JPH10105198A - Speech encoding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the speech encoding device which can perform efficient input speech level control making good use of existent software and hardware resources for a high-compressibility speech encoding process. SOLUTION: An input speech control means 21 performs a block flooding process for a digital speech signal inputted from a digital speech signal input device 1 to calculate scheduling bits, which are stored in a shift register 25. After quantization, encoding is carried out. A level detection part 26 detects whether or not a frame having a maximum frame energy code calculated by a frame energy quantizing means 22 is inputted successively more than a specific number of times on the basis of control information stored in a control information storage part 27. When the level detection part detects the successive high- level input state, the shift register 25 subtracts '1' from the value of the stored scaling bits.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speech coding apparatus, and more particularly to a speech coding apparatus having an input level adjusting function.

[0002]

2. Description of the Related Art Conventionally, in this type of speech coding apparatus, level adjustment of input speech has been realized by adding dedicated hardware before the encoding process, regardless of the speech coding algorithm. doing.

As shown in FIG. 9, for adjusting the level of an input audio signal, a threshold value for adjusting an audio signal is adjusted to a predetermined value at the time of analog / digital conversion. There is a level adjustment system. This level adjustment system comprises an analog audio signal input means 7;
An input audio signal level adjusting device 8 and digital audio signal output means 9 are provided.

The input audio signal level adjusting device 8 includes a variable gain amplifier 81 for adjusting an input signal, and a variable gain amplifier 81.
And decay timing generator 8 for generating control voltage to
2, a peak value detection comparator 83 for generating a trigger output of the response / decay timing generator 82, and a threshold value recovery timer 84 for measuring the duration of a signal exceeding the threshold value of the peak value detection comparator 83. And a drive amplifier 85 for amplifying the power gain with respect to the input signal.

In the above level adjustment system, the peak value detection comparator 83 is initially set to the high threshold state by the threshold value recovery timer 84. When the signal level and the peak value of the input signal increase, the peak value detection comparator 83 starts outputting a trigger pulse to the response / decay timing generator 82.

At the same time, the threshold recovery timer 84 begins measuring the duration of the signal exceeding the threshold of the peak value detection comparator 83, and the level of the output, which has been boosted over a preselected period of time, is used to detect the peak value. When the threshold value recovery timer 84 detects that the threshold value of the comparator 83 has been exceeded,
The compression threshold value of the peak value detection comparator 83 is switched to a low level.

[0007] The threshold value of the peak value detection comparator 83 remains at a low level until there is no continuous signal. At this point, the circuit returns to a normal operating state. This technique is disclosed in JP-A-2-278945.

FIG. 10 shows an example in which the above-described level adjustment system is applied to a high-compression-rate speech coding apparatus. In FIG. 10, a high-compression-rate audio encoding device includes an analog audio signal input unit 11, an input audio signal level adjusting device 8 corresponding to the above-described level adjustment system, a digital audio signal input unit 12, and an audio encoding device 13. And digitally encoded audio signal output means 14.

The analog audio signal input by the analog audio signal input means 11 is input to the audio encoding device 13 from the digital audio signal input means 12 after the input level is adjusted by the input audio signal level adjusting device 8. This analog audio signal is subjected to arbitrary high compression rate audio encoding processing in the audio encoding device 13 and then output from the digitally encoded audio signal output means 14.

As another level adjustment of input audio, there is an input audio level adjustment control method for adjusting a digital audio signal to a predetermined level and outputting the same as shown in FIG. The system configuration based on the input audio level adjustment control system includes a digital audio signal input unit 15 and
An input audio signal level adjusting device 16 and digital audio signal output means 17 are provided.

The input audio signal level adjusting device 16 includes an amplifier circuit 161 for amplifying the level of the digital audio signal, a level detection circuit 162 for detecting the average level of the digital audio signal, and a gain setting for setting the gain of the amplifier circuit 161. And a comparator 163 for comparing the average level of the digital audio signal with the threshold set in the threshold setting register 165.

In this system, the gain setting section 164
Sets the gain based on the comparison result by the comparison circuit 163, and amplifies the digital audio signal by the amplification circuit 161 so as to have a predetermined level.

The average level of the digital audio signal output from the amplification circuit 161 is detected by the level detection circuit 162, and the threshold set in the threshold setting register 165 is sequentially updated.

