JP2005173632A - Performance data generating apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a performance data generating apparatus capable of automatically editing performance parameters by modularizing expression adding rules and procedures and more particularly capable of adding expressions to the same timbre parts by control of volume parameters for the same timbre parts of the performance data. <P>SOLUTION: When the performance data (the former performance data) is supplied, the same timbre parts and the number of the parts are detected (S 141). Next, the volume values of the respective parts are calculated so as to reduce the volume of the parts of the same timbre (S 142) based on the detected number of the parts and the calculated musical tone control information representing the calculated volume value is imparted to the part of the former performance data (S 143). As a result, such an expression addition as to lower the volume of the respective parts can be performed at the time of a part division (to simulate the performance method of performing the parts of the same timbre by dividing the part of the same timbre to the plurality of parts and performing the respective parts with a plurality of persons). <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a performance data creation device for creating performance data to which facial expressions are given, in other words, parameter values for adding various facial expressions to performance data based on characteristics in the supplied performance data. The present invention relates to an automatic parameter editing system that automatically edits a parameter.

  If MIDI data is composed only of note information, the performance becomes mechanical and expressionless. In order to obtain various performance outputs with expressions such as more natural performances, beautiful performances, lively performances, hazy individual performances, etc., it is necessary to add various musical expressions and instrumentality as control data is there. As a system for adding various expressions, a method of adding expressions through musical scores can be considered. However, there are various facial expressions as described above. It can be said that it is a meaningful system if it can handle various facial expressions.

  Also, if you don't devise a system that automatically adds the number of facial expressions, you will be confused about what facial expressions to add. Is preferably modularized. Therefore, it can be seen that it is preferable to prepare a module in which various features of music expression are ruled in order to create musical MIDI data.

  Conventionally, the following are known as performance data creation apparatuses or parameter editing apparatuses for editing parameter values for adding various expressions to performances reproduced from supplied performance data.

That is,
(1) A mechanical input that is reproduced from the performance data by manually changing the value of a parameter already set or adding a new parameter to the supplied performance data. An expression that can add facial expressions (musical expressions such as natural performances, beautiful performances, live performances, etc.) to an expressionless performance (2) The performance data within the range specified by the user in the supplied performance data It is known that a performance expression (for example, crescendo or decrescendo effect) can be automatically added.

  However, in the above-described conventional performance data creation device or parameter editing device (1), the user has to select the parameter to be corrected or added and determine its value, and in particular the user If you are a beginner, it is difficult to select the parameters to add your favorite facial expression and determine the optimal parameter value for that facial expression. Become.

  Further, in the above-described conventional performance data creation apparatus or parameter editing apparatus (2), the range that can be specified by the user is a partial range that is local to the entire song data. The performance expression added based on the performance data belonging to the group must be simple (for example, a performance expression that linearly increases or decreases the volume of the part), and therefore, various expressions desired by the user can be obtained. In order to add a performance expression, the same operation had to be repeated many times while sequentially changing the range to which the performance expression should be added, and various performance expressions could not be added with simple operations. In addition, since the user must specify the range to which the performance expression should be added, the user knows what kind of expression should be added to what kind of performance point to obtain the optimal expression for the song. This is a difficult task for a novice user.

  In view of such circumstances, this invention pays attention to various rules and procedures when performing facial expression by temporal changes such as tempo and timing, and provides a novel facial expression converter in which this is modularized. It is an object of the present invention to provide a performance data creation system (parameter automatic editing system) capable of automatically converting original expressionless musical sound data into performance data that enables various musical performances.

  In addition, the present invention has been made paying attention to the above-mentioned points, and even if the user is a beginner, a performance data creation system (automatic parameter automatization) that can give various expressions to a song with a simple operation. The purpose is to provide an editing system.

  The present invention further automatically edits various control parameters related to performance such as time change parameters such as tempo and timing, volume parameters, pitch parameters such as pitch, etc. An object of the present invention is to provide a performance data creation system (parameter automatic editing system) capable of adding various expressions to a song with a simple operation.

  In particular, the present invention provides a performance data creation device that can easily perform effective expression on the same tone color part by volume parameter control for the same tone color part of the performance data, and according to the main features of the present invention. And performance data supply means (AB: FIG. 2) for supplying performance data composed of a plurality of parts each including note information and musical tone control information representing the volume, and detecting the parts of the same tone color and the number of parts from the performance data. Part detection means [S141], volume value calculation means [S142] for calculating the volume value of each part so that the volume of the part of the same tone color is reduced based on the number of parts, and tone control representing the volume value There is provided a performance data creation device comprising volume value giving means [S143] for giving information to a corresponding part [Embodiment (N): FIG. Corresponding] to. Note that the parentheses indicate the corresponding symbols, terms, and reference parts of the examples added for convenience of understanding, and the same applies to the following.

  According to the main feature of the present invention, when performance data (original performance data) consisting of a plurality of parts each including note information and musical tone control information indicating volume is supplied [AB: FIG. 2], the performance data has the same tone color. The part and the number of parts are detected [S141]. Then, based on the number of detected parts, the volume value of each part is calculated so that the volume of the detected same tone color part is reduced [S142], and new musical tone control information representing the calculated volume value is obtained. The performance data assigned to the corresponding part [S143] and expressed with new musical tone control information is used for reproduction [SB: FIG. 2].

  Therefore, according to the present invention, a volume value that extracts the same tone color part and the number of parts from the supplied performance data and reduces the volume of the extracted part to a value corresponding to the number of extracted parts. Since the musical tone control information is generated and added to the performance data, part division (simulating a performance method in which parts of the same tone color are divided into multiple parts and each part is played by multiple players) In addition, it is possible to add expressions to lower the volume of each part.

  In this specification, the “musical sound control amount” refers to a variable for controlling a musical tone in performance, such as a temporal musical sound control amount, a musical musical sound control amount, and a volume control amount. “Musical sound control information” refers to temporal musical sound control information (temporal parameters) such as tempo, gate time, and sounding start timing, musical musical tone control information (pitch parameters, etc.), and volumetric control information (volume parameter). The variable information for controlling the musical tone of the performance, such as “performance parameter” or simply “parameter”. The “performance data creation device” according to the present invention can be called a “parameter automatic editing device” from the viewpoint of editing performance parameters.

[Other features]
Moreover, it can comprise as follows by another characteristic regarding implementation of the various invention described in this specification.
[1] Supply means for supplying performance data, acquisition means for acquiring characteristic information of the supplied performance data, and a generation method for generating musical tone control information, which is information corresponding to predetermined characteristic information From the storage means for storing the generation method information to be represented, the acquired feature information, and the generation method information corresponding to the feature information, the generation means for generating musical sound control information, and the generated musical sound control information are supplied. A performance data creation device comprising an adding means for adding to the performance data (in the embodiment, FIG. 2 and FIGS. 52 and 53 correspond):
[1a] The performance data creating apparatus further includes an output means for outputting the performance data to which the musical tone control information is added by the adding means, and the output performance data is evaluated, and the musical tone generated according to the evaluation result Adjusting means for adjusting control information (corresponding to the embodiments (1) to (24) and (27) to (30)).

  In another feature [1], generation method information is stored in advance, musical tone control information is generated based on the generation method information corresponding to the characteristic information obtained from the performance data, and this is added to the performance data. That is, the correspondence between the predetermined feature information and the musical tone control information for facial expression is set as a rule in the facial expression module (facial expression algorithm), and generation method information representing the facial expression rule is stored in the storage means in advance. When the feature information of the supplied performance data is acquired, the musical tone control information (time parameter, pitch parameter, volume) is obtained according to the expression module (expression algorithm) based on the generation method information corresponding to the acquired feature information. Various performance parameters such as parameters) are generated and added to the performance data, so that various expressions can be added to the song with simple operations even if the user is a beginner according to the acquired feature information. More musical performance data can be automatically created. Further, according to another feature [1a], the performance data output by adding the musical tone control information is evaluated, and the musical tone control information is adjusted according to the evaluation result. It can be performed.

[2] Supply means for supplying performance data, extraction means for extracting feature information corresponding to a note generation time interval from the supplied performance data, and musical sound corresponding to feature information corresponding to a note generation time interval Storage means for storing generation method information representing a generation method for generating control information, generation means for generating musical sound control information based on the extracted feature information and generation method information, and generated musical sound control information A performance data creating device comprising an adding means for adding to the supplied performance data (corresponding to the embodiments (1), (3), (C)):
[2a] The characteristic information is information representing the number of notes per predetermined unit time, and the generation method information indicates the value of the reproduction tempo of the performance data when the number of notes per predetermined unit time exceeds the predetermined number. This is information representing a generation method for generating musical tone control information to be made faster [corresponding to the embodiment (1)],
[2b] The characteristic information is information representing the sound generation time interval of the notes, and the generation method information is generated as tone control information that slows down the playback tempo value of the performance data when the sound generation time interval of the notes is narrow. Information indicating the generation method to be performed (corresponding to the embodiment (3)), or
[2c] The characteristic information is information representing the number of notes per predetermined unit time, and the generation method information gradually decreases the volume of performance data when the number of notes per predetermined unit time exceeds the predetermined number. This is information representing a generation method for generating such tone control information (corresponding to the embodiment (C)).

  In another feature [2], feature information (note time information) corresponding to the note generation time interval is extracted from the supplied performance data, and the musical tone control information is based on the generation method information corresponding to the feature information. Is generated and added to the performance data, so that it is possible to obtain a performance output with various expressions based on the note time information (note density, note interval, etc.).

  Here, note density information (for example, “number of notes in one measure / number of beats in one measure”) representing the number of notes per predetermined unit time is extracted as feature information by another feature [2a] If the number of notes per hour exceeds the specified number, the performance data playback tempo value can be increased to add a facial expression that changes the tempo according to the note density (tempo acceleration by increasing the number of notes). It can be performed. In this case, the tempo change amount calculated for each section based on the set tempo strength value α, the note density and the tempo coefficient K obtained from the table, and the current set tempo value (current set tempo value × α × K) To MIDI data can be taken. Also, the original performance data is displayed on the display, and the assigned result can be confirmed by displaying the parameter value and the position provided on the displayed original performance data in an overlapping manner. Further, when performance data composed of a plurality of parts is supplied, the playback tempo may be changed based on the note density information extracted from the performance data of the selected predetermined part, or the plurality of parts The playback tempo may be changed according to the note density information evaluated as described above.

  In addition, if another feature [2b] is used to extract information representing the sound generation time interval of the notes as feature information, and when the sound generation time interval of the notes is narrow, the playback tempo value of the performance data is slowed down. When the notes are fine, you can add a facial expression that will slow down the tempo.

  Furthermore, information representing the number of notes per predetermined unit time is extracted as feature information by another feature [2c], and the volume of the performance data is gradually increased when the number of notes per predetermined unit time exceeds the predetermined number. If it is configured so as to decrease, it is possible to perform facial expression that changes the volume according to the fineness of the performance sound. For example, a part where the note density is equal to or higher than a predetermined value is detected as a difficult part, and a volume change is given to the part. Furthermore, the volume can be changed according to the change range of the maximum pitch value and the minimum pitch value of the difficult part. This method can also be applied to expressions that add instability to the pitch. In addition, it calculates the frequency of hammering-on in parts where the note density is high, and automatically performs hammering-on on performance data accordingly. By adding, it can also be applied to give a facial expression that hammers on when the notes are fine.

[3] Supply means for supplying performance data, storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to the progress state of the performance data, and the progress state of the supplied performance data And a performance data creation device comprising a generation means for generating musical tone control information based on the generation method information and an adding means for adding the generated musical tone control information to the supplied performance data [Embodiment (2), (B) corresponds]: where
[3a] The generation information information is information indicating a generation method for generating musical tone control information that gradually increases the reproduction tempo of the performance data as the performance data progresses (corresponding to the embodiment (2)), or ,
[3b] The generation information information is information representing a generation method for generating musical tone control information that gradually increases the volume as the performance data progresses (corresponding to the embodiment (B)).

  In another feature [3], the musical tone control information is generated and added to the performance data based on the generation method information corresponding to the progress state of the supplied performance data. Based on the evaluation, performance output with various expressions can be obtained.

  Here, according to another feature [3a], if the playback tempo of the performance data is gradually increased according to the progress of the performance data, for example, the tempo is gradually increased at a predetermined magnification for each measure as the music progresses. For example, it is possible to add a facial expression that gradually accelerates the tempo throughout the song. In this case, specifically, it is effective to calculate the tempo change amount for each section from the tempo strength value α, the tempo coefficient K, and the current tempo value for each section (the tempo coefficient K is calculated from the table for each section). The tempo gradually increases as the song progresses). Further, the tempo strength value α is named as the degree of excitement and can be selected with words such as “strong”, “medium”, “weak”, “no change”, and the usefulness can be further enhanced.

  Further, if another volume [3b] is used to gradually increase the value of the volume parameter as the performance data progresses, it is possible to add a facial expression that enhances the excitement level and produces a climax as the music progresses. it can. In this case, specifically, for example, a change value for each volume change section is calculated on the basis of pattern data representing the volume change tendency of the entire song, and is inserted into each corresponding position of the performance data. In this case, in a system having a plurality of tracks, the change imparting pattern used in each track may be the same or different. Note that this method can also be applied when changing a pitch parameter. For example, as the music progresses, the value of the pitch parameter can be gradually increased to increase the degree of excitement.

[4] Supply means for supplying performance data, extraction means for extracting from the supplied performance data portions where the sound length is greater than or equal to a predetermined length and to which trill or vibrato is applied, and the sound length is greater than or equal to a predetermined length And storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to a portion to which trill or vibrato is given, and musical tone control based on the extracted portion and the generation method information A performance data creation device comprising generation means for generating information and addition means for adding the generated musical tone control information to the supplied performance data (corresponding to the embodiments (4) and (G)): ,
[4a] The generation method information is information representing a generation method for generating musical tone control information that gradually and gradually increases the reproduction tempo value of the performance data according to the elapsed time of the trill or vibrato [ Corresponding to Example (4)], or
[4b] The generation method information is information representing a generation method for generating musical tone control information in which the volume is initially large, midway small, and finally large (corresponding to Example (G)).

  In another feature [4], from the supplied performance data, a portion (microvibration sound information) to which the sound length is equal to or longer than a predetermined length and to which trill or vibrato is attached is extracted as feature information, Based on the corresponding generation method information, musical tone control information is generated and added to performance data, so performance output with various expressions based on minute vibration sound information (long tone trill, vibrato, etc.) Can be obtained.

  Here, according to another feature [4a], if the playback tempo value of the performance data is initially slow and gradually increased according to the elapsed time of the trill or vibrato, the trill / vibrato portion of the long tone Can do facial expressions with a slow tempo at first. In this case, it is effective to obtain a tempo coefficient K corresponding to the time lapse of the trill or vibrato from a table prepared in advance and generate a pitch bend sequence for each section based on the value, the tempo strength value α, and the current tempo value. It is.

  Further, when another volume [4b] is configured such that the volume is initially large, small in the middle, and large in the end, the volume when trill or vibrato is added to the long tone is increased from small to large. It is possible to add facial expressions that change. This method can also be applied to changing the speed of pitch change.

[5] Supply means for supplying performance data, extraction means for extracting a portion to which pitch bend is applied from the supplied performance data, and generating tone control information corresponding to the portion to which pitch bend is applied Storage means for storing the generation method information representing the generation method, generation means for generating musical tone control information based on the extracted portion and the generation method information, and the generated musical tone control information in the supplied performance data An apparatus for creating performance data comprising additional means for adding [corresponding to the embodiments (20), (23), (Q)]:
[5a] The generation method information is information representing a generation method for generating musical tone control information that gradually decreases the value of the playback tempo of performance data based on a change in pitch bend [corresponding to Example (20). ],
[5b] The generation method information is information representing a generation method for generating musical tone control information that gradually decreases the value of the playback tempo of the performance data when the pitch bend is deep or fast [Example ( 23)] or
[5c] The generation method information is information representing a generation method for generating musical tone control information that reduces the volume as the rate of change in pitch bend increases (corresponding to Example (Q)).

  In another feature [5], a portion to which pitch bend is applied is extracted as feature information from the supplied performance data, and tone control information is generated based on the generation method information corresponding to the feature information. Since it is added to the data, it is possible to obtain a performance output with various expressions based on minute vibration sound information (pitch bend, etc.).

  Here, according to another feature [5a], if the playback tempo value of the performance data is gradually decreased based on the change in pitch bend, for example, a tempo change based on the detected pitch bend is given. Thus, it is possible to add a facial expression that touches the tempo when pitch bend is applied.

  Another feature [5b] is that when the pitch bend is deep or fast, if the playback tempo value of the performance data is gradually slowed down, the vibrato is deeply applied or fast applied. It is possible to add a facial expression that sets the tempo according to the depth and speed of the vibrato, such as slowing down the tempo. In this case, for example, it is effective to detect a vibrato, draw a table from the depth and speed, calculate a tempo coefficient K, and add a tempo change based on the calculated value to the original performance data.

  Further, according to another feature [5c], when the pitch bend change rate is increased, the volume is decreased (for example, greatly decreased), and the expression for changing the volume when the pitch bend is changed can be performed.

[6] Supplying means for supplying performance data, extraction means for extracting a phrase delimiter from the supplied performance data, and a generation method representing a generation method for generating musical tone control information corresponding to the phrase delimiter Storage means for storing information, generation means for generating musical tone control information based on the extracted part and generation method information, and addition means for adding the generated musical tone control information to supplied performance data Performance data creation device (corresponding to the embodiments (5), (F), (f)):
[6a] The generation method information is information representing a generation method for generating musical tone control information that slows the value of the playback tempo of the performance data at the phrase delimiter part (corresponding to the embodiment (5)).
[6b] Information indicating a generation method for generating musical tone control information for gradually decreasing the volume at the phrase delimiter [corresponding to the embodiment (F)], or
[6c] The generation method information is information representing a generation method for generating musical tone control information in which information on a damper on is inserted at a phrase start position and information on a damper off is inserted at a phrase end position in a phrase delimiter. [Corresponding to Example (f)].

  In another feature [6], the musical tone control information is generated based on the generation method information corresponding to the phrase delimiter extracted from the supplied performance data, and is added to the performance data. Performance output with various expressions can be obtained at the breaks (phrase end portion, etc.).

  Here, due to another feature [6a], the playback tempo value of the performance data is slowed down at the phrase delimiter (for example, a tempo change that slows down the tempo by detecting the break position of the phrase from the original performance data). (Append to the original performance data), the tempo can be slowed down at the end of the phrase section, and a slow expression can be applied at the phrase passage part.

  Another feature [6b] is that if the volume is gradually reduced at the phrase delimiter, the volume is reduced at the end of the phrase, and the volume is gradually reduced at the end of the phrase. Facial expression that gives a feeling of termination can be performed. In this case, the volume at the end position is preferably calculated based on the tempo set for the phrase. Note that the volume of the note at the beginning of the phrase may be increased.

  Further, according to another feature [6c], information on the damper on is inserted at the start position of the phrase, and information on the damper off is inserted at the end position of the phrase (for example, by interpreting the phrase, If the break is detected, the damper is turned on at the start position of the phrase and the damper is turned off at the end position of the phrase), it is possible to add a facial expression with the sustain pedal of the piano.

[7] Supply means for supplying performance data, calculation means for calculating average pitch information or smoothed pitch information for each predetermined section from the supplied performance data, average pitch information or smoothed Storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to the pitch information, and based on the calculated average pitch information or smoothed pitch information and generation method information A performance data creation device comprising generation means for generating musical sound control information and addition means for adding the generated musical sound control information to the supplied performance data (corresponding to the embodiments (6) and (M)). :here,
[7a] The generation method information is information representing a generation method for generating musical tone control information that increases the reproduction tempo value of the performance data as the pitch of the average pitch information or the smoothed pitch information is higher. [Corresponding to Example (6)].

  In another feature [7], average pitch information for each predetermined section or smoothed pitch information is calculated from the supplied performance data, and a musical tone is calculated based on generation method information corresponding to the pitch information. Since the control information is generated and added to the performance data, it is possible to obtain a performance output with various expressions corresponding to predetermined pitch information. By this method, for example, it is possible to give a change in volume based on a pitch average value for each section [see the embodiment (M)], and this method can also be applied to a pitch.

  Here, according to another feature [7a], the higher the pitch of the average pitch information or the smoothed pitch information, the faster the reproduction tempo value of the performance data becomes. By adopting the rule to speed up the expression, it is possible to add facial expressions that change the tempo according to the sound range. Specifically, for example, a table is subtracted from the average pitch to obtain a tempo coefficient K, and a tempo change for each section is added to the original performance data based on the value, the tempo strength value α, and the current tempo value. The pitch change may be smoothed by filtering, and the tempo coefficient K may be obtained from each pitch value and table on the pitch change curve after filtering.

[8] Supply means for supplying performance data, extraction means for extracting feature information corresponding to the change tendency of the pitch from the supplied performance data, and musical sound corresponding to the feature information corresponding to the change tendency of the pitch Storage means for storing generation method information representing a generation method for generating control information, generation means for generating musical sound control information based on the extracted feature information and generation method information, and generated musical sound control information , A performance data creation device comprising an adding means for adding to the supplied performance data (corresponding to the embodiments (7) and (A)):
[8a] The characteristic information is information indicating the switching position between the upward trend and the downward trend of the pitch, and the generation method information is the reproduction of the performance data at the switching position of the upward trend and the downward trend of the pitch. Information indicating a generation method for generating musical tone control information that slows down the tempo value (corresponding to the embodiment (7)),
[8b] The characteristic information is information indicating a switching position between the upward trend and the downward tendency of the pitch, and the generation method information is the position of the switching position at the switching position of the upward trend and the downward trend of the pitch. Information indicating a generation method for generating musical tone control information that gives an accent to the volume of a note (corresponding to the embodiment (A)), or
[8c] The characteristic information is information representing an upward trend portion of the pitch, and the generation method information is information representing a generation method for generating musical tone control information that gradually increases the volume in the upward trend portion. [Corresponding to Example (A)].

  In another feature [8], pitch change tendency information corresponding to a pitch change tendency is extracted as feature information from the supplied performance data, and based on generation method information corresponding to the pitch change trend information. Since the musical tone control information is generated and added to the performance data, it is possible to obtain a performance output with various expressions according to the pitch change tendency.

  Here, according to another feature [8a], the pitch change tendency information represents a switching position between an upward trend and a downward tendency of the pitch, and the performance data at the switching position of the upward trend and the downward trend of the pitch. If the playback tempo value is configured to be slow, the tempo is once slowed when switching from the rising part to the lowering part of the pitch, and a facial expression is added to the point where the pitch rises and falls. be able to. Specifically, for example, a method is used in which a tempo change in which the tempo gradually decreases near the tip of the pitch increasing portion is added to the original performance data using the pitch change curve and table of the original performance data. be able to.

  Further, according to another feature [8b], the pitch change tendency information indicating the switching position between the upward trend and the downward trend of the pitch is similarly used, and the switching position between the upward trend and the downward trend of the pitch is used. , If you add an accent to the volume of the note at the switching position (for example, add an accent to the note event at the change point when changing from ascending to descending) It can be performed.

  Further, according to another feature [8c], the pitch change tendency information representing the upward trend portion of the pitch is used as the feature information from the supplied performance data, and the volume is gradually increased in the upward trend portion. Then, when the pitch of the note event sequence tends to increase, facial expression that gradually increases the volume can be performed. Specifically, for example, the user sets an analysis section in the performance data, and searches for a partial section in which the pitch of the note event sequence tends to rise and fall from the analysis section (in this case, the whole trend tends to rise) The part is assumed to be an upward trend part.), The change speed of the pitch is calculated for each searched partial section, and according to the result, the change application pattern of the volume parameter to be given to the note event is determined and determined. A method of changing the volume parameter value in the partial section based on the change giving pattern is adopted. Note that this method can also be applied when changing a pitch parameter.

