JPH0711753B2 - Tone control method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a tone color control method suitable for application to a so-called electronic musical instrument that produces various musical instrument sounds by an electronic circuit, and in particular, a tone color during strong playing is a natural musical instrument. We propose a tone color control method that approximates the tone color.

"Prior art" Generally, it is known that the natural instrument sound has a different power spectrum in the stationary part when playing strongly and when playing weakly, and has a richer spectrum when playing strongly than when playing weakly. .

Therefore, in the conventional electronic musical instrument, the high-order component is attenuated by low-pass filtering the strong sound, and this is used as an approximation of the weak sound.

"Problems to be solved by the invention" The conventional method has an advantage that a tone color that gently shifts between strong and weak tones can be obtained by changing the cut-off frequency of the low pass filter. ing.

However, there is a drawback that the change in tone color from the strong sound to the weak sound and the change from the weak sound to the strong sound is monotonous, which is unsatisfactory in terms of hearing.

[Means for Solving the Problem] This invention proposes a tone color control method for producing a strong sound by passing a weak sound through a non-linear function, which is the reverse of the conventional method of making a weak sound from a strong sound. It is a thing.

That is, the non-linear function used in the tone color control method proposed by the present invention has a characteristic that it works as a linear function for inputs up to a predetermined level and as a non-linear function for inputs exceeding a predetermined level. Is set appropriately to obtain strong tones of various tones.

According to the tone color control system of the present invention, since a higher-order spectrum component is generated as a non-linear distortion of a weak tone, the tone color can be controlled by simply changing the level of the input signal without the need for a filter.

In addition, a complicated timbre change can be obtained in the process of transitioning between the strong sound and the weak sound.

"Embodiment" The weak sound of an electronic musical instrument is x (t) (Xminx (t) Xmax)
And After this weak sound x (t) is multiplied by α (| α | 1), it is distorted by passing it through the nonlinear function y = f (x) shown in FIG. 1, and y (t) = f (α · x (T)) ... (1). α is a parameter for controlling the timbre.

Here, the nonlinear function f (x) is defined in the range of XminxXmax as shown in FIG. 1, and is linear in the range of χmin | x | χmax.

In other words, y (t) = c · α · x (t) (2) (Xmin | x | Xmax) c is the slope of the linear part.

When α is small, χmin | α · x (t) | χmax, so that nonlinear distortion does not occur in the output according to the equation (2). An output obtained by simply multiplying by c · α is obtained. In other words, only the volume changes, and the timbre remains weak.

When α is larger than a certain value, αx (t) exceeds the linear range (χmin | x | χmax) described above and the nonlinear range Xmi
Since it falls within n to Xmax, the output is distorted and the timbre changes.

Since this higher-order harmonic component is generated by this non-linear distortion, the timbre becomes brighter and a strong approximation effect is obtained.

In this way, according to the tone color control method of the present invention, it is possible to obtain a signal having a tone color close to a strong tone by changing α.

FIG. 2 shows a specific device configuration when the tone color control method according to the present invention is put into practical use.

In FIG. 2, the weak performance input signal x (n) is applied to the input terminal 1. This weak input signal x (n) is input to the multiplier 2, and is input to the nonlinear function generator 3 in which the nonlinear function is written by being multiplied by the tone color parameter α in the multiplier 2.

The non-linear function generator 3 is composed of, for example, a storage device such as a RAM, and distorts the output signal according to the level of the input signal and outputs the output signal. That is, the input signal x (n), the output signal y (n) and the tone color parameter α are all digital signals, and the multiplier 2 and the non-linear function generator 3 are formed by digital circuits.

The non-linear function generator 3 is provided with a switch 5, and by switching the switch 5, the non-linear characteristic in the non-linear region Xmin to Xmax can be appropriately selected.

That is, the switch 5 has, for example, a plurality of push button switches 5A to 5E, and one of the push button switches 5A to 5E is selected and pressed to select one nonlinear function written in the RAM. The conversion operation is performed according to the nonlinear function.

Next, a method of constructing a variable part (non-linear part) of the non-linear function will be described.

Here, the method of constructing the variable part of the nonlinear function so that the output power spectrum is as close as possible to the strong tone signal of the natural musical instrument when the given weak tone signal is passed through the nonlinear function. Will be described.

It is assumed that the periodic signal x (t) as shown in FIG. 3 is input, passes through the nonlinear function f (x), and becomes the output y (t),
The Fourier coefficients of x (t) and y (t) are written as X (k) = F [x (t)] Y (k) = F [y (t)]. F [x] is a Fourier series expansion of x.

Similarly, the target signal is U (k) = F [u (t)].

The variable part of f (x) is set so that | Y (k) | ² ≈ | U (k) | ² .

For this purpose, in this example, a piecewise linear approximation of f (x) is performed,
The method of optimizing each line segment is adopted.

The domain [Xmin, Xmax] of the non-linear function f (x) is appropriately divided into sections, and each section is R _L , R _{L + 1} , ... R _-1 , R ₀ ,
_{Name them} R ₊₁ , ... R _{R -1} , R _R.

Ri = [Xi, Xi ₊₁ ] where X _L <X _{L + 1} … X _-1 <X ₀ <X ₊₁ …… <X _R-1 <X _R <
X _{R + 1} and X _L = Xmin, X ₀ = χmin, X ₁ = χmax, X _{R + 1} = Xmax, f (x) = Pi · x + qi (where f (x) is a linear function in each interval Ri x ∈ Ri).

On the other hand, consider a signal hi (t) that becomes 1 only when the amplitude level of the input signal α · x (t) to f (x) enters the section Ri as shown in FIG. 4B and becomes 0 elsewhere.

Similarly, α ・ x (t) And divide and put.

Can be expressed as

Considering the amplitude direction in this way, Write like Here, the Fourier coefficient Y (k) of the distorted signal y (t) such as Hi (k) = F [hi (t)]
Can be represented by the linear sum of Xi (k) and Hi (k),
By removing the non-linearity of f (x) from the optimization problem, the global convergence of the optimization calculation can be improved.

In fact, as an evaluation function for optimization, It may be solved by an appropriate nonlinear optimization method so as to minimize ε.

By adopting the above-described construction method, it is possible to obtain a non-linear function that gives various tones by a relatively simple arithmetic process. By the way, if you take an ordinary processing method, it takes several days to calculate one non-linear function continuously by the computer day and night, but by using the above processing method, it takes about 10 minutes. You can get the result with.

[Advantages of the Invention] As described above, according to the present invention, a strong signal can be obtained from a weak signal. In particular, distortion is applied in a non-linear part when it exceeds the linear function, and this distortion generates high-order harmonic components, which brightens the timbre and obtains a signal with a timbre that is close to the strong sound of the original natural instrument. You can

Further, since the present invention adopts a method of generating a strong signal by distorting the weak signal, the strong signal contains a large amount of high-order harmonic components. When changing from to a weak performance, the tone color also changes greatly, and an electronic musical instrument sound with rich expression can be obtained.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a graph for explaining an example of a non-linear function which is a main part of the present invention, and FIG. 2 shows an example of a specific configuration that can be adopted when implementing the present invention. Block diagrams, FIGS. 3 and 4 are graphs for explaining a calculation processing method when constructing a non-linear function.

Claims (1)

[Claims] 1. Tones of various timbres are provided which have a characteristic of acting as a linear function for inputs up to a predetermined level and acting as a non-linear function for inputs exceeding a predetermined level, and the characteristics of the non-linear function are appropriately set. A tone color control method for obtaining a strong sound.