The value and the average level detected by the average level detection circuit 162 are compared by a comparison circuit 163, and the amplification circuit 161 is determined based on the number of times the average level exceeds the threshold.
Set the gain of

The average level of the digital audio signal output from the amplifying circuit 161 is detected by a level detecting circuit 162, and a threshold value set in a threshold value setting register 165 is fixed.
The comparison circuit 163 compares the average level with the average level according to the above, and sets the gain of the amplification circuit 161 so that the number of times the average level exceeds the threshold value becomes a predetermined number. This technique is disclosed in Japanese Patent Application Laid-Open No. 4-127758.

FIG. 12 shows an example in which the above system is applied to a high-compression-rate speech coding apparatus. In FIG. 12, a high-compression-ratio speech coding apparatus includes digital speech signal input means 18.
And an input audio signal level adjusting device 16 corresponding to the above system, an audio encoding device 19, and digitally encoded audio signal output means 20.

The digital audio signal input from the digital audio signal input means 18 is input to the audio encoding device 19 after the input level is adjusted by the input audio signal level adjusting device 16. The digital audio signal is subjected to an arbitrary high compression rate audio encoding process in the audio encoding device 19, and then output from the digital encoded audio signal output means 20.

[0018]

SUMMARY OF THE INVENTION For a speech coding apparatus such as a digital cellular phone which performs speech coding at a high compression rate, the coding is matched with the coding theory while the size and power consumption of the apparatus are reduced. Improving voice quality is an important issue. In particular, in high-compression-rate audio coding, if the level of the input audio is higher than the limit of the encoding, the quality of the encoded audio is significantly deteriorated.

However, in the above-mentioned conventional technology,
Since the processing content of the audio encoding processing program is not used in the input audio level adjustment method, the quality of the encoded audio cannot be improved by a method that matches the audio encoding method.

Further, in the above-mentioned conventional technique, the input audio level adjusting method is configured to newly realize hardware for input control processing in the audio encoding apparatus, so that the apparatus can be downsized. In addition, power consumption cannot be reduced.

Therefore, an object of the present invention is to solve the above-mentioned problems and to provide a speech codec capable of performing efficient input speech level control using existing software and hardware resources as speech processing at a high compression rate. To provide a chemical conversion device.

[0022]

According to the present invention, there is provided an audio encoding apparatus for encoding input audio data by high-compression-rate audio encoding processing. Monitoring means for monitoring a frame energy code which is calculated and indicates the power of the input audio data in analysis frame units; and a high compression rate audio encoding process based on the monitoring result of the monitoring means. Means for performing level control.

As described above, in the speech coding apparatus of the present invention, the input speech level control is realized by software using the parameters calculated in the existing speech coding processing program. More specifically, a frame energy code (hereinafter, referred to as an R0 code), which is a calculation result of a frame energy quantizing unit that calculates and quantizes frame energy in a speech encoding processing program, is monitored by a level adjusting unit. The level of the input audio data is controlled by increasing or decreasing the input audio scaling bit value or shifting the input audio data based on the monitoring result.

The input audio level control is realized by using an existing audio coding processing program and hardware. More specifically, the R0 code, which is the calculation result of the frame energy quantization means, is monitored by the level adjustment means,
Based on the monitoring result, the level control of the input audio data is performed by adjusting the level of the next data of the input audio data at the analog level.

The level adjusting means monitors the continuous state of the maximum value of the R0 code generated in the high-compression-rate audio encoding processing program, and uses the R0 code as an index indicating the continuous high-level input state of the audio. When the maximum value is continuous, the level adjusting means adjusts the level in the form of an external command issued to the high-compression-rate audio encoding processing program or to hardware as audio input means. This makes it possible to perform efficient input audio level control using an existing software program.

[0026]

Next, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of one embodiment of the present invention. In the figure, a level adjustment system according to an embodiment of the present invention includes a digital audio signal input means 1 and a VSELP (Vector Sum Exci).
and a digital coded voice signal output means 3 which is an input to a transmission path error coding means (not shown), which realizes a high compression rate voice coding such as ted Linear Prediction. It is configured.

The speech encoding device 2 has input speech control means 21
, A frame energy quantization means 22, a speech processing means 23 for performing processes such as linear prediction analysis, pitch extraction, sound source codebook search, and gain quantization, a code multiplexer 24, a shift register 25, and a level detection section 26.
And a control information storage unit 27.