[9] Supply means for supplying performance data, extraction means for extracting a portion where the same or similar data string is continuously present from the supplied performance data, and the same or similar data string are continuously present Storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to a portion to be generated, generation means for generating musical tone control information based on the extracted feature information and generation method information, and generation A performance data creation device (corresponding to the embodiments (8) and (D)), characterized in that it comprises an adding means for adding the musical tone control information to the supplied performance data:
[9a] The generation method information is a generation method for generating musical tone control information that changes the value of the reproduction tempo of the performance data with respect to the subsequent data sequence of the same or similar data sequence that exists continuously. Information representing [corresponding to Example (8)], or
[9b] The generation method information represents a generation method for generating musical tone control information that changes the sound volume according to the degree of similarity with respect to the data sequence on the back side of the same or similar data sequence that exists continuously. Information [corresponding to Example (D)].

  In another feature [9], a portion where the same or similar data string is continuously present is extracted as feature information from the supplied performance data, and the musical sound is extracted based on the generation method information corresponding to the feature information. Since the control information is generated and added to the performance data, a performance output with various expressions can be obtained when the same or similar data string exists continuously.

  Here, by another feature [9a], it is the same or similar if the playback tempo value of the performance data is changed with respect to the subsequent data sequence of the same or similar data sequence that exists continuously When the patterns are continuous, the second pattern can be expressed with a slow or fast tempo.

  Further, when another volume [9b] is configured so that the volume is changed in accordance with the degree of similarity with respect to the data sequence on the back side of the same or similar data sequence that is continuously present, the same or similar When the pattern appears continuously, it is possible to perform facial expression that changes the volume parameter according to the similarity of the pattern. For example, when similar phrases repeatedly appear, the volume parameter of the second and subsequent similar phrases is changed according to the similarity and appearance mode. In this case, when phrases with high similarity appear continuously, the volume parameter value of the second and subsequent similar phrases is set to be smaller than that of the first similar phrase. In addition, when similar phrases repeatedly appear without being consecutive, the value is similar to that of the first similar phrase, and is changed to a value corresponding to the degree of similarity. Note that this method can be applied to facial expression that changes a pitch parameter.

[10] A reproduction tempo value of performance data is changed based on a difference between the supply means for supplying performance data, the extraction means for extracting a similar data string from the supplied performance data, and the similar data string Storage means for storing generation method information representing a generation method for generating musical tone control information, generation means for generating musical tone control information based on the extracted data string and generation method information, and A performance data creation device comprising an adding means for adding musical tone control information to the supplied performance data (corresponding to the embodiment (9)):

  In another feature [10], a similar data string is extracted from the supplied performance data, and the playback tempo value of the performance data is changed based on the difference between the similar data strings. Similar phrases can be expressed with the same or similar tempo. Specifically, for example, a difference between similar phrases is detected, and a tempo change based on the difference is added to the original performance data.

[11] A supply means for supplying performance data, an extraction means for extracting a pre-registered sound form from the supplied performance data, and a generation method for generating musical tone control information corresponding to the sound form Storage means for storing generation method information, generation means for generating musical tone control information based on the extracted sound form and generation method information, and addition means for adding the generated musical tone control information to the supplied performance data The performance data creation device comprising the above [corresponding to the embodiment (10)]:

  In another feature [11], a pre-registered sound form is extracted from the supplied performance data, and tone control information is generated based on the generation method information corresponding to the sound form and added to the performance data. Therefore, it is possible to add a facial expression that sets a tempo corresponding to the registered sound form. Specifically, for example, a predetermined tempo is set for a pre-registered tone form = rhythm pattern (phrase) portion. The phrase to be registered is not limited to the sound form, but may be an expression event sequence. Note that not only the tempo change but also the pitch parameter change can be applied in the same manner. In addition, this method can also add slur according to the type of musical instrument, as described in the embodiments.

[12] Supply means for supplying performance data, calculation means for calculating average volume information or smoothed volume information for each predetermined section from the supplied performance data, and average volume information or smoothed volume information Based on the storage means for storing the generation method information representing the generation method for generating the musical tone control information corresponding to, and the calculated average volume information or smoothed volume information and generation method information for each predetermined section A performance data creation device comprising generation means for generating control information and addition means for adding the generated musical tone control information to the supplied performance data (corresponding to the embodiment (11)):

  In another feature [12], average sound volume information for each predetermined section or smoothed sound volume information is calculated from the supplied performance data, and the musical tone control information is calculated based on the generation method information corresponding to the volume information. Since it is generated and added to the performance data, for example, it is possible to perform expression such that the tempo is made slower as the volume is lower. Specifically, for example, an average volume value is obtained for each section, and a predetermined tempo change is given to the original performance data based on a table in which the tempo is slower as the previously prepared volume is lower. Alternatively, each volume value on a volume change curve obtained by filtering volume changes may be used as a basis.

[13] Data supply means for supplying performance data, parameter input means for inputting tension parameters, and storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to the tension parameters And generating means for generating musical tone control information based on the supplied performance data, inputted tension parameter and generation method information, and adding means for adding the generated musical tone control information to the supplied performance data A performance data creation device (corresponding to Example (12)):

  In another feature [13], a tension parameter is input using a parameter input unit, and musical tone control information is generated based on the generation method information corresponding to the tension parameter and added to the performance data. Therefore, it is possible to input a parameter called tension according to the feeling of the entire song or the feeling of a specific part such as rust, and perform facial expression that changes the tempo according to the tension parameter. In order to change the tempo with tension, for example, a tension parameter is set using an operator, a corresponding tempo coefficient K is generated from the set tension parameter value, and the generated tempo coefficient K is set. A tempo change corresponding to the current tempo value is added to the original performance data.

[14] A supply means for supplying performance data, an extraction means for extracting a predetermined note group sequence from the supplied performance data, and a generation method for generating musical tone control information corresponding to the predetermined note group sequence Storage means for storing generation method information, generation means for generating musical tone control information based on the extracted note group sequence and generation method information, and addition for adding the generated musical tone control information to the supplied performance data A performance data creation device comprising the means [corresponding to the embodiments (13), (14), (E)]:
[14a] The generation method information represents a generation method for generating musical tone control information that lengthens the first note length of the note group sequence and corrects the remaining note length to a length that fits within the time length of the note group sequence. Information [corresponding to Example (13)], or
[14b] The generation method information is information representing a generation method for generating musical tone control information that divides a note group sequence into a plurality of group sequences and makes the first note length of each group sequence longer. Corresponding to (14)], or
[14c] The generation method information is information representing a generation method for generating musical tone control information that emphasizes the volume of the note appearing at the first timing of the note group sequence (corresponding to the embodiment (E)).

  In another feature [14], a predetermined note group sequence is extracted from the supplied performance data, and musical tone control information is generated based on the generation method information corresponding to the note group sequence and added to the performance data. Therefore, it is possible to obtain performance output with various expressions for a group of notes.

  Here, according to another feature [14a], when the first note length of the note group sequence is lengthened and the remaining note length is modified to be within the time length of the note group sequence, A predetermined temporal expression can be given to. For example, the first note of a group of notes grouped in a certain block, such as an enumeration of quarter notes of three time signatures, is lengthened, and an expression is given to regain the time used by the remaining notes. For this purpose, for example, a parameter called group note strength is set, and a tempo change is given to the original performance data based on the parameter value. It is also effective to apply only to the left hand of the piano or only to the accompaniment part.

  In addition, if the note group sequence is divided into a plurality of group sequences and the first note length of each group sequence is made longer by another feature [14b], an expression that emphasizes a predetermined sound in the lotus Can be attached. That is, the detected lotus mark is divided into a plurality of small lotus marks, and a tempo change corresponding to the divided small lotus marks is given to the performance data. In this case, a long tuplet of 5 or more lots is divided into small tuplets, for example, “double tuplet + triplet” or “triplet + double tuplet”. Emphasize the sound over time. Furthermore, a technique such as slowing down the tempo for the first small tuplet and increasing the tempo of the remaining small tuplets is also effective.

  Furthermore, if the volume of the note appearing at the first timing of the note group sequence is enhanced by another feature [14c], it is possible to add a facial expression to the note group sequence by enhancing the volume. For example, in the case of performance data of three beats, the first beat is emphasized when notes having the same note length appear continuously, and the volume change is corrected according to the number of notes at the first beat position. Furthermore, it is possible to express the character of a lotus by detecting the tuplet and increasing the volume of the first sound. Further, in the case of 5 beats, it is considered that 2 beats and 3 beats are alternately arranged, and the volume of each first beat is emphasized. The lotus mark can be considered in the same way.

[15] Supply means for supplying performance data, extraction means for extracting a portion where a plurality of sounds are simultaneously generated from the supplied performance data, and generating musical tone control information corresponding to a portion where a plurality of sounds are simultaneously generated Storage means for storing generation method information representing a generation method for generating, generation means for generating musical tone control information based on the extracted part and generation method information, and the generated performance control information generated by the generated musical tone control information. A performance data creation device comprising an adding means for adding data (corresponding to the embodiments (15) and (J)):
[15a] The generation method information is information representing a generation method for generating musical tone control information that changes the value of the reproduction tempo of the performance data in accordance with the number of sounds that are sounded simultaneously [Embodiment (15) Or)
[15b] The generation method information specifies a generation method for generating musical tone control information that defines the importance levels of sounds that are simultaneously generated and changes the volume of each sound according to the specified importance levels. Information [corresponding to Example (J)].

  In another feature [15], a portion where a plurality of sounds are simultaneously generated is extracted from the supplied performance data, and tone control information is generated based on the generation method information corresponding to this portion and added to the performance data. Therefore, it is possible to obtain a performance output with various expressions for a portion where a plurality of sounds are pronounced at the same time.

  Here, according to another feature [15a], the reproduction tempo value of the performance data is changed according to the number of sounds that are simultaneously generated [for example, the tempo according to the number of simultaneous pronunciations of chords (note density of chords) If it is configured so that the tempo can be changed according to the number of chord constituent sounds, expression can be performed.

  In addition, when another feature [15b] is used to define the importance level of sounds that are simultaneously generated, and to change the volume of each sound according to the specified importance level, multiple sound simultaneous sounding performance It is possible to apply an effective expression by volume. For example, the degree of importance is prepared in advance as a template, and the volume of the chord constituent sound is changed according to the degree of importance so as to increase the volume of the high importance sound. In this case, only the volume of the lowest tone and the highest tone of the chord may be set higher than the volume of the other component sounds, or only the volume of the root tone of the chord may be increased. Furthermore, the same applies to octave unison. It is also possible to apply this method to the pitch and automatically change the chord to a genuine tone.

[16] Supply means for supplying performance data, extraction means for extracting timbre information from the supplied performance data, and generation method information representing a generation method for generating musical tone control information corresponding to the timbre information are stored. Performance data comprising storage means, generation means for generating musical tone control information based on the extracted tone color information and generation method information, and addition means for adding the generated musical tone control information to the supplied performance data Creation device (corresponding to Examples (16) and (P)):
[16a] The generation method information is information representing a generation method for generating musical tone control information that changes the reproduction tempo value of the performance data in accordance with the timbre type or timbre parameter [corresponding to Example (16). Or
[16b] The generation method information is smooth from the volume corresponding to the timbre before the change to the volume corresponding to the timbre after the change at the timbre change position based on the volume for each timbre type defined in advance. This is information representing a generation method for generating musical tone control information to be changed to [corresponding to Example (P)].

  In another feature [16], timbre information is extracted from the supplied performance data, and tone control information is generated and added to the performance data based on the generation method information corresponding to the timbre information. Thus, it is possible to obtain a performance output having various expressions with respect to the predetermined tone color information in the performance data.

  In this case, if another playback tempo value of the performance data is changed according to the timbre type or timbre parameter according to another feature [16a], an expression in which the tempo automatically changes slightly every time the timbre changes is expressed. Can be obtained.

  Further, according to another feature [16b], the volume corresponding to the tone after the change from the volume corresponding to the tone before the change at the position where the tone is changed, based on the volume for each tone type specified in advance. If it is configured to change smoothly, it is possible to make a facial expression with a volume expression corresponding to a timbre change and to make a performance with a specific timbre stand out. .

[17] Supply means for supplying performance data, extraction means for extracting information relating to fingering from the supplied performance data, and generation representing a generation method for generating musical tone control information corresponding to information relating to fingering Storage means for storing method information, generation means for generating musical tone control information based on the extracted fingering information and generation method information, and addition for adding the generated musical tone control information to the supplied performance data A performance data creation device comprising the means [corresponding to the embodiments (17) to (19), (S)]:
[17a] The generation method information defines information related to fingering corresponding to a portion that is difficult to perform, and a generation method that generates musical tone control information that slows the value of the playback tempo of the performance data in the portion that is difficult to perform [Corresponding to Example (17)],
[17b] The generation method information defines information related to fingering corresponding to the position movement part, and a generation method for generating musical tone control information that gives fluctuation to the value of the playback tempo of the performance data in the position movement part. Information (corresponding to Example (18)),
[17c] The generation method information defines information related to fingering corresponding to a portion played at a low position, and musical tone control information that slows the reproduction tempo value of the performance data at the portion played at the low position. Information indicating the generation method to be generated (corresponding to Example (19)),
[17d] The generation method information is information indicating a generation method for defining fingering information corresponding to a portion that is difficult to perform and generating musical tone control information that reduces the volume in the portion that is difficult to perform. Corresponds to the embodiment (S)], or
[17e] The generation method information is information representing a generation method that defines information related to fingering corresponding to the position moving portion and generates musical tone control information that changes the pitch in the position moving portion. Corresponding to (S)].

  In another feature [17], information related to fingering is extracted from the supplied performance data, and tone control information is generated based on the generation method information corresponding to the information related to fingering and added to the performance data. Therefore, it is possible to obtain performance output with various expressions related to fingering.

  Here, another feature [17a] defines information related to fingering corresponding to a portion that is difficult to play, and in such a portion that is difficult to play, it is difficult to move if the playback tempo value of performance data is slowed down. It is possible to add a facial expression that changes the tempo according to the performance finger, such as slowing down the tempo of performance with a finger (a finger that is difficult to perform). For this purpose, for example, fingering fingers are detected, and a tempo change is calculated using a table prepared in advance.

  Another feature [17b] is that if information related to fingering corresponding to the position moving portion is used and fluctuation is given to the value of the playback tempo of the performance data in the position moving portion, the lower the position, the faster the performance. In order to reproduce the tendency to be unsuitable, it is possible to add a facial expression that slows down the tempo at a low position.

  According to another feature [17c], if information related to fingering corresponding to a portion played at a low position is used and the playback tempo value of performance data is slowed down at the portion played at the low position, the finger By setting the tempo coefficient according to the position of the ring, it is possible to express non-uniformity performed by humans.

  Furthermore, if another configuration [17d] is used to reduce the sound volume in the difficult-to-play part using information related to the fingering corresponding to the difficult-to-play part, it is difficult to perform the performance operation according to the fingering. It is possible to perform expression such that the volume of a possible pitch is relatively smaller than the volume of other pitches.

  According to another feature [17e], if information about the fingering corresponding to the position moving portion is used and the pitch is changed in the position moving portion, the pitch is automatically changed according to the position movement of the fingering. Facial expression can be performed. In consideration of fingering, a method of adding a noise sound can be applied when the volume is low at a low position.

[18] Supply means for supplying performance data, extraction means for extracting a portion corresponding to a specific musical instrument performance from the supplied performance data, and generation for generating musical tone control information corresponding to the specific musical instrument performance A storage unit that stores generation method information representing a method; a generation unit that generates musical tone control information based on the extracted portion corresponding to the specific musical instrument performance method and the generation method information; and a generated musical tone control information. A performance data creation device comprising an adding means for adding to the performance data [corresponding to the embodiments (21), (22), (24), (H), (T)]:
[18a] The specific musical instrument playing method is a hammering-on playing method or a pulling-off playing method, and the generation method information is a musical tone that speeds up the reproduction tempo value of the performance data in the portion corresponding to the hammering-on playing method or the pulling-off playing method. Information indicating a generation method for generating control information (corresponding to the embodiment (21)),
[18b] The specific musical instrument playing method is a piano sustain pedal playing method, and the generation method information is generated to generate musical tone control information that slows down the playback tempo value of the performance data in the portion corresponding to the piano sustain pedal playing method. Information representing the method (corresponding to Example (22)),
[18c] The specific musical instrument playing method is a string trill playing method, and the generation method information divides the performance data in the portion corresponding to the string trill playing method into a plurality of parts, and the reproduction tempo of the performance data, which is different for each part. Information indicating a generation method for generating musical tone control information for setting a value (corresponding to the embodiment (24)),
[18d] The specific musical instrument playing method is a trill or drum roll playing method, and the generation method information is a generating method for generating musical tone control information that makes the volume of notes in the portion corresponding to the trill or drum roll playing method non-uniform. [Corresponding to Example (H)], or
[18e] The specific musical instrument playing method is a bow playing method of a stringed instrument, and the generation method information is a generating method for generating musical tone control information that changes the volume of a nearby note in the portion corresponding to the bow playing method of the stringed instrument. Information to be represented (corresponding to Example (T)).

  In another feature [18], a portion corresponding to a specific musical performance method is extracted from the supplied performance data, musical tone control information is generated based on the generation method information corresponding to the specific musical performance method, and the performance data Therefore, it is possible to obtain a performance output with various expressions corresponding to a specific musical instrument playing method.

  Here, due to another feature [18a], the specific musical instrument playing method is a hammering-on playing method or a pulling-off playing method, and in the portion corresponding to the hammering-on playing method or the pulling-off playing method, the playback tempo value of the performance data is increased. If comprised in this way, the expression which makes a tempo quick can be performed at the time of guitar hammering-on or pulling-off. In order to accelerate the tempo at the time of pulling off in this manner, hammering on and pulling off are detected, and a tempo change based on the detection result is added to the performance data.

  According to another feature [18b], if the specific musical instrument playing method is a piano sustain pedal playing method, and the portion corresponding to the piano sustain pedal playing method is configured to slow down the playback tempo value of the performance data, the piano sustaining method is achieved. It is possible to add a facial expression that sets the tempo according to the sustain pedal operation, such as slightly slowing down the tempo according to the pedal operation. For this, for example, a sustain bedal-on is detected, and a tempo change based on the detection result is added to the performance data.

  Further, according to another feature [18c], the specific musical instrument playing method is a string trill playing method, and the performance data of the portion corresponding to the trill playing method that maintains a minute fluctuation (vibration) sound in the string is divided into a plurality of parts. If a different value is set for each playback tempo, expression of the string trill can be performed by using a plurality of parts and slightly shifting the timing of these parts. To this end, for example, the toll part is detected, the part is copied to a plurality of parts, and the MIDI data timing is changed for each part. In this case, different timbre characteristics can be set for each part. This method can also be applied to change the pitch parameter. In this case, the tolyl pitch in each decomposed part may be slightly shifted.

  Further, according to another feature [18d], when the specific musical instrument playing method is a trill or drum roll playing method and the volume of the notes corresponding to the trill or drum roll playing method is made non-uniform, It is possible to add an expression to the roll performance part so that the volume of each note becomes non-uniform.

  Further, according to another feature [18e], if the specific musical instrument playing method is a bowing method of a stringed instrument bow, and the volume of the neighboring notes corresponding to the bowing method of the stringed instrument bow is changed, the bowing of the stringed instrument is returned. It is possible to add a facial expression that gives a change in volume near the position.

[19] Supply means for supplying performance data, extraction means for extracting lyrics information from the supplied performance data, and generation method information representing a generation method for generating musical tone control information corresponding to the lyrics information are stored. Storage means; generation means for generating musical tone control information based on the extracted lyric information and generation method information; and addition means for adding the generated musical tone control information to the supplied performance data. Characteristic performance data creation device (corresponding to the embodiments (27) and (U)):
[19a] The generation method information defines a tempo control value for a specific word, and represents a generation method for generating musical tone control information that changes the reproduction tempo value of performance data based on the specified content. Information [corresponding to Example (27)], or
[19b] The generation method information defines the volume change for a specific word, and is information representing a generation method for generating musical tone control information that changes the volume based on the specified content [Example ( Corresponding to U)].

  In another feature [19], the lyrics information is extracted from the supplied performance data, and the musical tone control information is generated and added to the performance data based on the generation method corresponding to the lyrics information. When there is information, performance output with various expressions can be obtained.

  Here, according to another feature [19a], if the playback tempo value of the performance data is changed based on the tempo control value for a specific word, facial expression setting for setting the tempo according to the lyrics is performed. Can do. For this facial expression, for example, a predetermined word is registered in advance together with a tempo coefficient, the predetermined word is detected from the lyrics data of the original performance data, and when the word appears, a tempo change based on the word is performed. Is added to the performance data to change the tempo. In this case, a fast tempo is set for bright words, and a slow tempo is set for dark words and important words.

  In addition, when another volume [19b] is used to change the volume based on the volume change for a specific word, it is possible to add a facial expression that gives a volume change to a predetermined word. Such a parameter processing method associated with the lyric information can be applied to a pitch change.

[20] Supply means for supplying performance data, acquisition means for acquiring waveform information related to the output waveform from the sound source from the supplied performance data, and tone control corresponding to the output waveform from the sound source according to the type of tone Storage means for storing generation method information representing a generation method for generating information, generation means for generating musical sound control information based on the acquired waveform information and generation method information, and generated musical sound control information, A performance data creation device comprising an adding means for adding to the supplied performance data (corresponding to the embodiment (28)):

  In another feature [20], the musical tone control information is obtained based on the generation method information corresponding to the output waveform from the sound source corresponding to the tone type, by acquiring information related to the output waveform from the sound source from the supplied performance data. Is generated and added to the performance data.For example, while observing the actual output waveform, the on-time of the note is corrected so as to differ depending on the type of timbre, and according to the output waveform of the sound source. Expression to correct the timing.

[21] Supply means for supplying performance data, extraction means for extracting information on performance symbols from the supplied performance data, and generation method information representing a generation method for generating musical tone control information corresponding to the performance symbols Storage means for storing, generation means for generating musical tone control information based on information on the extracted performance symbols and generation method information, and addition means for adding the generated musical tone control information to the supplied performance data A performance data creation device (corresponding to the embodiments (29), (30), (K)):
[21a] The information about the performance symbol is a piano symbol and a forte symbol, and the generation method information is attached when the forte symbol is attached to the note immediately before the note with the piano symbol attached. This is information representing a generation method for generating musical tone control information that shortens the pronunciation length of the current note (corresponding to the embodiment (29)),
[21b] The information regarding the performance symbol is a staccato symbol, and the generation method information is information indicating a generation method for generating musical tone control information that changes the pronunciation length of the note immediately before the note to which the staccato symbol is attached. Yes [corresponding to Example (30)], or
[21c] The information regarding the performance symbol is a staccato symbol, and the generation method information is information indicating a generation method for generating musical tone control information that reduces the volume of the note immediately after the note to which the staccato symbol is attached. [Corresponding to Example (K)].

  In another feature [21], information on performance symbols is extracted from the supplied performance data, and musical tone control information is generated based on the generation method information corresponding to the performance symbols and added to the performance data. Therefore, a performance output with various expressions can be obtained corresponding to the performance symbols in the performance data.

  Here, according to another feature [21a], when the performance symbol is a piano symbol and a forte symbol, and the forte symbol is attached to the note immediately before the note to which the piano symbol is attached, the forte symbol is attached. If the note length of the note is shortened, the forte symbol immediately before the piano symbol can be set to be shorter and can be expressed.

  According to another feature [21b], if the performance symbol is a staccato symbol and the pronunciation length of the note immediately before the note to which the staccato symbol is attached is changed, the pronunciation time of the sound immediately before the staccato symbol is lengthened. You can make facial expressions. For this, for example, staccato is detected, the gate time of the sound just before is expanded according to the strength value α, and a tempo change based on the result is given to the performance data.

  Further, according to another feature [21c], if the performance symbol is similarly a staccato symbol and the volume of the note immediately after the note with the staccato symbol attached is reduced, the staccato symbol is emphasized to enhance the staccato tone. It is possible to add a facial expression that reduces the volume of the note immediately after the pronunciation. In such a staccato performance, it is preferable to adjust the degree of change in volume according to the note value and tempo of the staccato note.

