JP3211646B2 - Performance information recording method and performance information reproducing apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a performance information recording method for reducing the amount of transmission and storage when transmitting digitized performance information and for storing the transmitted information in a medium, and the performance information. Related to a reproducing apparatus.

[0002]

2. Description of the Related Art Generally, MIDI (Musical Instruments) is used as a means for transmitting performance information for generating musical tones.
Trument Digital Interfac
e) is widely used. MIDI is a hardware and software standard that is defined to enable information exchange by connecting to a sound source such as a musical instrument such as a synthesizer or an electronic piano. When MIDI is used, an input to a player's keyboard or the like is converted to MIDI information and output to a transmission path. On the other hand, a sound source that has a MIDI receiving capability and actually generates a sound is connected to the transmission path, and interprets the MIDI information to generate a sound.

[0003] Such MIDI information is roughly divided into the following information. (1) Note-like information (hereinafter referred to as note information) representing sounding start by keying (pressing the keyboard) or the like, and sounding stop by key release (stopping the keyboard pressing) or the like. This note information includes information on the pitch of the sound represented by the keyboard number. (2) Accent information indicating the strength of pronunciation. In MIDI, it is transmitted and received along with note information. (3) Control information for transmitting an inflectional expression, such as crescendo or pitch fluctuation, performed by a player using a pedal or lever to a sound source. Specifically, when the MIDI converter on the player side detects a change in the lever position, the position information is transmitted. Since these pieces of information are transmitted immediately as the performance progresses, the information flow is in a state in which the various kinds of information are mixed.

Since all MIDI information is digitized, recording, editing, and reproduction are performed using computer equipment such as a synthesizer and a recording medium. That is, the MIDI information is recorded on the medium as a data file. Here, SMF (Standard Midi Fi) is usually used as a commonly used file format.
le), but since MIDI information is real-time performance information, each information element (hereinafter MIDI) such as sound generation start and sound generation stop
The event is recorded in the SMF with time information added. Also, since recording is performed according to the order of occurrence of MIDI events,
Note information and control information are recorded in a mixed form.

When storing data files, not limited to SMF, it is desirable that the data files be compressed as much as possible due to the limited capacity of the recording medium. Is usually L used in programs such as LHA and GZIP.
A compression method based on pattern matching called the Z method is used. The principle of the data compression will be briefly described. The generation of the compressed file from the original file is performed by copying the data read from the processing position to the compressed file while moving the processing position in the original file from the head of the file. If two identical data areas exist in the original file, the processing position is set to two of the same data areas.
At the beginning of the first data area, instead of simply copying the data, the distance from the beginning position of the matching first data area to the processing position and the length of the matching data area Is recorded in the compressed file, the processing position in the original file is moved to the end of the second data area, and the processing is continued. As a result, it is not necessary to copy data for the second same data area.
Data may be compressed by adding only a small amount of distance information and length information.

As is apparent from the above description, the compression ratio of this compressed data increases as the size of two identical data areas increases. Also, when searching for a data pattern after a certain processing position from the data area before the processing position, the size of the search target area is limited. It has the characteristic that the rate is high. An example of using such a performance information file such as SMF is a communication karaoke apparatus. Among the non-storage type communication karaoke apparatuses, the terminal apparatus receives a MIDI performance information file distributed from a distribution center via a public line or the like, and the terminal apparatus reproduces the file and uses it as an accompaniment for singing. In this configuration, every time a user selects a music piece, information on the selected music piece is distributed from the center side. In such a case, if the performance information file in the distribution center is compressed and then distributed, the distribution time can be shortened, and the line usage fee can be reduced.

A terminal device of a communication type karaoke apparatus called a storage type has a built-in recording medium such as a large-capacity hard disk and stores therein a performance information file distributed in the past in an updatable manner. Also in this case, by storing the performance information file in a compressed state, more music pieces can be stored in the recording medium.

[0008]

By the way, the terminal device of the storage type karaoke apparatus as described above needs to hold as many songs as possible on a limited recording medium in order to respond to a user's request. Therefore, it is desirable that the performance information file be subjected to data compression. SM
In a performance information file such as F, a considerably higher compression ratio can be obtained than in a normal text file or the like, but because of the nature of the music performance information, the same information is often repeatedly recorded. is there.

