JP3498319B2 - Automatic performance device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic performance device for storing performance data in the order of occurrence, reading the data, and automatically playing the data. .

[0002]

2. Description of the Related Art Some automatic performance devices store note data relating to sound generation / silence of musical tones, as well as control data for controlling musical tones such as pitch bend, aftertouch, and volume as the music progresses. . During playback, this is read out sequentially and input to a sound source or effect circuit to perform automatic performance. The note data determines the pitch and duration of the musical tone to be pronounced. On the other hand, the control data controls subtle pitch movements and volume changes of the generated musical tones and contributes to musical expression. As described above, the control data is the data for controlling the subtle changes in the musical sound, and is the data that is generated each time the pitch or volume changes. Therefore, in the case of a musical tone which is input in real time (actual performance) and has a rich expression in which the pitch and the volume fluctuate well, a large amount of control data exists for one musical tone.

Therefore, the automatic performance data in which the control data is stored together has a very large amount of data. In order to store such automatic performance data, a large-capacity memory is required, which causes a cost increase of the automatic performance device.

Therefore, conventionally, there has been proposed a method of thinning (compressing) automatic performance data (control data) having a large data amount to reduce the data amount.

[0005]

However, since the conventional compression method is a method of evenly thinning out such as "leaving one in three", necessary data is thinned out and the change becomes unnatural. There were times when it became. On the contrary, depending on the control state, unnecessary data may remain, which is a disadvantage that proper compression cannot be performed.

It is an object of the present invention to provide an automatic performance device capable of reducing the amount of data while maintaining the variation of musical sound.

[0007]

According to the present invention, a storage means for storing performance data in time series, a reading means for sequentially reading the performance data from the storage means, and a performance data read by the reading means are thinned out. a compression means, compares the data stored in the read data and the previous by the reading means, the change in the value de to save
Data is greater than a predetermined value greater than the quantization step of the data, or the data read by the reading means.
Data read timing and the read timing of the data saved immediately before.
And a compression means for storing the data read by the reading means when the time difference is equal to or larger than a predetermined value larger than the sampling interval . This invention is a sampling timing
Storage means for storing performance data for each song in time series.
Reading means for sequentially reading the performance data from the storage means
And the performance data read by the reading means is thinned out.
And storing the compressed data by the reading means.
Each time the performance data is read, this read data
Sampling timing and the data saved immediately before.
Whether the time difference of sampling timing is more than a specified value
Whether or not this is the value of the read data
The amount of change in the value of the data saved immediately before is equal to or greater than the specified value
It is judged whether or not, and either judgment is affirmative
Compression means for storing this read data.
It is characterized by being provided. This invention can
Storage means for storing in sequence, and the performance from the storage means
Reading means for sequentially reading data and reading means for reading
With compression means that thins out the saved performance data and saves it
The data read by the reading means.
The distance to the previously saved data (value change and time difference
If the distance is equal to or greater than a specified value,
The data read by the reading means is stored in
And a compressing unit that is provided. This invention
Is a storage means for storing performance data in time series, and
Reading means for sequentially reading the performance data from the storing means,
Thinning out the performance data read by the reading means
It is a compression means to save, and is a section that consists of multiple data.
Seeking a Hitoshi Taira slope (variation / time) between, based on the inclination
A compression means for determining the thinning-out number in that section based on
And are provided. This invention is a performance day
Storage means for storing the data in time series, and
The reading means for sequentially reading the musical performance data and the reading means.
Compression that thins and saves the performance data read by
Means, the average slope of a section consisting of multiple data
(Change amount / time) is calculated, and the
Only the data whose slope from the data is steeper than the average slope is stored.
The slope from the previous data is less than or equal to the average slope.
And a compression means for thinning out the data.
It

[0008]

[Action] automatic performance apparatus of the present invention, compares the data stored in the musical data and the previous read out from the storage means
However, if the change in value or the passage of time is greater than or equal to the specified value,
Save the read data . Above mentioned was a predetermined value
Is set to a value and time greater than the minimum
It As a result, data generated at short intervals or data with little change can be thinned out, and the shape of the abrupt change can be saved as it is.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An automatic performance device according to an embodiment of the present invention will be described with reference to the drawings. This automatic performance device includes a keyboard 10 as a performance operator, a pedal controller, a modulation wheel,
A control operator 11 such as a pitch bend wheel is provided. The operation amount of the control operator 11 is taken in as control data for pitch control and volume control. The keyboard 10 and the control operator 11 are connected to the writing unit 12. The writing unit 12 is a circuit that allocates appropriate addresses to performance data (note data and control data) input from the keyboard 10 and the control operator 11 and writes the performance data in the memory 13. The memory 13 is composed of a backed up RAM. With the above circuit, the contents played on the keyboard 10 and the control operator 11 are
It is stored in the memory 13 in real time as automatic performance data.

