JPS5851307B2 - Hakei Hatsei Souchi - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a waveform generator that generates a waveform signal through calculations based on input data.

2. Description of the Related Art Waveform memories have conventionally been used as devices that generate waveform signals by sequentially sending out waveform amplitudes in response to input.

This waveform memory divides one cycle of the waveform into a necessary number of parts, stores all the amplitude values of the waveform corresponding to each division point in advance, and reads out the amplitude values of each division point sequentially according to address data. This is what I did.

However, in order to obtain an accurate waveform, it is necessary to divide one period into a considerably large number, and the address data for reading out this period requires at least about 10 to 11 bits.

For example, if the address data is 11 bits, the number of divisions of one waveform cycle is 2048, and the memory must store the waveform amplitude values of each of the 2048 division points, which requires a fairly large capacity memory.

Furthermore, when using a memory that stores amplitude values digitally, such as a read-only memory, the amplitude value data requires at least 7 to 8 bits, and the total number of memory bits exceeds at least 10,000 bits. Because of this large capacity, conventional devices had various drawbacks such as slow response speed required for reading out waveforms, high memory power consumption, and high memory costs. .

This invention was made to eliminate the above-mentioned drawbacks.
It is an object of the present invention to realize miniaturization, cost reduction, high responsiveness, etc. of the device by sequentially calculating the amplitude values of a waveform by calculation without using a large-capacity memory as in the past.

To achieve this objective, the present invention provides a first memory in which a basic amplitude value A for each sample point is stored in advance, and a difference value (
A second memory storing B-A) is provided, and a basic amplitude value A and a difference value (B-A) are generated from the first memory and second memory in response to an integer value of input data. ,
Function X(c
) and the above values A and (BA) to form the formula A+(BA
) X

First, the basic principle of this invention will be explained based on FIG.

Calculated basic amplitude values A and B corresponding to amplitude values at any two points
When y is given, in order to interpolate the amplitude value y between these two points, it is sufficient to operate the following equation (1).

Y "A+ (B - A) X X(c)...
... (1) However, X(C) is an arithmetic function, C is decimal part data, and the above formula (1) is X(c) when C=O.
= O, and when C=1, the condition that X(c)=1 is satisfied.

Therefore, when O≦C≦1, the arithmetic function X(c) is O≦X(C
)≦1.

Further, this calculation function X(c) determines the waveform shape between A and B, and any function form can be used depending on this waveform shape.

Since the calculation basic amplitude values A and B are determined by the integer part data, the decimal part data C between A and B gradually increases between 0 and 1.

In addition, input data including integer part data and decimal part data is O.
The waveform signal gradually increases from 0 to 0, and when it reaches a predetermined value, returns to O again and repeats the gradual increase, and the period of the generated waveform signal is determined according to this repetition period.

Therefore, the decimal part data C itself also satisfies the conditions of the function X(c), and when it is set to the linear function form X(c)=C, linear interpolation is performed.

Therefore, if the calculation basic amplitude value A and the difference value (B-A) are obtained in advance as calculation basic values based on integer data, and the calculation of equation (1) is executed based on this calculation basic value and decimal part data. , the amplitude value y is calculated sequentially according to changes in the decimal part data.

The invention will now be described in detail with reference to the embodiments of the accompanying drawings.

FIG. 2 is a block diagram showing an embodiment of the waveform generator according to the present invention, in which the input data ID is a plurality of bits (for example, 10
bit) digital data, and this input data ID
The amplitude value Y at each point in time of the waveform is sequentially calculated and output according to the waveform.

The upper digits (for example, the upper 6 bits) of this input data ID are added as integer data M to the operation basic value generation section 1, and the lower digits (for example, the lower 4 bits) are added as decimal data L to the operation section 2. ing.

The calculation basic value generation unit 1 generates a calculation basic value memo 1 which stores in advance the basic amplitude values of each division point obtained by roughly dividing one period of an arbitrary desired waveform, and a calculation basic value memo 1 which is based on the amplitude values stored in the memory 11. It has a difference value memory 12 that stores in advance the difference between two basic amplitude values (hereinafter referred to as difference values) of adjacent dividing points (in the order of generation), and calculates the difference according to integer data M. Basic amplitude value A and difference value B as calculation basic value
-A is configured to be sent to the calculation unit 2.

. The integer data M is added to the memories IJ 11 and 12 via decoders 13 and 14, respectively, and serves as an address for reading out the recorded contents of the memories 11 and 12.

The address for reading the amplitude value in the calculation basic value memory 11 and the address for reading the difference value between the amplitude value to be read next and the amplitude value from the difference value memory 12 are the same. ing.