The input voice control means 21 performs block loading processing on the digital voice signal input from the digital voice signal input device 1, and performs scaling for each data of one frame (= 160 samples) as a unit of voice processing. Calculate (normalized) bits.

Here, the block loading processing is a numerical value having a bit precision equal to or greater than the bit width of input data of a data memory or a multiplier and having a dynamic range correlated with each other as represented by, for example, an audio signal. This is a process of normalizing a data group by a common scale value (block scale value) and keeping the calculation accuracy of numerical data when saving the data in a memory or using it for an input of a multiplier.

The input voice control means 21 stores the calculated scaling bits in the shift register 25. The scaling bits stored in the shift register 25 are used in the frame energy quantization means 22 and the audio processing means 23.

The frame energy quantization means 22 calculates the power (= frame energy) of the input voice data for each frame, and quantizes and then encodes. That is,
The energy value R (0) is given by: R (0) = {[φ (0,0) + φ (NP, NP)] / 2
(NA-NP)}. Here, φ (j, k) indicates a covariance (autocorrelation) matrix, NA indicates the number of samples, and NP indicates the order of prediction.

The frame energy quantizing means 22 converts the energy value R (0) obtained by the above equation into the following equation based on the full scale Rmax (the full scale Rmax is defined as the square of the maximum sample amplitude). ). That is, RdB is calculated by the following equation: RdB = ₁₀ log ₁₀ [R (0) / Rmax], and this RdB is quantized to 32 levels to obtain a frame energy code (hereinafter referred to as an R0 code).

The above frame energy quantization means 2
2 for calculating the frame energy is described in “Digital Car Phone System Standard RCR STD-27D
1st Volume "(published by the Radio System Development Center, revised on June 27, 1995, pp. 594 to 604).

The voice processing means 23 performs a series of processes such as linear prediction analysis, pitch extraction, sound source codebook search, and gain quantization, which are the core of voice processing of high compression rate voice coding such as VSELP.

The code multiplexer 24 integrates the codes output from the frame energy quantizing means 22 and the sound processing means 23 to form coded sound data as input data for error correction processing and the like.

The shift register 25 is provided for the input voice control means 2
The scaling bit calculated in step 1 is stored. Also,
When the level detection unit 26 detects a continuous high level input, the shift register 25 subtracts “1” from the stored scaling bit value. The frame energy quantizing unit 22 and the audio processing unit 23 use the restored scaling bit value for calculation.

The level detecting section 26 detects that a frame whose R0 code has the maximum value (= 31) is continuously input more than a predetermined number of times. Level detector 2
When 6 detects a continuous high level input state, the shift register 25 subtracts "1" from the stored scaling bit value.

The control information storage section 27 stores control information (during level control / during level non-control) by the level detection section 26. These pieces of information are used to determine whether or not the change in the input level detected by the level detection section 26 causes level control.

FIG. 2 is a flowchart showing the operation of the speech coding apparatus 2 of FIG. The operation of the speech encoding device 2 will be described with reference to FIGS. First,
The input signal given from the digital audio signal input means 1 is supplied to the input audio control means 21.

The input voice control means 21 performs block floating processing (calculation of scaling bits for collectively normalizing the input voice data group) on the input signal, and stores the calculated scaling bits in the shift register 25. (Step S1 in FIG. 2).

The frame energy quantizing means 22 calculates the frame energy of the input voice, quantizes the frame energy from 0 to 31.
(Step S2 in FIG. 2).

The level detector 26 identifies the value of the control information (control state flag) stored in the control information storage 27 (step S3 in FIG. 2). That is, the control information in the control information storage unit 27 is Flag = 1 when the current state is under level control, and FL1 when the current state is not under level control.
ag = 0. The initial value of this Flag is 0.
And

The level detector 26 determines whether the value of the R0 code is the maximum value 31 even if the value of the flag is 1 or 0 (step S4 in FIG. 2).
S10).

When the value of the R0 code is less than the maximum value of 31 in step S4, the level detector 26 clears the value of the R0 counter (step S5 in FIG. 2). Note that R
The initial value of the 0 counter is set to 0.

