JPH0833743B2 - Waveform synthesis method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial application] The present invention relates to waveform synthesis using the ADPCM method, and in particular, by synthesizing one representative waveform segment and repeatedly using the representative waveform segment, a waveform is obtained. The present invention relates to a waveform element synthesizing method for synthesizing.

[Conventional technology]

Conventionally, PCM method, ADPCM method, etc. are known as speech synthesis methods. Especially, ADPCM method is very effective in terms of data volume compression because it can synthesize waveforms with half or less data volume compared to PCM method. It is a method. The ADPCM method is a data compression method that utilizes the strength of correlation between adjacent samplings of voice. The ADPCM method attempts to code and quantize the difference between amplitude values at consecutive sampling points, and adaptively changes the quantization width to reduce noise. As the quantization width used for quantization, the next quantization width is predicted and used based on the current quantization width and the ADPCM code. An example of this prediction calculation is Δ _nt1 = Δ _n × M
(L _n ), L _n is the nth ADPCM code data, Δ _n is the size of the quantization width for the nth sampling point, and M (L _n ) is the prediction coefficient and the prediction coefficient is the current quantization width. The value multiplied by is used as the quantization width of the next sampling point. When the voice level is low, the prediction coefficient is smaller than 1 and the quantization width is small. On the contrary, when the voice level increases, the prediction coefficient becomes larger than 1 and the quantization width increases. However, if the quantization width becomes too large, noise increases on the contrary, so it is necessary to determine the upper limit value of the quantization width. Further, if the value of the quantization width becomes 0, it will remain 0 even if any prediction coefficient is applied next time, so the lower limit of the quantization width is also necessary. The method of performing the processing including the multiplication with the prediction coefficient in the ROM as a table of the quantization width data and performing the prediction calculation using the ROM is the most suitable for the LSI.

Quantization table is ROM with ADPCM code and quantization width pointer
The address of is specified and the table of difference value data is created. Increasing the value of the quantization width pointer means multiplying the quantization width by a prediction coefficient greater than 1 if the difference value for inserting the ROM according to the quantization width pointer is set to a value that increases at a constant rate. Meaning, the multiplication of the prediction coefficient can be replaced by the addition and subtraction for the quantization width pointer.

[Problems to be solved by the invention]

In general, the voice waveform is divided into the voice noise of the consonant part, the tone waveform of the vowel part, and the connecting part between them.In particular, the tone waveform of the vowel part has almost the same period, but the waveform of the same shape continuously changes its shape gradually. It is normal to arrange them while changing the envelope, which is the maximum amplitude of one waveform. In addition, for instrument sounds, alarm sounds, and sound effects, the same waveform is arranged in a row with a constant cycle with the envelope changed. In the waveform segment synthesis method, if several continuous waveforms are almost the same waveform, one waveform is selected as a representative waveform and repeatedly used to compress the data amount of speech synthesis. When trying to compress the amount of data for speech synthesis by repeating such waveforms in the ADPCM system, ADPC
The M method encodes the voice at the Nyquist frequency and encodes the difference value of the amplitude value of the voice waveform between adjacent sampling points with an appropriate quantization width, and the quantization width is the difference value of each sampling point. This is a method of adaptively changing according to the size. Therefore, when the representative waveform segment is synthesized by the ADPCM method, the quantization width is not constant within one waveform, and the quantization width at each sampling point depends on the immediately preceding ADPCM data. Also, in the ADPCM system, the sampling frequency is usually 5 kHz to 8 kHz due to the relationship between sound quality and bit race.

When synthesizing the representative waveform segment using the ADPCM method as described above, the pitch period of the actual original waveform and the pitch period of the synthesized waveform cannot be perfectly matched, and they are synthesized due to the influence of the pitch period error. The amplitude of the last sampling point of the waveform is not zero. Even when the pitch period of the original sound and the pitch period of the synthesized waveform completely match, the amplitude value at the last sampling point is the amplitude value at the previous sampling point plus the difference value determined by the quantization width and ADPCM code. Therefore, depending on the value of the quantization width at the immediately preceding sampling point, the amplitude value at the last sampling point in one waveform may not be 0.

Therefore, when synthesizing the representative waveform segment using the conventional ADPCM method as it is, the amplitude value at the last sampling point of the representative waveform segment cannot be set to 0. Therefore, such a representative waveform segment is used multiple times. If repeated, the amplitude center of the composite waveform will change. This is ADPCM
Since the method is a differential encoding method in which the difference value is basically added to the amplitude value of each sampling point to obtain the amplitude value of the next sampling point, the final amplitude value of one waveform is 0.
Otherwise, the error will be accumulated and the amplitude center of the composite waveform will change. If the amplitude center of the synthesized waveform changes, the D / A converter that converts the waveform data to voice may overflow by the amount of the change of the amplitude center, and if it overflows, a serious drawback will occur that a normal voice waveform cannot be synthesized. . Further, since it is a differential encoding method, when the same ADPCM code data is used repeatedly, there is a fatal drawback that the same amplitude value is only repeated and an envelope cannot be added for each waveform.