[22] Supply means for supplying performance data, storage means for storing the relationship between predetermined characteristic information of musical performance data already supplied and musical tone control information, and characteristic information of newly supplied performance data are extracted An extracting means for generating musical tone control information according to the relationship stored in the storage means based on the extracted feature information, and an addition for adding the generated musical tone control information to newly supplied performance data A performance data creation device comprising means (corresponding to the embodiment (25)):

  In another feature [22], the relationship between the predetermined feature information of the performance data already supplied and the musical tone control information is stored, and when the feature of the newly supplied performance data is extracted, the relationship is stored in the storage means. According to the relationship, musical tone control information is generated and added to newly supplied performance data, so that a facial expression for setting a tempo can be performed by a learning function. In this method, a system for automatically predicting the relationship between the pitch change and the tempo change is configured, and the remaining tempo change after the tempo change is manually input halfway through a certain song is automatically input by the learning function. For example, the relation between the phrase and the tempo change is learned from the MIDI data to which the tempo change has already been applied in the learning routine, and the result is stored. A tempo change based on the stored learning result is added to the performance data for the MIDI data of the phrase to which the tempo has not yet been assigned.

[23] A library storing a plurality of relationships between predetermined feature information and musical tone control information regarding performance data, supply means for supplying performance data, extraction means for extracting feature information of the supplied performance data, and extraction A performance data creation device comprising: generation means for generating musical tone control information by referring to the library based on the feature information, and addition means for adding the generated musical tone control information to the supplied performance data Example (26) corresponds]:

  In another feature [23], the relationship between predetermined feature information and musical tone control information regarding performance data is stored in the library, and the characteristic information of the supplied performance data is extracted. Since the control information is generated and added to the performance data, it is possible to add a facial expression for setting the tempo using the library. In this method, tempo changes once generated corresponding to various pieces of characteristic information of performance data are cut out in a certain time interval and made into a library, and similarly applied to other portions. For example, a tempo change is extracted from MIDI data to which a tempo change has already been applied in a library creation routine, converted to a relative value, and made into a library. Then, a tempo change is selected from this library corresponding to predetermined feature information of the MIDI data, and the selected tempo change is expanded and contracted in the time direction and the tempo value direction, and given to the performance data. This method can also be applied to change the pitch parameter.

[24] Supply means for supplying performance data, generation means for generating musical tone control information based on predetermined characteristic information of the supplied performance data, generated musical tone control information, and musical tone control information of the supplied performance data A performance data creation device comprising a comparison means for comparing the performance data as a whole and a correction means for correcting the musical tone control information generated based on the comparison result (corresponding to the embodiments (31) and (V)). :

  In another feature [24], musical tone control information is generated based on predetermined characteristic information of the supplied performance data, and the generated musical tone control information and the musical tone control information of the supplied performance data are compared with the entire performance data. Since the generated tone control information is corrected based on the comparison result, it is possible to obtain a balanced performance output with an optimal expression by looking at the entire performance data. For example, by looking at the result of changing the tempo through one song, correct the overall tempo so that the average tempo is the originally set tempo value (correct the overall tempo uniformly, By preferentially correcting the tempo of a section having a large amount of time, the tempo can be determined with a view to the whole (see the embodiment (31)). It is also possible to calculate an average value of the volume of the entire performance data and add an offset to the overall volume so that the average value becomes a desired value (see the embodiment (V)).

[25] Supply means for supplying performance data, extraction means for extracting performance data for instructing sound generation from the supplied performance data and having a sound generation length equal to or longer than a predetermined length, and a sound generation length equal to or longer than a predetermined length Storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to the portion of the sound, wherein the generation method information changes the volume of a portion whose pronunciation length is a predetermined length or more in a non-uniform manner Storage means that represents the generation method of the musical sound control information, generation means for generating musical sound control information based on the extracted part and the generation method information, and the generated musical sound control information A performance data creation device comprising an adding means for adding to performance data (corresponding to the embodiment (L)):

  In another feature [25], from the supplied performance data, a portion of performance data for instructing pronunciation is extracted, and a portion whose pronunciation length is a predetermined length or longer is extracted, and a generation method corresponding to the pronunciation instruction data portion having a predetermined length or longer is extracted. Based on the information, the generation of musical tone control information that changes the volume of the portion where the pronunciation length is longer than the predetermined length is generated and added to the performance data. It is possible to add a random expression. In this case, the fluctuation is preferably determined based on a random number counter and a predetermined change imparting pattern. This method can also be applied to pitch parameters.

[26] Supply means for supplying performance data, extraction means for extracting a melody part from the supplied performance data, and generation method information representing a generation method for generating musical tone control information corresponding to the melody part is stored. The storage means, the generation method information is a storage means that represents a generation method for generating musical tone control information that changes the volume of the melody part to be larger than the volume of the other parts; Performance data comprising: generation means for generating musical tone control information based on the generated melody part and generation method information; and addition means for adding the generated musical tone control information to the supplied performance data Creation device (corresponding to Example (R)):

  In another feature [26], a melody part is extracted from the supplied performance data, and the volume of the extracted melody part is made larger than the volume of other parts based on the generation method information corresponding to the melody part. Since the musical tone control information to be changed to is generated and added to the performance data, expression according to the part can be performed. That is, the amount of change is determined so that the volume of the melody part is relatively higher than the volume of the other parts, and the melody line is raised. This technique can also be applied to pitch parameters, and the basic pitch is changed according to the melody part or accompaniment part. Furthermore, the present invention can be applied to changing the perspective according to the part, changing the depth of reverb, and the like.

[27] Supply means for supplying performance data, extraction means for extracting a portion to which a volume change is given from the supplied performance data, and generating musical tone control information corresponding to the portion to which the volume change is given Storage means for storing the generation method information representing the generation method of the musical sound, and the generation method information includes musical tone control information that applies a pitch change corresponding to the applied volume change to a portion to which the volume change is applied. Storage means representing the generation method for generating sound, generation means for generating musical tone control information based on the extracted part and generation method information, and adding the generated musical tone control information to the supplied performance data A performance data creation device comprising the additional means to be provided (corresponding to the embodiment (a)):

  In another feature [27], a portion with volume change is extracted from the supplied performance data, and a pitch change corresponding to the extracted volume change is based on the generation method information corresponding to the volume change addition portion. Is generated and added to the performance data, so that the pitch change corresponding to the volume change is applied to the part to which the volume change (accent) is added. The facial expression can be made to slightly increase the accent sound.

[28] Supply means for supplying performance data, extraction means for extracting a portion where double choking is performed from the supplied performance data, and generating musical tone control information corresponding to the portion where double choking is performed Storage means for storing generation method information representing a generation method for performing, and this generation method information divides the performance data of the portion where double choking is performed into two parts, upper sound and lower sound. Generating musical tone control information based on the storage means representing the generation method for generating musical tone control information that gives different volume changes to each of them, the extracted portion of the double choking, and the generation method information A performance data generating apparatus comprising: generating means for generating and adding means for adding the generated musical tone control information to the supplied performance data; b) the corresponding]:

  In another feature [28], the portion where the double choking is performed is extracted from the supplied performance data, and the performance data of the extracted double choking portion is obtained based on the generation method information corresponding to the double choking portion. It is divided into two parts, upper sound and lower sound, and musical tone control information that gives different volume changes to each is generated and added to the performance data. When double choking, double choking The upper sound and the lower sound can be separated into separate parts, and the expression that shifts the time change timing of the sound volume can be intentionally performed in another part.

[29] Supply means for supplying performance data, extraction means for extracting a portion where choking is continuously performed from the supplied performance data, and a musical sound corresponding to the portion where choking is continuously performed Storage means for storing generation method information representing a generation method for generating control information, and this generation method information is stored in the pitch of performance data for each choking in a portion where choking is continuously performed. Generation of musical tone control information based on the storage means that represents the generation method for generating musical tone control information that gives non-uniformity, the extracted portion where choking is continuously performed, and the generation method information A performance data generating apparatus comprising: a generating means for generating and an adding means for adding the generated tone control information to the supplied performance data; ) Is a corresponding]:

  In another feature [29], a portion where choking is continuously performed is extracted from the supplied performance data. Based on the generation method information corresponding to the continuous choking portion, in the extracted continuous choking portion, Musical tone control information that gives non-uniformity to the pitch of performance data for every choking is generated and added to the performance data, so that it is possible to give a delicate expression during continuous choking. In other words, in a portion where choking appears continuously, non-uniformity is added to the pitch so that the pitch change of every choking does not become the same. For example, based on the number of consecutive chokings and the changing tendency of the pitch, you can select a change-giving template, determine the changing tendency of the pitch to be given, and make it slightly higher than the pitch of the middle sound with choking. Or lower.

[30] Supply means for supplying performance data, extraction means for extracting a portion corresponding to the arpeggio performance from the supplied performance data, and generation for generating musical tone control information corresponding to the portion corresponding to the arpeggio performance A storage means for storing generation method information representing a method, wherein the generation method information detects common overtones in the arpeggio performance, and generates musical tone control information that causes the detected overtones to be pronounced in another part. A generating means for generating musical tone control information on the basis of the portion corresponding to the extracted arpeggio performance and the generating method information, and the performance supplied with the generated musical tone control information. A performance data creation device comprising an adding means for adding data (corresponding to the embodiment (d)):

  In another feature [30], a portion corresponding to the arpeggio performance is extracted from the supplied performance data, and a common overtone in the extracted arpeggio performance is detected based on the generation method information corresponding to the portion corresponding to the arpeggio performance. Since the musical tone control information that causes the detected harmonics to be generated by another part is generated and added to the performance data, the common harmonics of the alveggio are detected and the volume of the harmonics is reduced by another part. You can make facial expressions.

[31] Supply means for supplying performance data, extraction means for extracting a portion corresponding to a predetermined music symbol for instructing change in tone color from the supplied performance data, and predetermined music symbol for instructing change in tone color Is a storage means for storing generation method information representing a generation method for generating musical tone control information corresponding to, wherein the generation method information is a portion corresponding to a predetermined musical symbol instructing a change in timbre. A storage means representing a generation method for generating musical tone control information for changing a set tone color to a tone color corresponding to a music symbol, a portion corresponding to a predetermined music symbol for instructing change of the extracted tone color, and Based on the generation method information, a performance data creation device comprising a generation means for generating musical tone control information, and an adding means for adding the generated musical tone control information to the supplied performance data. Example (e) the corresponding]:

  In another feature [31], from the supplied performance data, a portion corresponding to a predetermined music symbol instructing a timbre change is extracted, and the timbre change is based on generation method information corresponding to the instruction music symbol, Musical tone control information that changes the set tone color is generated and added to the performance data in the portion corresponding to the extracted tone color change instruction music symbol. Expression can be performed by selecting a timbre by a symbol. For example, when “pizz.” Is displayed, the timbre is automatically changed to a pizzicato string and returned to the original string timbre at the display position of “arco”.

[Various features]
Also, the following features (1) to (23) can be configured by various features related to the implementation of various inventions described in this specification.
(1) Means for inputting performance data, means for obtaining characteristic information of the input performance data, and means for supplying a facial expression module in which correspondence between predetermined characteristic information and musical tone control information regarding the performance data is ruled And means for setting the tone control information according to the rules of the supplied facial expression module based on the acquired feature information, a means for adding the set tone control information to the input performance data, and the tone control information A performance data creating apparatus comprising: means for outputting performance data to which [1] is added [FIG. 2]. In other words, according to the configuration of (1), various facial expression modules in which the procedure for setting musical tone control information for the characteristics of the input performance data, which is a factor for musical tone control, are prepared, and based on this facial expression module. Since musical tone control information is set, more musical performance data can be automatically created.

(2) means for inputting performance data; means for acquiring characteristic information of the input performance data; means for setting a musical sound control amount according to a predetermined rule based on the acquired characteristic information; Means for adjusting the control parameter of the musical sound control amount, means for determining the musical sound control information based on the set musical sound control amount and the adjusted control parameter, and the determined musical sound control information as input performance data And a performance data creating apparatus comprising means for outputting performance data to which musical tone control information is added, and in the performance data creating apparatus, a parameter adjusting means evaluates the output performance data. The control parameters are adjusted again according to the evaluation results [Examples (1) to (24), (27) to (30)]. That is, according to the configuration of (2), when various musical tone control information is set, musical tone control amounts (various musical tone control amounts in performance such as time, pitch, and volume) set according to the rules based on the characteristic information. Control parameters can be adjusted, and musical tone control information (various performance parameters such as time parameters, pitch parameters, volume parameters, etc.) is determined based on the musical tone control amount and the adjusted control parameters. It is possible to perform facial expression using control information.

(3) A method of extracting note time information from input performance data [Examples (1) and (3)], a method of evaluating the progress of performance of input performance data [Example (2)], Method of extracting minute vibration sound information from input performance data [Examples (4), (20), (23)], Method of recognizing phrase breaks from input performance data [Example (5)] A method of calculating pitch information for each predetermined section or smoothed pitch information from the input performance data [Example (6)], a method of acquiring pitch change direction change information from the input performance data [ Embodiment (7)], a method of detecting the same or similar pattern from input performance data [Embodiment (8), (9)], a pre-registered tone form from input performance data Detection method [Example (10)], input A method of calculating volume information for each predetermined section or smoothed volume information from performance data [Example (11)], a method of acquiring atmosphere information of input performance data [Example (12)], and input A method of extracting predetermined note group sequence information from performance data [Examples (13) and (14)], a method of extracting chord number information of input performance data [Example (15)], and the like. A method for extracting timbre information from performance data [Embodiment (16)], a method for extracting fingering information from input performance data [Embodiments (17) to (19)], and a specific method from input performance data. A method of extracting rendition style information corresponding to a musical instrument performance method (Examples (21) and (22)), a method of extracting minute fluctuation sound information such as strings from input performance data (Example (24)), and Performance A method of acquiring lyrics information of the data [Example (27)], a method of acquiring output waveform information of the sound source from the input performance data [Example (28)], and extracting predetermined performance symbol information from the input performance data Based on the feature information acquired by the method [Embodiment (29), (30)] etc., the time parameter, the pitch parameter, etc. To set the musical tone control information. As described above, the characteristics of the input performance data include note time information (note density, interval between two notes), performance progress, minute vibration sound information (long tone trill / vibrato, pitch bend, etc.), phrase delimiter (phrase end) Part), pitch information, pitch change direction change information (up / down part, etc.), same or similar pattern (same pattern continuous, similar phrases, etc.), registered sound form (phrase template etc.), volume information, atmosphere information (Such as “tension”), note group information (grouped notes, long tuplets, etc.), chord information, tone information, fingering information (finger, position movement, position, etc.), performance information (guitar pulling) Off, hammering on, piano sustain pedal, etc.), minute fluctuating sound information (multi-part trill, etc.), lyric information, predetermined performance symbols (dynamic symbols, stackers) Etc.) to acquire the, by setting the tone control information in accordance with these features, it is possible to obtain a performance output with versatile expression based on these characteristic information.

(4) means for inputting performance data, means for storing the relationship between the predetermined characteristic information of the performance data already input and the musical tone control information, means for acquiring characteristic information of the newly input performance data, Based on the acquired feature information, means for setting musical tone control information according to the stored relationship, means for adding the set musical tone control information to newly inputted performance data, and musical tone control information are added A performance data creation device comprising a means for outputting performance data [Example (25)], a library in which a plurality of sets of relations between predetermined feature information and musical tone control information are recorded, and performance data are input. Means for acquiring characteristic information of the input performance data, and means for setting the tone control information by referring to the library based on the acquired characteristic information. When the musical tone control information set, and means for adding to the input performance data, performance data creation and means for outputting the performance data tone control information is added Example (26)]. In this way, the relationship between the predetermined characteristic information of the performance data already input and the musical tone control information is stored, and the learning result according to the stored relationship is used based on the characteristic information of the newly input performance data. The musical tone control information is set by referring to a library in which a plurality of sets of relations between predetermined characteristic information and musical tone control information are recorded with respect to performance data based on characteristic information obtained from input performance data. By setting so that the expression is set, the adaptability of the facial expression can be improved by the learning function and the library.

(5) A means for inputting performance data, a means for setting musical tone control information based on predetermined characteristic information of the inputted performance data, and playing the set musical tone control information and the musical tone control information of the inputted performance data A performance data creating apparatus comprising a means for comparing the entire data and a means for correcting the set tone control information based on the comparison result [Example (31)]. That is, according to the configuration of (5), the set tone control information is corrected based on the result of comparing the set tone control information and the tone control information of the input performance data with the entire performance data. Therefore, the musical tone control information can be set to an optimum value by looking at the performance data.

(6) Supply means for supplying performance data, and the supplied performance data are analyzed, and a partial section in the entire section including all the performance data to which any one of a plurality of types of facial expressions can be assigned is extracted. Analyzing means, determining means for selecting and determining facial expressions to be added to performance data included in the extracted partial section, and parameters of performance data included in the extracted partial section are determined. A parameter automatic editing device having parameter editing means for automatically editing according to a facial expression algorithm corresponding to a facial expression, a supply module for supplying performance data from the supply means, and analyzing the supplied performance data, and a plurality of types An analysis module that extracts a partial section of the entire section including all performance data, which can be assigned any of the facial expressions A determination module that selects and determines facial expressions to be added to the performance data included in the selected partial section, and the parameters of the performance data included in the extracted partial sections correspond to the determined facial expressions A parameter storage medium storing a program that can be realized by a computer, including a parameter editing module that automatically edits according to a facial expression algorithm [FIGS. 52 and 53]. Here, the performance data is assumed to be sequence data. For this reason, performance data can be arranged in time series, and the concept of “section” can be entered there. The “parameter” is variable information for controlling the musical tone on performance such as temporal musical tone control information such as tempo and timing, pitch musical tone control information, volume control information, etc., and is referred to as “performance parameter”. The same applies to the following.

(7) Supply means for supplying performance data, extraction means for extracting a data area whose pitch tends to increase from the supplied performance data, and the volume or pitch of the performance data included in the extracted data area Parameter editing means for editing the volume or pitch parameter value of each performance data so that the performance data gradually increases or deviates from the performance data located at the beginning of the data area to the performance data located at the end. A parameter automatic editing apparatus having the following (Example (A)). Here, the rising tendency is not a simple rising sound system, but includes a thing that can be said to be a rising sound system when the whole is evaluated, and the same applies to the following.

(8) Supply means for supplying performance data, and from the supplied performance data, a data area in which the pitch tends to increase and a data area in which the pitch tends to decrease are extracted, and the pitch is calculated from the extracted data area. An extracting means for further extracting performance data at the changing point that changes from an upward trend to a downward trend, and an accent is given to the volume of the performance data at the changing point in the performance data included in the extracted data region Thus, the parameter automatic editing apparatus having the parameter editing means for editing the volume parameter value [Embodiment (A)]. Here, contrary to the upward trend, the downward trend is not a simple downward sound system, but includes a thing that can be said to be a downward sound system when the whole is evaluated, and the same applies to the following.

(9) Supply means for supplying performance data, and a volume or pitch change pattern that defines a change tendency of the volume or pitch parameter value from the performance data located at the beginning to the performance data located at the end of the supplied performance data. A storage means for storing a plurality of types, a selection means for selecting any one of a plurality of stored volume or pitch change patterns, and a volume or a pitch change pattern in which the volume or pitch parameter value of the supplied performance data is selected. A parameter automatic editing apparatus (embodiment (B)) having parameter editing means for editing the volume or pitch parameter value of each performance data so as to have a prescribed change tendency. Here, as the change tendency, for example, there is a tendency to gradually increase the volume or pitch parameter value as the music progresses. By using such a change tendency, the degree of excitement can be increased as the music progresses. The increase characteristic is desirably a characteristic that converges to a predetermined finite value as the music progresses. This is because the output range of the sound source is not exceeded, and more natural expression is achieved. The same applies to the following.

(10) Supply means for supplying performance data, extraction means for extracting from the supplied performance data a data area in which the appearance density of the performance data instructing pronunciation is a predetermined value or more, and the extracted data area A parameter automatic editing apparatus (example (C)) having parameter editing means for editing the volume or pitch instability parameter value of performance data into a value corresponding to the appearance density. Here, the appearance density means the appearance frequency per unit time. When the appearance density is equal to or higher than the predetermined value, it means that the performance is difficult. Therefore, the value corresponding to the appearance density is usually a value smaller than the original volume parameter value, that is, in a direction in which the volume decreases. This means a change, and the pitch instability parameter value means a change in an increasing direction. The same applies to the following.

(11) calculating means for calculating a pitch based on performance data included in the extracted data area, and calculating a pitch change width between the minimum value and the maximum value of the calculated pitch; A parameter automatic editing device that edits the volume or pitch instability parameter value of performance data included in the extracted data area into a value corresponding to the appearance density and the calculated pitch change range [Embodiment (C)]. Here, the pitch change range can also be an index for changing the volume parameter value or the pitch instability parameter value. Normally, as the pitch or the pitch instability change range is larger, the volume parameter value is changed in a decreasing direction. The pitch instability parameter value is changed in the increasing direction. The same applies to the following.

(12) Supply means for supplying performance data, extraction means for extracting a similar phrase area from the supplied performance data, and calculation means for calculating the similarity between similar phrases included in the extracted similar phrase area When the similar phrase region appears continuously, the volume parameter value of the similar phrase appearing after the second is smaller than the volume parameter value of the similar phrase appearing first, and depends on the calculated similarity On the other hand, when the similar phrase region appears discretely, the volume parameter value of the similar phrase appearing after the second is similar to the volume parameter value of the first similar phrase, A parameter automatic editing device having a parameter editing means for editing the value according to the calculated similarity (Example (D)). Here, the similarity is, for example, (1) all the performance data is the same, (2) only a part of the performance data is different, (3) only a part of the performance data is the same, (4 ) Although it is easy to understand that the performance data are all different, it is easy to understand that the value is 4 levels, but this level may be finer or coarser. The same applies to the following.

(13) Supply means for supplying performance data; Extraction means for extracting, from the supplied performance data, a data area having a measure length in which all the pronunciation lengths of the performance data for instructing the pronunciation are three beats and the same tone length; A determining means for determining a beat position to which strength is to be added in the extracted data area, and a volume parameter value of the performance data at the beat position whose strength is determined by the determining means in the performance data included in the extracted data area And parameter editing means for editing so as to decrease the volume parameter value of the performance data at the beat position whose weakness is determined by the determination means (Embodiment (E)). Here, the criteria for determining the beat position to which strength is to be given include, for example, the style, era and composer of the song selected as performance data. The same applies to the following.

(14) Supply means for supplying performance data, extraction means for extracting a phrase length data area from the supplied performance data, calculation means for calculating a tempo value set in the extracted data area, The performance data located at the end included in the extracted data area, the performance data for instructing the sound generation, the detection means for detecting the sound generation length, and the volume of the detected performance data calculated as the tempo value And a parameter automatic editing device having parameter editing means for editing the volume parameter value of the performance data so as to gradually attenuate for a duration corresponding to the detected pronunciation length. [Example (F)]

(15) Supply means for supplying performance data, and extraction means for extracting performance data for instructing pronunciation from the supplied performance data, to which trill or vibrato is given and whose pronunciation length is equal to or greater than a predetermined length A volume change pattern or a pitch change speed pattern that regulates the speed of a change in volume or a change in pitch when playing performance data having a pronunciation length of a predetermined length or more, and performance data having a pronunciation length of a predetermined length or more. A storage means for storing a volume change pattern or a pitch change speed pattern that defines the speed of a change in volume or a pitch change during vibrato performance according to the pronunciation length, respectively, and in accordance with the extracted performance data Reading means for reading the volume change pattern or pitch change speed pattern from the storage means, and the volume change or pitch change speed of the extracted performance data The volume parameter value or the pitch change speed parameter value of the performance data is edited so that the volume change or the pitch change speed specified in the read volume change pattern or the pitch change speed pattern is obtained. A parameter automatic editing apparatus having parameter editing means. [Example (G)]

(16) Supply means for supplying performance data, extraction means for extracting a data area in which a toll performance or roll performance is performed from the supplied performance data, and a volume parameter value of the performance data included in the extracted data area A parameter automatic editing device having parameter editing means for editing the values into non-uniform values. [Example (H)]

(17) Supply means for supplying performance data, extraction means for extracting a data area composed of a plurality of performance data for instructing pronunciation from the supplied performance data, and the position of the performance data to be emphasized Storage means for storing the number of performance data for instructing pronunciation simultaneously and the type corresponding to the pitch, and storing the number of performance data included in the extracted data area and the pattern corresponding to the pitch Reading means read from the means, and parameter editing means for editing the volume data or the pitch parameter value of the performance data so that the performance data at the position indicated by the read pattern is emphasized in the extracted performance data. A parameter automatic editing device [Example (J)]. Here, the plurality of performance data for instructing pronunciation simultaneously is typically performance data constituting a chord, but is not limited to this, for example, octave unison. The same applies to the following.