If the value of the R0 code is the maximum value of 31 in step S4,
The value of the 0 counter is incremented (step S in FIG. 2).
6) It is determined whether the value of the R0 counter is equal to or greater than a predetermined value Xh (step S7 in FIG. 2). Here, the predetermined value Xh is a threshold value of the R0 code for detecting that the input voice is constantly at a high level.

When the value of the R0 counter exceeds a predetermined value Xh, the level detection unit 26 sets the value of Flag to 1 (step S8 in FIG. 2) and activates the level adjustment function. That is, the level detector 26 decrements the value of the scaling bit stored in the shift register 25 by 1 (−
1) Do (Step S9 in FIG. 2). Since the scaling bits that have been decremented and re-stored in the shift register 25 are used in the audio processing means 23, as a result, audio processing in which the input audio level is suppressed is performed.

The control state flag F is continuously output at a high level.
After the lag becomes 1, if the value of the R0 code further continuously indicates 31, the level detector 26 keeps the value of the R0 counter at the predetermined value Xh (step S11 in FIG. 2) and continues the level adjustment (step S11 in FIG. 2). FIG. 2 step S9).

On the other hand, the level detector 26 determines in step S10
When it is detected that the R0 code is less than the maximum value 31 while the control state flag Flag is 1, the value of the R0 counter is decremented by 1 (step S1 in FIG. 2).
2) Compare with a predetermined value Xl (step S13 in FIG. 2).
The predetermined value XI is a threshold value of the R0 code for detecting that the input voice is constantly at a low level.

When the value of the R0 counter is equal to or larger than the predetermined value X1, the level detector 26 continues the level adjustment (step S9 in FIG. 2). When the level detection unit 26 detects in step S13 that the low level input has continued and the value of the R0 counter has become less than the predetermined value Xl,
The value of Flag is set to 0, and the R0 counter is cleared. In this state, the level detection unit 26 stops the level adjustment function and does not perform the level adjustment.

As described above, the continuous high-level input state is determined by the level detection unit 26 using the R0 code calculated in the high-compression-rate audio encoding process, and the result is similarly used in the audio encoding process. By reflecting the result on the scaling bits used in the above, efficient input voice level control using an existing program can be performed. In addition, since conventional input control processing hardware is not required, the size and power consumption of the device can be reduced.

FIG. 3 is a diagram for explaining the level control operation according to one embodiment of the present invention. The operation of the level control according to the embodiment of the present invention will be described in detail with reference to FIGS. The predetermined values Xh and Xl can be set arbitrarily.

Before (1) in FIG. 3, the level of the input voice frame is less than 31 of the R0 code, and the processing at this time is a route of steps S3 → S4 → S5 in FIG. Subsequently, in (1) of FIG. 3, continuous input of frames in which the R0 code is the maximum value of 31 is started. The process in this case is performed in steps S3 → S4 → FIG.
The route is S6 → S7.

FIG. 3 (2) shows a state in which the R0 code becomes less than 31 again before the value of the R0 counter reaches the predetermined value Xh. The process at this time is performed again in step S3 in FIG.
Return to the route of S4 → S5.

FIG. 3 (3) shows that the value of the R0 counter has reached the predetermined value Xh due to the continuous high level input. The processing at this time is performed in steps S3 → S4 → S6 in FIG.
Go to the route of → S7 → S8 → S9. That is, F
lag is set to 1 and the scaling bit of the current frame stored in the shift register 25 is decremented by 1 as a level control function. After this, step S3 →
The processing is continued along the route of S10 → S11 → S9.

FIG. 3D shows a case where the R0 code temporarily becomes less than 31 in a continuous high level input state. The process at this time is performed in steps S10 → S12 in FIG.
It is executed in the route of → S13 → S9.

FIG. 3 (5) shows a state in which the R0 code becomes 31 again before the value of the R0 counter reaches the predetermined value X1. The process at this time is performed again in step S3 → S1 in FIG.
Return to the route of 0 → S11 → S9.

FIG. 3 (6) shows that the value of the R0 counter has reached the predetermined value Xl due to the continuous low level input. The process at this time is performed in steps S3 → S10 → S in FIG.
It goes to the route of 12 → S13 → S14 → S15. That is, Flag is set to 0, and the level control function is not performed. Thereafter, the processing is continued along the route of steps S3 → S4 → S5.