[Means for solving problems]

The waveform synthesis method of the present invention sequentially accumulates difference values.
The ADPCM waveform synthesis method has means for storing and reading the quantization width pointer, and means for multiplying the amplitude value to be added and accumulated by the envelope data, and AD for one waveform.
PCM code data is used repeatedly, the difference value is added and accumulated at the beginning of the repeated waveform, the amplitude value register is reset to 0, the value of the quantization width pointer stored in advance is read, and the added and accumulated amplitude value is read. Has a function of synthesizing waveforms by multiplying envelope data set for each number of waveforms designated in advance.

〔Example〕

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a circuit configuration diagram of the first embodiment of the present invention.
The M code data is latched in the L register, and the address of the difference value data table ROM is designated by the content of the L register.
The value obtained by decoding the contents of the L register and the value of the A register containing the current value of the quantization width pointer are added, and the value of the quantization width pointer at the next sampling point is set to A.
The address of the difference value data table ROM is specified by the contents of the register A register. Insert a limiter so that the value of the quantization width pointer does not exceed the specified value. The contents of the A register containing the value of the quantization width pointer are transferred to and stored in the A'register so that the contents of the A'register can be read out to the A register when necessary.
The difference value data is output from the difference value data table ROM, and the contents of the X register containing the previous waveform amplitude value at each sampling point are added to the difference value data, and the next waveform amplitude value is entered in the X register. ing. The contents of the X register and the difference value data are added and the next waveform amplitude value is stored in the X register. The contents of the X register can be set to 0 by a reset signal. The value of the Z register obtained by shifting the 1-bit contents and the contents of the X register containing the waveform amplitude value are added by the contents of the E register containing the envelope data. The contents of the Z register become the amplitude data of the final output waveform and are converted into analog signals by the D / A converter.

A flowchart for waveform synthesis is shown in FIG.

4 and 5 show waveforms obtained by sampling one waveform into 16 divisions with an amplitude of 7 bits. Select one waveform from the similar waveform repetition of the original sound, normalize the amplitude so that the maximum amplitude is 40H, and then set the value of the quantization width pointer to the optimum value.
Create ADPCM coded data. As shown in Fig. 6, the envelope data of the entire waveform is coded in 4 bits for each envelope division section. However, in the case of musical instrument sounds, the pitch frequency hardly changes. Are uniformly divided. If the number of waveforms for each section is set separately, it is also possible to temporally delimit it with an arbitrary width. When creating ADPCM code data and envelope data for one waveform, an appropriate combination can be selected so that the amplitude of the finally synthesized waveform is the same as the original sound waveform.

When waveform synthesis is started, first the initial value is set, the number of envelope repetition waveforms, the number of waveform repetitions, the 4-bit A'register that stores the value of the quantization width pointer, and the initial value of the 4-bit E register that stores the envelope data. Are set according to the waveform to be combined.

At the beginning of synthesis of repeated waveforms, the amplitude value register X is set to 0, and the contents of the A'register containing the value of the quantization width pointer are transferred to the A register. Next, the ADPCM code data is latched in the L register for each sampling point, the address of the differential value data table ROM is specified by the L register and the A register of the quantization width pointer, and the differential value data table R is specified.
The output of OM and the X register are added and the result is stored in the X register. When the least significant bit of the E register containing the envelope data is 0, the Z register is set to 0, and when it is 1, the value of the X register is transferred to the 7-bit Z register. Next, if the 2-bit date from the bottom of the E register is 0, shift the Z register to the smaller 1-bit value.
The value obtained by shifting the register so that the 1-bit value becomes smaller and the value in the X register are added and stored in the Z register. By repeating this operation for the third bit from the lower bit of the E register and the most significant bit, the value obtained by multiplying the contents of the X register and the E register by the Z register is finally obtained. For example, if the X register value is 40H and the E register value is FH, the Z register value is 78H. When the value of the X register or the E register is 0, the value of the Z register is 0. An analog waveform is output by the D / A converter using the 7-bit value of this Z register as amplitude data. After performing the above calculation processing for one sampling point, A is obtained by decoding the value of the L register latching the ADPCM code data.
To limit the result of adding or subtracting the value of the quantization width pointer of the register from -1 to +3 within the range of 0 to FH and use it as the new quantization width pointer in the A register at the time of the arithmetic processing of the next sampling point. Stored in.