A read-out unit 14 is connected to the memory 13. The reading unit 14 is a circuit that sequentially reads the automatic performance data stored in the memory 13 according to the clock. The read automatic performance data is the sound source 15.
Alternatively, it is input to the compression unit 18. The sound source 15 forms a musical tone signal based on the input note data, and
The delicate pitch and volume of the tone signal are controlled according to the input control data. This tone signal is input to the sound system 16. The sound system 16 amplifies the input musical tone signal and outputs it as a musical tone from the speaker. With the above circuit, the memory 13
The automatic performance data stored in can be played (automatic performance).

On the other hand, the compression section 18 extracts only a plurality of control data designated from the automatic performance data input from the reading section 14 and outputs a plurality of control data (2 in this embodiment).
Buffer). The designated control data is data designated by the operator when starting the compression work, and is, for example, pitch bend data or volume data. The compression unit 18 of this embodiment generates (reads) timing t of two buffered data.
And the size x of the data are compared to determine whether to save or thin the data (details will be described later with reference to FIG. 2). The data determined to be stored is input to the writing unit 12 (the data other than the designated control data is directly input to the writing unit 12 without being buffered). At the time of this compression work, the memory 13 is provided with a storage area for compressed data in addition to the automatic performance data to be read. The writing unit 12 sequentially stores the input data in this storage area. By this work, the automatic performance data in which the designated control data is compressed is newly stored.

2, FIG. 3 and FIG. 4 are diagrams for explaining the compression processing executed by the same automatic musical instrument. In this automatic performance device, compression is performed by an algorithm that thins out a time difference (t2-t1) or an operation amount change amount | x2-x1 | of a fixed value (t0, x0) or less. Here, the sound source 15 has a function of, when control data relating to volume and pitch bend is input, holding the value until the next control data is input. Therefore, when the control data is thinned out in a portion where the change is drastic, a discontinuous stepwise change occurs, and therefore such a portion is not thinned out. On the other hand, in the portion where the change is gradual, even if one control data is held for a long time, there is no unnatural change, so the data is thinned out and compressed. That is, in FIG. 2, assuming that the time threshold value is t0 and the operation amount change threshold value is x0, the data A is not thinned out because the operation amount change exceeds the threshold value with respect to the immediately preceding data t1 and x1. Is not thinned out because the time data exceeds the threshold value t0. On the other hand, the B data is subject to thinning because both the time data and the change data are within the threshold.

FIG. 3 is a diagram showing an example of control data input in real time. In this figure, 58 pieces of data are represented. FIG. 4 shows an example in which the data of FIG. 3 is compressed by the method of the present invention. The threshold value (t0, x0) in this case is shown at the left end of FIG. When data in a range not exceeding this is thinned out, what is indicated by ○ in the figure is the remaining data that is not thinned out, and data indicated by × is the thinned data. In this example, 41 data of 58 data are thinned out and 17 data are stored.
That is, the compression rate (decimation rate) is 70% or more. However, the compression rate of the portion where the change is drastic (the 11th to 18th in the raw data (data of FIG. 3)) is low, and the smooth change is stored. On the other hand, the compression rate of the portion where the change is gradual (for example, the 42nd raw data and thereafter) is extremely high, and the remaining data is about 20%. But the changes are not so discontinuous.

On the other hand, FIG. 5 is a diagram showing an example in which the data of FIG. 3 is thinned out by a conventional method (uniform method). In this example, a method of saving every third data and thinning out two out of three data is adopted. Therefore, 20 pieces of data remain out of 58 pieces of data, and the compression rate is about 65%. That is, the compression rate is lower than that of the compression method of the present application. However, since it is a uniform thinning, it can be seen that in a portion where the change is severe, the change is extremely rough and discontinuous.

FIG. 6 is a flow chart showing the data compression operation of the compression section 18 of the automatic musical instrument. First, the threshold values t0 and x0 are input (n1, n2). This threshold may be input by the operator or may be automatically generated. After this, the compression operation is executed, but first,
The first data in the compression range is read out and set as t1, x1 (n3). The t1 and x1 are stored in the memory as save data (n4). Convert this data to logarithmic value (n
5). That is, since the volume sensed by the listener is proportional to the logarithmic value of the physical quantity, the logarithmic value is used for calculation to match this sensitivity. Next, the next data is read out and designated as t2 and x2 (n6). This data is also logarithmically converted for calculation (n7). After this, (t2-t1)
And | x2-x1 | are below the threshold values t0 and x0 (n8, n9). Here, only x2-x1 takes an absolute value because the manipulated variable data may decrease. If both conditions are satisfied, this data (t2, x
2) is deleted because it can be thinned (n10), and the next data is read (n6). On the other hand, (t2-t
If either one of 1) and | x2-x1 | exceeds the thresholds t0 and x0, this data is stored in the memory as it should be saved (n11). Since the next comparison is based on this data, set t2 and x2 to t1 and x1 (n1
2) Then, the next data is read (n6). If it is determined in n13 that the data has ended, the operation ends.

