TWI765772B - Method for equalizing input signal to generate equalizer output signal and parametric equalizer - Google Patents

Abstract

A parametric equalizer includes a first parametric equalizer circuit, a second parametric equalizer circuit, a first multiplication circuit, a second multiplication circuit, an addition circuit, and a weighting control circuit. The first parametric equalizer circuit processes an input signal to output a first output signal. The second parametric equalizer circuit processes the input signal to output a second output signal. The first multiplication circuit multiplies the first output signal and a first weighting value to generate a first adjusted output signal. The second multiplication circuit multiplies the second output signal and a second weighting value to generate a second adjusted output signal. The addition circuit combines the first adjusted output signal and the second adjusted output signal to generate an equalizer output signal. The weighting control circuit dynamically adjusts the first weighting value and the second weighting value according to the equalizer output signal.

Description

Method for equalizing input signal to generate equalizer output signal and parametric equalizer

The present invention relates to the design of an equalizer, and more particularly, to a method for dynamically adjusting a weighting value to equalize an input signal to generate an output signal of an equalizer, and a related parameter equalizer.

A parametric equalizer is an equalizer that can configure parameters such as center frequency and quality factor (Q) to modulate the output frequency response of an input signal. By using the parametric equalizer, when the level of the signal input to the parametric equalizer is too large, clipping of the signal can be avoided. However, when the signal input to the parametric equalizer is If the level is too small, a good hearing effect cannot be achieved by using the parametric equalizer. In addition, when the level of the signal input to the parametric equalizer is too large, gain compensation may be lost. Therefore, in order to To solve the above problems, a method for compensating the output frequency response according to the level of the input signal and a related parameter equalizer are extremely needed.

Therefore, one of the objectives of the present invention is to provide a method for dynamically adjusting a weight value to equalize an input signal to generate an output signal of an equalizer and a related parameter equalizer to solve the above problems.

According to an embodiment of the present invention, a parameter equalizer is disclosed. The parametric equalizer of the present invention dynamically adjusts the weight value according to the equalizer output signal, instead of adjusting the gain value of the parametric equalizer. The parameter equalizer includes a first parameter equalizer circuit, a second parameter equalizer circuit, a first multiplication circuit, a second multiplication circuit, an addition circuit and a weight control circuit. The first parameter equalizer circuit has a first transfer function with a positive gain, and is used for processing an input signal according to the first transfer function and outputting a first output signal by processing the input signal. The second parameter The equalizer circuit has a second transfer function with negative gain, and is used for processing the input signal according to the second transfer function and outputting a second output signal by processing the input signal, the first multiplying circuit is coupled to in the first parameter equalizer circuit and used for multiplying the first output signal and a first weight value to generate a first adjusted output signal, the second multiplication circuit is coupled to the second parameter equalizer circuit, and used for multiplying the second output signal and a second weight value to generate a second adjustment output signal, the adding circuit is coupled to the first multiplication circuit and the second multiplication circuit, and is used for the first adjustment signal and the second adjustment signal is combined to generate an equalizer output signal, the weight control circuit is coupled to the first multiplying circuit, the second multiplying circuit and the adding circuit, and is used for dynamically adjusting the equalizer output signal according to the equalizer output signal the first weight value and the second weight value.

According to an embodiment of the present invention, a method for dynamically adjusting a weight value to equalize an input signal to generate an equalizer output signal is disclosed. The method includes: processing the input signal by a first parametric equalizer circuit having a first transfer function with a positive gain to generate and output a first output signal; by having a first transfer function with a negative gain A second parameter equalizer circuit of the second transfer function processes the input signal to generate and outputs a second output signal; multiply the first output signal and a first weight value to generate a first adjustment output signal; multiply the second output signal and a second weight value to generate a second adjusted output signal; combine the first adjusted output signal and the second adjusted output signal to generate the equalizer output signal; and dynamically adjusting the first weight value and the second weight value according to the output signal of the equalizer.

FIG. 1 is a schematic diagram of a parameter equalizer 10 according to an embodiment of the present invention. The parameter equalizer 10 dynamically adjusts the weight value according to the equalizer output signal instead of adjusting the gain value. For example, the weight value includes There is a first weight value and a second weight value, and the sum of the first weight value and the second weight value is equal to one. For the sake of brevity, the first weight value and a second weight value are marked as " β " and "1-beta".

The parameter equalizer 10 receives an input signal A_IN such as an audio signal, and equalizes the input signal A_IN to generate an equalizer output signal EQ_OUT. As shown in FIG. 1, the parameter equalizer 10 includes a parameter equalizer circuit 12, a parameter equalizer circuit 14, a multiplication circuit 16, a multiplication circuit 18, an addition circuit 20, and a weight control circuit 22, and the parameter equalizer circuit 12 has a transfer function H ₁ with a positive gain G ₁ (G ₁ >0), and is used to process the input signal A_IN according to the transfer function H ₁ and output a first output signal A_OUT_1 by processing the input signal A_IN, parameter The equalizer circuit 14 has a transfer function H ₂ with a negative gain G ₂ (G ₂ <0), and is used to process the input signal A_IN according to the transfer function H ₂ and output a second output by processing the input signal A_IN In the signal A_OUT_2, the center frequency of the parameter equalizer circuit 12 is the same as the center frequency of the parameter equalizer circuit 14 .

The multiplying circuit 16 is coupled to the parameter equalizer circuit 12 and used to multiply the first output signal A_OUT_1 by the first weight value β to generate a first adjusted output signal AD_OUT_1 (ie AD_OUT_1 = A_OUT_1 * β). The multiplying circuit 18 is coupled to the parameter equalizer circuit 14 and used to multiply the second output signal A_OUT_2 by the second weight value 1-β to generate a second adjusted output signal AD_OUT_2 (ie AD_OUT_2 = A_OUT_2 * ( 1-β)). The adding circuit 20 is coupled to the multiplying circuit 16 and the multiplying circuit 18 , and is used for combining the first adjustment output signal AD_OUT_1 and the second adjustment output signal AD_OUT_2 to generate the equalizer output signal EQ_OUT (ie EQ_OUT = AD_OUT_1 + AD_OUT_2 ) . The weight control circuit 22 is coupled to the multiplying circuit 16, the multiplying circuit 18 and the adding circuit 20, and is used for dynamically adjusting the first weight value β and the second weight value 1-β according to the equalizer output signal EQ_OUT.