In this way, the difference value is read from the ADPCM code data and the value of the quantizer width point value, the value obtained by cumulative addition is multiplied by the envelope data, and the analog waveform is output by the D / A converter.
DPCM code Code data is decoded and the value of the quantization width pointer is moved repeatedly for 16 sampling points for one waveform. After outputting the data of 16 sampling points for one waveform, the number of waveforms is counted and the match with the number of repeated waveforms in the envelope is detected. If they do not match, the envelope data is not updated. Store in E register. After detecting a match with the number of repeated waveforms in the envelope, a match with the number of repeated waveforms is detected. If there is no match, the value of the amplitude value register X is set to 0, the contents of the A'register are read into the A register, and one waveform is read again. Waveform data is output, and if they match, the initial value of the next waveform is set and the next waveform is output.

By performing such arithmetic processing, the same waveform can be repeated to synthesize a waveform with an envelope as shown in FIG. In FIG. 7, the number of repeated waveforms of the envelope is 3, the number of repeated waveforms 1 is 8, and the envelope data is a combined waveform when changed to 2/4/6/8/7.

[Embodiment 2] FIG. 2 is a circuit diagram of a second embodiment of the present invention. Same as the first embodiment up to the point where the address of the difference value data table ROM is specified by the ADPCM code data of the register and the quantization width pointer of the A register, and the difference value output of the difference value data table ROM is cumulatively added to the X register. In the configuration, the envelope data of the E register and the amplitude value data of the X register are directly multiplied and stored in the Z register. Any multiplication circuit can be used as long as the number of bits in each register is the same, the signal flow becomes simple when viewed from the circuit configuration, and the result can be obtained faster than repeating addition and register shift. By following the waveform synthesizing flow shown in FIG. 3, the repeated waveform with the envelope shown in FIG. 7 can be output in the same manner as in the first embodiment.

〔The invention's effect〕

As described above, the present invention is based on the amplitude 0 as the difference value of the first sampling point of the repeated waveform,
The value of the quantization width pointer at the first sampling point of the repeated waveform after the second time is used as the initial value of the repeated waveform by using the value of the first time, and the ADPCM code data of the representative waveform segment is used as is in the ADPCM method. However, the same waveform can be repeatedly synthesized, and the D / A converter outputs the result of multiplying the value obtained by cumulatively adding the difference value and the envelope data, thereby making it possible to synthesize a repeating waveform with varying envelope. It is possible.

In general, the ADPCM method can perform high-quality voice synthesis, but if you try to synthesize the waveform of an instrument sound, an alarm sound, a sound effect, etc. with the ADPCM method, the synthesis time is long because the same waveform is repeated many times and the entire time is long. The amount of ADPCM code data is required to be proportional to, and if the waveform has a changing envelope, the discrete difference value is selected by moving the quantization width pointer.
The shape of each waveform may change and the timbre may differ,
With the ADPCM method, it was difficult to synthesize waveforms such as instrument sounds, warning sounds, and sound effects other than voice. However, if the configuration of the present invention is used, high-quality voice synthesis can be performed by the conventional ADPCM method, and more than 10 waveforms of the same waveform are often repeated and the entire time is long. Waveform synthesis
It can be done with a very small amount of data, less than 1/10,
It is very effective in compressing the amount of combined data. It is obvious that the number of bits of each register described in this embodiment is related to the accuracy of the composite waveform and may be set to any value.

[Brief description of drawings]

FIG. 1 is a circuit diagram showing a first embodiment of the waveform synthesizing method of the present invention, FIG. 2 is a circuit diagram showing a second embodiment of the waveform synthesizing method of the present invention, and FIG. 4 is a waveform diagram showing sampling points of waveform 1, FIG. 5 is a waveform diagram showing sampling points of waveform 2, FIG. 6 is a waveform diagram showing envelope data, and FIG. 7 is a synthesized output. It is a waveform diagram showing a waveform. 101,201 ... L register that latches ADPCM code data,
102,202 …… Decoder that decodes the contents of the L register and determines the movement amount of the quantization width pointer, 103, 203 …… Limiter that prevents the value of the quantization width pointer from exceeding the specified range, 104, 204 …… Quantization width pointer Register 105 for storing the value of A, 105, 205 ... A'register for storing the contents of the A register and for reading the contents stored in the A register, 106, 206 ... Difference value data table ROM, 107, 207 ...
... X register that stores the amplitude value that cumulatively adds the difference value data sequentially, 108,208 ... Z that stores the amplitude value of the output waveform
Register, 109,209 ... E to enter envelope data
register.

Claims (1)

[Claims] 1. A quantization width in an ADPCM waveform synthesis method for selecting difference value data required for synthesis with ADPCM code data and a value of a quantization width pointer and sequentially accumulating the difference values to synthesize a waveform. A means for storing the value of the pointer and a means for multiplying the amplitude value to be added and accumulated and the envelope data are used, and the difference value data for one waveform is repeatedly used, and the difference value accumulated at the beginning of the repeated waveform. Is reset, the cumulative addition data is generated from the difference value data read based on the value of the quantization width pointer stored in advance, and this is multiplied by the envelope data. Waveform synthesizing method characterized by synthesizing waveforms by