[0009] However, with the recent diversification of karaoke users' requests, it has become necessary to hold more songs. Here, if the recording medium is further increased in capacity, not only does the cost rise, but also the line usage fee for distribution increases as the number of distributed songs increases. On the other hand, at the stage of production of performance information, the performance is further simplified,
By distributing a performance information file with a small amount (information amount), the problem of the data amount and the distribution cost can be solved, but the quality of the performance provided to the user is greatly reduced, which is not practical. For this reason, it has been desired to solve the problems of the data amount and the distribution cost while compressing the file and keeping the quality of the performance the same.

[0010] The present invention focuses on the above problems,
The purpose of the present invention is to solve this problem effectively, and the purpose is to record the performance information to greatly exceed the compression ratio of the compressed performance information file achieved by using the conventional technology without deteriorating the performance quality. It is to provide a method. That is, according to the present invention, the capacity of the recording medium, the distribution cost,
It is possible to hold more songs on a recording medium while maintaining the quality of performance as before.

[0011]

The present inventor has conducted intensive studies on a compression method of performance information, and as a result, focused on the following points in the form of information in a performance information file.
This has led to the present invention. (1) Performance information is mainly recorded in a form in which note information, accent information, and control information are mixed. (2) When the accent information and the control information are removed from the performance information, the data pattern of the note information matches well between repeated portions of the music. (3) The data patterns of the accent information and the control information do not always match well between repeated parts. This is because the control information is information relating to the intonation expression of the performance, and may greatly change as the music progresses. Therefore, if this performance information file is compressed as it is, the same pattern will be fragmented and cannot be compressed efficiently because control information is inserted between note information, but if these are separated, the same pattern will be repeated. Increase.

That is, the performance information recording method of the present invention comprises:
At least one of the two information start of sounding tone and the musical note information representing a sound stop, and control information indicating the intonation of the performance and accent information indicating the intensity of sound for each sound
A method for recording performance information bets are mixed, the Starring
Said accent information and / or the control information is a response rate information collected and the note information for each information, the collected thereof
Information is stored in different and independent areas on the recording medium .
Compress data by pattern matching
Those were by that records Unishi as file. The reproducing apparatus of the performance information of the present invention is a reproducing apparatus of the stored performance information by the performance information recording method according to claim 1, wherein the performance information is recorded as the zipped file
Decoding means for decoding the compressed file from the compressed file recording medium to obtain the performance information; a performance information storage medium for temporarily storing the obtained performance information; A reproduction control means for sequentially reproducing the performance information in the storage medium, and a sound source section for reproducing the performance information output from the reproduction control means as a musical tone.

As described above, according to the performance information recording method of the present invention, when performance information is recorded in a file, note information, accent information, and control information are recorded in independent areas in the file. After data compression of the performance information file, the data area of the note information is largely compressed, and the overall compression ratio can be greatly increased as compared with the conventional performance information recording method. In this case, control information may not be included. In this case, note information and accent information are recorded in different areas. According to the reproducing apparatus of the present invention, Compressed files data compressed by the pattern matching technique performance information is recorded in the zipped file recording medium, already this compression
The compressed file read from the file recording medium is decompressed by decoding means to obtain performance information.
The obtained performance information is temporarily stored in a performance information storage medium, and subsequently processed and output under the control of the playback control means. For example, it is input to the tone generator section as MIDI information to emit a musical tone. .

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a performance information recording method and a performance information reproducing apparatus according to the present invention will be described below in detail with reference to the accompanying drawings. Figure 1 shows MID with time information
FIG. 2 is a diagram showing an example of the description of an I event. FIG. 2 shows the event shown in FIG.
FIG. 3 is a diagram showing a data string when described in the first embodiment of the method of the present invention. In the method of the present invention, for note information indicating the start and stop of sound generation of a musical tone, a performance in which at least one of accent information indicating the strength of sound generation and control information indicating inflection of the performance is mixed for each sound generation. When information is recorded, the note information, the accent information, and the control information are collected for each piece of information, and are recorded in different and independent areas.

As described above, the recording area is made different depending on the type of the information because the data pattern of the note information matches well in the repeated portion of the music due to the characteristics of the music, while the accent information and the control information are different. This is because the information data pattern does not always match between repeated parts. Now, description will be given by taking performance information as shown in FIG. 1 as an example. Here, in order to make the description easy to understand, an example is described in which control information is not provided and only note information and accent information are represented.