In addition, the parametric equalizer circuit 12 may be a peaking filter implemented by a biquad filter, and the parametric equalizer circuit 14 may be implemented by a biquad filter is a notch filter, but the present invention is not limited to this. For example, when the parametric equalizer circuit 12 or the parametric equalizer circuit 14 is implemented by a second-order filter, the transfer function of the parametric equalizer circuit 12 or the parametric equalizer circuit 14 can be represented by the following equation:

The coefficients b ₀ , b ₁ , b ₂ , a ₁ , and a ₂ can be determined by the following matlab code: K = tan( (pi * fc) / fs);

V0 = 10^(G / 20);

if ( V0 < 1 ) % invert gain if a cut

V0 = 1 / V0;

end

if (G > 0)

b0 = ( 1 + ( V0 / Q ) * K ) + K^2 ) / ( 1 + ( ( 1 / Q ) * K ) + K^2 );

b1 = ( 2 * ( K^2 - 1 ) ) / ( 1 + ( (1 / Q) * K ) + K^2 );

b2 = ( 1 - ( V0 / Q ) * K ) + K^2 ) / ( 1 + ( ( 1 / Q ) * K ) + K^2 );

a1 = b1;

a2 = ( 1 - ( 1 / Q ) * K ) + K^2 ) / (1 + ( ( 1 / Q ) * K ) + K^2 );

else

b0 = ( 1 + ( 1 / Q ) * K ) + K^2 ) / ( 1 + ( ( V0 / Q ) * K ) + K^2 );

b1 = ( 2 * ( K^2 - 1 ) ) / ( 1 + ( ( V0 / Q ) * K ) + K^2 );

b2 = ( 1 - ( 1 / Q ) * K ) + K^2 ) / ( 1 + ( ( V0 / Q ) * K ) + K^2 );

a1 = b1;

a2 = ( 1 - ( V0 / Q ) * K ) + K^2 ) / ( 1 + ( ( V0 / Q ) * K ) + K^2 );

end

In the above matlab code, fc is the center frequency, fs is the sampling rate, G is the logarithmic gain in decibels (dB), and Q is the quality factor.

=

For example, the parametric equalizer circuit 12 is a spike filter with +12 dB gain (ie, G _{1 =} +12 dB) and quality factor QP implemented by a second-order filter, and the parametric equalization The filter circuit 14 is a notch filter with a gain of -6 dB (ie, G ₂ = -6 dB) and a quality factor Q _N implemented by a second-order filter, but the invention is not limited thereto. Using the above equation and matlab code, the transfer function H ₁ and the transfer function H ₂ can be determined. Furthermore, the combination of the weight values of the transfer function H ₁ and the transfer function H ₂ can be used to achieve different gain values, which are expressed as follows:

=

，藉此取得以下分數：

=

In order to reduce the above two fractions to a common denominator, it is assumed that

, to obtain the following scores:

=

係為一新的線性增益，當β係等於1時，參數等化器10可視為一尖峰濾波器，其由參數等化器電路12所實現，而當β係等於0時，參數等化器10可視為一陷波濾波器，其由參數等化器電路14所實現。此外，權重值組合所需要的品質因數Q _P以及品質因數Q _N之間的關係（例如

）可藉由將該兩個分數化簡為一個公分母的計算過程中來決定，因此，參數等化器10可藉由轉移函數H ₁以及轉移函數H _２的權重值組合來達到+12分貝以及-6分貝之間的不同增益值，其中品質因數Q _N與品質因數Q _P之比值係為一預定常數（例如

）。由於參數等化器10僅應用參數等化器電路12、參數等化器電路14以及權重值（亦即第一權重值β以及第二權重值1-β）組合，因此本發明的硬體成本不高而且本發明相當容易配置，此外，參數等化器電路12以及參數等化器電路14可由可程式化(programmable)的二階濾波器來實現，因此，當參數等化器電路12以及參數等化器電路14沒在使用時，參數等化器電路12以及參數等化器電路14可以被重複使用(reuse)於其它用途。 in

is a new linear gain, when β is equal to 1, the parameter equalizer 10 can be regarded as a spike filter, which is realized by the parameter equalizer circuit 12, and when β is equal to 0, the parameter equalizer 10 can be regarded as a notch filter, which is implemented by the parametric equalizer circuit 14 . In addition, the relationship between the quality factor Q _P and the quality factor Q _N required for the combination of weight values (e.g.

) can be determined by reducing the two fractions to a common denominator in the calculation process, therefore, the parameter equalizer 10 can achieve +12 dB by combining the weight values of the transfer function H ₁ and the transfer function H ₂ and different gain values between -6 dB, where the ratio of the quality factor Q _N to the quality factor Q _P is a predetermined constant (e.g.

). Since the parameter equalizer 10 only applies the parameter equalizer circuit 12, the parameter equalizer circuit 14, and the combination of weight values (ie, the first weight value β and the second weight value 1-β), the hardware cost of the present invention is not high and the present invention is relatively easy to configure. In addition, the parameter equalizer circuit 12 and the parameter equalizer circuit 14 can be implemented by programmable second-order filters. Therefore, when the parameter equalizer circuit 12 and the parameter equalizer circuit 14, etc. When the equalizer circuit 14 is not in use, the parameter equalizer circuit 12 and the parameter equalizer circuit 14 can be reused for other purposes.