(18) Supply means for supplying performance data, extraction means for extracting performance data for instructing pronunciation from the supplied performance data and immediately after the performance data for staccato pronunciation, and extracted performance data A parameter automatic editing device comprising parameter editing means for editing the volume parameter value of the performance data so that the volume of the performance data is reduced. [Example (K)]

(19) Supply means for supplying performance data, extraction means for extracting performance data for instructing pronunciation from the supplied performance data and having a pronunciation length equal to or longer than a predetermined length, and outputting a non-uniform value Output means that changes the variation range of the non-uniform value according to the elapsed time from the start of sound generation, and the volume or pitch of the extracted performance data is non-uniform and the variation range changes A parameter automatic editing apparatus (embodiment (L)) having parameter editing means for editing the volume or pitch parameter value of the performance data into a non-uniform value output by the output means so as to maintain the pronunciation length. Here, the performance data to be extracted is so-called long tone performance data, and the volume is changed while the amplitude of the long tone is being sounded unevenly and gradually changing the amplitude. Will be granted. The same applies to the following.

(20) Supply means for supplying performance data, detection means for detecting parts of the same tone color and the number of parts from the supplied performance data, and the performance volume value of each part according to the detected number of parts A parameter automatic editing apparatus having a calculating means for calculating and a setting means for setting each calculated performance volume value as a volume parameter value of each part [Embodiment (N)]. Here, when reproducing the performance of a score for which a part division is specified, the conventional parameter editing device does not make any changes to the volume parameter. In order to solve the problem of increasing the performance, “the performance volume value of each part is calculated according to the number of detected parts”. This situation is the same in the following.

(21) Supply means for supplying performance data, extraction means for extracting performance data instructed to change the pitch by pitch bend from the supplied performance data, and calculating a change tendency of pitch bend in the extracted performance data The volume parameter value of the performance data is edited so that the volume change of the extracted performance data becomes the determined volume change according to the calculation result. A parameter automatic editing apparatus having parameter editing means for performing [Example (Q)]

(22) Supply means for supplying performance data; extraction means for extracting a melody part from the supplied performance data; comparing the extracted melody part with other parts; A parameter automatic unit having a determining unit that determines a change amount of the performance parameter so as to stand out from the part, and a parameter editing unit that edits the performance parameter value of the extracted melody part based on the determined change amount of the performance parameter Editing device [Example (R)]. Here, it is assumed that the performance data consists of a plurality of parts. However, if the top note included in the performance data is extracted and used as the melody part for performance data such as a piano that is assigned from one melody to accompaniment, the structure of (22) Can be applied as is.

(23) Supply means for supplying performance data, extraction means for extracting lyric information from the supplied performance data, and detecting a word to which volume or pitch change should be given from the words included in the extracted lyric information A storage means for storing a volume or a pitch change pattern indicating a volume or a pitch change pattern to be applied to the word for each word, and a volume or a pitch change pattern corresponding to the detected word from the storage means. Edit the volume or pitch parameter value of the performance data so that the change of the volume or pitch of the read word and the detected word becomes the volume or pitch change indicated by the read volume or pitch change pattern. A parameter automatic editing device having a parameter editing means for performing the embodiment (U). According to the parameter automatic editing apparatus having the above configurations (6) to (23), the performance data supplied from the supply means is analyzed, and any one of a plurality of types of facial expressions can be given. A partial section in the entire section including all data is extracted, and facial expressions to be added to the performance data included in the extracted partial sections are selected and determined from the plurality of types of facial expressions, and the extracted partial sections The parameters of the performance data included are automatically edited according to the expression algorithm corresponding to the determined expression, so that even a user who is a beginner can add various expressions to a song with simple operations. It becomes.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. The following embodiments are merely examples, and various modifications can be made without departing from the spirit of the present invention.

[Hardware configuration]
FIG. 1 is a block diagram showing a hardware configuration of a performance data creation apparatus (that is, a parameter automatic editing apparatus) according to an embodiment of the present invention. In this example, the system includes a central processing unit (CPU) 1, a read only memory (ROM) 2, a random access memory (RAM) 3, first and second detection circuits 4 and 5, a display circuit 6, and a sound source circuit 7. , An effect circuit 8, an external storage device 9, etc., and these devices 1 to 9 are connected to each other via a bus 10, a performance data creation system for performing performance data creation processing, or automatic parameter editing An automatic parameter editing system for processing is configured.

  The CPU 1 that controls the entire system (that is, controls the entire apparatus) includes a timer 11 that is used to generate a tempo clock and an interrupt clock (that is, measures an interrupt time and various times in timer interrupt processing). Various controls are performed in accordance with a predetermined program, and in particular, various processes for tempo change, volume change, pitch change, etc., which will be described later, are centrally performed. The ROM 2 stores predetermined control programs for controlling the performance data creation (parameter automatic editing) system. These control programs include basic performance information processing and various tempo / programs according to the present invention. Conversion processing programs such as timing, volume change and pitch change, various tables, and various data can be included (that is, control programs executed by the CPU 1, various table data, etc. are stored). The RAM 3 stores data and parameters necessary for these processes, and temporarily stores various registers and flags and various data being processed (that is, temporarily stores performance data, various input information, calculation results, and the like). Used as a work area.

  The first detection circuit 4 is connected to a performance operation device 12 having performance operators such as a keyboard, and the operation switch device 13 connected to the second detection circuit 5 sets various modes, parameters, and the like. For this purpose, operators such as numeric / character keys and knobs are provided. For example, the performance operating device 12 has a keyboard for mainly inputting voice information and character information, the operation switch device 13 has a mouse as a pointing device, and the first and second detection circuits 4 and 5. The key operation detection circuit for detecting the operation state of each key of the keyboard and the mouse operation detection circuit for detecting the operation state of the mouse are provided correspondingly.

  The display circuit 6 includes a display 14 [for example, large liquid crystal display (LCD) or CRT (Cathode Ray Tube) display] and various indicators [light emitting diode (LED), etc.] for displaying various information. 14 and the indicator can be juxtaposed on various operators on the operation panel of the switch device 13. The display 14 can also display various setting screens and various operation buttons to set and display various modes and parameter values.

  The sound system 15 connected to the effect circuit 8 composed of a DSP or the like constitutes a musical sound output unit together with the sound source circuit 7 and the effect circuit 8, and based on various performance data created during various processes according to the present invention. It is possible to listen to the performance using the output performance data with a facial expression. For example, the tone generator circuit 7 converts performance data input from the keyboard (12) or pre-recorded performance data into a musical tone signal, and the effect circuit 8 gives various effects to the musical tone signal from the tone generator circuit 7, The sound system 15 includes a DAC (Digital-to-Analog Converter), an amplifier, a speaker, and the like, and converts a musical sound signal from the effect circuit 8 into sound.

  The external storage device 9 includes a hard disk drive (HDD), a compact disk read only memory (CD-ROM) drive, a flexible disk drive (FDD), a magneto-optical (MO) disk drive, a digital multipurpose disk (DVD) drive, etc. It is possible to store various control programs and various data. Therefore, various programs and data necessary for performance data processing can be read not only from the ROM 2 but also from the external storage device 9 into the RAM 3, and the processing results can be externally stored via the RAM 3 as necessary. It can also be recorded in the device 9. For example, the FDD drives a flexible disk (FD) that is a storage medium, the HDD drives a hard disk that stores various application programs including control programs and various data, and the CD-ROM drive includes a control program. A CD-ROM for storing various application programs and various data is driven.

  As described above, the hard disk of the HDD (9) can also store the control program executed by the CPU 1. If the control program is not stored in the ROM 2, the control program is stored in the hard disk. By reading into the RAM 3, the CPU 1 can be caused to perform the same operation as when the control program is stored in the ROM 2. In this way, control programs can be easily added and upgraded.

  The control program and various data read from the CD-ROM of the CD-ROM drive (9) are stored in the hard disk in the HDD (9). As a result, a new installation or version upgrade of the control program can be easily performed. In addition to the CD-ROM drive, an apparatus for using various types of media such as a magneto-optical disk (MO) apparatus may be provided as the external storage device 9.

  In this example, a MIDI interface (I / F) 16 is connected to the bus 10, the system can communicate with other MIDI devices 17, an external MIDI (Musical Instrument Digital Interface) signal is input, or MIDI Output the signal to the outside. Further, a communication interface 18 is also connected to the bus 10, and data can be transmitted to and received from the server computer 20 via the communication network 19, and control programs and various data can be stored in the external storage device 9 from the server computer 20. In addition to the server computer 20, other client computers can be connected to the communication network 19.

  The MIDI I / F 16 is not limited to a dedicated one, and may be constituted by a general-purpose interface such as RS-232C, USB (Universal Serial Bus), IEEE 1394 (I Triple 1394) or the like. In this case, data other than MIDI messages may be transmitted and received simultaneously.

  As described above, the communication I / F 18 is connected to a communication network 19 such as a LAN (Local Area Network), the Internet, or a telephone line, and is connected to the server computer 20 via the communication network 19. When the above programs and various parameters are not stored in the hard disk in the HDD (9), the communication I / F 18 is used to download the programs and parameters from the server computer 20. A computer serving as a client (a performance data creation device or parameter automatic editing device in this embodiment) transmits a command for requesting a program or parameter download to the server computer 20 via the communication I / F 18 and the communication network 19. To do. Upon receiving this command, the server computer 20 distributes the requested program and parameters to the computer via the communication network 19. The computer receives these programs and parameters via the communication I / F 18 and receives the HDD ( 9) The download is completed by accumulating in the internal hard disk.

  In addition, an interface for directly exchanging data with an external computer or the like may be provided.

  The performance data creation device or parameter automatic editing device of the present embodiment is constructed on a general-purpose personal computer as can be seen from the above-described configuration, but is not limited to this, and the present invention is implemented. You may construct | assemble on the exclusive apparatus comprised only from the minimum possible element.

  The performance data processing (parameter automatic editing) system according to the present invention can be implemented in the form of an electronic musical instrument, but can also be implemented in a form in which a musical sound processing application program is incorporated into a personal computer (PC). Further, the tone generator circuit 7 does not need to be configured by hardware, and can be configured by a software tone generator (in this embodiment, the tone generator circuit 7 is all configured by hardware as the name indicates, However, the present invention is not limited to this, and part of the software may be configured, and the remaining part may be configured of hardware, or all may be configured of software. This function can be entrusted to another MIDI device 17.

Embodiment 1
First, as specific performance parameters, temporal musical tone control information such as timing shift such as tempo change, gate time, sounding start timing, etc. (in the specific example, other performance parameters such as pitch parameter and volume parameter are included. "Embodiment 1" will be described in which a facial expression is added using some of them.

[Overview of system functions]
FIG. 2 shows an outline of functions by the performance data processing (parameter automatic editing) system of the present invention. The system functions include an original performance data fetch block AB for fetching the original performance data OD, an expression block EB for performing a time-related expression on the original performance data OD supplied from the block AB, mainly using tempo conversion, A facial expression module EM that stores facial expression rules and methods corresponding to various facial expressions supplied to the visual attachment block EB, and a post-expression performance that sends the post-expression facial performance data ED to the sound source unit, etc. It consists of a data transmission block SB. Here, the facial expression module EM is a rule mainly for characteristics of tempo changes in various music expressions, and in accordance with these rules, tempo changes, gate times, pronunciation start timings, etc. are applied to the original performance data OD. The timing expression is mainly used for temporal expression, and the original performance data OD is converted into musical data ED after musical expression. In this expression module EM, various expression modules EM1, EM2,..., EMn are prepared in advance, and any one or a plurality of expression modules are selected by the user, and the expression module EB is displayed. Supplied with.

  In the performance data processing (parameter automatic editing) system shown in FIG. 1, the function of the expression module EM is performed by operating various tempo / timing conversion processing programs in the ROM 2 or from the external storage device 9 to a desired tempo / timing conversion. This can be realized by loading a processing program. Since the expression module can be supplied from the external storage device 9 in this way, only the expression module desired by the user from among the many expression modules is provided for the performance data processing (parameter automatic editing) system of FIG. Can be installed, or a new facial expression module supplied from a manufacturer or the like can be installed. The original performance data OD can be arbitrarily input from the keyboard-type controller device 12, the external storage device 9, the MIDI device 17 such as another sequencer or performance device, etc., by the system function original performance data fetch block AB. The post-expression performance data ED can be optionally output to the musical tone output unit, the external storage device 9, other MIDI devices, etc. by the post-expression performance data transmission block SB. The performance data is typically MIDI data.

In other words, the expression module EM according to the present invention prepares the characteristics of performance data as a rule in advance as a factor for temporal control of musical sounds. When the original performance data OD is input to the expression block EB, temporal control information (tempo and timing) is set according to the rule based on the feature information extracted from the original performance data OD. This musical tone control information is added to the original performance data OD and output as performance data ED with a facial expression added. The temporal musical sound control amount can be adjusted by a control parameter, and an optimal temporal expression can be applied to the musical sound. Furthermore, the range of facial expression addition can be expanded by learning functions and library creation, and the musical tone control information can be corrected in comparison with the entire performance data, so that the performance data can be set as a prospect.

  The facial expression module EM can also prepare in advance the characteristics of performance data that serve as other performance control factors such as the volume and pitch of the musical tone as rules. Accordingly, when the original performance data OD is input to the expression block EB, performance control information (musical control information, performance, etc.) that changes the volume or pitch according to the rules based on the feature information extracted from the original performance data OD. Parameter). The performance control information is added to the original performance data OD and can be output as performance data ED with a facial expression added. Further, the musical tone control amount based on the performance control information can also be adjusted by the control parameter, and an optimal expression can be applied to the musical tone. Furthermore, it is possible to expand the range of facial expression by using a learning function or a library, or to correct the performance control information in comparison with the entire performance data, and to set the performance data in view.

[Various examples of facial expression modules]
Hereinafter, various rules and methods such as various tempo changes executed by various expression modules EM (EM1, EM2,..., EMn) according to an embodiment of the present invention will be described for each of the embodiments (1) to (31). explain. In these embodiments, the module EM is described as converting the original MIDI performance data OD to the MIDI performance data ED after expression, but the present invention can be applied regardless of the data format such as MIDI. it can. Note that there are a method of adding a tempo change to performance data, a method of inserting a tempo change event, and a method of shifting a note time. In the following description, a method of inserting a tempo change event is mainly used, but it may be replaced with a method of shifting the time of a note. Conversely, the method of shifting the time of a note may be replaced with a method of inserting a tempo change event.

(1) “The tempo increases as the number of sounds increases”:
In general, as the number of reference sounds increases, the performance becomes difficult and the tempo is likely to become slower, so it is easier. However, in actual performance, there is an aspect that the consciousness of maintaining the tempo is strongly expressed from such a psychological state, and conversely, it becomes faster. In addition, there is a tendency that the number of sounds is often a more active performance phase. Therefore, it is both human and natural to play with an increased tempo as the number of reference sounds increases. Therefore, in this tempo change rule, in order to reproduce such an expression, for example, performance data is evaluated in measure units according to the density of notes, and the tempo is changed according to the number of notes per beat. The technique of letting it be adopted is adopted.

  For example, in the musical score example shown in FIG. 3, the number of notes per beat is one at the first measure, but the number of notes per beat is two at the second measure. Therefore, this numerical value [(number of notes in one bar) / (number of beats in one bar)] is used as a note density, and the tempo is changed according to the note density. In the first measure, the note density is “1”, so the quarter note “132” is obtained at the same tempo. On the other hand, since the note density is “2” in the second measure, the tempo value is set from this note density through a table. As such a table, for example, a “note density-tempo coefficient (magnification)” table as shown in FIG. 4 can be used. Therefore, for the second measure of the musical score example of FIG. 3, for the note density “2”, for example, a result of tempo coefficient K = 1.02 (2% speed increase) can be obtained from the table of FIG. . Then, the tempo coefficient K is multiplied by the tempo value of the original performance data to determine a desired tempo value.

  FIG. 5 is a flowchart showing an example of a note density response process for accelerating the tempo according to the note density by the expression module EM. In the first step A1 of this processing flow, a target part for performing tempo determination is selected from a plurality of performance parts from the original MIDI data OD, and in the next step A2, a strength value α is set for the tempo ( At first, when the process first proceeds from step A1 to step A2, for example, “1” is set as the initial value of the value α), and the process proceeds to step A3.

  In the next step A3, the selected part is divided into designated sections (for example, sections for each measure), and in the next step A4, the tempo coefficient K of each designated section is obtained, thereby obtaining the note density in each designated section. evaluate. The evaluation of the number of notes in steps A3 to A4 may be performed in units of measures, for example, in units of measures, with the designated section being a section of one measure, and may be performed in units of beats, or a time serving as a tempo reference (Tick = minimum time resolution) , Clock or step time). Further, the tempo coefficient K of each designated section can be obtained from, for example, the “note density-tempo coefficient” table of FIG. 4 described above.

  When obtaining the tempo coefficient K in step A4, when the number of interval notes is “0”, the tempo coefficient K is exceptionally set to “1” as indicated by ◯ → ● in FIG. In general, rests are not treated as musical notes in MIDI data, and as a result, a measure composed only of rests results in a slow tempo. For this reason, when there are no notes (rests) in the section to be evaluated, it is necessary to perform an exceptional process in which the tempo coefficient K is evaluated as “1”.

  Proceeding from step A4 to step A5, the tempo is based on the tempo coefficient K, strength value α, and currently set tempo value (for example, “132” in the example of FIG. 3) for each specified section. A change amount (for example, “132” × α × K) is calculated, and this tempo change is added to the MIDI data of each designated section. In the next step A6, the entire MIDI data to which the tempo change has been applied or only a necessary part is reproduced, and for example, the musical performance is output via the musical sound output unit and the performance based on the MIDI data is auditioned. Thereafter, the process proceeds to step A7, and if it is determined that the MIDI data to which the tempo change is given is good as a result of the MIDI data reproduction, the note density response process is terminated. On the other hand, if it is determined that the playback MIDI data has a defect in tempo setting, the process returns to step A2, the strength value α is reset, and the processing of steps A3 to A7 is repeated.

  In the pass / fail judgment step A7 to the strength / weakness setting step A2, reproduction MIDI data is automatically judged according to a predetermined judgment criterion, and the strength value α is automatically increased / decreased by a predetermined value in a direction to obtain a good result. It can also be automated. The strength value α may be set to a different value for each section. Further, a method for setting the strength value α in common to each section and a method for setting the strength value α individually for each section may be shared. For example, when this tempo change rule is applied through all songs with the strength value α = 1 initially set, and MIDI data with a facial expression is reproduced (step A6), when this returns NG to step A2. For each section with excessive expression, the strength value α is reset to less than “1” according to each excess degree, and conversely, for each insufficient section, the value exceeds “1” according to each lack degree. Editing can be done such as re-setting. These aspects can also be implemented in the same manner when the performance parameter value assignment is performed by applying the rules regarding the tempo change / timing shift and the like in the embodiment (2).

  In this case, the original performance data (MIDI data) OD is displayed on the display 14, and the assigned parameter value (tempo change amount, etc.) and its position are displayed on the displayed performance data OD in an overlapping manner. The user can check the parameter value setting result. Further, at the time of confirmation, the user may be allowed to specify a position where the parameter change giving process is not desired.

  When there are a plurality of performance parts (rhythm / accompaniment, melody, etc.), the tempo is generally common. Therefore, in the above-described note density response process of FIG. 5, a specific part for determining the tempo is selectively selected in step A1. Although it is set, as will be described next, a method of determining the tempo by evaluating the information of a plurality of parts as a whole can also be adopted. Instead of changing the tempo event, the tempo change is reproduced by shifting the time of the note, so that independent tempo setting can be performed among a plurality of parts.

  FIG. 6 is a flowchart showing another example of a note density response process for accelerating the tempo in accordance with the note density by the expression module EM. Applied when the tempo is determined by the evaluation. In this processing flow, a strong value α is set for the tempo in the first step B1, and the process proceeds to step B2. In step B2, each part is divided into designated sections (for example, sections for each bar). In the next step B3, the average note density of each designated section of all parts is calculated. In step B4, the calculated note density is calculated. The tempo coefficient K of each designated section is obtained from the above, and the note density in each designated section is evaluated. Note that the evaluation of the number of notes in steps B2 to B4 may be performed in units of measures, in units of beats, or may be performed in time (Tick) as a tempo reference, as described above. Of course, the tempo coefficient K is set to “1” in the designated section without notes.

  After proceeding from step B4 to step B5, as in steps A5 and A6 described above, in step B5, the tempo change corresponding to the tempo coefficient K, the strength value α and the current tempo value obtained in step B5 is set for each designated section. This is added to the MIDI data, and the MIDI data is reproduced in the next step B6. Thereafter, if the reproduced MIDI data is good in step B7, the note density response process is terminated. If there is a defect, the process returns to step B1 and the processes of steps B1 to B7 are repeated.

(2) “Slowly, the tempo gradually increases throughout the song”:
In live performances, it is a common phenomenon that the tempo increases as the degree of excitement increases. In this tempo change rule, attention is paid to the fact that the degree of excitement increases as the performance of one song progresses, and this phenomenon is used for song production. In other words, by gradually increasing the tempo as the song progresses, a so-called “run” -like expression with a high degree of excitement is produced. In this case, for example, it is preferable to increase the tempo at a predetermined magnification for each measure.

  FIG. 7 is a flowchart of a process for accelerating the tempo according to the progress of music by the expression module EM according to the embodiment of the present invention. In this processing flow, a strong value α is set for the tempo in the first step C1, and in the next step C2, the original MIDI data OD is divided into designated sections (for example, sections for each measure), and the process goes to step C3. Evaluate the progress of the advance song. In step C3, for example, a tempo coefficient (magnification) K for each designated section is obtained using a table as shown in FIG. In this table, the period from the start to the end of one song is divided into, for example, designated sections for each measure, and the value of the tempo coefficient K that gradually increases as the song progresses is determined in advance for each measure. Therefore, by using such a table, the value of the tempo coefficient K of the corresponding measure can be sequentially extracted as the music progresses.

  In the next step C4, the tempo change corresponding to the obtained tempo coefficient K, the strength value α and the current tempo value is given to the MIDI data of each designated section, and in the next step C5, the MIDI data is reproduced. Thereafter, if it is determined in step C6 that the reproduced MIDI data is good, the progress state response process is terminated. If there is a defect, the process returns to step C1 and the processes of steps C1 to C6 are repeated.

  As a requirement of the table used in step C3, if the tempo is increased unnecessarily, the natural expression is impaired. Therefore, as shown in FIG. Preferably, the coefficient K is saturated and has an upwardly convex curve so that the coefficient K cannot be increased beyond a certain value (for example, K = 1.2). Of course, it is not necessary to change the tempo simply in bars, and it may be in beat units or in unit time (Tick) for determining the tempo.

  Further, in the above processing flow, when using the tempo coefficient K determined for each designated section in one song, using an adjustable parameter named “excitability” as the strength value α set in step C1, If the tempo coefficient K, which becomes faster for each measure, can be adjusted, the operability of the tempo determination procedure can be improved. This excitement level parameter is a numerical value representing how many times the vertical axis (tempo coefficient K) of the table as shown in FIG. 8 is to be calculated. , Words such as “strong”, “medium”, “weak”, “no change” of excitement are displayed in a selectable manner, and in the actual processing, using the numerical value corresponding to the selected word Make tempo changes. As a result, the user is provided with an interface for performing setting by words, so that the user becomes friendly.