In the illustrated example, there is shown an example of performance information for performing [Dremidremidremi] such that its volume gradually increases. In the SMF, MIDI events with time information are serially recorded in a file as shown in FIG. Here, if the general description of the performance information is expressed as (DT, A, B, C), DT indicates time information of the event, and here, a relative time to the immediately preceding event is recorded. (A, B, C) is MI
It represents a DI event, and the meaning of each code is as follows. A: MIDI event type identification data ON: sound generation start event OF: sound generation stop event When the code A is a sound generation start event or sound generation stop event, the codes B and C have the following contents. B: Keyboard number (pitch) C: Accent (volume) information (The greater the numerical value, the stronger it is played. The accent information for the sound generation stop event has no meaning and a fixed value is recorded.) Here, DT, A , B form note information.

In the above example, a performance is recorded in which each sound of [Dremidremidremi] is played at intervals of time 10 with a sounding length of 8 and an accent (volume) of 40 to 80, which is progressively stronger. FIG. 2 shows an actual data sequence in the SMF. Here, recording is performed from the upper data row to the lower data row. That is, the MI with time information shown in FIG.
The DI event information is simply recorded in a time-series manner, and note information and accent information are mixed and recorded in the event order. In this case, since the occurrence rate of the same data pattern is very low, the compression efficiency becomes very poor when data is compressed.

Therefore, in the present invention, the information is recorded in different areas for each type of information. As shown in FIG.
Accent information is extracted from each event information with time, and only note information is recorded in chronological order first, and thereafter, accent information is continuously recorded, so that both recording areas are distinguished independently. That is, the first stage is a note region 1 for recording note information, and the subsequent stage is an accent region 2 for recording accent information. The division of such an area is described in an abstract manner in FIG. In this case, information on the start position 2A of the accent area 2 in which the accent information is recorded is recorded as accent start position information 3 in the start area of the entire file, so that the boundary between the two areas can be determined. I have.

When such a data string is subjected to data compression by a compression technique based on pattern matching, the note area 1 has the same data pattern with a large length, so that it is compared with the case where SMF is compressed. Thus, the data compression efficiency can be greatly improved. In the illustrated example, from the first [ON] in the first row to the last [Mi] in the same row.
Are the patterns from the first [ON] in the second column to the last [MI] in the second column and the first [O] in the third column.
N] to the last [mi] in the same row are the same patterns, and are the targets of compression. In the first embodiment described above, performance information not including control information has been described. Next, as a second embodiment, a recording method when performance information is also included will be described.

FIG. 5 shows an example of a description of a MIDI event in which a performance similar to that of the first embodiment is converted into data by another expression method. FIG. 6 shows the event shown in FIG. 5 actually described in the SMF. FIG. 7 is a diagram showing a data sequence at the time, and FIG. 7 is a diagram showing a data sequence at the time of description according to the second embodiment of the method of the present invention. The performance shown in FIG. 5 is a recording example in which the strength of playing the keyboard is constant, but the volume control lever is moved simultaneously with the performance to gradually increase the volume. Here, if the general description of the performance information is expressed as (DT, A, B, C), DT indicates the time information of the event, and here, the relative time to the immediately preceding event is recorded. I have.

(A, B, C) indicates a MIDI event, and the meaning of each code is as follows. A: MIDI event type identification data CL: Control event If A is a control event, B and C
Is as follows. B: Control type identification data EX: Volume control C: Volume data That is, when A is a control event, (DT, A,
B, C) as a whole becomes control information.

When the data sequence in the actual SMF is illustrated,
It is as shown in FIG. Also in this case, recording is performed in accordance with the lower data row from the upper data row. That is, the MIDI event with time information shown in FIG.
The information is simply recorded in chronological order, and note information, accent information, and control information are recorded in the order of events in a mixed state. In this case, the compression efficiency is not good because the appearance frequency of the same data pattern is low as in the case described above. Therefore, in the second embodiment, recording is performed as shown in FIG. 7 with different areas for each of the three types of information. As shown in the figure, for events other than the control event, accent information, for example, [64] is extracted from the event, and only the note information, which is the remaining part, is recorded in chronological order first (3 in the figure, from the upper row). Up to the stage). Thereafter, only the extracted accent information is continuously recorded (fourth row in the figure). Further, following the accent information, control information of the control event is recorded in chronological order, so that each recording area is distinguished independently.

