of a voice signal for each frequency band of the specific section for each frequency band of the voice signal divided for each frequency band from the voice signal of the specific section among the voice signals for each frequency band, Determining a type of a voice signal according to a frequency band,
Determining an amplitude range amplification value for each of the frequency bands according to the determination result of the voice signal interval determination step and the determination result of the voice signal type determination step for each frequency band. / RTI > 9. The method of claim 8,
The step of determining a voice signal type according to the frequency band includes:
Calculating a signal strength statistic value including at least one of a maximum value, a minimum value, an average value and a standard deviation of the signal intensity for each voice signal in each frequency band of the specific section on the basis of the signal strength of the voice signal for each frequency band of the specific section And determines a speech signal type according to the calculated signal strength statistic value for a speech signal of each frequency band of the specific section based on a speech signal type according to a pre-designated signal intensity statistical value. How it works. 9. The method of claim 8,
Wherein the amplification value determination step comprises:
Outputting a speech signal for each of the frequency bands of the predetermined period from the plurality of dynamic range input / output curves whose increase rate of the dynamic range amplification value is different from the signal strength increase of the input signal, A dynamic range input / output curve selection step for selecting a dynamic range input / output curve for each frequency band according to a frequency band,
A maximum absolute value determining step of determining a maximum absolute value of the absolute values of the signal intensities of the respective voice signal samples included in the voice signal of the specific period and determining the determined maximum absolute value as the maximum absolute value of the specific period,
And determining a dynamic range amplification value for each of the frequency bands by using the determined maximum absolute value as the signal strength of the input signal based on the dynamic range input / output curve selected for each of the frequency bands. A method of operating a dynamic range processor for each frequency band.

Images