CN101567190A - Speech gain control method and device - Google Patents

Abstract

The invention embodiment relates to a speech gain control method, which uses the first feedback signal obtained by adjusting the second speech frame before the first speech frame to adjust the first speech frame, and the first feedback signal is corresponding to the differential value of low pass energy and reference energy after adjustment of the second speech frame. In addition, the invention embodiment also provides a speech gain control device. The invention embodiment uses simple and high-efficient algorithm, which adjusts the original inappropriate speech gain to appropriate level fast, for keeping speech sound volume reasonable and comfortable, and the receiver can obtain better hearing effect, especially in sound speech environment with larger noise. The adjusted speech gain does not have obvious fluctuation and is not affected by the background noise, thereby obtaining better sampling accuracy in analog-digital conversion.

Description

Speech gain control method and device

Technical field

The present invention relates to field of voice signal, relate in particular to a kind of speech gain control method and device.

Background technology

In speech communication, often go out the volume balance phenomenon of the improper appearance of realize voice gain control, if the too for a short time volume that causes of voice gain is too little, voice quality descends, and when noise ratio is serious, can make reciever to the voice indigestion; If the voice gain causes volume too big too greatly, the voice degree of comfort decreased, even cause playback equipment unsweet sound phenomenon to occur.

Summary of the invention

Embodiment of the invention technical matters to be solved is, a kind of speech gain control method and device are provided, can adopt simple and high-efficient algorithm, original inappropriate voice gain is adjusted to suitable level very soon, keep the reasonable and comfortable of speech volume, obtain better auditory effect at reciever, especially in the bigger speech sound environment of noise, the voice gain that regulates can obviously not fluctuateed, also can not be subjected to the influence of ground unrest, better sampling precision can be arranged in analog-digital conversion.

For solving the problems of the technologies described above, the embodiment of the invention adopts following technical scheme:

A kind of speech gain control method, described method comprises:

Import first speech frame;

Second speech frame of employing before to described first speech frame adjusted first feedback signal of gained, and described first speech frame is adjusted, and described first feedback signal is corresponding to the difference of adjusted low pass energy of described second speech frame and reference energy.

A kind of speech gain control device comprises:

Input block is used to import first speech frame;

Processing unit, be used to adopt first feedback signal of second speech frame before described first speech frame being adjusted gained, described first speech frame is adjusted, and described first feedback signal is corresponding to the difference of adjusted low pass energy of described second speech frame and reference energy;

Output unit is used for the result output with described processing unit.

The beneficial effect of the embodiment of the invention is:

By a kind of speech gain control method and corresponding device are provided, can adopt simple and high-efficient algorithm, original inappropriate voice gain is adjusted to suitable level very soon, keep the reasonable and comfortable of speech volume, obtain better auditory effect at reciever, especially in the bigger speech sound environment of noise, the voice gain that regulates can obviously not fluctuateed, also can not be subjected to the influence of ground unrest, better sampling precision can be arranged in analog-digital conversion.

Below in conjunction with accompanying drawing the embodiment of the invention is described in further detail.

Description of drawings

Fig. 1 is the main process flow diagram of the speech gain control method of the embodiment of the invention;

Fig. 2 is the specific embodiment synoptic diagram of speech gain control method of the present invention;

Fig. 3 is the primary structure figure of the speech gain control device of the embodiment of the invention;

Fig. 4 is the specific embodiment synoptic diagram of speech gain control device of the present invention.

Embodiment

The embodiment of the invention provides a kind of speech gain control method, mainly comprises flow process as shown in Figure 1:

101, import first speech frame;

102, second speech frame of employing before to described first speech frame adjusted first feedback signal of gained, described first speech frame is adjusted, and described first feedback signal is corresponding to the difference of adjusted low pass energy of described second speech frame and reference energy.