FIG. 8 shows such a division of the area. The preceding stage is a note area 1 for recording note information.
The middle section following this is an accent area 2 for recording accent information, and the subsequent section is a control area 4 for recording control information. Here, it should be noted that, since only control events are extracted from one column of events in the SMF and collected in another area of the file, the relative events of both the extracted and remaining event columns are used. The point is that the time must be recalculated. In the example of the present invention, the time information of the control event is rewritten to the relative time with respect to the immediately preceding control event.
That is, the time of an event other than the control event in the middle,
That is, the relative time between the control events is obtained by adding the time 2 and the time 8, and this is rewritten.

In this example, an accent area 2 for recording accent information is provided immediately after a note area 1 for recording note information, and a control area 4 for recording control information follows this. For this reason, information relating to the head positions 2A and 4A of the regions 2 and 4 for recording the accent information and the control information is described as the accent start position information 3 and the control start position information 5 in the head region of the file. It can be determined. In this embodiment,
Since the volume data is included in the control information and is controlled by the control information, the accent information described in the accent area 2 is simply necessary for the format, and the numerical value here can be set to any numerical value on a program. In performance, it becomes a meaningless numerical value.

Then, as shown, for example, [64]
If the values are set to the same value, the efficiency can be improved at the time of data compression. When such a data string is subjected to data compression by a compression technique based on pattern matching, the note area 1 and the accent area 2 have the same data pattern with a large length, respectively. Significant data compression is possible. In the illustrated example, the pattern from the first [ON] in the first column to the last [MI] in the same column is the first [ON] in the second column.
, And the pattern from the first [ON] in the third row to the last [Mi] in the same row are the same, and the accent area 2 is all [64]. Since they are the same pattern, they are subject to compression.

Therefore, the note area 1 and the accent area 2 in the file each have the same pattern with a large length, and when the file is compressed, the compression efficiency can be greatly improved as compared with the compressed SMF. Next, S
A method of converting an MF file into a file format according to the second embodiment will be described with reference to the flow shown in FIG.

First, in S1, the target SMF is opened, and the processing position is set to the head of the file. Then, AT,
AT1 and AT2 are initialized to 0. Here, AT is the absolute time from the beginning of the song of the event, and AT1 is
4 is the absolute time from the beginning of the song of the event written in AT4, and AT2 is the absolute time from the beginning of the song in the event previously written in S5. In FIG. 5, the absolute time of each event when the beginning of the music is set to the absolute time 0 is obtained by cumulatively adding time information indicated by the leftmost DT.

Next, in S2, an event (DT, A, B, C) with time information is read from the SMF processing device. The relative time DT is added to the AT so that the AT becomes the absolute time of the event just read. Next, in S3, the process branches depending on the value of A. A = ON or O
In the case of F, the result is YES, and in the case of A = CL, the result is NO. Next, in S4, if the value of A is ON or OF in S3, the contents of the event are written in the note area 1 holding note information and the accent area 2 holding accent information. At this time, since the DT written in the note area 1 is a relative time from the event written in the note area 1 in the closest past, the DT is recalculated and recorded here. At this time, the value of AT is written in AT1.

In S5, when the value of A is other than ON or OF in S3, that is, since the control event is A = CL, the content of the event is written to the control area 4 holding the control information. At this time, since the DT written in the control area 4 is a relative time from the event written in the control area 4 in the closest past, the DT is recalculated and recorded here. That is, the time difference from the immediately preceding control event is obtained. At this time, the value of AT is written in AT2.

Next, in S6, the processing position of the SMF is advanced by the amount read in S2. Then, in S7, it is determined whether or not the processing position is at the end of the SMF. If there is no more processing target (YES), the processing proceeds to S7.
Proceed to 8. If there is a processing target (NO), S2
And the same process is repeated. And S8
In FIG. 7, a note area 1 holding note information, an accent area 2 holding accent information, and a control area 4 holding control information are combined, and File.

Next, a description will be given of an apparatus for reproducing performance information compressed using the above-described performance information recording method. FIG. 10 is a block diagram showing a reproducing apparatus according to the present invention, and FIG. 11 is a flowchart showing a reproducing procedure performed by the reproducing apparatus. As shown in the figure, the reproducing apparatus 10 includes a compressed file recording medium 11 for recording a compressed file obtained by compressing data of a performance information file recorded by the above-described performance information recording method by a pattern matching method, and decoding the file. Decoding means 12 for so-called decompressed performance information; performance information storage medium 13 for temporarily storing the decompressed performance information; and reproduction control means 14 for sequentially processing the decompressed performance information. It mainly comprises a sound source section which actually reproduces and outputs the processed output information as a musical tone, for example, a MIDI sound source section 15.