、

、

、…等等），但是本發明不限於此。 FIG. 2 is a schematic diagram of a weight control circuit 200 according to an embodiment of the present invention. The weight control circuit 22 in the parameter equalizer 10 shown in FIG. 1 can be implemented by the weight control circuit 200 shown in FIG. 2 . As shown in FIG. 2 , the weight control circuit 200 includes an energy detection circuit 24 and a weight value generator 26, the energy detection circuit 24 is used to measure an energy level EL of the equalizer output signal EQ_OUT, the weight value generator 26 is coupled to the energy detection circuit 24, and is used to measure an energy level EL of the equalizer output signal EQ_OUT. The energy level EL of the output signal EQ_OUT dynamically adjusts the first weight value β and the second weight value 1-β. In this embodiment, the energy detection circuit 24 includes an absolute value circuit 28 and an alpha filter 30. The absolute value circuit 28 is used to generate the absolute value calculation of the equalizer output signal EQ_OUT. An absolute value AV of the equalizer output signal EQ_OUT, the alpha filter 30 is coupled to the absolute value circuit 28 and the weight value generator 26, and is used to generate the energy level according to the absolute value calculation result (ie the absolute value AV) quasi EL, and transmit the energy level EL to the weight value generator 26, for example, the alpha filter 30 may be a second-order alpha filter, and the coefficient of the second-order alpha filter is a power of 2 (ie

,

,

, ... etc.), but the present invention is not limited thereto.

For directly controlling the weight value (eg, one of the first weight value β and the second weight value 1-β; for better understanding, the first weight value β is controlled) to be controlled according to the slope adjustment The energy level EL of the equalizer output signal EQ_OUT dynamically adjusts the first weight value β and the second weight value 1-β. An attack control and a release are used in the weight value generator 26. control, wherein when the energy level EL is less than a threshold value (threshold value) TH, the release control is used to release the equalizer output signal EQ_OUT, and when the energy level EL is not less than the threshold value TH, the attack control is used to suppress ( suppress) equalizer output signal EQ_OUT. Therefore, the weight generator 26 is further configured to receive a reference threshold RTH, an attack rate setting AR, and a release rate setting RR, wherein the threshold TH used by the weight generator 26 is obtained by combining the reference threshold RTH and the release rate setting RR. A predetermined coefficient is multiplied to obtain (eg TH = RTH * 0.635, where 0.635 is the predetermined coefficient), and the attack rate setting AR may be greater than the release rate setting RR.

After receiving the energy level EL from the alpha filter 30, the weight generator 26 compares the energy level EL with the threshold value TH, and if the energy level EL is less than the threshold value TH (ie within the release control range), Then the weight value generator 26 controls the first weight value β according to the slope adjustment specified by the release rate setting RR; and if the energy level EL is not less than the threshold value TH (that is, within the attack control range), the weight value generator 26 26 The first weight value β is controlled according to the slope adjustment specified by the attack rate setting AR. The parameter equalizer 10 directly controls the first weight value β to dynamically adjust the first weight value β and the second weight value 1-β through slope adjustment instead of linear to logarithm transfer To dynamically adjust the gain value, it should be noted that the process of dynamically adjusting the first weight value β and the second weight value 1-β does not require linear to logarithmic conversion, so that the parameter equalizer 10 has a lower hardware cost.

、S ₂=

、S ₃=

、S ₄=

）用以設定釋放率設置RR以供斜率調整之用，但是本發明不限於此。如果第一權重值β大於或等於C ₁（例如0.666），則藉由將分數S ₁（例如

）與第一權重值β相加（亦即β = β + S ₁）來調整／更新第一權重值β；如果第一權重值β大於或等於C ₂（例如0.43）並且小於C ₁（例如0.666），則藉由將分數S ₂（例如

）與第一權重值β相加（亦即β = β + S ₂）來調整／更新第一權重值β；如果第一權重值β大於或等於C _３（例如0.26）並且小於C ₂（例如0.43），則藉由將分數S ₃（例如

）與第一權重值β相加（亦即β = β + S ₃）來調整／更新第一權重值β；以及如果第一權重值β小於C _３（例如0.26），則藉由將分數S ₄（例如

）與第一權重值β相加（亦即β = β + S ₄）來調整／更新第一權重值β。 In the release control (ie the energy level EL is less than the threshold value TH), 3 coefficients C ₁ , C ₂ , C ₃ from the largest to the smallest (eg C ₁ = 0.666, C ₂ = 0.43, C ₃ = 0.26) used to classify the first weight value β, and 4 scores S ₁ , S ₂ , S ₃ , S ₄ from largest to smallest (eg S ₁ =

, S ₂ =

, S ₃ =

, S ₄ =

) is used to set the release rate setting RR for slope adjustment, but the present invention is not limited thereto. If the first weight value β is greater than or equal to C ₁ (eg 0.666), then by dividing the fraction S ₁ (eg 0.666)

) and the first weight value β (ie β = β + S ₁ ) to adjust/update the first weight value β; if the first weight value β is greater than or equal to C ₂ (eg 0.43) and less than C ₁ (eg 0.666), then by dividing the fraction S ₂ (eg

) and the first weight value β (ie, β = β + S ₂ ) to adjust/update the first weight value β; if the first weight value β is greater than or equal to C ₃ (eg 0.26) and less than C ₂ (eg 0.43), by dividing the fraction S ₃ (eg

) and the first weight value β (ie, β = β + S ₃ ) to adjust/update the first weight value β; and if the first weight value β is less than C ₃ (eg, 0.26), by adding the fraction S ₄ (e.g.

) and the first weight value β (ie, β = β + S ₄ ) to adjust/update the first weight value β.

、B ₂=

、B ₃=

、B ₄=

）用以設定攻擊率設置AR以供斜率調整之用，但是本發明不限於此。如果第一權重值β大於或等於C ₁（例如0.666），則藉由從第一權重值β減去分數B ₁（例如

），亦即β = β - B ₁，來調整／更新第一權重值β；如果第一權重值β大於或等於C ₂（例如0.43）並且小於C ₁（例如0.666），則藉由從第一權重值β減去分數B ₂（例如

），亦即β = β - B ₂，來調整／更新第一權重值β；如果第一權重值β大於或等於C _３（例如0.26）並且小於C ₂（例如0.43），則藉由從第一權重值β減去分數B ₃（例如

），亦即β = β – B ₃，來調整／更新第一權重值β；以及如果第一權重值β小於C _３（例如0.26），則藉由從第一權重值β減去分數B ₄（例如

），亦即β = β – B ₄，來調整／更新第一權重值β。 In attack control (that is, the energy level EL is not less than the threshold value TH), three coefficients C ₁ , C ₂ , and C ₃ from the largest to the smallest (for example, C ₁ = 0.666, C ₂ = 0.43, C ₃ = 0.26 ) to classify the first weight value β, and 4 scores B ₁ , B ₂ , B ₃ , B ₄ from largest to smallest (eg B ₁ =