FIG. 4 is a block diagram showing the configuration of another embodiment of the present invention. In the figure, a level adjustment system according to another embodiment of the present invention is a digital audio signal input means 1.
, A voice coding device 4 and a digitally coded voice signal output means 3.

The speech encoding device 4 is provided with an input speech control means 41.
, A frame energy quantizing unit 42, an audio processing unit 43, a code multiplexer 44, a shift register 45, a level detecting unit 46, and a control information storage unit 47.

In the above configuration, the speech encoding apparatus 4 has the same configuration as that of the speech encoding apparatus 2 shown in FIG. 1 except that the level detection unit 46 directly controls the input speech control means 41. The operation is the same.

The level detecting section 46 detects that a frame in which the R0 code has the maximum value (= 31) is continuously input more than a predetermined number of times. Level detector 4
When a continuous high-level input state is detected, the input voice control means 41 performs level adjustment by shifting the input data group of the frame next to the analysis frame by one bit to the left.

FIG. 5 is a flowchart showing the operation of the speech coding apparatus 4 of FIG. The operation of the speech encoding device 4 will be described with reference to FIGS. In FIG. 5, steps S21 to S24 and S38 and S39 are performed.
The operation of each of the digital voice signal input means 1, voice coder 4 and digital coded voice signal output means 3 shown in FIG. Since the operation in each of the signal output means 3 (the operation shown in steps S1 to S4 and S16 and S17 in FIG. 2) is the same,
The description is omitted.

When the R0 code is less than the maximum value of 31 in step S24, the level detector 46 clears the value of the R0h counter (step S25 in FIG. 5). still,
The R0h counter is a counter for counting the number of high-level input frames, and its initial value is 0.

If the value of the R0 code is the maximum value of 31 in step S24, the level detector 46 sets
The value of the counter is incremented (+1) (step S26 in FIG. 5), and it is determined whether or not the value of the R0h counter is equal to or more than a predetermined value Xh (step S27 in FIG. 5). The predetermined value Xh is a threshold value of the R0 code for detecting that the input voice is constantly at a high level.

When the value of the R0h counter is equal to or more than the predetermined value Xh, the level detector 46 sets the value of Flag to 1 (step S28 in FIG. 5) and clears the value of the R0h counter (step S29 in FIG. 5). . Subsequently, the level detector 46 activates the level adjustment function by determining the value of Flag in step S36.

In one embodiment of the present invention, the shift register 25
The level is adjusted by decrementing the value of. On the other hand, in another embodiment of the present invention, the determination result of the level detection section 46 is fed back to the input voice control means 41, and the input data group of the frame next to the analysis frame is shifted by one bit to the right to adjust the level. (Step S37 in FIG. 5).

The level detector 46 detects that the value of the R0 code is a predetermined value R0 Low after the control state flag Flag is set to 1 after the high level input continues (FIG. 5).
Step S30). The predetermined value R0 Low is an arbitrary R0 code value for recognizing that the input voice is at a low level.

The level detector 46 outputs R0 by a low level input following a continuous high level input state (Flag = 1).
When code <R0 Low (step S3 in FIG. 5)
0), the value of the R01 counter is incremented (step B12), and it is determined whether the value of the R01 counter is equal to or more than a predetermined value X1 (step S33 in FIG. 5). The predetermined value Xl is a threshold value of the R0 code for detecting that the input audio signal is constantly at a low level.

When the value of the R01 counter is equal to or larger than the predetermined value X1, the level detector 46 sets the value of the Flag to 0 (step S34 in FIG. 5) and clears the value of the R01 counter (step S35 in FIG. 5). Subsequently, the level detector 4
6 stops the level adjustment function by determining the value of Flag in step S36.

As described above, in another embodiment of the present invention, in addition to the effect of the embodiment of the present invention in which the R0 code is used, the level of the input audio data is directly adjusted without using the shift register 45. In this configuration, efficient input voice level control can be performed only by using an existing program configuration as it is and adding only a conditional branching routine to the program configuration.

FIG. 6 is a diagram for explaining an operation of level control according to another embodiment of the present invention. These FIGS. 4 to 6
The operation of the level control according to another embodiment of the present invention will be described in detail with reference to FIG. FIG. 6 shows how the R0 code, the control state flag Flag, the value of the R0h counter, and the value of the R01 counter change over time. The values of R0 Low and the predetermined values Xh and Xl in FIGS. 5 and 6 can be set arbitrarily. The timings (11) to (16) in FIG. 6 are shown in (1) to (6) in FIG.
Timing.