The compressed file recording medium 11 is composed of, for example, a large-capacity recording disk or the like, and records therein a compressed file distributed via a communication line such as a telephone line. Further, the performance information recording medium 13 is constituted by, for example, a RAM or the like having a high reading / writing speed, and can correspond to a reproducing speed at the time of reproducing MIDI information. Further, the decoding means 12 and the reproduction control means 14 are integrated into, for example, the microprocessor 16 and perform arithmetic processing specified by software on the performance information.

Next, the reproduction process of the performance information will be described with reference to FIG. The compressed file on the compressed file recording medium 11 is expanded by decoding and decompressing by the decoding means 12 to obtain performance information. It is assumed that the obtained performance information is stored in the performance information storage medium 13 in a format as shown in FIG. 8, for example. The reproduction control means 14 reproduces the MIDI information by processing the performance information file according to the following processing flow, and emits the sound from the MIDI sound source section 15.

First, in S11, the performance information storage medium 13
, Read the accent start position information 3 and the control start position information 5 from the top of the
And the accent area 2 and the control area 4 can be read from the beginning. The head of the note area 11 is next to the position of the control start position information 5. Here, two flags f
flag1 and flag2 are prepared, and one of the flags flag
1 indicates a playback state from the note area and the accent area,
The other flag flag2 indicates a state of reproduction from the control area. Prior to playback, these two flags are initialized to unfinished. Further, in addition to the above two flags, here, further, two flags oflag1 and ofl
ag2 is prepared, one flag offlag1 indicates the output state of each MIDI event read from the note area 1 and the accent area 2, and the other flag oflag2 indicates the output state of each MIDI event read from the control area 4. Here, both flags offlag1,
oflag2 is initialized to a state where output has been completed.

Next, in S12, it is determined whether or not all the contents of the note area have been read for reproduction. When there is no content to be read next in the note area 1 (YE
In S), the process proceeds to S13, and the flag flag1 is ended. If the content of the note area 1 remains in S12 (NO), the process proceeds to S14, where the state of the flag offlag1 is checked, and the output state of each MIDI event read from the note area 1 and the accent area 2 is changed. to see.
Here, in the case of NO, that is, in the case of already output, S1
Go to 5. Here, the next information (DT1, A1, B1) described above with reference to FIG. 1 is read from the note area 1, and the next accent information (C1) is read from the accent area 2. Further, information (DT1, A1, B
1) and information (C1) correspond to MIDI events (A1, B
In order to output to the MIDI tone generator 15 as (1, C1), information (DT1, A1, B1, C1)
Is reconstructed. Further, the flag oflag1 is set to a non-output state. Reading is performed sequentially from the beginning of each area, but note information (DT1, A1, B1) and accent information (C1) have a one-to-one correspondence. At the same time, the determination as to whether or not all the contents have been read may be made only for the note area.

Next, in S16, S17, S18, and S19, the same processing as that in S12, S13, S14, and S15 is performed on the control area 4. That is, if the contents have not been read from the control area 4 to the end (N0), the state of the flag oflag2 is checked (S18). If NO here (output has been completed), S1 is checked.
At step 9, the control information (DT2, A2, B2, C2) described with reference to FIG.
lag2 is set to a non-output state. Then, if the reading has been completed to the end in S16 (YES), the flag fla
g2 is terminated (S17). Here, the number given to each information element is for identifying the information (DT1, A1, B1, C1) reconstructed from the note information and the accent information. Next, the process proceeds to S20, and it is determined whether or not all the contents of the note area 1, the accent area 2, and the control area 4 have been read. This consists of two flags, flag1 and flag2.
Are determined by determining whether both indicate end. Both flags flag1 and flag2 end (Y
If it is (ES), the flow shifts to S21, where all the reproduction processes are ended. If NO, the process proceeds to S22.

At S22, the state of the flag flag1 is checked. If the flag flag1 indicates the end (YES), the MI from the note area 1 and the accent area 2
Since the reproduction of the DI event has already been completed, the subsequent processing relating to the output is skipped, and the process branches to S25. If the flag flag1 has not been completed (N0), there is a MIDI event (A1, B1, C1) that has been read and is waiting to be output, so that the output processing after S23 is executed. That is, in S23, the value of DT1 indicating the time information of the event is checked. DT1 represents the time until output of information (A1, B1, C1), and when DT1 = 0 (Y
In ES), the current time indicates information (A1, B1, C1) by MI.
Since it is time to output to the DI sound source unit 15, S
Branch to MIDI event (A1, B1, C1)
To the MIDI tone generator 15 and the flag o
It is assumed that flag1 has been output. . DT1 in S23
Is not 0 (NO), the information (A1, B1, C
Time elapse is required until the output of 1). Then, S
In response to inspecting DT1 again at 23, S28
DT1 is changed to a value smaller by one. And S2
At 8, the unit time wait is executed.