Fig. 2 shows a specific embodiment of speech gain control method of the present invention, and with reference to this figure, this method mainly comprises:

201, to the adjusted low pass energy of the second speech frame i x before the first speech frame i+1 _iQuantize, particularly, can be to the adjusted low pass energy of second speech frame i x _iQuantize with predefined 25 quantification gradations, wherein, predefined quantification gradation comprises 25 quantification gradations as shown in table 1 below, wherein:

First quantification gradation is 1 decibels, and corresponding low pass energy is for being less than or equal to 44;

Second quantification gradation is 2 decibels, and corresponding low pass energy is 54;

The 3rd quantification gradation is 3 decibels, and corresponding low pass energy is 66;

The 4th quantification gradation is 4 decibels, and corresponding low pass energy is 82;

The 5th quantification gradation is 5 decibels, and corresponding low pass energy is 100;

The 6th quantification gradation is 6 decibels, and corresponding low pass energy is 122;

The 7th quantification gradation is 7 decibels, and corresponding low pass energy is 150;

The 8th quantification gradation is 8 decibels, and corresponding low pass energy is 184;

The 9th quantification gradation is 9 decibels, and corresponding low pass energy is 224;

The tenth quantification gradation is 10 decibels, and corresponding low pass energy is 274;

The 11 quantification gradation is 11 decibels, and corresponding low pass energy is 334;

The 12 quantification gradation is 12 decibels, and corresponding low pass energy is 410;

The 13 quantification gradation is 13 decibels, and corresponding low pass energy is 500, and described the 13 quantification gradation is reference levels, and described reference energy is 13 decibels;

The 14 quantification gradation is 14 decibels, and corresponding low pass energy is 612;

The 15 quantification gradation is 15 decibels, and corresponding low pass energy is 748;

The 16 quantification gradation is 16 decibels, and corresponding low pass energy is 916;

The 17 quantification gradation is 17 decibels, and corresponding low pass energy is 1120;

The 18 quantification gradation is 18 decibels, and corresponding low pass energy is 1370;

The 19 quantification gradation is 19 decibels, and corresponding low pass energy is 1674;

The 20 quantification gradation is 20 decibels, and corresponding low pass energy is 2048;

The 21 quantification gradation is 21 decibels, and corresponding low pass energy is 2506;

The 22 quantification gradation is 22 decibels, and corresponding low pass energy is 3068;

The 23 quantification gradation is 23 decibels, and corresponding low pass energy is 3740;

The 24 quantification gradation is 24 decibels, and corresponding low pass energy is 4596;

The 25 quantification gradation is 25 decibels, and corresponding low pass energy is more than or equal to 5626;

Table 1

Quantification gradation (Y)	Low pass energy (y i)	Quantification gradation (Y)	Low pass energy (y i)
				1	Be less than or equal to 44	14	612
2	54	15	748
				3	66	16	916
4	82	17	1120
				5	100	18	1370
6	122	19	1674

7	150	20	2048
				8	184	21	2506
9	224	22	3068
				10	274	23	3740
11	334	24	4596
				12	410	25	More than or equal to 5626
13, reference levels	500

And quantification can be adopted the handling function y of above-mentioned table 1 correspondence _i=F (x _i);

202, the quantized result according to 201 obtains the adjusted low pass energy of second speech frame i x _iProbability density P _i(Y), particularly, can realize as follows:

At first, to the adjusted low pass energy of second speech frame i x _iCorresponding quantitative grade y _iCarry out the increment counting, specifically can adopt following handling function:

$A_i\left(Y\right)=\left\{\begin{array}{c}1,y_i=Y\\ 0,y_i\≠Y\end{array}\right.$

Wherein, A _i(Y) the expression second speech frame i quantification gradation is the increment count results of Y;

Secondly, the A as a result that counts according to increment _i(Y), adopt the preceding adjusted low pass energy of the 3rd speech frame i-1 count results C of the level and smooth second speech frame i of first order pole filtering technique _I-1(Y), obtain the adjusted low pass energy of second speech frame i count results C _i(Y), specifically can adopt following handling function:

C _i(Y)＝C _i-1(Y)+[A _i(Y)-C _i-1(Y)]/32

Wherein, the feedback factor that the first order pole filtering technique is adopted is 1/32, is equivalent to a short-time average effect;

Then, according to the low pass energy count results C after the smoothing processing _i(Y), calculate the adjusted low pass energy of second speech frame i x _iProbability density P _i(Y), specifically can adopt following handling function:

$P_i\left(Y\right)=\frac{C_i\left(Y\right)}{\underset{Y=1}{\overset{25}{\mathrm{\Σ}}}{C}_i\left(Y\right)}$

203, according to low pass energy x _iProbability density P _i(Y), obtain the adjusted low pass energy of second speech frame i x _iDifference ε with reference energy (energy that REF is corresponding) _i, specifically can adopt following handling function:

${\mathrm{\ϵ}}_i=\mathrm{REF}-\underset{Y=1}{\overset{25}{\mathrm{\Σ}}}{\mathrm{YP}}_i\left(Y\right)$

Wherein, according to table 1, REF=13;