FIG. 6 is different from FIG. 3 in that the level adjustment is performed for which analysis frame and the method 2 for detecting the low level input in the continuous high level input state.
Is a point. The other details are the same as those of the embodiment of the present invention, and the description thereof is omitted.

First, for a frame to be subjected to level adjustment, in one embodiment of the present invention, "level adjustment" in FIG. 3 is performed on the current frame. On the other hand, in another embodiment of the present invention, the “level adjustment” in FIG. 6 is performed for the next frame (step S37 in FIG. 5).

In a method of detecting a low-level input in a continuous high-level input state, in one embodiment of the present invention, the state of the R0 counter for counting the number of frames of the high-level input (R0 code = 31) is determined according to the state. There are provided predetermined values Xh and Xl for judging. On the other hand, in another embodiment of the present invention, the high level input (R0 code = 3)
A predetermined value Xh for determining a steady state is provided for the value of the R0h counter for counting the number of frames in 1),
In addition, a predetermined value Xl for determining a steady state is provided for a value of an R01 counter for counting the number of frames of a low-level input (R0 code = R0 Low).

FIG. 7 is a block diagram showing the configuration of another embodiment of the present invention. In the figure, a level adjustment system according to another embodiment of the present invention includes digital audio signal input means 5.
, An audio encoding device 6 and digitally encoded audio signal output means 3.

The speech encoding device 6 includes an input speech control means 61
, A frame energy quantizing unit 62, an audio processing unit 63, a code multiplexer 64, a shift register 65, a level detecting unit 66, and a control information storing unit 67.

In the above configuration, the configuration of the speech encoding apparatus 6 is the same as that of the speech encoding apparatus 4 shown in FIG. 4 except that the level detecting section 66 directly controls the digital speech signal input means 5. The operation is the same.
In another embodiment of the present invention, it is assumed that the digital audio signal input means 5 has a level adjusting means (not shown) by external command control or the like.

The level detecting section 66 determines whether or not a frame in which the R0 code has the maximum value (= 31) is continuously input more than a predetermined number of times. When detecting a continuous high-level input state, the level detection unit 66 performs level adjustment by directly controlling the level adjustment unit included in the digital audio signal input unit 5.

FIG. 8 is a flowchart showing the operation of the speech encoding device 6 of FIG. The operation of the speech encoding device 6 will be described with reference to FIGS. The operations of the digital audio signal input means 5, the audio encoding device 6, and the digitally encoded audio signal output means 3 shown in FIG. 8 except for step S57 are the same as those of the digital audio signal of the other embodiment of the present invention. The operations are the same as those of the input means 1, the voice coding device 4, and the digitally coded voice signal output means 3, and the description thereof will be omitted.

In the embodiment of the present invention, when a continuous high level input is recognized, the value of the shift register 25 is decremented by 1 as a level adjusting function (step S in FIG. 2).
9). Further, in another embodiment of the present invention, in such a case,
The result is fed back to the input voice control means 41,
The level adjustment is performed by shifting the input data group of the frame next to the analysis frame to the right by one bit (step S37 in FIG. 5).

On the other hand, in another embodiment of the present invention, the result is fed back to an external command of the digital audio signal input means 5, and the input audio level is adjusted at the analog level for the next frame of the analysis frame. (Step S57 in FIG. 8).

As described above, in another embodiment of the present invention, in addition to the effect of the embodiment of the present invention in which the R0 code is used, the level adjusting means by external command control or the like of the digital audio signal input means 5 is used. By using this function in conjunction with, it is possible to effectively utilize the hardware functions and to perform efficient input audio level control without deteriorating the coding performance using the existing program configuration as it is. it can.

The operation of the level control according to another embodiment of the present invention can be described with reference to FIG. That is, the detailed contents of the operation of the level control according to another embodiment of the present invention are the same as those of the other embodiments of the present invention, and the description thereof will be omitted.

In one embodiment of the present invention, "level adjustment" in FIG. 3 is performed on the current frame. On the contrary,
In another embodiment of the present invention, "level adjustment" in FIG. 6 is performed on the next frame (step S57 in FIG. 8). At this time, the level adjustment method is the digital audio signal input means 5.
Depends on the function and configuration of level adjustment by input of an external command.