, B ₂ =

, B ₃ =

, B ₄ =

) is used to set the attack rate setting AR for slope adjustment, but the present invention is not limited to this. If the first weight value β is greater than or equal to C ₁ (eg 0.666), then by subtracting the fraction B ₁ (eg 0.666) from the first weight value β

), that is, β = β - B ₁ , to adjust/update the first weight value β; if the first weight value β is greater than or equal to C ₂ (eg 0.43) and less than C ₁ (eg 0.666), then by A weight value β minus the fraction B ₂ (eg

), that is, β = β - B ₂ , to adjust/update the first weight value β; if the first weight value β is greater than or equal to C ₃ (eg 0.26) and less than C ₂ (eg 0.43), then by A weight value β minus the fraction B ₃ (eg

), ie β = β - B ₃ , to adjust/update the first weight value β; and if the first weight value β is less than C ₃ (eg, 0.26), by subtracting the fraction B ₄ from the first weight value β (E.g

), that is, β = β – B ₄ , to adjust/update the first weight value β.

、

、

、

而不是

、

、

、

；以及攻擊率設置AR包含有B ₁、B ₂、B ₃、B ₄，可以被設置為

、

、

、

而不是

、

、

、

。應注意的是，在本實施例中，攻擊率設置AR通常大於釋放率設置RR。 In attack control or release control, the numerators of the fractions are run from largest to smallest rather than all the same (eg, the numerators of the fractions are all 1) in order to have a relatively linear rate control over the entire control range of the first weight value β, and in addition , the control speed of the first weight value β can be made uniform. For example, the release rate setting RR includes S ₁ , S ₂ , S ₃ , S ₄ , and can be set as

,

,

,

instead of

,

,

,

; and the attack rate setting AR contains B ₁ , B ₂ , B ₃ , B ₄ , and can be set as

,

,

,

instead of

,

,

,

. It should be noted that, in this embodiment, the attack rate setting AR is generally greater than the release rate setting RR.

、S ₂=

、S ₃=

、S ₄=

）的分母，尤指分母的次方項（例如2的次方項諸如17），用以控制釋放率，分母的次方項越大，則釋放率愈大，舉例來說，當分母的次方項係為18時的控制率大於當分母的次方項係為16時的控制率。相似地，在攻擊控制中，4個分數B ₁、B ₂、B ₃、B ₄（例如B ₁=

、B ₂=

、B ₃=

、B ₄=

）的分母的次方項也用以控制攻擊率。 In release control, 4 fractions S ₁ , S ₂ , S ₃ , S ₄ (eg S ₁ =

, S ₂ =

, S ₃ =

, S ₄ =

), especially the power term of the denominator (for example, the power term of 2 such as 17) is used to control the release rate. The larger the power term of the denominator, the greater the release rate. The control rate when the square term is 18 is greater than that when the power term of the denominator is 16. Similarly, in attack control, the 4 fractions B ₁ , B ₂ , B ₃ , B ₄ (eg B ₁ =

, B ₂ =

, B ₃ =

, B ₄ =

) to the power of the denominator is also used to control the attack rate.

It should be noted that the weight control circuit 200 is only used as an example to illustrate the weight control circuit 22 in the parameter equalizer 10. In fact, any device that can dynamically adjust the first weight value β according to the equalizer output signal EQ_OUT And the circuit structure of the second weight value 1-β can be adopted by the weight control circuit 22, and these alternative designs all fall within the scope of the present invention.

赫茲(hertz, Hz)與參數等化器電路14的中心頻率

赫茲相同。藉由上述轉移函數H ₁以及轉移函數H _２的權重值組合，參數等化器10可以達到任何介於+12分貝至-6分貝的增益值，其中品質因數Q _N與品質因數Q _P之比值係為一預定常數（例如

）。 FIG. 3 is a schematic diagram illustrating that different gain values of the parameter equalizer 10 shown in FIG. 1 are set under different weight value combinations according to an embodiment of the present invention. The parametric equalizer circuit 12 and the parametric equalizer circuit 14 are respectively configured as a spike filter with a transfer function H1 with _a positive gain of +12 dB and a quality factor _QP and a transfer function with a negative gain of -6 dB H ₂ and a notch filter of quality factor Q _N , and the center frequency of the parametric equalizer circuit 12

Hertz (Hz) and the center frequency of the parametric equalizer circuit 14

Hertz is the same. Through the combination of the above-mentioned weight values of the transfer function H ₁ and the transfer function H ₂ , the parameter equalizer 10 can achieve any gain value between +12 dB and -6 dB, wherein the ratio of the quality factor _{Q N} to the quality factor QP is a predetermined constant (such as

).

As shown in Fig. 3, the weight value (for example, one of the first weight value β and the second weight value 1-β is directly controlled by the slope adjustment; for better understanding, the first weight value is controlled The weight value β) is used to obtain 64 frequency response curves. The number of frequency response curves can be determined according to the accuracy. Therefore, the 64 frequency response curves are only used for illustration purposes, and the present invention is not limited thereto. When β is equal to 1, the parameter equalizer 10 can be regarded as a spiking filter with a transfer function H ₁ with a positive gain of +12 dB (ie, the uppermost frequency response curve in Fig. 3); and when β is equal to 0 , the parameter equalizer 10 can be regarded as a notch filter with a transfer function _H2 with a negative gain of -6 dB (that is, the lowermost frequency response curve in Figure 3). In addition, the response curves in Figure 3 are all To smooth the curve, it results in less signal variation and less power variation when parametric equalizer 10 is used.