(3) “When the notes are fine, slow down the tempo”:
This tempo change rule seems to contradict the tempo change rule of (1), but when a note is fine, a person cannot play it, so it is effective for giving such an expression. For this purpose, simply use a table in which the tempo coefficient K decreases as the note density increases, contrary to the “note density-tempo coefficient” table of FIG. 4 used in the tempo change rule of (1). The processing can be performed by the same method as the processing flow of FIGS. For example, when the number of notes is large as shown in the second measure of FIG. 2, if the number of notes / beat = 2 is passed through this table, for example, the result is “0.98” (2% slower). What should I do.

  If you want to simulate the difficulty of playing and increase your sense of reality, as shown in FIG. 9, during a low tempo, the “note time interval” between two consecutive notes Thus, it is preferable to use a table that changes the tempo T later than a certain tempo Ts. By using this tempo threshold value Ts as an adjustable parameter, the poor simulation can be adjusted.

(4) “Long-tone trill / vibrato is slow at first”:
A trill that repeatedly plays different pitches in a minute cycle, or a vibrato that slightly shakes the sound, for example, for a long tone of 1 second or longer (such as a half note with a tempo value of “120”) It is preferable to slow down at first. Therefore, this tempo change rule is to set a slow tempo at first for the long-tone trill / vibrato. Trills and vibratos are generally expressed as note events and pitch bend events, but tempo events (tempo events) are generated for note events and pitch bend events generated at equal intervals in time. By changing the event, the speed of the trill or vibrato can be adjusted. In this case, the tempo of the song itself and the tempo of the trill or vibrato are changed.

  FIG. 10 is a flowchart showing an example of processing for a long tone trill / vibrato by the expression module EM of one embodiment of the present invention. In this processing flow, after setting a strong and weak value α for the tempo in the first step D1, in step D2, a trill and long tone portion or a vibrato and long tone portion is detected from the original MIDI data OD. When the portion where the tril or vibrato is applied to the long tone is detected in step D2, the process proceeds to step D3. For example, the tempo change curve as shown in FIG. A corresponding tempo coefficient K is obtained.

  In the next step D4, a trill note string (the toll note string can also be expressed as a pitch bend string) or a vibrato pitch bend string is generated according to the obtained tempo coefficient K, the strength value α and the current tempo value, In the subsequent step D5, the generated trill note string or vibrato pitch bend string is added to the original MIDI data OD, and in the next step D6, the MIDI data is reproduced. Thereafter, if it is determined in step D7 that the reproduced MIDI data is good, the long sound trill / vibrato process is terminated. If there is a defect, the process returns to step D1 and the processes of steps D1 to D7 are repeated.

  The tempo change curve prepared in step D3 is as shown in FIG. In this curve, the tempo change process is started at time to, the tempo coefficient K decreases from the initial value (Ko = 1), the tempo change decreases most at time ta to become the value Kd, and then gradually increases to increase the time. At tb, the target of tempo change is reached and the target value Kt is reached. Here, the change amount ΔKt from the initial value Ko to the target value Kt is preferably set to 10 to 20% of the tempo change width ΔK in the entire tempo change time to to tb, and the increase time ta to tb is the decrease time. It is preferably at least twice as large as to-ta.

  For the tempo change curve as shown in FIG. 11, not only the magnification (strong value α) of the tempo coefficient K but also the time (horizontal axis) and other change amounts (vertical axis) can be adjusted. It is better to prepare parameters so that appropriate tempo changes can be selected. In this case, for example, parameters such as the total time to to tb, the decrease time to to ta, the tempo change width ΔK, and the target change amount ΔKt are expressed in terms of “strong”, “weak”, “fast”, “slow”, etc. It is convenient if it can be set.

(5) “Slowly at the passage of the phrase”:
According to this tempo change rule, the phrase interval is interpreted with respect to the original MIDI data OD, or the user specifies a phrase break, and at the end of the phrase interval, the phrase ends and the next phrase starts soon. In order to express what is to be done, a facial expression of slowing down the tempo is automatically performed according to the setting. FIG. 12 shows an example of changes in the tempo in this case, and each phrase 1, 2,... Comprises, for example, a section of about 8 bars. Specifically, for example, the late part can be about the last two beats, and the slowing tempo change amount can be about 95%. However, since the effect of such a tempo change depends on the original tempo and the time signature setting, a considerable degree of freedom is required. For example, the time signature setting is 2 beats for 4/4 time, but 1 beat for 3/4 time.

  FIG. 13 is a flowchart showing an example of a process of slowing down the tempo in the phrase passage part by the expression module EM of one embodiment of the present invention. In this processing flow, after setting a strong and weak value α for the tempo in the first step E1, in step E2, the portion of the phrase break position is detected from the original MIDI data OD, and the process proceeds to step E3. In step E3, a tempo change that gradually slows as shown in FIG. 12 is applied to the end of the phrase corresponding to the break portion according to the strength value α and the current tempo value, and in the next step E4, the break portion is applied to the break portion. A tempo change for returning to the original tempo is given to the beginning of the corresponding next phrase. Thereafter, after the MIDI data is reproduced in the next step E5, if the reproduced MIDI data is determined to be good in step E6, the phrase end process is terminated. If there is a defect, the process returns to step E1. The processes E1 to E6 are repeated.

(6) “Change the tempo according to the range”:
For example, a musical instrument that produces a higher sound can cope with faster fingering, and the sound rises faster, so it is appropriate to make the tempo faster for higher notes and slower for lower notes. Therefore, in this tempo changing method, a rule for changing the tempo according to the sound range and making the tempo faster as the pitch becomes higher is employed for expression. For this purpose, the tempo coefficient K is determined based on the pitch information, and the tempo change amount is determined by drawing the “pitch-tempo coefficient” table with the tempo coefficient K. For this table, the correlation between pitch and tempo change may be used as an adjustable parameter. Such processing may be performed for each sound, but it is also effective to determine the processing from the average of the pitch for each specified section. Here, the designated section may be in units of phrases, or in units of measures or beats.

  FIG. 14 is a flowchart showing an example of a process of changing the tempo according to the sound range and increasing the tempo as the pitch increases, according to the expression module EM of one embodiment of the present invention. In this processing flow, after setting the strength value α for the tempo in the first step F1, in step F2, the original MIDI data OD is divided into designated sections (for example, sections for each bar), and further, in step F3. Then, after calculating the average pitch of the notes in each designated section, the process proceeds to Step F4.

  In the next step F4, a tempo coefficient K is obtained from the calculated average pitch using a “pitch-tempo coefficient” table as shown in FIG. In this case, when the number of notes in the section is “0”, the value of the tempo coefficient K is “1”. Further, in the next step F5, the obtained tempo coefficient K, the strength value α, and the tempo change according to the current tempo value are given to the MIDI data of each designated section, and this MIDI data is reproduced in the next step F6. . Thereafter, if it is determined in step F7 that the reproduced MIDI data is good, the sound range corresponding process is terminated. If there is a problem, the process returns to step F1 and the processes of steps F1 to F7 are repeated.

  In the processing flow described above, the “pitch-tempo coefficient” table as shown in FIG. 15 is used in step F4, and an adjustable parameter of strength value α is set in step F1 so that the tempo coefficient K can be adjusted. Processing is decided from the average of the pitch every time. In this case, the operability of the tempo determination procedure can be improved by using predetermined words such as “strong”, “medium”, and “weak” as setting parameters. Further, if the parameter value α can be individually designated for each designated section for performing the processing in steps F1 to F4, efficient editing can be performed.

  For example, at the beginning of startup, the parameter is set to “medium” and α = 1, and this rule is applied to all songs, and new song data is generated in step F5. Then, when the song data is reproduced in step F6 and there is a problem with the expression according to this rule in step F7, when returning to step F1, the parameter α is set to “weak” for the portion where the expression is extreme. On the contrary, it is possible to perform editing in such a manner that the insufficient part is set to “strong”. When the sections are individually indicated in this way, it is preferable to change the tempo with reference to a “pitch-tempo coefficient” table (for example, FIG. 15) for each sound.

  Also, if changing the tempo for each note results in an unnatural conversion, the pitch change filter is used to smooth the change in pitch and then the “pitch-tempo coefficient” It is a good technique to refer to the table and change the tempo. FIG. 16 is a flowchart showing another example of the process of changing the tempo according to the sound range and increasing the tempo as the pitch increases, according to the expression module EM of one embodiment of the present invention. In this processing flow, after a strong value α is set for the tempo in the first step G1, the pitch change is filtered in step F2 to obtain a pitch change curve.

  In the next step G3, a tempo coefficient K is obtained by referring to the “pitch-tempo coefficient” table on the basis of the pitch value of each point on the pitch change curve. Further, in the next step G4, a tempo change corresponding to the obtained tempo coefficient K, the strength value α and the current tempo value is given to the MIDI data, and in the next step G5, the MIDI data is reproduced. Thereafter, if it is determined in step G6 that the reproduced MIDI data is good, the sound range corresponding process is terminated. If there is a defect, the process returns to step G1 and the processes of steps G1 to G6 are repeated.

(7) “Pause when the pitch changes from rising to falling”:
According to the tempo change rule of (6), the tempo expression that the rising sound system is gradually faster and the falling sound system is gradually slower can be realized. If there is a tempo gap, a natural connection may be obtained. Therefore, in this tempo changing method, a rule is adopted in which an interval is given when the pitch changes from rising to falling. For example, this rule is applied to all songs using the degree of opening as a parameter, and new song data is generated. Then, when playback is performed, editing is performed such that the portion where the expression according to this rule is extreme is reset to “weak”, and the portion where the expression is not enough is set to “strong”.

  FIG. 17 is a flowchart showing an example of processing to be performed at the up / down change portion of the pitch by the expression module EM of one embodiment of the present invention. In this processing flow, after setting a strong and weak value α for the tempo in the first step H1, the pitch change curve is obtained by filtering the pitch change in step H2. Here, for example, as shown in the lower part of FIG. 18, the filtering is performed when there is a minute pitch drop from the sound n1 to the sound n2 in the tone group that is generally in a pitch increasing trend. It has a function to remove the descent phenomenon as noise. In the next step H3, the polarity sign of the obtained pitch change curve [a positive (+) value when rising and a negative (-) value when falling]. ], The time position at which the pitch changes from rising to falling is detected.

  In the next step H5, for example, as shown in the upper part of FIG. 18, the sound near the end of the rising portion of the MIDI data is given a tempo change that gradually decreases according to the strength value α and the tempo value. Proceed to H6. In step H6, the MIDI data is reproduced. Thereafter, if it is determined in step H7 that the reproduced MIDI data is good, the pitch up / down changing section process is terminated. If there is a defect, the process returns to step H1 and the processes of steps H1 to H7 are repeated.

(8) “When the same pattern continues, the second time is slow / fast”:
In general, playing the same pattern with the same expression is disliked as not musical. There are various methods for avoiding such performance, but there is also a method for changing the tempo between the first time and the second time. In this tempo changing method, when the same / similar pattern continues, a rule of playing late or fast for the second time is adopted. In order to execute this method, initially, for example, “the second time is fast” (the second strength value α = α1> 1) is set, and this rule is applied through all the music to generate new music data. Then, when the song data is played back and the expression according to this rule is not suitable, “no expression”, that is, “same as the first” (α = 1) or “slow the second” ( α <1), or edit with a method such as “Second time faster than the previous time” (1 <α <α1) or “Second time faster than the previous time” (α> α1) I do.

  Note that if the second time is played with multiple sounds, the tempo tends to be slow since the multiple sounds are more heavy. Similarly, when playing the same phrase with a solo instrument and an accompaniment instrument, the solo instrument tends to be faster. Moreover, there is a similar tendency when similar phrases are continued. Therefore, these tendencies can also be adopted as rules in this tempo change procedure.

  FIG. 19 is a flowchart showing an example of processing when the same or similar pattern is continued by the expression module EM of one embodiment of the present invention. In this processing flow, after setting the strength value α for the tempo in the first step J1, the repetition of the same or similar pattern is detected in step J2, and the strength is detected for each detected pattern in the MIDI data in the next step J3. A different tempo change corresponding to the value α and the current tempo value is applied, and the process proceeds to step J4 to reproduce the MIDI data. If it is determined in step J5 that the reproduced MIDI data is good, the process when the same / similar pattern is continued is terminated. If there is a defect, the process returns to step J1 and the processes of steps J1 to J5 are repeated.

(9) “Similar phrases set the same or similar tempo expression”:
After setting a tempo expression that gives a subtle tempo change to a phrase, if the same or similar phrase is reproduced in a different location, this tempo change rule will cause the first tempo expression (subtle tempo A good expression can be given by copying (change) as it is or by setting a similar but slightly different tempo expression. If they are not exactly the same phrase but are similar, it is better to set a tempo that is similar (slightly different) to the initial tempo expression. In the first and second times, the tempo of the entire phrase is changed according to the example of (8).

  FIG. 20 is a flowchart showing an example of processing for setting similar tempo expressions to similar phrases by the expression module EM of one embodiment of the present invention. In this processing flow, after setting a strong value α for the tempo in the first step K1, similar phrases are detected in step K2, and differences between the detected similar phrases are detected in the next step K3. To do. That is, in both steps K2 and K3, the original MIDI data OD is sequentially interpreted as phrases, the interpreted phrases are sequentially stored, and the newly interpreted phrases are compared with the already stored phrases. Make a similar judgment. And if you find that there is a similar phrase, find out what is different.

  In the next step K4, a different tempo change corresponding to the detected difference, the currently set strength value α and the current tempo value is given to each detected phrase of the MIDI data, and the process proceeds to step K5. This MIDI data is reproduced. If it is determined in step K6 that the playback MIDI data is good, the similar tempo expression setting process for this similar phrase is terminated. If there is a defect, the process returns to step K1 and the processes of steps K1 to K6 are repeated.

(10) “Set tempo for registered sound form”:
In this tempo expression method, when a pre-registered sound form (for example, a rhythm pattern) appears, a predetermined tempo is set in this portion. Although this method is very similar to the method (9), in order to specify the same phrase, a sound form or the like is registered in advance as a phrase template, and the sound form of the MIDI data is compared with this. Since this is also a good method, this method is adopted. FIG. 21 shows an example of a phrase template. As illustrated in FIGS. 21 (1) and 21 (2), there is a break (↑) at a position corresponding to a bow return in a violin performance or a breath (breathing) in a wind instrument performance. As shown in (3), when there are a plurality of delimiter position candidates (↑), these candidates may be randomly selected after weighting each candidate.

  FIG. 22 is a flowchart showing an example of processing for setting a predetermined tempo to a registered sound form by the expression module EM of one embodiment of the present invention. In this processing flow, after setting a strong and weak value α for the tempo in the first step L1, in step L2, a phrase that matches the registered sound form is detected, and in the next step L3, the MIDI data A predetermined tempo change corresponding to the registered tone form, the strength value α, and the current tempo value is given to the detected phrase, and the process proceeds to step L4 to reproduce the MIDI data. If it is determined in step L5 that the reproduced MIDI data is good, the registered sound form application process is terminated. If there is a problem, the process returns to step L1 and the processes of steps L1 to L5 are repeated.

  It is also a good method to evaluate the tempo by referring to other types of events such as a registered expression event instead of using only the sound form as a phrase template. For example, a violin can produce a realistic expression by generating an expression event that responds to the bow return. However, when the bow is returned, the tempo is disturbed, such as the tempo is too slow. Is likely to occur. In order to automatically add such a situation, if the tempo change is linked to an expression event that reproduces a bow return (for example, the volume decreases because the bow stops for a moment), a more realistic performance is reproduced. be able to. Similarly, in wind instruments, an expression event (the volume decreases due to breathing) may be used to express a feeling of breathing, but the same is true. Therefore, change the tempo to express it.

  According to one embodiment of the present invention, the processing using the registered tone form (template) described with reference to FIGS. 21 and 22 is similarly applied to the pitch parameter, and performance data (MIDI data etc.) is registered. Based on the comparison with the phrase template, a predetermined pitch change can be set. In this case as well, in order to identify the same phrase, it is better to register the sound form as a template, as well as the tempo, and instead of using the sound form as a reference, the registered expression event It is also a good method to evaluate the pitch with reference to other events.

  For example, when an expression event corresponding to the return of a bow is generated in a violin or the like, a realistic expression can be created. However, when the bow is returned, unevenness in the pressure of the bow occurs and the pitch tends to change. In order to automatically add such a situation, a more realistic performance can be reproduced by exercising a change in tempo in an expression event that reproduces bow return.

  In wind instruments, the expression event may be used to express the breath, but the same is true, and the pitch changes as the wind pressure changes.

  It is also effective to register portamento times by phrases and event strings and set them automatically. It is also effective to automatically set a pitch change and a portamento time by selecting a tone color instead of a phrase or event sequence.

  According to an embodiment of the present invention, another example of changing the pitch expression method depending on the musical instrument is changing the slur processing method depending on the type of musical instrument. As illustrated below, even if the same slur symbol is specified, it is desirable to change the reproduction method depending on the type of musical instrument.

  In a keyboard musical instrument such as a piano, as shown in FIG. 23 (upper side), when slurring is applied to two consecutive sounds w1 and w2, the continuous sounds may overlap in time. Since it is preferable to make the length longer, for example, slur processing as shown in FIG. 23 (lower side) is performed.

  On the other hand, in the case of a wind instrument such as a clarinet or trumpet, in the case of a slur, it is unnatural if there is a temporal overlap between successive sounds, and it is better that the successive sounds on the back side have a small attack sound. In order to realize this, in performance data such as MIDI data, the two continuous sounds w1 and w2 in FIG. 24 (upper) are generated as one long note as shown in FIG. 24 (lower). In the middle of the process, it is preferable to use a processing method for changing the pitch according to data such as a pitch bend event.

(11) “Change tempo with volume”:
As with changing the pitch, changing the tempo with the volume is also an effective expression method, so in this tempo change rule, the tempo is changed with the volume. The fact that the tempo becomes slower as the volume is lower tends to occur on the performance side, so this rule is adopted. Of course, the setting can be reversed. Other processing in progress of the procedure can be executed in the same manner as the tempo change rule of (6).

  FIG. 25 is a flowchart showing an example of processing for changing the tempo according to the sound volume by the expression module EM of one embodiment of the present invention. In this processing flow, after setting the strength value α for the tempo in the first step M1, in step M2, the original MIDI data OD is divided into designated sections (for example, sections for each measure), and further in step M3. After calculating the average volume of the notes in each designated section, the process proceeds to step M4. In step M4, for example, a tempo coefficient is obtained from the calculated average volume by using a “volume-tempo coefficient” table set in advance so that the tempo becomes slower as the volume decreases. In this case, when the number of notes in the section is “0”, the tempo coefficient value is “1”. In the next step M5, the tempo change corresponding to the obtained tempo coefficient, the strength value α and the current tempo value is given to the MIDI data of each designated section, and this MIDI data is reproduced in the next step M6. Thereafter, if it is determined in step M7 that the reproduced MIDI data is good, the sound volume response process is terminated. If there is a problem, the process returns to step M1 and the processes of steps M1 to M7 are repeated.

  FIG. 26 is a flowchart showing another example of processing for changing the tempo according to the volume. After setting the strength value α for the tempo in the first step N1, the volume change is filtered by filtering the volume change in the step N2. A curve is obtained, and in the next step N3, a “tempo coefficient” is obtained by referring to the “volume-tempo coefficient” table based on the volume value of each point on the volume change curve. Further, at step N4, a tempo change according to the tempo coefficient, the strength value α and the current tempo value is given to the MIDI data, and at the next step N5, the MIDI data is reproduced. Thereafter, if it is determined in step N6 that the reproduced MIDI data is good, the sound volume response process is terminated. If there is a problem, the process returns to step N1 and the processes of steps N1 to N6 are repeated.

(12) “Change the tempo with the word tension”:
Tempo is closely related to the tension of performance. For example, when the feeling of tension increases, the tempo tends to be slow due to the feeling of playing each sound carefully. For this reason, changing the tempo by the parameter “tension” is effective as an interface. Therefore, in this tempo change rule, the tempo is changed in accordance with the parameter represented by the word “tension” in accordance with the feeling of the entire song or the feeling of a specific part such as “rust”. For example, as shown in FIG. 27, this parameter can be set to an arbitrary value by providing a “tension” knob 13k on the operation switch device 13 and moving the knob 13k. The tempo setting by the “tension” knob 13k sets the tempo change to about “+ 10%” and “−10%” with respect to the upper and lower maximum values corresponding to the positions where the tension is “high” and “low”, respectively.

  FIG. 28 is a flowchart showing an example of processing for changing the tempo according to the tension by the expression module EM according to one embodiment of the present invention. In the first step P1 of this processing flow, the “tension” knob 13k is operated to set the tension parameter value α, and then in step P2, a corresponding tempo coefficient is generated from the tension parameter value α of the knob 13k. . In the next step P3, a tempo change corresponding to the generated tempo coefficient and the current tempo value is given to the MIDI data, and in the next step P4, the MIDI data is reproduced. Thereafter, if it is determined in step P5 that the reproduced MIDI data is good, the tension response process is terminated. If there is a defect, the process returns to step P1 and the processes of steps P1 to P5 are repeated.

(13) “In a series of three-note quarter notes, take a long time for the first beat”:
In a group of notes that are not limited to a series of quarter-notes of three time signatures but are organized in some blocks, the expression of making the first note longer and recovering the used time with the remaining notes may be effective. Therefore, in this tempo change rule, a predetermined temporal expression is given to a group of notes. For example, a parameter α such as “strength of group of notes” is prepared, and first, this rule is applied through all songs to generate new song data. Then, when playback is performed, editing is performed so that a portion where the expression according to this rule is extreme is reset to “weak”, and a portion where the expression is insufficient is reset to “strong”.

  FIG. 29 is a flowchart showing an example of processing for changing the tempo and timing for a group of notes such as a series of quarter notes of three beats by the expression module EM of one embodiment of the present invention. In this processing flow, after setting a parameter value α representing the strength of the group of notes in the first step Q1, a predetermined group of notes is detected in step Q2, and in the next step Q3, the detected group of notes is detected. A tempo change corresponding to the group parameter value α and the current tempo value is given to the corresponding MIDI data, or the timing of the note is changed according to the group parameter value α. Then, after reproducing the MIDI data in step Q4, if the reproduced MIDI data is determined to be good in step Q5, the note group correspondence processing is terminated, and if there is a defect, the process returns to step Q1 and returns to steps Q1 to Q1. Repeat the process of Q5.

  Here, for the note group detection in step Q2, the method of handling the notes as a group may be in units of beats, and if a musical score is used, it is also possible to use the linkage of the flag of the notes. is there. Naturally, the tuplet shows a strong group feeling. As for step Q3, as described above, as a method of taking time, not only slowing down the tempo but also shifting the time (timing) of the note is an effective method. In this tempo change rule, it would be easier to handle shifting the time of a note.

  It is also effective to apply only to the left hand of the piano or only the accompaniment part. For example, when playing a waltz-type song on a piano, change the tempo by spending a lot of time on "Bun"-"Chat"-"Cha" "Bun" and playing the rest lightly. realizable.

(14) “For a tuplet of five or more tuplets, it is converted into a simple combination of numbers such as 2 + 3, and the first sound is emphasized over time”:
In the rule (13), it is described that the tuplet is handled as a note group. However, in the case of a note group of five or more tuplets, in order to obtain a performance with a clear string of notes, a tuplet + triplet is used. It is better to break it up into a smaller number of tuplets. For the tuplet decomposed in this way, the note group correspondence processing according to the rule (13) may be performed, but it is preferable to strongly set the first one among the decomposed tuplets. Therefore, in this tempo change rule, a long tuplet of five or more tuplets is divided into small tuplets such as double tuplet + triplet or triplet + double tuplet. Not only to emphasize the first note in each tuplet over time, but also to view long tuplets as a group of tuplets, the first tuplet slows down the tempo and the rest The tuplet of will make the tempo faster and give more effective expression.