As described above, R0 calculated in the high-compression-rate speech encoding process executed by the software program
Using the code as a parameter for input level control,
By monitoring the 0 code, a continuous high-level input state is determined, and the result is reflected in the scaling bits used in the voice coding process, thereby providing a means for monitoring the input level to the voice code. It is possible to perform efficient input sound level control using an existing software program without separately providing the same outside the coding device. This eliminates the need for conventional hardware for input control processing, thereby reducing the size and power consumption of the device.

Also, by monitoring the R0 code, a continuous high level input state is determined, and based on the result, the level is directly adjusted for the input audio data without passing through the shift register 45. With this configuration, it is possible to perform efficient input voice level control by adding a conditional branch routine to an existing program configuration including scaling bit calculation.

Further, a continuous high-level input state is determined by monitoring the R0 code, and based on the result,
By using the level adjustment means by external command control or the like of the digital audio signal input means 5 in cooperation with the R0 code, the hardware functions can be effectively utilized,
In addition, it is possible to perform efficient input audio level control without deteriorating the encoding performance using the existing program configuration as it is.

[0088]

As described above, according to the present invention, in a speech encoding apparatus for encoding input speech data by high-compression-rate speech encoding processing, the speech data is calculated during the high-compression-rate speech encoding processing and input. Analysis of audio data The frame energy code indicating the power of each frame is monitored, and based on the monitoring result, the level control of the input audio data is performed in the high-compression-rate audio encoding processing, whereby the high-compression-rate audio encoding processing is performed. As a result, there is an effect that efficient input voice level control can be performed using existing software and hardware resources.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

FIG. 2 is a flowchart illustrating an operation of the speech encoding device in FIG. 1;

FIG. 3 is a diagram for explaining an operation of level control according to one embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of another embodiment of the present invention.

FIG. 5 is a flowchart showing an operation of the speech encoding device in FIG. 4;

FIG. 6 is a diagram for explaining an operation of level control according to another embodiment of the present invention.

FIG. 7 is a block diagram showing a configuration of another embodiment of the present invention.

FIG. 8 is a flowchart showing an operation of the speech encoding device in FIG. 7;

FIG. 9 is a block diagram showing an example of a configuration of a conventional example.

FIG. 10 is a diagram showing an example in which the level adjustment system of FIG. 9 is applied to a high-compression-rate audio encoding device.

FIG. 11 is a block diagram showing another example of the configuration of the conventional example.

FIG. 12 is a diagram showing an example in which the system of FIG. 11 is reflected on a high-compression-rate audio encoding device.

[Explanation of symbols]

 1,5 Digital audio signal input means 2,4,6 Audio encoding device 3 Digitally encoded audio signal output means 21,41,61 Input audio control means 22,42,62 Frame energy quantization means 23,43,63 Audio Processing means 24, 44, 64 Code multiplexer 25, 45, 65 Shift register 26, 46, 66 Level detector 27, 47, 67 Control information storage

Claims (5)

[Claims] 1. An audio encoding apparatus for encoding input audio data by a high-compression-rate audio encoding process, comprising: Monitoring means for monitoring a frame energy code indicating electric power; and means for controlling the level of the input audio data in the high-compression-rate audio encoding process based on the monitoring result of the monitoring means. Audio coding device. 2. The level control unit according to claim 1, wherein the level control unit controls the level of the input voice data by increasing or decreasing an input voice scaling bit value used in the high compression rate voice coding process based on a monitoring result of the monitoring unit. 2. The speech encoding device according to claim 1, wherein 3. The level control unit according to claim 1, wherein the level control unit controls the level of the input audio data by shifting the input audio data based on a monitoring result of the monitoring unit. A speech encoding device according to claim 1. 4. A means for performing the level control, wherein the input sound data is reflected by a command for instructing a level adjustment at an analog level with respect to the next data of the input sound data on the basis of a monitoring result of the monitoring means. 2. The speech encoding apparatus according to claim 1, wherein level control is performed. 5. The level control unit, wherein the frame energy code is set to be equal to or more than a first predetermined value and equal to or less than a second predetermined value by the monitoring unit. 5. The speech encoding apparatus according to claim 1, wherein a level control of the input speech data is performed when the speech data is continuous.