赫茲與參數等化器電路14的中心頻率

赫茲相同。藉由上述轉移函數H ₁以及轉移函數H _２的權重值組合，參數等化器10可以達到任何介於+12分貝至-6分貝的增益值，其中品質因數Q _N與品質因數Q _P之比值係為一預定常數（例如

）。 FIG. 4 is a schematic diagram illustrating the relationship between the gain value achieved by the parameter equalizer 10 shown in FIG. 1 and the weight value used by the parameter equalizer 10 shown in FIG. 1 according to an embodiment of the present invention. Assume that the parametric equalizer circuit 12 and the parametric equalizer circuit 14 are respectively configured as a spike filter with a transfer function H1 with _a positive gain of +12 dB and a quality factor _QP and a transfer with a negative gain of -6 dB A notch filter of function _H2 and quality factor _QN , and the center frequency of the parameter equalizer circuit 12

center frequency of hertz and parametric equalizer circuit 14

Hertz is the same. Through the combination of the above-mentioned weight values of the transfer function H ₁ and the transfer function H ₂ , the parameter equalizer 10 can achieve any gain value between +12 dB and -6 dB, wherein the ratio of the quality factor _{Q N} to the quality factor QP is a predetermined constant (such as

).

、

、

、

（或

、

、

、

）用以在第4圖中為斜率調整來設置釋放率設置RR（或攻擊率設置AR），此外，3個係數0.666、0.43以及0.26用以分類權重值（例如第一權重值β以及第二權重值1-β之中的其中一個；為了更好地理解，在此為第一權重值β）。在本實施例中，在4個區間中（亦即0≦β＜0.26、0.26≦β＜0.43、0.43≦β＜0.666以及0.666≦β＜1）大約取得4個斜率值，再者，該些分數的分子由這4個斜率值來決定，如第4圖所示，藉由將釋放率設置RR以及攻擊率設置AR的分數之分子設置為4到1而不是全部相同（例如分數之分子均為1），代表第1圖所示之參數等化器10所達到的自+12分貝至-6分貝的增益值以及權重值（例如第一權重值β）之間的關係之曲線可以為一平滑曲線，因此，權重值（例如第一權重值β）的整體控制範圍有著相對線性速度控制，此外，可以使權重值（例如第一權重值β）的控制速度均勻。 In the above release control (or attack control), 4 points

,

,

,

(or

,

,

,

) is used to set the release rate setting RR (or attack rate setting AR) for slope adjustment in Figure 4, in addition, 3 coefficients 0.666, 0.43 and 0.26 are used to classify weight values (eg the first weight value β and the second weight value β and the second One of the weight values 1-β; for better understanding, here is the first weight value β). In this embodiment, approximately four slope values are obtained in four intervals (ie, 0≦β<0.26, 0.26≦β<0.43, 0.43≦β<0.666, and 0.666≦β<1). The numerator of the fraction is determined by these 4 slope values, as shown in Figure 4, by setting the numerator of the fraction for the release rate setting RR and the attack rate setting AR to be 4 to 1 instead of all being the same (e.g. is 1), the curve representing the relationship between the gain value from +12 dB to -6 dB and the weight value (for example, the first weight value β) achieved by the parameter equalizer 10 shown in FIG. 1 can be a Smoothing the curve, therefore, the overall control range of the weight value (eg, the first weight value β) has a relatively linear speed control, and in addition, the control speed of the weight value (eg, the first weight value β) can be made uniform.

赫茲與參數等化器電路14的中心頻率

赫茲相同。藉由上述轉移函數H ₁以及轉移函數H _２的權重值組合，參數等化器10可以達到任何介於+12分貝至-12分貝的增益值，其中品質因數Q _N與品質因數Q _P之比值係為一預定常數（例如

）。為了簡潔起見，在此不再重複詳細描述本實施例。 FIG. 5 is a schematic diagram illustrating that different gain values of the parameter equalizer 10 shown in FIG. 1 are set under different weight value combinations according to another embodiment of the present invention. Assume that the parametric equalizer circuit 12 and the parametric equalizer circuit 14 are respectively configured as a spike filter with a transfer function H1 with _a positive gain of +12 dB and a quality factor _QP and a transfer with a negative gain of -12 dB A notch filter with function _H2 and quality factor _QN , and the center frequency of the parameter equalizer circuit 12

center frequency of hertz and parametric equalizer circuit 14

Hertz is the same. Through the combination of the above-mentioned weight values of the transfer function H ₁ and the transfer function H ₂ , the parameter equalizer 10 can achieve any gain value between +12 dB and -12 dB, wherein the ratio of the quality factor _{Q N} to the quality factor QP is a predetermined constant (such as

). For the sake of brevity, the detailed description of this embodiment is not repeated here.

As shown in Fig. 5, the weight value (for example, one of the first weight value β and the second weight value 1-β is directly controlled by the slope adjustment; for better understanding, the first weight value is controlled The weight value β) is used to obtain 64 frequency response curves. The number of frequency response curves can be determined according to the accuracy. Therefore, the 64 frequency response curves are only used for illustration purposes, and the present invention is not limited thereto. When β is equal to 1, the parameter equalizer 10 can be regarded as _a spiking filter with a transfer function H1 with a positive gain of +12 dB (ie, the uppermost frequency response curve in Fig. 5); and when β is equal to 0 , the parameter equalizer 10 can be regarded as a notch filter with a transfer function _H2 with a negative gain of -12 dB (ie, the lowermost frequency response curve in Figure 3). In addition, the response curves in Figure 5 are all To smooth the curve, it results in less signal variation and less power variation when parametric equalizer 10 is used.

赫茲與參數等化器電路14的中心頻率

赫茲相同。藉由上述轉移函數H ₁以及轉移函數H _２的權重值組合，參數等化器10可以達到任何介於+12分貝至-12分貝的增益值，其中品質因數Q _N與品質因數Q _P之比值係為一預定常數（例如

）。 FIG. 6 is a schematic diagram showing the relationship between the gain value achieved by the parameter equalizer 10 shown in FIG. 1 and the weight value used by the parameter equalizer 10 shown in FIG. 1 according to another embodiment of the present invention. . Assume that the parametric equalizer circuit 12 and the parametric equalizer circuit 14 are respectively configured as a spike filter with a transfer function H1 with _a positive gain of +12 dB and a quality factor _QP and a transfer with a negative gain of -12 dB A notch filter of function _H2 and quality factor _QN , and the center frequency of the parameter equalizer circuit 12

center frequency of hertz and parametric equalizer circuit 14

Hertz is the same. Through the combination of the above-mentioned weight values of the transfer function H ₁ and the transfer function H ₂ , the parameter equalizer 10 can achieve any gain value between +12 dB and -12 dB, wherein the ratio of the quality factor _{Q N} to the quality factor QP is a predetermined constant (such as

).