Such compression can be performed on any control data. For example, if this compression is performed by a sound source such as a physical model sound source in which all of tone generation / silence and tone color are controlled by parameters having continuity, the amount of data can be extremely reduced.

The maximum and minimum values of the change curve may be preferentially left regardless of the threshold value. By doing so, the shape of the change curve can be better preserved.

7 to 9 show another embodiment of the present invention. FIG. 7 shows a method of saving data when the distance between data (sum of squares of time difference and variation) exceeds a certain value (threshold value), and thinning out the data when the distance is less than a certain value. This method is effective when the sound source is not a method of storing the value of the previously input parameter but a method of interpolating between the parameters. In this embodiment, when the value is equal to or more than the threshold value and is saved when the threshold value is less than the threshold value, it is thinned out.

Further, FIG. 8 shows a method in which an average slope (change amount / time) is obtained from a plurality of data and the thinning-out number in the section is determined based on the slope. That is, if the inclination is normal as shown in FIG. If the inclination is steep as shown in FIG. 6B, only one sample is thinned out of the five samples. In addition, the same figure (C)
If the slope is gentle as shown in FIG. Can be adopted. In this case, one, two, three for the slope of the section
It is sufficient to provide individual thresholds and a plurality of thresholds.

Further, in FIG. 9, the average slope of a certain section may be obtained, and only data steeper than the average slope may be stored in that section. In this case, the average slope of the section functions as the threshold.

[0021]

As described above, according to the present invention, the amount of data can be extremely reduced, and the change curve based on this data can be well saved.

[Brief description of drawings]

FIG. 1 is a block diagram of an automatic performance device according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining a data compression method of the automatic musical instrument.

FIG. 3 is a diagram showing an example of compressed data.

FIG. 4 is a diagram showing an example in which the above data is compressed by the method of the present invention.

FIG. 5 is a diagram showing an example in which the above data is compressed by a conventional method.

FIG. 6 is a flowchart showing a data compression operation of the automatic musical instrument.

FIG. 7 is a diagram for explaining a data compression method according to another embodiment of the present invention.

FIG. 8 is a diagram for explaining a data compression method according to another embodiment of the present invention.

FIG. 9 is a diagram for explaining a data compression method according to another embodiment of the present invention.

[Explanation of symbols]

18-Compressor

Claims (5)

(57) [Claims] 1. A storage means for storing performance data in time series, a reading means for sequentially reading the performance data from the storage means, and a compression means for thinning and storing the performance data read by the reading means. Then, the data read by the reading means is compared with the data stored immediately before, and when the change in the value is equal to or more than a predetermined value larger than the quantization step of the data , or when the reading means reads the data.
The timing of reading the extracted data and the data saved immediately before
Data read timing, and the time difference is sampled.
An automatic performance device comprising: compression means for storing the data read by the reading means when the value is equal to or larger than a predetermined value larger than the ring interval . 2. Performance data for each sampling timing
And a reading means for sequentially reading the performance data from the storage means.
And the performance data read by the reading means are thinned out.
A compression means for storing, which is played by the reading means.
Each time the data is read, the
Sampling timing and sampling of the data saved immediately before
Check whether the time difference of ring timing is more than a predetermined value.
Judgment and immediately before the value of this read data
Whether the amount of change in the value of the data stored in is greater than or equal to a predetermined value
If any of the judgments are positive, the
And an compressing means for storing the read data of the automatic performance device. 3. Storage means for storing performance data in time series
And a reader for sequentially reading the performance data from the storage means.
And the performance data read by the reading means are thinned out.
Compression means for storing, which is read by the reading means
The distance between the output data and the data saved immediately before (value of
Calculate the sum of squares of change amount and time difference, and calculate this distance as a predetermined value.
If it is above, it is read by the reading means.
An automatic performance device characterized in that it is provided with a compression means for storing the data . 4. Storage means for storing performance data in time series
And a reader for sequentially reading the performance data from the storage means.
And the performance data read by the reading means are thinned out.
It is a compression means to save, and is a section that consists of multiple data.
Calculate the average slope (change amount / time) between
A compression means for determining the thinning-out number in that section based on
Automatic performance apparatus is characterized by providing the, the. 5. Storage means for storing performance data in time series
And a reader for sequentially reading the performance data from the storage means.
And the performance data read by the reading means are thinned out.
It is a compression means to save, and is a section that consists of multiple data.
Calculate the average slope (variation / time) between
And the slope from the previous data is steeper than the average slope.
Save only the data, and the slope from the previous data is
An automatic performance device , comprising: compression means for thinning out data having an inclination or less .