  FIG. 30 is a flowchart showing an example of processing for changing the tempo for a long tuplet such as a tuplet of five or more tuplets by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step R1 of this processing flow, a predetermined tuplet group is detected in step R2, and the detected tuplet is divided into a plurality of small tuplets in the next step R3. . Further, in step R4, a tempo change according to the strength value α and the current tempo value is given to the MIDI data corresponding to the divided tuplet, or the timing of the note is changed according to the strength value α. To do. Then, after reproducing the MIDI data in step R5, if it is determined that the reproduced MIDI data is good in step R6, the note group handling process is terminated. If there is a defect, the process returns to step R1 and returns to steps R1 to R1. Repeat the process of R6.

(15) “For chords, slow down the tempo by the number”:
When playing chords, as the number of simultaneous sounds increases, there is a tendency to slow down the tempo in order to use more time to emphasize those sounds. Therefore, in this tempo change rule, a predetermined temporal expression is given such that the tempo is slowed according to the number of simultaneous pronunciations in the chord (the note density of the chord). For example, a parameter α such as “chord emphasis” is prepared, and this rule is applied to all songs to generate new song data. Then, when playback is performed, editing is performed such that the portion where the expression according to this rule is extreme is reset to “weak”, and the portion where the expression is not enough is set to “strong”.

  FIG. 31 is a flowchart showing an example of processing for changing the tempo according to the number of simultaneous chords generated by the expression module EM according to one embodiment of the present invention. After setting the strength value α in the first step S01 of this processing flow, in step S02, a chord is detected and the note density of the chord is evaluated to detect the number of simultaneous pronunciations. In the next step S03, a tempo change corresponding to the number of simultaneous pronunciations, the strength value α, and the current tempo value is calculated, and the calculated tempo change is added to the MIDI data corresponding to the detected chord. Then, this MIDI data is reproduced in step S04, and if the reproduced MIDI data is determined to be good in the next step S05, the chord simultaneous pronunciation number response process is terminated. If there is a defect, the process returns to step S01. The processes in steps S01 to S05 are repeated.

(16) “Set tempo coefficient in conjunction with tone”:
According to this tempo change rule, a “tone-tempo coefficient” table having a tempo coefficient value corresponding to each timbre is prepared in advance, and each time the timbre changes, this table automatically changes the tempo slightly. Can be obtained. Since it is unnatural to change the tempo for each part for the performance of the ensemble, this rule can be applied to the case of solo or the melody part of the ensemble. In this rule, the tempo is not changed based on the timbre itself, but the tempo is changed by a timbre parameter such as EG attack time (envelope generator attack time). If it is fast, the tempo is made faster, and if the attack is slow like a violin, the tempo is made slower.

  It is also effective for some timbres not to simply change the tempo, but to shift the note in front of the tempo beat. In other words, it is said that the chords are stabilized if the bass is pronounced quickly, and even the same instrument may be shifted differently depending on the pitch. For example, such a shifting process is effective in a low musical instrument system such as a bass guitar or a bass drum. Therefore, if a timbre for enabling such processing is set in advance, it is possible to provide a mechanism that can automatically adjust the timbre only by selecting a timbre. In this case, it is not necessary to limit to solo or melody part.

  FIG. 32 is a flowchart showing an example of processing for setting a tempo coefficient in conjunction with a tone color and tone color parameters by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step T1 of this processing flow, the timbre type or the predetermined timbre parameter is detected in step T2, and in the next step T3, the previously prepared “tone type / tone parameter-tempo” is prepared. Using the “coefficient” table, a tempo change corresponding to the detected timbre type or predetermined timbre parameter, strength value α, and current tempo value is obtained, and the obtained tempo change is added to the MIDI data. Alternatively, the timing of the note is changed according to the strength value α. Then, this MIDI data is reproduced in step T4, and if the reproduced MIDI data is good in the next step T5, this chord simultaneous sound number response processing is terminated. If there is a defect, the process returns to step T1. The processes in steps T1 to T5 are repeated.

(17) “Considering fingering, change tempo coefficient according to finger”:
Human fingers include easy-to-move fingers and difficult-to-move fingers. For example, ring fingers and little fingers are generally considered to be difficult to move. Although this is a problem that can be improved by learning performance, the tempo is slower for fingers that do not move easily. Therefore, in this tempo change rule, the non-uniformity performed by humans can be expressed by setting the tempo coefficient according to the finger of the fingering.

  FIG. 33 is a flowchart showing an example of processing for setting a tempo coefficient according to a finger of the fingering by the expression module EM according to one embodiment of the present invention. After setting the strength value α in the first step U1 of this processing flow, in step U2, fingering fingers are detected from the finger data in the song data, and in the next step U3, the “finger-tempo” prepared in advance is detected. A tempo change according to the detected fingering finger, the strength value α and the current tempo value is calculated using the “coefficient” table, and the calculated tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step U4, and if the reproduced MIDI data is determined to be good in the next step U5, this fingering finger response process is terminated, and if there is a defect, the process returns to step U1 and returns to step U1. Repeat the process of ~ U5.

(18) “Slow down the tempo if there is position movement for fingering”:
Fingering involves position movement, with and without position movement. Naturally, the ease of playing is different. In particular, if the position moves with a wide pitch interval, the difficulty of playing further increases. Therefore, in this tempo change rule, by setting the tempo coefficient according to the amount of position movement including the presence or absence of position, the fluctuation of the tempo accompanying the movement of position is reproduced and the non-uniformity played by humans is expressed. Can do.

  FIG. 34 is a flowchart showing an example of processing for setting a tempo coefficient in accordance with finger position movement by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step V1 of this processing flow, fingering and position movement are sequentially detected in steps V2 and V3. In the next step V4, “position movement amount—tempo coefficient” The table is used to calculate a tempo change corresponding to the detected position movement, the strength value α and the current tempo value, and the calculated tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step V5, and if the reproduced MIDI data is determined to be good in the next step V6, this fingering position movement response process is terminated. If there is a defect, the process returns to step V1. The process of V1 to V6 is repeated.

(19) “Consider fingering and slow down the tempo at low positions”:
With a keyboard instrument, the lower the keyboard, the heavier the keyboard might feel, and with a string instrument, the lower the position, the wider the fingers must be opened, so the lower the position, the lower the tendency. In addition, thick strings must be pressed hard, making them difficult to play fast. Therefore, in this tempo change rule, non-uniformity performed by humans can be expressed by setting a tempo coefficient in accordance with the fingering position.

  FIG. 35 is a flowchart showing an example of processing for setting a tempo coefficient in accordance with the fingering position by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step W1 of this processing flow, fingering and position are sequentially detected in steps W2 and W3, and in the next step W4, a “position-tempo coefficient” table prepared in advance is used. Then, a tempo change corresponding to the detected position, the strength value α and the current tempo value is calculated, and the calculated tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step W5, and if the reproduced MIDI data is determined to be good in the next step W6, this fingering position response process is terminated, and if there is a defect, the process returns to step W1 and returns to step W1. Repeat the process of ~ W6.

(20) “When you apply pitch bend, touch the tempo automatically”:
Although it is a kind of fingering evaluation, applying pitch bend is often accompanied by position movement (slide or glissando), so it is effective to slow down the tempo using the place where pitch bend is applied. There is a case. Therefore, according to this tempo change rule, when pitch bend is applied, a tempo change that automatically slows down the tempo when the bend is applied is automatically applied.

  FIG. 36 is a flowchart showing an example of processing for setting the tempo according to the pitch bend by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step X1 of this processing flow, the pitch bend is detected in step X2, and in the next step X3, the detected pitch bend, the strength value α and the tempo change according to the current tempo value are performed. The calculated tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step X4, and if the reproduced MIDI data is determined to be good in the next step X5, this pitch bend handling process is terminated, and if there is a defect, the process returns to step X1 and returns to steps X1 to X5. Repeat the process.

(21) “Increase the tempo when the guitar is pulled off”:
Contrary to the pitch bend of (20), the hammering-on of the guitar (hammering)
on), pulling off, etc., there is a tendency that the tempo is faster because the strings are not newly repelled. Therefore, in this tempo change rule, the tempo is increased when pulling off or the like. For this purpose, for example, it is sufficient to detect hammering on, pulling off, etc. at the pitch bend detection part. For this detection, for example, the pitch bend is suddenly changed by more than a semitone or the hammer is turned on with a musical score symbol. Etc. are used.

  FIG. 37 is a flowchart showing an example of processing for speeding up the tempo at the time of guitar pull-off or the like by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step Y1 of this processing flow, hammering on, pulling off, etc. are detected in step Y2, and in the next step Y3, the detected hammering on, pulling off, etc. are detected. A tempo change corresponding to α and the current tempo value is calculated, and the calculated tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step Y4, and if the reproduced MIDI data is determined to be good in the next step Y5, the response process such as pull-off is terminated. If there is a defect, the process returns to step Y1 and returns to step Y1. Repeat the process of ~ Y5.

(22) “Set tempo coefficient according to piano sustain pedal operation”:
In a piano, the sound becomes richer when the sustain pedal is depressed, but in this case, the separation of the sound becomes worse when played fast, so there is a tendency to slightly slow down the tempo. Therefore, in this tempo change rule, a predetermined tempo coefficient is set according to the sustain pedal operation of the piano. For this purpose, it is only necessary to detect sustain bedallon or the like in the pitch bend detection part.

  FIG. 38 is a flowchart showing an example of processing for setting a tempo coefficient according to the sustain pedal operation of the piano by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step Z1 of this processing flow, the sustain pedal is detected in step Z2, and in the next step Z3, the detected sustain pedal, the strength value α, and the tempo corresponding to the tempo data are detected. The change is calculated, and the calculated tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step Z4, and if the reproduced MIDI data is determined to be good in the next step Z5, this sustain pedal response process is terminated, and if there is a defect, the process returns to step Z1 and returns to steps Z1 to Z1. Repeat the process of Z5.

(23) “Set tempo coefficient according to vibrato depth and speed”:
Applying vibrato deeply is when you want to produce a lot of reverberation, so the tempo tends to be slow. Also, when applying vibrato quickly, it is when trying to produce a sound with a sense of tension, so the tempo is also slowed to perform so as to confirm each sound. On the other hand, when you slow down the vibrato, you are trying to produce a calm sound, so this also tends to slow down the tempo.

  Therefore, in this tempo change rule, the vibrato depth and speed are determined using a “vibrato depth-tempo coefficient” table as shown in FIG. 39 and a “vibrato speed-tempo coefficient” table as shown in FIG. Accordingly, a tempo coefficient K is set. That is, from the above, the relationship between vibrato and tempo change is as follows. As for the depth of vibrato, as shown in FIG. 39, the tempo becomes slower as it is deeper, whereas as for the speed of vibrato, as shown in FIG. In addition, it is preferable that the curve has a moderate blur, in which the tempo is slow regardless of whether it is fast or slow. 39 and 40, the tempo coefficient K is exceptionally set to "1" when the vibrato depth and speed values are "0".

  In implementation, it is only necessary to detect vibrato in the bend detection portion and to further search for the setting of the vibrato depth and speed parameters in each table. If there is no vibrato parameter, information on the vibrato depth and speed may be obtained from the output waveform.

  FIG. 41 is a flowchart showing an example of processing for setting a tempo coefficient according to the depth and speed of vibrato by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step Aa1 of this processing flow, the vibrato and the depth and speed of the vibrato are sequentially detected in steps Aa2 and Aa3. In the next step Aa4, the detected vibrato depth and speed using the “vibrato depth-tempo coefficient” table as shown in FIG. 39 and the “vibrato speed-tempo coefficient” table as shown in FIG. The tempo coefficient K corresponding to is calculated, a tempo change corresponding to the tempo coefficient K, the strength value α, and the current tempo value is calculated, and the calculated tempo change is added to the MIDI data. Then, the MIDI data is reproduced at step Aa5, and if the reproduced MIDI data is satisfactory at the next step Aa6, the vibrato speed / depth response process is terminated. If there is a defect, the process proceeds to step Aa1. The processing of return steps Aa1 to Aa6 is repeated.

(24) “Strings' trills use multiple parts to slightly shift timing”:
In general, a trill such as a string has a plurality of long notes as a “musical score on a performance” shown in FIG. 42 (2). This is achieved by breaking it down into short notes. However, in an actual string spurt, a plurality of players perform, and it is strange that it is perfectly synchronized with the string trill, so it is difficult to think that all the players will play the trill fine notes at the same timing. Therefore, by using this timing change rule, a plurality of parts are used and the timings of these parts are slightly shifted. In other words, in order to avoid such a synchronous performance state and reproduce a natural performance with MIDI sound source strings, effective results are obtained by using multiple parts and slightly shifting the timing of the trill in each part. Can be obtained.

  Note that it is preferable that the plurality of parts have slightly different timbres (for example, violin and viola). Also, as a method of slightly shifting the timing of the trill, it is effective to change the on time and duration of the trill's fine note using random numbers.

  FIG. 43 is a flowchart showing an example of a process of dividing into a plurality of parts according to the string trill and shifting the timing of each part by the expression module EM of one embodiment of the present invention. After setting the strength value α in the first step Bb1 of this processing flow, the tolyl portion is detected in step Bb2, and the tolyling portion is copied to a plurality of parts in the next step Bb3. In this case, it is preferable to change the tone color slightly for each part. Thereafter, in step Bb4, the timing of the MIDI data of the trill portion is changed so as to differ for each part according to the strength value α. Then, the MIDI data is reproduced in step Bb5. If the reproduced MIDI data is determined to be good in the next step Bb6, the string stroil timing process is terminated. If there is a defect, the process returns to step Bb1 and returns to step Bb1. Repeat the process of ~ Bb6.

  According to one embodiment of the present invention, the method of multi-part processing of trill described with reference to FIGS. 42 and 43 can be applied to change in pitch parameters. In other words, it is strange that string trills are perfectly synchronized, so they are divided into multiple parts and given subtle shifts. In order to reproduce this with MIDI sound source strings, in step Bb1, a strong and weak value α is set for the pitch shift, and in steps Bb2 and Bb3, the same processing as in FIG. Later, in Step Bb4, an effective result can be obtained by slightly shifting the pitch of the trill in each decomposed part. In this case, it is preferable that the shifted parameter includes not only the pitch but also the timing.

(25) “Set tempo by learning function”:
This tempo assignment method has a learning function for tempo setting, and configures a system that automatically predicts the relationship between pitch change and tempo change, etc., and allows users to automatically build tempo change rules It is. As an example, after manually inputting a tempo change to the middle of a certain song, the remaining tempo change is automatically input by a learning function. In this case, when the remaining song data is input, the phrase is interpreted and compared with the already processed phrase, the same phrase is given the same tempo change as the processed phrase, and the new phrase is changed. Randomize and so on. As another example, a tempo change of a certain song is analyzed, the other song is interpreted, and a corresponding analyzed tempo change is given according to a comparison result with the analyzed song.

  FIG. 44 is a flowchart showing an example of processing based on the learning function of the expression module EM according to one embodiment of the present invention. In this process, the learning routine CcS learns the relationship between the phrase and the tempo change from the MIDI data to which the tempo change has already been applied, and stores the learning result in a predetermined area in the RAM 3.

  Next, when assigning the tempo change, after setting the strength value α in step Cc1, in step Cc2, based on the learning result, the MIDI data of the phrase to which no tempo change has been given yet. The tempo change according to the strength value α and the current tempo value is given. Then, the MIDI data is reproduced in step Cc3, and if the reproduced MIDI data is determined to be good in the next step Cc4, the tempo assigning process by this learning is terminated. If there is a defect, the process returns to step Cc1 and returns to step Cc1. Repeat the process of ~ Cc4.

(26) “Set tempo change using library”:
If a tempo change once generated corresponding to various pieces of characteristic information of performance data is cut out in a certain time section and made into a library, it may be applicable to other parts as well. This tempo assigning method provides an easy-to-use tempo change method by creating a library of tempo changes for feature information. When creating a library, if you name the library and save it, it will be even easier to use. Moreover, it is good to memorize | store with a relative tempo value, and to be able to expand-contract in a time change direction or a tempo value change direction.

  FIG. 45 is a flowchart showing an example of processing for setting a tempo change using a library of the expression module EM according to one embodiment of the present invention. In this processing, in the library creation routine DdL, a tempo change is extracted from MIDI data to which a tempo change has already been applied, the extracted tempo change is converted into a relative value, converted into a library, and stored in a predetermined area of the external storage device 9. deep.

  Next, in giving a tempo change, in step Dd1, a tempo change is selected from the library corresponding to predetermined feature information of the MIDI data, and the selected tempo change is changed in the time direction and / or tempo value direction. Is expanded and contracted at a predetermined magnification α. In the next step Dd2, the selected and expanded tempo change is converted into an absolute value according to the current tempo value, and the converted tempo change is added to the MIDI data. Then, this MIDI data is reproduced in step Dd3, and if the reproduced MIDI data is determined to be good in the next step Dd4, the tempo assigning process using this library is terminated. On the other hand, if there is a problem, the process returns to step Dd1 to change the magnification α, or another tempo change is selected and the processes of steps Dd2 to Dd4 are repeated.

  According to one embodiment of the present invention, the processing method using the library described with reference to FIG. 45 can be applied to change in pitch parameters to create a pitch change library. That is, if a pitch change once generated by the same processing as in FIG. 45 is cut out into a library in a certain time interval, it may be applicable to other portions in the same manner. In this case, it goes without saying that if the library is named and saved, it will become even easier to use.

(27) “Set tempo coefficient according to lyrics (poem)”:
In the case of a song with a song, the tempo may change depending on the contents of the lyrics even if the melody is the same. Therefore, in this tempo changing method, a tempo coefficient is set according to the lyrics (poetry) using a procedure in which a word is registered in advance together with a tempo coefficient and the tempo is changed when the word appears. When setting the tempo coefficient in advance, the tempo is assumed to be fast for bright words, while the tempo is assumed to be slow for dark words, and the tempo is slow for important words. The coefficient value should be registered. It is also possible to change only the word portion, change the tempo of the entire section including the word, or designate the target portion of the tempo change.

  FIG. 46 is a flowchart showing an example of processing for setting a tempo coefficient according to lyrics (poetry) in the expression module EM according to one embodiment of the present invention. In this process, after a strong value α is set in the first step Ee1, a predetermined word is detected from the lyrics data of the original MIDI data OD in step Ee2. In the next step Ee3, a tempo coefficient set corresponding to the detected word, a set strength value α, and a tempo change corresponding to the current tempo value are added to the MIDI data. Then, this MIDI data is reproduced in step Ee4, and if the reproduced MIDI data is determined to be good in the next step Ee5, this lyrics response processing is terminated, and if there is a malfunction, the process returns to step Ee1 and steps Ee1 to Ee5. Repeat the process.

(28) “Correction of timing according to output waveform of sound source”:
Even with the same sound source, the audible sounding timing differs depending on the type of timbre. Therefore, in this timing change rule, the timing is corrected according to the output waveform of the sound source using a mechanism for correcting the on time of the note while observing the actual output waveform. FIG. 47 is a diagram illustrating the relationship between the sound source output waveform and the correction timing. In the figure, if the output waveform TW of a sound source of a certain tone color starts to sound at timing ts and rises and reaches the maximum volume value (1.00) at timing tm, for example, the sound source output waveform TW By shifting the time point tc at which the volume value reaches 80% of the maximum value to the original sounding start timing, the time point tc can be adjusted to the sounding timing on hearing. Further, the correction timing tc is not limited to 80% of the maximum volume value, but is selected to be an appropriate value according to circumstances such as the type of sound source, and may be 100% (timing tm), for example.

  As a specific procedure, for example, in the case of music data production, timing correction data is generated while playing once. Thereafter, if the timing corrected by the generated correction data is reproduced, the timing shift due to the timbre can be corrected, and further the timing shift due to the sound source can be corrected. On the other hand, in the case of real-time performance (playback), the first note can be corrected only after the pronunciation begins, but for the second and subsequent notes, the correction amount is predicted and the predicted value is corrected by learning. It is sufficient to correct the timing. Furthermore, if the output waveform of all timbres and all pitches is observed before the music data is reproduced, correction can be made from the first sound without reproduction.

  FIG. 48 is a flowchart showing an example of processing for correcting the timing according to the output waveform of the sound source in the expression module EM according to one embodiment of the present invention. In this process, after setting the strength value α in the first step Ff1, the output waveform TW of the sound source is observed in step Ff2. In the next step Ff3, a timing correction amount (ts-tc) is calculated according to the observation result. Further, in step Ff4, the timing of the MIDI data is determined according to the calculated correction amount and the strength value α. to correct. Then, this MIDI data is reproduced in step Ff5, and if the reproduced MIDI data is satisfactory in the next step Ff6, this output waveform correspondence processing is terminated, and if there is a defect, the process returns to step Ff1 and returns to steps Ff1 to Ff1. The process of Ff6 is repeated.

(29) “Forte (f) just before piano (p) should be short”:
When the piano (p) appears after the forte (f) of the strength symbol is followed, if the sound of the forte (f) immediately before the piano (p) is shortened, the sound of the piano (p) is easily heard. In this tempo change rule, the tempo change of a sound accompanied by such a dynamic symbol is controlled.

  FIG. 49 is a flowchart showing an example of processing for dynamic symbols in the expression module EM according to one embodiment of the present invention. In this process, after setting the strength value α in the first step Gg1, in step Gg2, the piano (p) after the forte (f) is detected from the original MIDI data OD is detected. In the next step Gg3, the gate time of the forte (f) sound immediately before the detected piano (p) is shortened according to the strength value α. Then, in step Gg4, the MIDI data subjected to the shortening process is reproduced, and in the next step Gg5, if it is determined that the reproduced MIDI data is good, the dynamic symbol corresponding process is terminated. Returning to Gg1, the processes of steps Gg1 to Gg5 are repeated.

(30) “Longer sound just before staccato”:
If the sound just before the staccato is made longer, it will sound like a staccato and will sound sharp. In this tempo change rule, the tempo change of the sound accompanied by this staccato is controlled.

  FIG. 50 is a flowchart showing an example of processing for staccato in the expression module EM according to one embodiment of the present invention. In this process, after setting the strength value α in the first step Hh1, in step Hh2, staccato is detected from the original MIDI data OD. In the next step Hh3, the gate time of the sound immediately before the detected staccato is extended according to the strength value α. In step Hh4, the decompressed MIDI data is reproduced. If the reproduced MIDI data is determined to be good in the next step Hh5, the staccato handling process is terminated. If there is a defect, step Hh1 is performed. The process of steps Hh1 to Hh5 is repeated.

(31) “Determine the tempo after looking at the whole”:
As explained so far, there are various factors for changing the tempo. As a result of adding various tempo changes throughout a song, it may be very different from the tempo before the change. In this tempo determination method, in preparation for such a situation, look at the result of changing the tempo through one song, so that the average of the result is the tempo value that was originally set (the method of taking the average is Any method can be adopted as necessary.) Performs overall tempo correction. In order to perform this overall tempo correction, for example, the entire tempo is corrected uniformly, or the tempo in the section where the number of tempo changes is high is preferentially corrected instead of the overall uniform correction. Note that an allowable amount that the tempo is not corrected if the difference between the original tempo and the average tempo is within a predetermined range may be selected in advance or may be selectable by the user.

  FIG. 51 is a flowchart showing an example of the overall review process of the expression module EM according to one embodiment of the present invention. In the first step Jj1 of this process, each tempo change is given to the original MIDI data OD based on a predetermined rule adopted as necessary from the tempo change rules (1) to (30) described so far. To do. In step Jj2, the entire tempo of the MIDI data to which individual tempo changes are applied is equal to or close to the original average tempo, or the total performance time is the original total performance time. The overall tempo is modified to match or be close.

  Thereafter, the process proceeds to step Jj3 to verify whether a desired tempo or total performance time has been obtained by automatic calculation or playback trial listening of one piece of MIDI data. The review process is terminated. On the other hand, if there is a problem, the process proceeds to step Jj4 to ask whether or not to perform individual tempo change processing. If it is determined that it is necessary to perform individual tempo change processing (YES), the processing returns to step Ji1 and the processing of steps Jj1 to Jj2 is performed again. Otherwise (NO), the processing of step Jj2 is performed again. Do.