、

、

、

（或

、

、

、

）用以在第6圖中為斜率調整來設置釋放率設置RR（或攻擊率設置AR），此外，3個係數0.666、0.43以及0.26用以分類權重值（例如第一權重值β以及第二權重值1-β之中的其中一個；為了更好地理解，在此為第一權重值β）。在本實施例中，在4個區間中（亦即0≦β＜0.26、0.26≦β＜0.43、0.43≦β＜0.666以及0.666≦β＜1）大約取得4個斜率值，再者，該些分數的分子由這4個斜率值來決定，如第6圖所示，藉由將釋放率設置RR以及攻擊率設置AR的分數之分子設置為4到1而不是全部相同（例如分數之分子均為1），代表第1圖所示之參數等化器10所達到的自+12分貝至-12分貝的增益值以及權重值（例如第一權重值β）之間的關係之曲線可以為一平滑曲線，因此，權重值（例如第一權重值β）的整體控制範圍有著相對線性速度控制，此外，可以使權重值（例如第一權重值β）的控制速度均勻。 In the above release control (or attack control), 4 points

,

,

,

(or

,

,

,

) is used to set the release rate setting RR (or the attack rate setting AR) for the slope adjustment in Figure 6, in addition, 3 coefficients 0.666, 0.43 and 0.26 are used to classify the weight values (such as the first weight value β and the second weight value β and the second One of the weight values 1-β; for better understanding, here is the first weight value β). In this embodiment, approximately four slope values are obtained in four intervals (ie, 0≦β<0.26, 0.26≦β<0.43, 0.43≦β<0.666, and 0.666≦β<1). The numerator of the fraction is determined by these 4 slope values, as shown in Figure 6, by setting the numerator of the fraction for the release rate setting RR and the attack rate setting AR to be 4 to 1 instead of all being the same (e.g. is 1), the curve representing the relationship between the gain value from +12 dB to -12 dB and the weight value (for example, the first weight value β) achieved by the parameter equalizer 10 shown in FIG. 1 can be a Smoothing the curve, therefore, the overall control range of the weight value (eg, the first weight value β) has a relatively linear speed control, and in addition, the control speed of the weight value (eg, the first weight value β) can be made uniform.

FIG. 7 shows the input signal A_IN and the output of the equalizer when the input signal A_IN is the in-band signal of the parameter equalizer 10 shown in FIG. 1 according to an embodiment of the present invention. A schematic diagram of the signal EQ_OUT, the energy level EL of the equalizer output signal EQ_OUT, and the first weight value β. In this embodiment, the frequency of the input signal A_IN is 200 Hz, and the center frequency of the parameter equalizer 10 is also 200 Hz. Therefore, the input signal A_IN is within the frequency band of the parameter equalizer 10 , as shown in FIG. 7 As shown, the parameter equalizer 10 can control the weight value (eg, one of the first weight value β and the second weight value 1-β; for better understanding, the first weight value β) by adjusting the slope. To generate the equalizer output signal EQ_OUT, when the energy level EL of the equalizer output signal EQ_OUT is less than the threshold TH, the equalizer output signal EQ_OUT can be released by the release control, or when the equalizer output signal EQ_OUT When the energy level EL is not less than the threshold value TH, the equalizer output signal EQ_OUT can be suppressed by the attack control.

FIG. 8 shows the input signal A_IN, the equalization signal under the condition that the input signal A_IN is the out-of-band signal of the parameter equalizer shown in FIG. 1, according to another embodiment of the present invention. A schematic diagram of the output signal EQ_OUT of the equalizer, the energy level EL of the output signal EQ_OUT of the equalizer, and the first weight value β. In this embodiment, the frequency of the input signal A_IN is 5000 Hz, and the center frequency of the parameter equalizer 10 is 200 Hz. Therefore, the input signal A_IN is the out-of-band signal of the parameter equalizer 10, as shown in the eighth As shown in the figure, although the parameter equalizer 10 can control the weight value (for example, one of the first weight value β and the second weight value 1-β) by adjusting the slope; A weight value β) is used to generate the equalizer output signal EQ_OUT, but the equalizer output signal EQ_OUT is the same as the input signal A_IN. Therefore, when the input signal A_IN is an out-of-band signal of the parametric equalizer 10, the parametric equalizer 10 has no effect on the input signal A_IN.

FIG. 9 is a schematic diagram of the parameter equalizer 10 shown in FIG. 1 operating on a sweep frequency according to an embodiment of the present invention. The frequency is swept from 20 kilohertz (kHz) to 20 Hz, and the center frequency of the parametric equalizer 10 is 2 kHz, as shown in FIG. 9, the parametric equalizer 10 operates only on 2 kHz.

FIG. 10 is a schematic diagram of the parameter equalizer 10 shown in FIG. 1 operating on a frequency sweep according to another embodiment of the present invention. The frequency is swept from 20 kHz to 20 Hz, and the center frequency of the parametric equalizer 10 is 5 kHz. As shown in FIG. 10, the parametric equalizer 10 operates only on 5 kHz.

Returning to FIG. 3 , the method for dynamically adjusting the weight value to equalize the input signal A_IN disclosed in the present invention can achieve two effects: type I adjustment and type II adjustment. FIG. 11 is a schematic diagram of the output frequency response of the Type I adjustment of the input signal after using the method of the present invention. FIG. 12 is a schematic diagram of the output frequency response of the Type II adjustment of the input signal after using the method of the present invention.

Please refer to FIG. 11, and the embodiment of Type I adjustment can refer to FIG. 2 and FIG. 3. In Type I adjustment, when the energy level EL is greater than the threshold value TH, the first weight value β is set to not less than 0.1432 , in addition, when the first weight value β is equal to 0.1432, the output gain is equal to 0 dB. Therefore, in Type I adjustment, when the input level is greater than a predetermined threshold value (marked as "threshold" in Figure 11), The output gain remains constant at a specific frequency (eg, the center frequency of the parametric equalizer of the present invention).