204, with difference ε _iDetermine the first feedback signal g _i, the first feedback signal g _iFor the second speech frame i being adjusted gained, the first feedback signal g _iCorresponding to the adjusted low pass energy of second speech frame i x _iWith the difference of reference energy, particularly, can realize as follows:

At first, according to difference ε _iWith the result of calculation of predefined linear scale factor, obtain feedback gain increment Delta g _i, specifically can adopt following handling function:

$\mathrm{\&Delta;}{g}_i=\left\{\begin{array}{c}0,\left|{\mathrm{\&epsiv;}}_i\right|\&le;2\\ ({\mathrm{\&epsiv;}}_i+2)/32,{\mathrm{\&epsiv;}}_i<-2\\ ({\mathrm{\&epsiv;}}_i-2)/32,{\mathrm{\&epsiv;}}_i>2\end{array}\right.$

Wherein, linear scale factor is 32, feedback gain increment Delta g _iWith difference ε _iBe directly proportional, except difference ε _iNear zero dead band;

Secondly, adopt feedback gain increment Delta g _iThe second feedback signal g of gained is adjusted in renewal to the 3rd speech frame i-1 _I-1, obtain the first feedback signal g _i, the second feedback signal g _I-1Corresponding to the difference of adjusted low pass energy of the 3rd speech frame i-1 and reference energy, the speed of renewal can be controlled by adjusting linear scale factor, specifically can adopt following handling function:

g _i＝g _i-1+Δg _i

205, import the first speech frame i+1;

206, adopt the first feedback signal g that the second speech frame i is adjusted _i, the first speech frame i+1 is adjusted the first feedback signal g _iCorresponding to the adjusted low pass energy of second speech frame i x _iDifference ε with reference energy _i

As a kind of embodiment, on average replace statistical average can further reduce computation complexity with the time.

Need to prove that the speech gain control method of the embodiment of the invention and device mainly are applicable to the speech sound environment, in silent environment or non-voice environment, then do not calculate feedback signal.

The embodiment of the invention provides a kind of speech gain control device, mainly comprises structure as shown in Figure 3:

Input block 301 is used to import first speech frame;

Processing unit 302, be used to adopt first feedback signal of second speech frame before described first speech frame being adjusted gained, described first speech frame is adjusted, and described first feedback signal is corresponding to the difference of adjusted low pass energy of described second speech frame and reference energy;

Output unit 303 is used for the result output with described processing unit 302,

Input block 301, output unit 303 link to each other with processing unit 302 respectively.

Fig. 4 shows a specific embodiment of speech gain control device of the present invention, and with reference to this figure, this device mainly comprises:

Input block 401 is used to import the first speech frame i+1;

Processing unit 402 is used to adopt the first feedback signal g that the second speech frame i before the first speech frame i+1 is adjusted _i, the first speech frame i+1 is adjusted the first feedback signal g _iCorresponding to the adjusted low pass energy of second speech frame i x _iDifference ε with reference energy _i

Output unit 403 is used for the result output with processing unit 402;

Input block 401, output unit 403 link to each other with processing unit 402 respectively;

Particularly, processing unit 402 comprises:

Quantifying unit is used for the adjusted low pass energy of second speech frame i x _iQuantize, particularly, can be to the adjusted low pass energy of second speech frame i x _iQuantize with predefined 25 quantification gradations, wherein, predefined quantification gradation comprises as above 25 quantification gradations shown in the table 1, repeat no more, and quantification can be adopted the handling function y of above-mentioned table 1 correspondence herein _i=F (x _i);

Low pass energy probability density computing unit is used for the quantized result according to quantifying unit, obtains the adjusted low pass energy of second speech frame i x _iProbability density P _i(Y), particularly, can comprise:

Increment counting unit is used for the adjusted low pass energy of second speech frame i x _iCorresponding quantitative grade y _iCarry out the increment counting, specifically can adopt following handling function:

$A_i\left(Y\right)=\left\{\begin{array}{c}1,y_i=Y\\ 0,y_i\&NotEqual;Y\end{array}\right.$

Wherein, A _i(Y) the expression second speech frame i quantification gradation is the increment count results of Y;

Low pass energy meter counting unit is used for the A as a result according to the increment counting _i(Y), adopt the preceding adjusted low pass energy of the 3rd speech frame i-1 count results C of the level and smooth second speech frame i of first order pole filtering technique _I-1(Y), obtain the adjusted low pass energy of second speech frame i count results C _i(Y), specifically can adopt following handling function:

C _i(Y)＝C _i-1(Y)+[A _i(Y)-C _i-1(Y)]/32

Wherein, the feedback factor that the first order pole filtering technique is adopted is 1/32, is equivalent to a short-time average effect;

Level and smooth back computing unit is used for according to the low pass energy count results C after the smoothing processing _i(Y), calculate the adjusted low pass energy of second speech frame i x _iProbability density P _i(Y), specifically can adopt following handling function:

$P_i\left(Y\right)=\frac{C_i\left(Y\right)}{\underset{Y=1}{\overset{25}{\mathrm{\&Sigma;}}}{C}_i\left(Y\right)}$

Low pass energy meter counting unit links to each other with increment counting unit, level and smooth back computing unit respectively;

Difference computational unit is used for according to low pass energy x _iProbability density P _i(Y), obtain the adjusted low pass energy of second speech frame i x _iDifference ε with reference energy (energy that REF is corresponding) _i, specifically can adopt following handling function:

${\mathrm{\&epsiv;}}_i=\mathrm{REF}-\underset{Y=1}{\overset{25}{\mathrm{\&Sigma;}}}{\mathrm{YP}}_i\left(Y\right)$

Wherein, according to table 1, REF=13;

Determining unit is used for difference ε _iDetermine the first feedback signal g _i, the first feedback signal g _iFor the second speech frame i being adjusted gained, the first feedback signal g _iCorresponding to the adjusted low pass energy of second speech frame i x _iWith the difference of reference energy, particularly, can comprise:

Feedback gain incremental computations unit is used for according to difference ε _iWith the result of calculation of predefined linear scale factor, obtain feedback gain increment Delta g _i, specifically can adopt following handling function:

$\mathrm{\&Delta;}{g}_i=\left\{\begin{array}{c}0,\left|{\mathrm{\&epsiv;}}_i\right|\&le;2\\ ({\mathrm{\&epsiv;}}_i+2)/32,{\mathrm{\&epsiv;}}_i<-2\\ ({\mathrm{\&epsiv;}}_i-2)/32,{\mathrm{\&epsiv;}}_i>2\end{array}\right.$

Wherein, linear scale factor is 32, feedback gain increment Delta g _iWith difference ε _iBe directly proportional, except difference ε _iNear zero dead band;

The feedback signal determining unit is used to adopt feedback gain increment Delta g _iThe second feedback signal g of gained is adjusted in renewal to the 3rd speech frame i-1 _I-1, obtain the first feedback signal g _i, the second feedback signal g _I-1Corresponding to the difference of adjusted low pass energy of the 3rd speech frame i-1 and reference energy, the speed of renewal can be controlled by adjusting linear scale factor, specifically can adopt following handling function:

g _i＝g _i-1+Δg _i

Quantifying unit, difference computational unit link to each other with low pass energy probability density computing unit respectively, and difference computational unit links to each other with determining unit;

Implement speech gain control method and the device of the invention described above embodiment, can adopt simple and high-efficient algorithm, original inappropriate voice gain is adjusted to suitable level very soon, keep the reasonable and comfortable of speech volume, obtain better auditory effect at reciever, especially in the bigger speech sound environment of noise, the voice gain that regulates can obviously not fluctuateed, also can not be subjected to the influence of ground unrest, better sampling precision can be arranged in analog-digital conversion; In addition, each feedback signal is (as first feedback signal, second feedback signal etc.) be to produce by the speech frame low pass energy that calculating is lower than 1kHz, because the size of low pass energy mainly depends on first resonance peak, it is relatively independent of specific speaker's voice attributes, another reason of using the low pass energy is that it can also be as the output of voice activation detection, the low pass energy is the short-time average attribute of voice, it to sound fluctuation cause not steady, the leakage of radio-frequency component, the interference of ambient noise etc. are very responsive, for level and smooth these disturbance factors, by calculating the mean value that low pass energy probability density obtains the low pass energy.

Need to prove that the speech gain control method of the embodiment of the invention and device mainly are applicable to the speech sound environment, in silent environment or non-voice environment, then do not calculate feedback signal.

In addition, one of ordinary skill in the art will appreciate that all or part of flow process that realizes in the foregoing description method, be to instruct relevant hardware to finish by program, described program can be stored in the computer-readable recording medium, this program can comprise the flow process as the embodiment of above-mentioned each side method when carrying out.Wherein, described storage medium can be magnetic disc, CD, read-only storage memory body (Read-OnlyMemory, ROM) or at random store memory body (Random Access Memory, RAM) etc.