  According to one embodiment of the present invention, the overall review processing method described with reference to FIG. 51 can be applied to change in pitch parameters, and the pitch can be adjusted again after looking at the whole. Adding facial expressions with various pitch changes may eventually result in the entire song being shifted by a certain pitch (eg, with pitch bend data). In such a case, it may be easier to handle it by looking at the whole and then grouping it into other parameters such as master tuning, pitch shift, and fine tuning.

[Embodiment 2]
Next, “Embodiment 2” will be described in which facial expression is performed mainly on the volume parameter as a specific performance parameter (some specific examples include other performance parameters such as a pitch parameter). . 52 and 53 are flowcharts illustrating an overview of functions of the expression module in the system shown in FIG. 2 as a processing flow executed by the CPU 1 from the viewpoint of automatic parameter editing. First, the outline will be described with reference to FIGS. 52 and 53, and then specific examples (A) to (V) of “Embodiment 2” will be described in detail with reference to FIGS. 54 to 84. And

The parameter automatic editing apparatus of the present embodiment mainly executes the following control processing. That is,
[1] Analyzing the supplied performance data, selecting a parameter to which a change is to be added and determining its content based on the analysis result, and a parameter change giving process for giving a parameter change to the performance data [2] Reproduction processing for reproducing performance data that has been expressed by the processing of [1] [3] Confirmation processing for confirming performance data that has been expressed by the processing of [1].

  In this embodiment, the performance data is assumed to be created in the SMF (Standard MIDI File) format, for example. However, the performance data may be in a different format. The confirmation of the performance data in the confirmation process [3] is specifically to display the performance data in some form on the display device (display) 14 and display the performance data on the displayed performance data [1]. It means that the user confirms the processing result of [1] by displaying the parameter value and the position given in the processing of FIG. Then, during this confirmation process, the user may be allowed to specify a position where the parameter change giving process of [1] is not desired.

  In addition, as described in the examples of “Embodiment 1”, as described in [2] and [3], the strength value α of a predetermined performance parameter such as volume and pitch is set (for example, Step A2), a method of reproducing performance data (MIDI data, etc.) to which the performance parameter is assigned, and resetting the strength value α according to the result of the quality determination of the performance data after the parameter assignment (for example, Step A7). It goes without saying that can be adopted.

  FIG. 52 is a flowchart showing a procedure of a main routine executed by the parameter automatic editing apparatus of the present embodiment, particularly the CPU 1.

  In the figure, first, initialization processing for clearing the RAM 3 and selecting performance data to be processed first (any one of the processes [1] to [3]) is performed (step S1).

  Next, the performance data selected by the user is retrieved from, for example, various storage media (the hard disk, FD, CD-ROM, etc.) or from the outside (the other MIDI device 17 and server computer 20, etc.). The performance data storage area secured at the position is stored (step S2).

  Next, the parameter change imparting process [1] (the specific process procedure will be described later with reference to FIG. 53) is performed on the performance data (step S3). The parameter change giving process in FIG. 53 is configured by extracting processes common to all of a plurality (19 in this embodiment) of parameter change giving processes to be described later. Therefore, in step S3, more specifically, at least one or more selected by the user among the parameter change assigning processes 1 to 19 is executed.

  Further, other processes other than the parameter change giving process are executed (step S4). In the other processes, the processes [2] and [3] are performed. In addition, for example, a new performance data creation process may be performed.

  Then, it is determined whether or not the user has performed an operation for terminating the main routine (step S5). When the operation is performed, the main routine is terminated, and when the operation is not performed, the process proceeds to step S2. Returning, the process of steps S2 to S5 is repeated.

  FIG. 53 is a flowchart showing the procedure of the parameter change giving process in step S3.

  In the figure, first, performance data stored in the performance data storage area is analyzed (step S11). Specifically, the performance data analysis refers to extracting a portion of the performance data where the parameter value should be changed.

  Next, an optimum parameter value (change value) is determined for each extracted location (step S12).

  Then, after each determined parameter value is assigned to the corresponding position in the performance data (step S13), the parameter change giving process is terminated.

  Note that the processing described with reference to FIGS. 52 and 53 gives changes to the time control information (time parameter) such as the tempo and timing described above in addition to the volume parameter, and changes to other parameters such as the pitch. Needless to say, it is used to give

  The present invention is characterized by this parameter change providing process, and the specific processing method will be described in detail below.

(A) Processing in a portion where a note is rising FIG. 54 is a flowchart showing the procedure of the volume parameter change giving process 1. In the volume parameter change giving process 1, the volume parameter of the performance data is set to the pitch of the performance data. Make a facial expression that changes according to the change. Specifically, the volume parameter is changed according to the following algorithm. That is,
(1) When the pitch (note number) of the note event sequence is increasing, the volume is gradually increased. (2) When the pitch of the note event sequence is changed from increasing to decreasing, the change point Here, note events are generally used to include both note-on events and note-off events. In this embodiment, note-off events are not taken into account. The event is used only for the note-on event.

  In FIG. 54, first, when the user designates a section in the performance data stored in the performance data storage area (hereinafter, this performance data is referred to as “selected performance data”) to which the volume change is to be given, the designated section is designated. The data is stored in the work area of the RAM 3 and is set as an analysis section (step S21). Needless to say, the entire section of the performance data may be set as the analysis section without specifying the section. Here, since the performance data is sequence data, each performance data has a spread with respect to the time axis. For this reason, the concept of a section can be brought there.

  Next, a partial section in which the pitch of the note event sequence has an upward tendency and a downward tendency is searched and cut out from the analysis section (step S22).

  FIG. 55 is a diagram showing an example of a note event sequence in which the pitch tends to rise. FIG. 55A shows an example of a simple rising sound system, and FIG. 55B is not a simple rising sound system. As a whole, an example of a rising sound system is shown. That is, in the present embodiment, not only the section of the note event sequence shown in FIG. 55 (a) is cut out as a partial section in which the pitch tends to rise, but also the section of the note event sequence shown in FIG. 55 (b). Then, the pitch is cut out as a partial section in which the pitch tends to rise. Hereinafter, a method for searching for this partial section will be described.

  First, only note events are extracted from the selected performance data, and note numbers included in the note events are arranged in time series to generate a time series note number string. By applying a high-pass filter (HPF) process to the note number sequence, a pitch change tendency is calculated. If the change in the calculated value is too small in time only by this HPF process, the calculated value (data string) is further subjected to a low-pass filter (LPF) process, and the change is temporally changed. To smooth. By applying such a filtering process to the note number row, a sound system that can be said to be an ascending sound system as a whole as shown in FIG. 55 (b) becomes a simple ascending sound system as shown in FIG. 55 (a). You can search as well.

  FIG. 56 is a diagram showing an example of time-series data representing the change tendency of the pitch calculated in this way, and the vertical axis shows a trend of increasing the pitch on the positive side and a tendency of decreasing the pitch on the negative side. The horizontal axis indicates time. That is, in the figure, among the analysis sections t0 to t7, sections t0 to t1, t2 to t3, t4 to t5, and t6 to t7 are cut out as partial sections in which the pitch tends to increase.

  Moreover, the partial section in which the pitch tends to decrease can be easily cut out in the process of cutting out the partial section in which the pitch tends to increase. That is, in FIG. 56, each of the sections t1-t2, t3-t4, t5-t6 where the time series data takes a negative value corresponds to a partial section where the pitch tends to decrease. What is necessary is just to cut out as a partial area where the pitch tends to fall.

  Note that the length of the cut-out section may be an integer multiple of the measure length or the beat length, or may be an arbitrary length. Then, as shown in FIG. 56, when there are a plurality of partial sections whose pitches tend to increase in one analysis section, all of the partial sections are cut out.

  Returning to FIG. 54, for each segmented segment, the pitch change speed of the note number sequence belonging to the segment is calculated, and given to each note event of the note event sequence belonging to the segment according to the calculation result. A volume parameter change giving pattern to be determined is determined (step S23). Here, the pitch change speed means the slope of the pitch change, that is, the pitch change amount per unit time. The change imparting pattern is a template that is created in advance and represents a change tendency of the volume parameter stored in the ROM 2, for example. Specifically, this template uses time series data of either data that replaces the original volume parameter value or data that corrects the original volume parameter value. In the present embodiment, the change tendency of the volume parameter means a change tendency that the value of the volume parameter is gradually increased in the partial section in which the pitch tends to increase. A template having a large increase rate is selected for a partial section that is changed every time and the pitch change speed is large.

  In addition, as a change provision pattern, you may make it calculate with a predetermined | prescribed arithmetic expression instead of using the said template.

  Next, the volume parameter value to be assigned to each note event of the note event sequence belonging to the section is changed for each of the cut out partial sections based on the determined change imparting pattern (step S24). If there is no volume parameter to be changed, the volume parameter may be added. Here, if the velocity included in the note event is specifically taken as the volume parameter, the change of the volume parameter value means the change of the velocity value. In addition, expression data may be inserted in association with each note event in the section.

  Next, the partial sections cut out in step S22 are arranged in time series, and when the pitch changes from an upward trend to a downward trend, that is, in FIG. 56, the sign of the time-series data after filtering is changed from positive to negative. The changing time point (time t1, t3, t5) is detected, and the volume data indicating the accent is given to the note at that point, that is, the beginning position or the last note (note event) of the cut out segment (step S25). ).

  Then, it is determined whether or not the user has performed an operation for ending the present volume parameter change providing process 1. If the user has performed the operation, the present volume parameter change providing process 1 is ended. If not, the process returns to step S21. The processes of steps S21 to S26 are repeated.

  According to one embodiment of the present invention, the technique of “volume parameter change applying process 1” described with reference to FIGS. 54 to 56 can be applied to change a pitch parameter. In other words, in the pitch increasing system, the pitch gradually shifts and a so-called “not rising” performance situation arises. Correspondingly, the portion where the pitch is increasing is searched from the performance data (MIDI data, etc.). (Step S22), the speed of pitch increase is evaluated using the HPF (also LPF in some cases) (Step S23), and a change in pitch is calculated based on a predetermined pitch assignment pattern (Step S24). Is added to the performance data (step S25), it is possible to realize the expression that the pitch is gradually lowered in units smaller than the key in the portion where the pitch of the note is increasing.

  In this way, when the speed of pitch increase is detected from performance data using HPF or the like, and the pitch is changed according to the detected value, the pitch shifts greatly lower when the pitch rises due to a drastic change, and On the contrary, when descending due to a drastic change, it is possible to realize that the pitch is greatly shifted. However, depending on the case, it may be desired to make the pitch shift lower as the change is larger, whether it is a decrease or an increase. In that case, the pitch change may be calculated according to the absolute value detected by the HPF.

  It should be noted that the expression processing for shifting the pitch lower in the pitch raising system in this way is more natural not to use this function for a keyboard instrument or the like that has few elements whose pitch is shifted. Therefore, it is necessary to check the musical instrument type by determining the timbre when analyzing performance data (for example, step S22).

  This expression processing system can be used to observe changes in the performance range, so if there is a significant change in the performance range, portamento (for strings instruments without frets) or slides (for strings instruments with frets) ) Is also possible.

(B) Process for Representing Excitement Accompanied by Progress of Song FIG. 57 is a flowchart showing the procedure of the volume parameter change giving process 2. In the volume parameter change giving process 2, the value of the volume parameter of the performance data is shown. By gradually increasing the expression, the expression is increased to increase the degree of excitement.

  In the figure, first, the total time length of the selected performance data is calculated (step S31).

  Next, a change imparting pattern is selected (step S32). The change imparting pattern refers to pattern data representing the volume change tendency of the entire song. For example, a plurality of types of pattern data are prepared in the form of table data in the ROM 2 in accordance with the user's selection. Or one of them is automatically selected. FIG. 58 is a diagram showing an example of a change imparting pattern, in which the vertical axis represents the volume coefficient and the horizontal axis represents the direction of progression of the music. As shown in the figure, the volume coefficient, that is, the coefficient for weighting the volume data (expression data or volume data) increases as the music progresses. However, increasing the volume data infinitely impairs the hardware limitations of the tone generator circuit 7 and natural expression of the music, so the volume coefficient is predetermined as the music progresses as shown in FIG. It is desirable to have characteristics that converge to a finite value of.

  Next, a change value for each volume change section is calculated based on the total time length data calculated in step S31 and the change applying pattern selected in step S32 (step S33). This is because the song length is arbitrarily different depending on the performance data, but the change imparting pattern is based on a predetermined fixed length, so when the selected performance data is determined, the change impart is given accordingly. This is because the entire length of the pattern needs to be expanded and contracted (scaling) to match the total time length of the selected performance data. In this way, from the change imparting pattern in which the total length is matched with the total time length of the selected performance data, the change for each volume change section (for example, for each measure, but for each performance phrase) Determine the value. That is, in the present embodiment, since the change imparting pattern has a table data format, data at positions corresponding to the respective sound volume change sections are read from the scaled table data. Since the table data is composed of volume coefficient values as described above, each read volume coefficient value is multiplied by volume data for each volume change section (weighting is performed). Therefore, it is necessary to convert it into a change value.

  Finally, each of the calculated change values for each volume change section is inserted into the corresponding position of the selected performance data (step S34). Specifically, the calculated change value for each measure (volume data weighted by the volume coefficient value) is recorded (inserted) at the head position of each measure in the selected performance data.

  When the selected performance data is composed of a plurality of tracks, the same change imparting pattern may be used for each track, or different change imparting patterns may be used. You may do it. Furthermore, a function that allows the user to create or edit the change imparting pattern may be provided. Specifically, when the change imparting pattern is the table data shown in FIG. 58, when the user selects a numerical value indicating the magnification, each data value is uniformly increased or decreased by the magnification, and the overall characteristics of the table data are determined. Do not change. At this time, instead of the numerical value, words such as “strong”, “medium”, “weak”, “no change” are selected, and when the user selects any of these words, the CPU 1 selects this word. The converted words may be converted into a magnification, and each data value may be uniformly increased or decreased by the magnification. Furthermore, the data value may be increased or decreased at a different magnification for each predetermined range.

  According to one embodiment of the present invention, the “volume parameter change applying process 2” described with reference to FIGS. 57 and 58 can be applied to assign not only a volume parameter but also a pitch parameter. A corresponding effect can be obtained for the pitch. That is, the length of the musical piece of performance data (MIDI data or the like) is evaluated (step S31), the parameter of the change imparting pattern (pitch change table) is determined (step S32), and the pitch change is calculated based on this (step S32). By performing facial expression (step S33) and performing facial expression (step S34), it is possible to perform facial expression that gradually increases the pitch parameter value to gradually increase the pitch and increase the degree of excitement as the music progresses.

  Here, as shown in FIG. 59, a “music progression-pitch coefficient” table (the vertical axis is “pitch coefficient”) whose coefficient value increases is used as the change imparting pattern, as in the table of FIG. Is preferred. In this case, for example, the pitch may be increased at a rate according to the “music progression-pitch coefficient” table of FIG. 59 for each measure. Such a “music progression-pitch coefficient” table is preferably configured so that an actual change curve is obtained by setting a target value of a pitch and multiplying that value by a normalized pitch coefficient change curve. . It should be noted that this pitch parameter change providing effect is not an effective function in any case, so it must be turned off.

(C) Processing according to the fineness of the performance sound FIG. 60 is a flowchart showing the procedure of the volume parameter change giving process 3. In the volume parameter change giving process 3, the volume is changed according to the fineness of the continuous performance. Make a facial expression.

  In the figure, first, a partial area having a note density of a predetermined value or more, that is, a difficult part is detected from the selected performance data (step S41). The note density means the number of notes per unit time. The partial area is detected, for example, in one bar unit, half bar unit, or in the vicinity of a place where a note having a predetermined note length or less exists.

  For example, in FIG. 3 showing an example of a musical score, if the note density is the number of notes per beat, the predetermined value is two or more eighth notes, and the partial area is one bar unit, in step S41, Two bars are detected as partial areas.

  Next, the pitch change width of the difficult part (detected part) is calculated (step S42). The pitch change width means a change width between the maximum value and the minimum value of the pitch. The reason why the pitch change range of the difficult part is calculated is that the volume during performance naturally changes depending on the pitch change range in addition to whether it is a difficult part.

  Next, a change imparting pattern corresponding to the note density and pitch change width is determined for each difficult part (step S43). As the change imparting pattern, a number of patterns corresponding to the respective values of the note density and the pitch change width are prepared in advance as templates. In step S43, the calculated note density and the pitch change width are calculated from the plurality of templates. In response, one template is selected.

  For example, as shown in FIG. 61, table data indicating the relationship between note density and volume (the volume coefficient described above) is prepared, and table data is searched for the calculated volume density. Thus, the corresponding volume coefficient value may be obtained, and the volume coefficient value may be corrected according to the calculated pitch change range. Furthermore, the volume coefficient may be calculated only by calculation using a predetermined calculation algorithm without using table data.

  Next, the volume parameter is corrected based on the determined change imparting pattern (step S44). Here, if the velocity included in the note event is specifically taken as the volume parameter, the correction of the volume parameter means that each velocity in the performance data belonging to the difficult part is corrected based on the change imparting pattern. It means to do. If the volume coefficient is used as data constituting the change imparting pattern, the volume parameter is corrected by multiplying the original velocity value by the volume coefficient value. It should be noted that the volume tends to decrease in the part where both the note density and the performance difficulty level are high, but if the player knows the tendency, the volume may increase conversely in order to correct the tendency. You may enable it to give the change which considered this. (In the difficult part, the volume is not always reduced.)

  According to one embodiment of the present invention, the method of “volume parameter change giving process 3” described with reference to FIGS. 3, 60 and 61 can be applied to facial expression that gives “pitch instability”. It is. That is, particularly in a stringed instrument or a wind instrument, in response to a performance aspect in which the pitch becomes unstable because it is difficult to play if the note is fine, a portion where the pitch of the note is fine is detected (step S41), and the range of the pitch change. (Step S42), the pitch instability is calculated from the evaluation (step S43), and the pitch parameter is corrected (step S44), thereby increasing the pitch instability in accordance with the performance details. You can make a facial expression.

  For example, in the musical score example shown in FIG. 3, the number of notes per beat is one for the first measure and two for the second measure, but the numerical value is used as a coefficient to change the instability of the pitch. . For this purpose, a “note time interval-pitch instability” table is prepared, and in the musical score example of FIG. 3, a pitch instability coefficient value “a” is obtained at the first measure. At the bar, an instability coefficient value “b” (b> a) of the pitch is obtained. Actually, since the number of notes per time is more important than the number of notes per beat or beat, it is essential to consider the tempo.

  In other words, in order to simulate the fact that the pitch becomes unstable as a result of being difficult to play, and to add a facial expression that enhances the sense of reality, as shown in FIG. The “note time interval-pitch instability” table, which changes so as to increase, is used as a change imparting pattern, and the processing flow of FIG. 60 may be applied to parameters relating to pitch instability. Also, the poorness simulation can be adjusted by using the threshold value of this table as a parameter.

  In applying the processing flow of FIG. 60, the note density in step S41 may be evaluated by counting the number of notes in units of measures and beats, and by evaluating the tempo, thereby counting the number of notes per unit time. The evaluation of the range of pitch change in step S42 is to check the maximum value and the minimum value of the note in key units per unit time. In live performances, the wider the range between the maximum and minimum values, the more intense the movement, so the pitch in units smaller than the key becomes unstable. Therefore, in addition to the evaluation of the density of notes, a natural expression is obtained by increasing the degree of pitch instability as the pitch range is wider.

  In the case of keyboard instruments, although there is a mistouch, the pitch does not basically become unstable, so this system should be applied to stringed instruments (especially those without frets) and wind instruments.

  Further, when the note density is increased, for example, in the case of a plucked string instrument, the repelling is likely to occur. In this case, the same effect as the pull-off performance method and the hammering-on performance method is obtained. Following such a case, automatically inserting a pull-off or hammering-on performance with a certain probability according to the note density contributes to reproduction of a natural performance. Therefore, according to one embodiment of the present invention, a portion having high note density is detected from performance data (MIDI data or the like) (step S41), and the frequency of hammering on (pulling off) in this portion is calculated (step S41). Corresponding to this, hammering on (pulling off) is automatically added to the performance data in response to this (corresponding to steps S43 and S44), so that “hammering on (pulling off) when the note is fine”. The system can be configured.

(D) Process associated with performance of same / similar phrases FIG. 63 is a flowchart showing the procedure of the volume parameter change giving process 4. In the volume parameter change giving process 4, the selected performance data in which similar phrases appear repeatedly is shown. In addition, the expression of changing the volume parameter of the second and subsequent similar phrases in accordance with the similarity and appearance mode is performed. Specifically, the volume parameter is changed according to the following algorithm. That is,
(1) When phrases having high similarity appear continuously, the volume parameter value of the second and subsequent similar phrases is made smaller than that of the first similar phrase. (2) Without similar phrases being consecutive. In the case of repeated appearance, the value of the volume parameter of the second and subsequent similar phrases is changed to a value that is similar to that of the first similar phrase and that corresponds to the similarity.

  In FIG. 63, first, the selected performance data is divided into phrases (step S51). In this embodiment, the phrase unit is, for example, one bar length. However, the phrase unit is not limited to this, and may be a plurality of bar lengths or other unit lengths. Moreover, you may employ | adopt what is called a performance phrase, such as an intro part, a fill-in part, and a main part, as a phrase.

  Next, similar phrases are detected by comparing performance data (specifically, note event, delta time, etc.) included in each of the divided phrases between the phrases (step S52).

  Then, the similarity is calculated for each similar phrase detected in this way (step S53). Here, the similarity is, for example, (1) all the performance data is the same, (2) only a part of the performance data is different, (3) only a part of the performance data is the same, (4 ) Although the performance data are all different, the four-stage value is used, but this stage may be finer or coarser. In step S53, since the similar phrase is detected in the previous step S52, the similarity cannot be (4).

  Next, it is determined whether or not the similar phrases are continuous (step S54). When the similar phrases appear continuously, the similar phrases are applied to the phrase to be given the change and the phrase based on the calculated similarity. A change imparting pattern to be determined is determined (step S55), and the volume parameter of the performance data included in the determined phrase is corrected based on the determined change imparting pattern (step S56).

  In step S55, the phrase that should be given a change refers to the phrase that appears second or later among similar phrases that appear continuously, and the change-given pattern refers to the volume of each piece of performance data included in the phrase that appears first. A volume coefficient value (ratio) pattern for weighting a parameter (specifically, velocity in a note event). The volume coefficient value is changed according to the similarity. For example, when the similarity is (1), the volume coefficient value is “0.8”, and when the similarity is (2), the volume coefficient value is “0.9”. In the case of (3) above, the volume coefficient value is set to “1.0”. The volume coefficient value may be determined in consideration of the appearance order in addition to the similarity. For example, when the similarity is (1) above, the volume coefficient value is set to “0.8” for the second similar phrase that appears, and the volume for the third similar phrase that appears. The coefficient value is set to “0.7”.

  In step S56, the volume coefficient determined in this way is multiplied by each velocity value in the phrase that appears first, and the corresponding velocity value in the second and subsequent similar phrases is replaced with each multiplication result. To correct the volume parameter.

  On the other hand, when the similar phrases do not appear continuously in step S54, the volume changes between the similar phrases are matched at a rate based on the similarity for each phrase (step S57). Specifically, when the similarity between similar phrases is (1) above, the volume parameter (specifically, the velocity in the note event) of each performance data included in the phrase that appears first is 2 If the volume parameter of each piece of performance data included in the phrase that appears after the first is replaced and the similarity between similar phrases is (2) above, for the same portion of the performance data, Replace the volume parameter of the corresponding part in the phrase appearing after the second with the volume parameter of the corresponding part of, and for the part where the performance data is different, adapt it to the replaced part Edit the volume parameter (adjust the velocity value ratio).