Please refer to FIG. 12, and the embodiment of the type II adjustment can refer to FIG. 2 and FIG. 3. In the type II adjustment, when the energy level EL is greater than the threshold value TH, the first weight value β is set to 0, so , in type II adjustment, when the input level is greater than a predetermined threshold value (marked as "threshold" in Figure 12), the output gain is at a specific frequency (such as the center frequency of the parametric equalizer of the present invention) can be attenuated to further avoid signal clipping.

FIG. 13 is a schematic diagram of the output frequency response when using the prior art parametric equalizer. As shown in Figure 13, the prior art parametric equalizer (marked as "PEQ") has a fixed gain attenuation regardless of the size of the input/output signal. Compared with the prior art parametric equalizer, the parametric equalizer of the present invention Equalizers have many variations that alter the output response at a particular frequency (eg, the center frequency of the parametric equalizer of the present invention).

FIG. 14 is a flowchart of a method for dynamically adjusting the weight value to equalize the input signal A_IN according to an embodiment of the present invention. If the same result can be obtained, the steps do not have to be executed in sequence according to the flow shown in Fig. 14. For example, the method shown in Fig. 14 can be performed by the parameter equalizer 10 shown in Fig. 1. be realized.

In step S60, the input signal A_IN is received by the parameter equalizer.

In step S62, the input signal A_IN is processed by the first parameter equalizer circuit to generate and output the first output signal A_OUT_1.

In step S64, the input signal A_IN is processed by the second parameter equalizer circuit to generate and output the second output signal A_OUT_2.

In step S66, the first output signal A_OUT_1 is multiplied by the first weight value β to generate the first adjustment signal AD_OUT_1.

In step S68, the second output signal A_OUT_2 is multiplied by the second weight value 1-β to generate a second adjustment signal AD_OUT_2.

In step S70, the first adjustment signal AD_OUT_1 and the second adjustment signal AD_OUT_2 are combined to generate the equalizer output signal EQ_OUT.

In step S72, the first weight value β and the second weight value 1-β are dynamically adjusted according to the equalizer output signal EQ_OUT.

Since those skilled in the art can easily understand the operation of each step shown in FIG. 14 through the description of the parameter equalizer 10 shown in FIG. 1, for the sake of brevity, similar content in this embodiment is omitted here. Repeat. The above descriptions are only preferred embodiments of the present invention, and all equivalent changes and modifications made according to the scope of the patent application of the present invention shall fall within the scope of the present invention.

10, PEQ: parameter equalizer 12,14: Parametric Equalizer Circuit 16,18: Multiplication circuit 20: Addition circuit 22: Weight control circuit A_IN: input signal A_OUT_1: The first output signal A_OUT_2: The second output signal AD_OUT_1: The first adjustment output signal AD_OUT_2: The second adjustment output signal β first weight value 1-β: The second weight value EQ_OUT: Equalizer output signal 200: Weight Control Circuit 24: Energy detection circuit 26: Weight value generator 28: Absolute value circuit 30: Alpha filter AV: absolute value EL: energy level TH: Threshold value RTH: Reference Threshold AR: Attack rate setting RR: Release rate setting S60~S72: Steps

FIG. 1 is a schematic diagram of a parameter equalizer according to an embodiment of the present invention. FIG. 2 is a schematic diagram of a weight control circuit according to an embodiment of the present invention. FIG. 3 is a schematic diagram illustrating that different gain values of the parameter equalizer shown in FIG. 1 are set under different weight value combinations according to an embodiment of the present invention. FIG. 4 is a schematic diagram illustrating the relationship between the gain achieved by the parameter equalizer shown in FIG. 1 and the weight value used by the parameter equalizer shown in FIG. 1 according to an embodiment of the present invention. FIG. 5 is a schematic diagram illustrating that different gain values of the parameter equalizer shown in FIG. 1 are set under different weight value combinations according to another embodiment of the present invention. FIG. 6 is a schematic diagram illustrating the relationship between the gain value achieved by the parameter equalizer shown in FIG. 1 and the weight value used by the parameter equalizer shown in FIG. 1 according to another embodiment of the present invention. FIG. 7 shows the energy of the input signal, the output signal of the equalizer, and the output signal of the equalizer when the input signal is the in-band signal of the parametric equalizer shown in FIG. 1 according to an embodiment of the present invention. Schematic diagram of level and weight value. FIG. 8 shows the input signal, the output signal of the equalizer, and the output signal of the equalizer when the input signal is an out-of-band signal of the parametric equalizer shown in FIG. 1 according to another embodiment of the present invention. Schematic diagram of energy levels and weight values. FIG. 9 is a schematic diagram of the parameter equalizer shown in FIG. 1 operating on a frequency sweep according to an embodiment of the present invention. FIG. 10 is a schematic diagram of the parameter equalizer shown in FIG. 1 operating on a frequency sweep according to another embodiment of the present invention. FIG. 11 is a schematic diagram of the output frequency response of the Type I adjustment of the input signal after using the method of the present invention. FIG. 12 is a schematic diagram of the output frequency response of the Type II adjustment of the input signal after using the method of the present invention. FIG. 13 is a schematic diagram of the output frequency response when using the prior art parametric equalizer. FIG. 14 is a flowchart of a method for dynamically adjusting weight values to equalize input signals according to an embodiment of the present invention.