Therefore, the difference value between the amplitude value read out from the memory 11 and the amplitude value at the next division point is read out from the memory 12, and the output of the memory 11 becomes the calculated basic amplitude value A. , and the output of the memory 12 is the difference value B −
Corresponds to A.

The calculation section 2 executes the calculation of the above-mentioned formula (1) based on the calculation basic amplitude value A, the difference value B-A, and the decimal data L. Through this calculation, the basic amplitude value A and the difference value B-
Amplitude value Y obtained by interpolating the waveform amplitude value between the two basic amplitude values substantially given by A
Output sequentially.

The decimal data L is added to the function section 21, and the value of a required arithmetic function X(c) is output in accordance with the decimal data L.

The output of this function section 21 becomes the multiplier input of the multiplier 22.

On the other hand, the difference value (B-A) becomes the multiplicand input of the multiplier 22.

The multiplication of (B − A ) X X (c) is performed in the multiplier 22 , and the multiplication result is added to the adder 23 .

The calculation basic amplitude value A is added to the other input side of the adder 23, and "A+(B-A)×X(c
)", the amplitude value Y is output as the calculation result.

Assuming that the decimal data L is the lower 4 bits of the human data ID, the integer data M maintains a constant value (calculation basic value A
1 difference value B-A does not change), the decimal data L is oo
It gradually changes to 16 values from oo to 1111.

Therefore, while the calculated basic amplitude value A1 and the difference value B-A maintain a constant value, 16 amplitude values Y are sequentially calculated and output in accordance with changes in the decimal data L, forming a waveform signal.

In addition, the calculation basic amplitude value A1 difference value B-A is the human data ID
There are 64 changes during one synchronous period of gradual change, so 64X
One waveform synchronization is formed by 161,024 amplitude values.

This waveform is obtained by finely interpolating between the basic amplitude values of the waveform roughly stored in the calculation basic value memory 11, and has an accurate waveform shape.

Note that when performing linear interpolation, the function section 21 is not necessary, and the value of the decimal data L may be directly input as a multiplier to the 5 multiplier 22.

Of course, each calculation device of the memo 1JIL, 12 and the calculation unit 2 can be configured as either digital or analog, and according to this configuration, calculates and outputs an analog amplitude value Y or a digital amplitude value Y. be able to.

As explained above, according to the present invention, the calculated basic amplitude value A and the difference value B - person are determined based on the integer data, and the waveform amplitude is calculated based on the calculated basic amplitude value A, the difference value B - A, and the decimal data. Since the waveform signal is generated by calculating and outputting the values, the memory is used only to obtain the calculated basic amplitude value and the difference value, and the capacity is extremely small.

Moreover, the overall size of the device has been significantly reduced, and since the amplitude value can be obtained through calculation, the long readout time required in the past is not required, the response speed can be increased, and power consumption has been reduced. It has various excellent effects such as considerably reducing manufacturing costs.

In addition, especially in this invention, Su? Since the difference value between two basic amplitude values having successive pulling orders is stored in the memory in advance, the configuration of the calculation section is simplified when executing the calculation of the above-mentioned equation (1).

Moreover, since the difference value memory can generally be configured with a relatively small number of bits, it is advantageous in terms of device scale, cost, calculation speed, etc., and has particularly significant advantages when configured in a digital format.

In addition, although the waveform shape is obtained by interpolation calculation, it is comparable to the waveform shape obtained by conventional large-capacity memory, and it can be used for various other waveforms such as musical sound waveforms in electronic musical instruments. This invention can be applied to the occurrence of.

[Brief explanation of the drawing]

Figure 1 is a graph for explaining the invention in detail, Figure 2 is a graph for explaining the invention in detail.
The figure is a block diagram showing one embodiment of a waveform generator according to the present invention. 1... Calculation basic value generation section, 2... Calculation section, 11... Calculation basic value memory, 12...
...Difference value memory, 20...Function section, 22...
...Multiplier, 23...Adder.

Claims (1)

[Scope of Claims] 1. In a waveform generator that sequentially generates waveform amplitude values based on input data consisting of an integer part and a decimal part and whose values change sequentially, Reading out the basic amplitude value A corresponding to the integer value of the human data from the stored first memory, and from the second memory in which the difference values between two basic amplitude values whose sampling order is consecutive are stored in advance, respectively. calculation basic value generating means for reading out a difference value (B-A) between the basic amplitude value A and the basic amplitude value B of the next sample point of the amplitude value A; 'A+ (B-A) X(c)", which sequentially calculates and outputs the waveform amplitude values between the basic amplitude values A and B according to the change in the decimal value of the input data, A waveform generator that generates waveform amplitude values densely.