The above is the specific embodiment of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also are considered as protection scope of the present invention.

Claims (8)

1, a kind of speech gain control method is characterized in that, described method comprises:

Import first speech frame;

Second speech frame of employing before to described first speech frame adjusted first feedback signal of gained, and described first speech frame is adjusted, and described first feedback signal is corresponding to the difference of adjusted low pass energy of described second speech frame and reference energy.

2, the method for claim 1 is characterized in that, described method also comprises:

The adjusted low pass energy of described second speech frame is quantized;

According to described quantized result, obtain the adjusted low pass energy of described second speech frame probability density;

According to described low pass energy probability density, obtain the difference of adjusted low pass energy of described second speech frame and reference energy;

Determine described first feedback signal with described difference.

3, method as claimed in claim 2 is characterized in that, described the adjusted low pass energy of described second speech frame is quantized to be specially:

The adjusted low pass energy of described second speech frame is quantized with predefined quantification gradation,

Described according to described quantized result, obtain the adjusted low pass energy of described second speech frame probability density and specifically comprise:

The adjusted low pass energy of described second speech frame corresponding quantitative grade is carried out the increment counting;

According to the result of described increment counting, adopt the preceding adjusted low pass energy of the 3rd speech frame count results of level and smooth described second speech frame of first order pole filtering technique, obtain the adjusted low pass energy of described second speech frame count results;

According to the low pass energy count results after the described smoothing processing, calculate the adjusted low pass energy of described second speech frame probability density,

Describedly determine that with described difference described first feedback signal specifically comprises:

According to the result of calculation of described difference and predefined linear scale factor, obtain the feedback gain increment;

Adopt described feedback gain incremental update that described the 3rd speech frame is adjusted second feedback signal of gained, obtain described first feedback signal, described second feedback signal is corresponding to the difference of adjusted low pass energy of described the 3rd speech frame and reference energy.

4, method as claimed in claim 3 is characterized in that, described predefined quantification gradation comprises 25 quantification gradations, wherein:

First quantification gradation is 1 decibels, and corresponding low pass energy is for being less than or equal to 44;

Second quantification gradation is 2 decibels, and corresponding low pass energy is 54;

The 3rd quantification gradation is 3 decibels, and corresponding low pass energy is 66;

The 4th quantification gradation is 4 decibels, and corresponding low pass energy is 82;

The 5th quantification gradation is 5 decibels, and corresponding low pass energy is 100;

The 6th quantification gradation is 6 decibels, and corresponding low pass energy is 122;

The 7th quantification gradation is 7 decibels, and corresponding low pass energy is 150;

The 8th quantification gradation is 8 decibels, and corresponding low pass energy is 184;

The 9th quantification gradation is 9 decibels, and corresponding low pass energy is 224;

The tenth quantification gradation is 10 decibels, and corresponding low pass energy is 274;

The 11 quantification gradation is 11 decibels, and corresponding low pass energy is 334;

The 12 quantification gradation is 12 decibels, and corresponding low pass energy is 410;

The 13 quantification gradation is 13 decibels, and corresponding low pass energy is 500, and described the 13 quantification gradation is reference levels, and described reference energy is 13 decibels;

The 14 quantification gradation is 14 decibels, and corresponding low pass energy is 612;

The 15 quantification gradation is 15 decibels, and corresponding low pass energy is 748;

The 16 quantification gradation is 16 decibels, and corresponding low pass energy is 916;

The 17 quantification gradation is 17 decibels, and corresponding low pass energy is 1120;

The 18 quantification gradation is 18 decibels, and corresponding low pass energy is 1370;

The 19 quantification gradation is 19 decibels, and corresponding low pass energy is 1674;

The 20 quantification gradation is 20 decibels, and corresponding low pass energy is 2048;

The 21 quantification gradation is 21 decibels, and corresponding low pass energy is 2506;

The 22 quantification gradation is 22 decibels, and corresponding low pass energy is 3068;

The 23 quantification gradation is 23 decibels, and corresponding low pass energy is 3740;

The 24 quantification gradation is 24 decibels, and corresponding low pass energy is 4596;

The 25 quantification gradation is 25 decibels, and corresponding low pass energy is more than or equal to 5626,

The feedback factor that described first order pole filtering technique is adopted is 1/32, and described linear scale factor is 32.