10: Parameter equalizer

12,14: Parametric Equalizer Circuit

16,18: Multiplication circuit

20: Addition circuit

22: Weight control circuit

A_IN: input signal

A_OUT_1: The first output signal

A_OUT_2: The second output signal

AD_OUT_1: The first adjustment output signal

AD_OUT_2: The second adjustment output signal

β: the first weight value

1-β: The second weight value

EQ_OUT: Equalizer output signal

Claims (26)

A parameter equalizer that includes: a first parametric equalizer circuit having a first transfer function with a positive gain and for processing an input signal according to the first transfer function and outputting a first output signal by processing the input signal; a second parametric equalizer circuit having a second transfer function with negative gain, and for processing the input signal according to the second transfer function and outputting a second output signal by processing the input signal; a first multiplying circuit, coupled to the first parameter equalizer circuit, and used for multiplying the first output signal and a first weight value to generate a first adjusted output signal; a second multiplying circuit, coupled to the second parameter equalizer circuit, and used for multiplying the second output signal and a second weight value to generate a second adjusted output signal; an adding circuit, coupled to the first multiplying circuit and the second multiplying circuit, and used for combining the first adjustment signal and the second adjustment signal to generate an equalizer output signal; and A weight control circuit is coupled to the first multiplying circuit, the second multiplying circuit and the adding circuit, and is used for dynamically adjusting the first weight value and the second weight value according to the output signal of the equalizer. The parametric equalizer as described in claim 1, wherein the first parametric equalizer circuit is a peak filter implemented by a second-order filter. The parametric equalizer as described in claim 1, wherein the second parametric equalizer circuit is a notch filter implemented by a second-order filter. The parameter equalizer as described in claim 1, wherein a sum of the first weight value and the second weight value is equal to 1. The parameter equalizer as described in claim 1, wherein the weight control circuit comprises: an energy detection circuit for measuring an energy level of the output signal of the equalizer; and A weight value generator is coupled to the energy detection circuit and used for dynamically adjusting the first weight value and the second weight value according to the energy level of the output signal of the equalizer. The parameter equalizer as described in claim 5, wherein the energy detection circuit comprises: an absolute value circuit for generating an absolute value of the output signal of the equalizer; and an alpha filter, coupled to the absolute value circuit and the weight value generator, and used for generating the energy level according to an output of the absolute value circuit. The parameter equalizer as described in claim 5, wherein the weight value generator is further used for receiving an attack rate setting and a release rate setting; when the energy level of the output signal of the equalizer is less than a threshold value, the weight value generator controls one of the first weight value and the second weight value according to the slope adjustment specified by the release rate setting; and when the energy level of the equalizer output signal is different When less than the threshold value, the weight value generator controls the one of the first weight value and the second weight value according to the slope adjustment specified by the attack rate setting. The parameter equalizer of claim 7, wherein the release rate setting includes a plurality of fractions, and a power term of the denominator of the plurality of fractions controls a release rate to which the release rate setting is applied. The parameter equalizer as described in claim 7, wherein the attack rate setting includes a plurality of fractions, and a power term of the denominator of the plurality of fractions controls an attack rate to which the attack rate setting is applied. The parameter equalizer as described in claim 7, wherein the weight generator is further used for receiving a reference threshold value, and obtaining the threshold value by multiplying the reference threshold value and a predetermined coefficient. The parameter equalizer as described in claim 1, wherein the parameter equalizer has a first adjustment mode and a second adjustment mode; in the first adjustment mode, when a level of the input signal is greater than a threshold value, a gain of the output signal of the equalizer remains unchanged; and in the second adjustment mode, when the level of the input signal is greater than the threshold value, the gain of the output signal of the equalizer attenuated. The parameter equalizer as described in claim 1, wherein when the input signal is an out-of-band signal of the parameter equalizer, the output signal of the equalizer is the same as the input signal. The parameter equalizer as described in claim 1, wherein the first transfer function further includes a first quality factor, the second transfer function further includes a second quality factor, and the second quality factor and the The ratio of the first quality factor is a predetermined constant. A method of equalizing an input signal to generate an equalizer output signal, comprising: processing the input signal by a first parametric equalizer circuit having a first transfer function with positive gain to generate and output a first output signal; processing the input signal by a second parametric equalizer circuit having a second transfer function with negative gain to generate and output a second output signal; multiplying the first output signal and a first weight value to generate a first adjusted output signal; multiplying the second output signal and a second weight value to generate a second adjusted output signal; combining the first adjusted output signal and the second adjusted output signal to generate the equalizer output signal; and The first weight value and the second weight value are dynamically adjusted according to the output signal of the equalizer. The method of claim 14, wherein the first parametric equalizer circuit is a spike filter implemented by a second-order filter. The method of claim 14, wherein the second parametric equalizer circuit is a notch filter implemented by a second-order filter. The method of claim 14, wherein a sum of the first weight value and the second weight value is equal to 1. The method of claim 14, wherein the step of dynamically adjusting the first weight value and the second weight value according to the output signal of the equalizer comprises: measuring an energy level of the equalizer output signal; and The first weight value and the second weight value are dynamically adjusted according to the energy level of the output signal of the equalizer. The method of claim 18, wherein the step of measuring the energy level of the equalizer output signal comprises: performing an absolute value calculation on the equalizer output signal to generate an absolute value calculation result; and The energy level is generated according to the absolute value calculation result. The method of claim 18, wherein the step of dynamically adjusting the first weight value and the second weight value according to the energy level of the output signal of the equalizer comprises: receiving an attack rate setting and a release rate setting; When the energy level of the equalizer output signal is less than a threshold value, controlling one of the first weight value and the second weight value according to a slope adjustment specified by the release rate setting; and When the energy level of the equalizer output signal is not less than the threshold value, the one of the first weight value and the second weight value is controlled according to the slope adjustment specified by the attack rate setting. The method of claim 20, wherein the release rate setting includes a plurality of fractions, and a power term of the denominator of the plurality of fractions controls a release rate to which the release rate setting is applied. The method of claim 20, wherein the attack rate setting includes a plurality of fractions, and a power term of the denominator of the plurality of fractions controls an attack rate to which the attack rate setting is applied. The method of claim 20, wherein the step of dynamically adjusting the first weight value and the second weight value according to the energy level of the output signal of the equalizer further comprises: receiving a reference threshold; and The threshold value is obtained by multiplying the reference threshold value by a predetermined coefficient. The method of claim 14, wherein the parameter equalizer has a first adjustment mode and a second adjustment mode; in the first adjustment mode, when a level of the input signal is greater than a threshold value when the method keeps a gain of the equalizer output signal unchanged; and in the second adjustment mode, when the level of the input signal is greater than the threshold value, the method attenuates the equalizer output signal of this gain. The method as described in claim 14, wherein since the input signal is an out-of-band signal, the method generates the output signal of the equalizer which is the same as the input signal. The method of claim 14, wherein the first transfer function further includes a first quality factor, the second transfer function further includes a second quality factor, and the second quality factor and the first quality factor The ratio of the factors is a predetermined constant.

Images