Next, in steps S25, S26, and S27, the MIDI events (A2, B2,
For C2), the same processing as the processing performed in S22, S23, and S24 is performed, and a state check is performed on the flag flag2 (S25).
Output to the sound source unit 15 and assume that the flag oflag2 has been output. When entering S28, a MIDI event (A1, B1, C1) that has been read and is waiting for output or (A1)
2, B2, C2) are present. In order to output these at the correct time, DT1 and DT2 are each decremented by 1 in S28, and the process proceeds to S22 after waiting for a unit time. Through the above steps, the performance information can be reproduced from the MIDI sound source section 15 and emitted.

In each of the embodiments described above, the SMF is used as the conventional performance information file. However, the present invention is not limited to the SMF, and the performance information file stores note information and at least one of accent information and control information. In the case where these pieces of information are recorded in the order of occurrence, the present invention can be applied to perform efficient data compression and reduce the data capacity.

[0040]

As described above, according to the performance information recording method and the performance information reproducing apparatus of the present invention, the following excellent operational effects can be obtained. According to the method of the present invention, note information and the like included in performance information are recorded in an independent area by being classified according to the type of information, so that it is possible to combine information with a high probability of occurrence of the same pattern. Therefore, when the performance information file is compressed by the pattern matching method, the data compression efficiency can be greatly improved. Therefore, when the compressed data is stored on a recording medium or transmitted through a line, the data compression efficiency is reduced. Can greatly reduce the data capacity of the device, so that a small-capacity recording medium can be used, and the line use time can be reduced, so that the line use fee can be reduced. According to the present invention, the compressed file of the performance information created as described above can be reproduced as a musical tone, and the capacity of the compressed file recording medium for recording the compressed file can be reduced.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of a description of a MIDI event with time information.

FIG. 2 is a diagram showing a data string when the event shown in FIG. 1 is actually described in an SMF.

FIG. 3 is a diagram showing a data string when described according to the first embodiment of the method of the present invention.

FIG. 4 is a diagram showing an abstraction of the data sequence shown in FIG. 3;

FIG. 5 is a diagram showing an example of a description of a MIDI event in which a performance similar to that of the first embodiment is converted into data by another expression method.

FIG. 6 is a diagram showing an example of data when the event shown in FIG. 5 is actually described in an SMF.

FIG. 7 is a diagram showing a data sequence when described according to a second embodiment of the method of the present invention.

8 is a diagram showing an abstraction of the data sequence shown in FIG. 7;

9 is a diagram showing a flow when forming the data string shown in FIG. 7;

FIG. 10 is a block diagram showing a reproducing apparatus of the present invention.

11 is a flowchart showing a playback procedure performed by the playback device shown in FIG.

[Explanation of symbols]

1 note area, 2 accent area, 3 accent start position information, 4 control area, 5 control start position information, 1
0: playback device, 11: compressed file recording medium, 12:
Decoding means, 13 ... performance information storage medium, 14 ... reproduction control means, 15 ... MIDI sound source section (sound source section).

Claims (4)

(57) [Claims] And 1. A note information indicating the start of sounding and pronunciation stopping of the musical tone, bets least one of the two information with the control information indicating the intonation of the performance and accent information indicating the intensity of sound for each sound is In the method for recording mixed performance information, the accent information and / or
Alternatively, the control information and the note information are collected for each information, and the collected information is stored in different areas on a recording medium independently by a pattern matching method.
Performance information recording method, comprising the Turkey be compressed and recorded as a compressed file. 2. The performance information recording method according to claim 1, wherein the performance information is MIDI information. 3. A playback apparatus for playing performance information stored by the performance information recording method according to claim 1, wherein the compressed file is read from a compressed file recording medium on which the performance information is recorded as the compressed file. Decoding means for decoding to obtain the performance information, a performance information storage medium for temporarily storing the obtained performance information, and reproduction control for sequentially reproducing the performance information in the performance information storage medium Means for reproducing the performance information output from the reproduction control means as a musical tone. 4. The apparatus according to claim 3, wherein the performance information is MIDI information.