5, a kind of speech gain control device is characterized in that, comprising:

Input block is used to import first speech frame;

Processing unit, be used to adopt first feedback signal of second speech frame before described first speech frame being adjusted gained, described first speech frame is adjusted, and described first feedback signal is corresponding to the difference of adjusted low pass energy of described second speech frame and reference energy;

Output unit is used for the result output with described processing unit.

6, device as claimed in claim 5 is characterized in that, described processing unit comprises:

Quantifying unit is used for the adjusted low pass energy of described second speech frame is quantized;

Low pass energy probability density computing unit is used for according to described quantized result, obtains the adjusted low pass energy of described second speech frame probability density;

Difference computational unit is used for obtaining the difference of adjusted low pass energy of described second speech frame and reference energy according to described low pass energy probability density;

Determining unit is used for determining described first feedback signal with described difference.

7, method as claimed in claim 6 is characterized in that, described quantifying unit specifically is used for the adjusted low pass energy of described second speech frame is quantized with predefined quantification gradation,

Described low pass energy probability density computing unit specifically comprises:

Increment counting unit is used for the adjusted low pass energy of described second speech frame corresponding quantitative grade is carried out the increment counting;

Low pass energy meter counting unit, be used for result according to described increment counting, adopt the preceding adjusted low pass energy of the 3rd speech frame count results of level and smooth described second speech frame of first order pole filtering technique, obtain the adjusted low pass energy of described second speech frame count results;

Level and smooth back computing unit is used for according to the low pass energy count results after the described smoothing processing, calculates the adjusted low pass energy of described second speech frame probability density,

Described determining unit specifically comprises:

Feedback gain incremental computations unit is used for the result of calculation according to described difference and predefined linear scale factor, obtains the feedback gain increment;

The feedback signal determining unit, be used to adopt described feedback gain incremental update described the 3rd speech frame to be adjusted second feedback signal of gained, obtain described first feedback signal, described second feedback signal is corresponding to the difference of adjusted low pass energy of described the 3rd speech frame and reference energy.

8, device as claimed in claim 7 is characterized in that, described predefined quantification gradation comprises 25 quantification gradations, wherein:

First quantification gradation is 1 decibels, and corresponding low pass energy is for being less than or equal to 44;

Second quantification gradation is 2 decibels, and corresponding low pass energy is 54;

The 3rd quantification gradation is 3 decibels, and corresponding low pass energy is 66;

The 4th quantification gradation is 4 decibels, and corresponding low pass energy is 82;

The 5th quantification gradation is 5 decibels, and corresponding low pass energy is 100;

The 6th quantification gradation is 6 decibels, and corresponding low pass energy is 122;

The 7th quantification gradation is 7 decibels, and corresponding low pass energy is 150;

The 8th quantification gradation is 8 decibels, and corresponding low pass energy is 184;

The 9th quantification gradation is 9 decibels, and corresponding low pass energy is 224;

The tenth quantification gradation is 10 decibels, and corresponding low pass energy is 274;

The 11 quantification gradation is 11 decibels, and corresponding low pass energy is 334;

The 12 quantification gradation is 12 decibels, and corresponding low pass energy is 410;

The 13 quantification gradation is 13 decibels, and corresponding low pass energy is 500, and described the 13 quantification gradation is reference levels, and described reference energy is 13 decibels;

The 14 quantification gradation is 14 decibels, and corresponding low pass energy is 612;

The 15 quantification gradation is 15 decibels, and corresponding low pass energy is 748;

The 16 quantification gradation is 16 decibels, and corresponding low pass energy is 916;

The 17 quantification gradation is 17 decibels, and corresponding low pass energy is 1120;

The 18 quantification gradation is 18 decibels, and corresponding low pass energy is 1370;

The 19 quantification gradation is 19 decibels, and corresponding low pass energy is 1674;

The 20 quantification gradation is 20 decibels, and corresponding low pass energy is 2048;

The 21 quantification gradation is 21 decibels, and corresponding low pass energy is 2506;

The 22 quantification gradation is 22 decibels, and corresponding low pass energy is 3068;

The 23 quantification gradation is 23 decibels, and corresponding low pass energy is 3740;

The 24 quantification gradation is 24 decibels, and corresponding low pass energy is 4596;

The 25 quantification gradation is 25 decibels, and corresponding low pass energy is more than or equal to 5626,

The feedback factor that described first order pole filtering technique is adopted is 1/32, and described linear scale factor is 32.

Images