  According to one embodiment of the present invention, the method of “volume parameter change applying process 4” described with reference to FIG. 63 can be applied as it is to facial expression for changing the pitch parameter of performance data. That is, since similar phrases have similar pitch expressions, if the same phrase is repeated continuously by applying the processing flow of FIG. 63 to correction of pitch parameters, it is reproduced in another place. In this case, it is better to copy the expression of the first pitch as it is, and if the phrase is “similar” rather than the exact same phrase, it is possible to set a pitch change similar to the expression of the first pitch. good.

FIG. 64 is a flowchart showing the procedure of the volume parameter change giving process 5. In the volume parameter change giving process 5, in the case of performance data of 3 beats, the same note length is used. When the musical notes appear continuously, a facial expression that emphasizes the first beat is performed.

  In the figure, first, a measure having the same note length in all measures in three beats is detected (step S61). The same note length means, for example, a quarter note.

  Next, the number of notes at the head position is detected for each detected bar (step S62). The number of notes at the beginning position means the number of notes belonging to the beginning beat (the number of notes simultaneously generated at the beginning beat timing).

  In this way, the number of notes that are simultaneously generated is detected by simply applying the volume parameter change giving process 5 to performance data including chord events. Since it is detected unconditionally by S61, if many chord events are included in one measure, the volume simply increases. In this case as well, in order to prevent the volume from increasing simply, in step S62. First, for each detected bar, the number of notes constituting a chord event is detected, and in the next step S63, processing for correcting the volume change is added.

  Next, a change value for each detected measure is calculated based on the number of detected notes (step S63). In the present embodiment, the change value indicates a volume coefficient value, and this volume coefficient value indicates a rate of change with respect to the original velocity value. In the present embodiment, the number of notes belonging to the first beat is considered in the calculation of the change value, but the number of notes may not be considered.

  Based on the calculated change value, the volume parameter of the first note in each detected measure is corrected (step S64). In the present embodiment, the volume parameter is the velocity of the note event, and the correction is a correction in the direction of increasing the original velocity value. Note that there may be a plurality of leading notes (note events). In this case, the velocity value of each note is similarly corrected. Of course, a weighting coefficient for weighting each velocity value at a different ratio may be set, and the increased velocity value may be weighted with this weighting coefficient and set as each velocity value.

  Thus, in the volume parameter change applying process 5, only the note of the first beat is emphasized. However, in Wiener Waltz, it is desirable that the first beat is standard, the second beat is strong, the third beat is weak, and such expression is also made by changing the processing of steps S62 to S64 slightly. It becomes possible. Moreover, you may make it change the beat position which should be strong / weak according to a style, the time, and a composer.

  Furthermore, if a tuplet (triplet, quintuplet, etc.) is detected and the volume of the first sound is increased, the tuplet-likeness can be expressed. This can also be realized by slightly changing the volume parameter change applying process 5. Specifically, the process of step S61 is not performed, the tuplet is detected in step S62, and the amount of change in the leading sound of the detected tuplet is determined in step S63 (for example, set in advance). Or the current volume is increased by a predetermined rate), and the volume parameter is changed based on the determined change amount in step S64 to increase the volume of the first sound of the tuplet. be able to.

  Further, for example, a quintuplet may be converted into a simple combination of numbers, such as 2 + 3, and an accent may be added.

  Time signatures that are easy to understand for ordinary people are “2 time signatures”, “3 time signatures”, and “4 time signatures”. With a time signature of 5 time signatures or more, it is easier to grasp the image by dividing it into smaller units. For example, in 5 beats, instead of counting "1, 2, 3, 4, 5", like "1, 2, 1, 2, 3" or "1, 2, 3, 1, 2" It is easy to imagine if it is divided into 2 beats + 3 beats, and at 6 beats, it is 2 beats +2 like “1, 2, 1, 2, 1, 2” or “1, 2, 3, 1, 2, 3”. It is easy to understand that it breaks down like time signature +2 time signature or 3 time signature +3 time signature.

  In order to express this time sensation in a specific sound, for example, in the case of five time signatures, it is assumed that the two time signatures and the three time signatures are alternately arranged, by emphasizing the volume of each first beat. Can be realized. Similarly, when the quintuplet, hexuplet, and more are decomposed into dyads and triplets, and the volume of the first sound of each is emphasized, the time signature becomes easier to understand. However, since the 5 beats are only 5 beats, it is better to distinguish the first note of the original time signature from the first note of the original beat even if it is decomposed to 2 + 3. Should be more emphasized.

(F) Processing corresponding to phrase end feeling FIG. 65 is a flowchart showing the procedure of the volume parameter change giving process 6. In the volume parameter change giving process 6, the expression of suppressing the volume at the end of the phrase is given. Do.

  In the figure, first, the selected performance data is divided into phrases as in step S51 (step S71).

  Next, for each divided phrase, a stop volume is calculated based on the tempo set for the phrase (step S72). Here, the stop volume does not decrease the volume of the note at the end position of the phrase, but gradually decreases the volume after the note is pronounced. Then, the time from sound generation to mute is determined according to the tempo set for the phrase and the note length. Note that in order to perform control to gradually reduce the volume of a note, the note needs to generate a continuous sound. However, if the note generates a decaying sound, the note needs to be generated. The end volume is calculated such that the volume of the note is simply reduced or the volume of some notes in front of the note is gradually reduced. Of course, such a stop sound volume may be calculated even when the note generates a continuous sound.

  Then, for each of the divided phrases, the volume at the end position of the phrase (specifically, the velocity value in the note event at the end position) is corrected based on the calculated end volume (step S73). .

  In this volume parameter change giving process 6, the volume at the end position of the phrase is suppressed in this way. Conversely, the note (note event) at the beginning position of the phrase is detected and the volume of the note (note event) is detected. Increasing the velocity value may be effective. This is especially true when the selected performance data is for a bass rhythm instrument.

(G) Processing for long tone trill / vibrato FIG. 66 is a flowchart showing the procedure of the volume parameter change giving process 7. In the volume parameter change giving process 7, the trill or vibrato is given to the long tone. When there is a facial expression, the volume is increased initially, decreased midway, and increased at the end.

  In the figure, first, a note to which a pitch data string having a trill or vibrato tendency is given and whose note length is longer than a predetermined length (long tone) is detected (step S81). Here, the predetermined length refers to, for example, a half note length. Of course, this is merely an example, and other note lengths may be used.

  Next, an applicable change imparting pattern is determined for each detected note (step S82). FIG. 67 is a diagram showing an example of this change imparting pattern, in which the vertical axis indicates the volume coefficient and the horizontal axis indicates time. In the figure, a solid line indicates a volume change curve, an arrow A indicates a volume change width, an arrow B indicates an overall time of the volume change, an arrow C indicates a target value of the volume change, and an arrow D indicates It shows the time until the volume change drops most. The values of these parameters may be set using words such as “strong”, “weak”, “fast”, and “slow”. Such a change imparting pattern is prepared in advance for each type of trill and vibrato (for each note length, for example), and an optimum change imparting pattern is automatically determined for each detected note. .

  Then, based on the change imparting pattern determined in this way, the volume parameter (velocity) at the time of playing the trill or vibrato is corrected (step S83).

  According to one embodiment of the present invention, the method of the “volume parameter change applying process 7” described with reference to FIGS. 66 and 67 is applicable to change the “pitch change speed”. That is, for example, a change curve pattern similar to the pattern of FIG. 67 as shown in FIG. 68 is prepared as a change imparting pattern of “pitch change speed”, and trill (vibrato) is obtained from performance data (MIDI data, etc.). ) And a long tone portion is detected (step S81), and a change in pitch is calculated from a change curve pattern as shown in FIG. 68 (steps S82 and S83), whereby a slow tone is initially set for a long tone trill (vibrato). It is possible to add facial expressions that give changes.

  Here, the symbol “A” in FIG. 68 represents the width of the speed change, the symbol “B” represents the total time of the trill (vibrato), and the symbol “C” represents the time to the inflection point of the speed change. Represent. In determining the change imparting pattern (step S82), it is preferable that an appropriate change can be selected based on parameters (A to C, etc.) that can change the time and the amount of change of the change curve pattern in FIG. Also, it is preferable that these parameters can be set by words such as “strong”, “weak”, “fast”, “slow”, etc., as already described.

  It should be noted that in the case of trill or vibrato, it becomes more natural that the pitch for each cycle of pitch change is non-uniform, so it is preferable to include a function that generates non-uniformity. In addition, in order to reproduce a soristic (solo performance-like) trill, it is preferable to change the pitch above the trill higher. Therefore, an interface that sets the pitch change higher by setting “Solitistic” is preferable. In order to automatically reproduce a more natural expression, the depth and speed of vibrato should be given a pitch correlation, volume correlation, and duration (pronunciation length) correlation.

(H) Process associated with roll performance of trill and percussion instrument FIG. 69 is a flowchart showing the procedure of the volume parameter change applying process 8. In the volume parameter change applying process 8, a natural expression is applied during the roll performance of the trill and percussion instrument. I do.

  In the figure, first, a partial section corresponding to a roll performance of a trill or percussion instrument is detected in the selected performance data (step S91).

  Next, for each detected section, a change value of the volume of each note in the section is determined (step S92). Here, as the change value, a value that is natural during roll performance of a trill or percussion instrument, specifically, a value at which the volume (velocity value) of each note becomes non-uniform is determined. As a method for determining such a change value, for example, a method of simply generating a random value to make a non-uniform value, a change application pattern (template) in which the flow of the non-uniform value is recorded in advance, The method of determining based on this etc. is mentioned.

FIG. 70 is a flowchart showing the procedure of the volume parameter change giving process 9. In the volume parameter change giving process 9, in the case of performance data containing chords, The expression of making the pronunciation clear for chords.

  In the figure, first, a portion where a plurality of sounds are pronounced simultaneously, that is, a chord portion is detected in the selected performance data (step S101).

  Next, a change value corresponding to the importance of the constituent sound is determined for each detected chord (step S102). Here, the change value is a value (for example, a volume coefficient value) that changes a value of a volume parameter (specifically, velocity in each note event) of each constituent sound (note event) of the chord. In the present embodiment, the corresponding change values are made different according to the importance of the constituent sounds. This is because it may only be noisy if you play each component of the chord at the same volume, so let's play a clean chord with only the main component of the chord set to a higher volume. It is because. As a criterion for determining whether or not it is a major one, that is, a degree of importance, for example, a chord main tone is most important and a subordinate tone is most important next. In addition, detection of importance for each chord is performed using a chord importance template prepared in advance, specifically, for each chord, each component sound of the chord has a priority (for example, the higher the priority, the more important the importance). This is done using the information recorded with the Examples of the method for determining the change value include a method of determining a value determined in advance for each importance level, and a method of increasing or decreasing the current volume value at a rate corresponding to the importance level.

  Then, the volume parameter is corrected based on the change value determined in step S102 (step S103).

  In the volume parameter change applying process 9, the volume of important notes is increased. Conversely, the volume of unimportant notes may be decreased.

  Also, without using the concept of importance, only the lowest and highest volume of the chord may be made larger than the volume of the other component sounds, or only the volume of the root of the chord The volume may be higher than the volume of the constituent sound.

  In addition, although it is not a chord, there is a so-called distributed chord that is often played on a guitar. However, emphasizing the main tone of this distributed chord gives a sense of tonality and a clear performance. The process in this case can also be easily realized by changing only a part of the volume parameter change applying process 9.

  In addition, when playing in octave unison, if all parts are played at the same volume, the timbres may be mixed and unison-likeness may not be expressed. For this reason, when octave unison is detected in the selected performance data, it may be effective to increase or decrease the volume in order from the top octave. This can be realized simply by changing a part of the volume parameter change applying process 9. It should be noted that not only octave unison but also the same octave unison may be adjusted depending on the tone, and the same process may be performed at 3 degrees or 5 degrees or other unison. Also good.

  According to one embodiment of the present invention, the method of “volume parameter change applying process 9” described with reference to FIG. 70 can be applied to the pitch, and the expression of automatically changing the chord to a genuine tone is performed. Can do. One of the pitch characteristics of string instruments and wind instruments without frets is that they can produce genuine chords. Correspondingly, first, the part that is pronounced with chords is searched ( Corresponding to step S101), the type of searched chord is evaluated, and a pitch change in the case of genuine tone is calculated corresponding to the evaluated chord (corresponding to step S102). Here, for calculating the pitch change, it is easy to refer to a pitch table of each chord. Then, based on this change in pitch, the pitch parameter is modified to change the pitch (corresponding to step S103), and a facial expression that produces a genuine tone is performed.

(K) Process Accompanied by Staccato Performance FIG. 71 is a flowchart showing the procedure of the volume parameter change giving process 10. In the volume parameter change giving process 10, a live staccato is shown in the case of performance data including a staccato. Facial expression imitating performance.

  In the figure, first, a note immediately after a note that produces staccato is detected in the selected performance data (step S111).

  Next, the amount of change is determined for each detected note (step S112). In live staccato performance, staccato not only shortens the length of the sound, but also emphasizes the sound. One way to emphasize the staccato sound is to reduce the volume of the note immediately after the note. In step S112, this volume control is performed. Therefore, determining the amount of change here means determining the amount by which the value of the volume parameter (specifically, velocity during the note event) of the detected note (note event) is changed in the decreasing direction. doing.

  Then, the volume parameter of the note detected in step S111 is corrected based on the amount of change determined in step S112 (step S113).

  In general, the degree of change in the staccato expression varies depending on the note value and tempo of the note, so the degree of change in the volume of the note immediately after the note that produces staccato depends on the note value and tempo of the note that produces staccato. It is preferable to adjust. In addition, when detecting the note immediately after the staccato, there is a time corresponding to a rest, and when this is excluded from the detection target, in step S111, “the note immediately after the predetermined time of the note that generates the staccato” May be detected.

(L) Process with Fluctuation / Randomness FIG. 72 is a flowchart showing the procedure of the volume parameter change giving process 11. In the volume parameter change giving process 11, the volume is fluctuated for a long tone. Perform facial expression.

  In the figure, first, in the same manner as in step S81, a note (long tone) having a predetermined pronunciation length or longer is detected (step S121).

  Next, for each detected note, a change amount is determined using a random number counter and a change imparting pattern (step S122). Live performance tends to increase gradually when playing long tones. In step S122, a change amount for simulating this phenomenon is determined. That is, the amplitude of the random number output from the random number counter is the amplitude of the random number generated by the random number counter in accordance with the change imparting pattern prepared in advance, specifically, as shown in FIG. Is changed based on the table data to be changed, and the random number value output from the changed random number counter is determined as the change amount. As a result, the amplitude of the output random number increases with the lapse of time after the long tone starts sounding, so that it is possible to determine the amount of change that gradually increases the fluctuation. If the amplitude of the random number is increased without limit as the counter value increases, the amount of change determined accordingly becomes unnatural. Therefore, as shown in FIG. 73, the amplitude of the random number increases as the counter value increases. It is desirable to converge to a predetermined value as it goes on.

  Then, the value of the volume parameter (specifically, the velocity during the note event) of the note (note event) detected in step S121 is corrected based on the amount of change determined in step S122 (step S123).

  In this volume parameter change giving process 11, a random number counter is used, but a fluctuation table may be used instead. It is also effective to use nonuniform parameter change data collected from live musical instrument performance data as the change imparting pattern.

  In the selected performance data, if there is a long tone in a section where f (forte) is added as a volume symbol, if the long tone continues to be played at the same volume, the presence will gradually disappear and it will not become f. come. Therefore, in such a case, it is effective to gradually increase the volume of the long tone. This volume control can also be realized by changing a part of the volume parameter change applying process 11. Specifically, an f-symbol detection process is added at the beginning of the process of step S121. In the process of step S122, a normal counter is used instead of the random number counter, and the process of step S123 is used as it is. . In the case of wind instruments, when the tone is too long, the volume decreases gradually. In this volume control, the detection process of the f symbol may be deleted, and the normal counter may be changed to a subtraction counter so that the volume gradually decreases.

  According to one embodiment of the present invention, the method of the “volume parameter change applying process 11” described with reference to FIGS. 72 and 73 is also applicable to the pitch, and the pitch varies with respect to the long tone (long note). You can make a facial expression. In human performances such as stringed instruments and wind instruments without frets, it is impossible to perform at a pitch that does not change, and there should be some degree of pitch fluctuation. Therefore, in order to realize such fluctuation, a long tone having a dynamic symbol (f) is detected from the musical score data (step S121), a pitch change is calculated (step S122), and a pitch parameter is changed (step S123). Thus, the “pitch” is changed by a random number as in the case of the volume.

  In this case, it may be more effective to apply a band limitation to the random number by using a band pass filter (BPF) instead of changing the pitch without using a simple random number. Further, the pitch may be changed by a table instead of by a random number. In particular, the effect becomes natural when a table in which the pitch change of a live musical instrument is recorded is used.

  Further, in live performance long tones, there is a tendency for pitch fluctuations to gradually increase with increasing tone length, and as in the case of volume fluctuations, the output of a counter reset with a new note causes the output of FIG. It is preferable to construct a system in which pitch fluctuation gradually increases during a long tone by indicating a random number table having such characteristics. It should be noted that such a pitch fluctuation is more natural not to use this function in a case where there are few elements whose pitches are shifted, such as a keyboard instrument. Therefore, it is necessary to configure the musical instrument type by determining the timbre when analyzing performance data (such as MIDI data) (step S121).

  Furthermore, according to one embodiment of the present invention, in addition to the above-described pitch fluctuation processing, it is possible to perform facial expression that increases the randomness of the pitch. For example, if there is a long tone in a section where the volume symbol is f (forte), etc., if the performance continues at a pitch that does not change, the presence will gradually disappear, and it will become less like f. Therefore, in such a case, it is effective to gradually increase the instability of the pitch. In other words, the long tone gradually increases the randomness of the pitch. In addition, it is preferable that the randomness of the pitch be changed in accordance with the tone color or by the melody and accompaniment (back).

  In order to perform such a process, first, in the dynamic symbol detection step, it is detected whether the dynamic symbol is f (or more than f), a long tone note within the range is detected, and then, Evaluate the length of the note and add a facial expression that gradually increases the instability of the pitch. Here, the length of the note is evaluated because it is desired to change the instability of the pitch effectively within the time of the note, that is, the instability of the pitch is not increased without limit. Therefore, it is necessary to adjust the curve of the change in pitch instability according to the length of the note.

  On the contrary, even when the strength symbol is p (piano), such an unstable phenomenon can occur. Therefore, in the above-described strength symbol detection step, a configuration in which the strength symbol is not moderate is detected. It is also good to give pitch instability. In addition, a method of ignoring any strong or weak symbol may be used.

  Some timbres are unnatural when the randomness is large, so it is necessary to control the degree of randomness according to the timbre. In particular, a keyboard instrument such as a piano is unnatural if there is randomness, and a clarinet or the like is unnatural even in a wind instrument. Furthermore, if the accompaniment part is given randomness more than necessary, it becomes a noisy accompaniment, so it is better to adjust the randomness between the melody part and the accompaniment part.

(M) Parameter Processing According to Sound Range FIG. 74 is a flowchart showing the procedure of the volume parameter change giving process 12. In the volume parameter change giving process 12, expression is given by volume expression according to the pitch average.

  In the figure, first, the selected performance data is divided into phrases as in step S51 (step S131).

  Next, an average value of pitches is calculated for each divided phrase, and a change imparting pattern is determined based on the average value (step S132). The change imparting pattern is stored in advance in the form of table data for each average value of pitches, and one of them is determined in step S132.

  Then, the volume parameter (specifically, expression data) is corrected for each divided phrase based on the determined change imparting pattern (step S133).

  According to one embodiment of the present invention, the technique of the “volume parameter change applying process 12” described with reference to FIG. 74 can be applied to a pitch, and can be expressed by changing the pitch expression depending on the pitch range. A pitch coefficient (in a unit smaller than the key) is determined based on the pitch information, and a pitch change amount is determined by drawing a certain table based on the pitch coefficient. However, adding such an expression to each sound is the same as the conventional key scale, but for each specified section (phrase, measure, beat, etc.), processing is performed from the average of the pitch. It is also effective to decide.

(N) Process Accompanied with Part Performance FIG. 75 is a flowchart showing the procedure of the volume parameter change giving process 13. In the volume parameter change giving process 13, the volume of each part is reduced during part division. Perform facial expression.

  Prior to explaining the expression processing, the gist of the present invention will be described. In the orchestra score, there is a designation of “Division” in which one string part is further divided into two or more parts. Assuming that the first violin part is played by 10 people, when the performance is divided into two parts according to the division symbol, it is generally divided into 5 parts each. Therefore, simply thinking, the volume of the Division part should be halved from 10 to 5 people. However, in the conventional DTM (desktop music), even if the division is designated by the orchestra score, it is common to make a song that allows polyphonic performance with the same sound source part. Then, the volume of each note is constant, and the number of notes that play simultaneously doubles, resulting in an increase in volume. Skilled users have noticed this point and have manually set the volume to decrease when Division is specified. The volume parameter change giving process 13 is intended to automatically perform this.

  In FIG. 75, first, the parts of the same tone color and the number of parts are detected from the selected performance data (step S141). Here, the same tone color may include a similar tone color. If detection is performed only under the condition of the same tone color, the first violin and the second violin may be handled as divisions. In such a case, the part may be selected manually instead of automatic detection.

  Next, the performance volume value of each part is calculated based on the detected number of parts (step S142). Specifically, the maximum volume value of multiple parts played in the same tone (same division) is detected, and the value calculated by dividing the maximum volume value by the number of parts is used as the performance volume value of each part. decide. The performance volume is a value for adjusting the overall volume of each part performance (specifically, volume data or expression data for each part).

  Then, the calculated volume value is recorded in each part (step S143). Specifically, the calculated volume value is recorded as the initial volume of each part.

  It should be noted that the volume control is preferably performed by the above method for each part (performance section) in which a plurality of parts are reproduced in parallel. In that case, the calculated volume value is validated only in the corresponding performance section.

(P) Process linked to timbre change FIG. 76 is a flowchart showing the procedure of the volume parameter change applying process 14. In this volume parameter change applying process 14, facial expression is expressed by volume expression corresponding to the timbre change.

  In the figure, first, a timbre change position and change contents are detected (step S151). That is, the position where the timbre is changed in the analysis target part in the selected performance data and the content of the timbre change (the timbre before the change and the timbre after the change) are detected for each change position.

  Next, the amount of change in volume corresponding to the change content is determined (step S152). Specifically, an optimal volume value for each tone is determined in advance in the form of table data, for example, and the respective volume values of the tone before the change and the tone after the change are read from the table data, and based on the volume Thus, the sound volume parameter (expression data) is calculated and determined so that the sound volume is smoothly changed at the timbre change position.

  Then, the volume parameter is corrected based on the amount of change determined in step S152 (step S153). Specifically, the calculated volume parameter is recorded for each change position.

  In this way, the volume parameter change applying process 14 can make a performance with a specific tone color stand out.

(Q) Process Associated with Pitch Bend Change FIG. 77 is a flowchart showing the procedure of the volume parameter change giving process 15. In the volume parameter change giving process 15, an expression is given that the volume changes during pitch bend.

  In the figure, first, a pitch change position by pitch bend is detected (step S161). Specifically, the position where the pitch is changed in the analysis target part in the selected performance data and the content of the pitch change (pitch bend data) are detected for each changed position.

  Next, the change tendency of the pitch bend is calculated for each detected pitch change position, and the amount of change in volume is determined according to the calculation result (step S162). Here, the change tendency of the pitch bend is detected by performing high-pass filter (HPF) processing on the detected pitch bend data. If necessary, the processing result is subjected to a low-pass filter (LPF) process to smooth the change. Then, by using the absolute value of the pitch bend data subjected to the filtering process, the amount of change (decrease rate) of the volume parameter (the velocity of the note event at the pitch change position) is calculated. That is, as the rate of change in pitch bend is larger, the amount of change that lowers the volume is calculated and determined. Thereby, a natural pitch bend feeling can be expressed.

  Then, the volume parameter is corrected according to the amount of change determined in step S162 (step S163). The volume parameter to be corrected is specifically the velocity of the note event at the pitch change position as described above, but is not limited to this, and the performance data to be detected in step S161 is the first string in the stringed instrument. Expression data may be used as the volume parameter.

(R) Processing according to part FIG. 78 is a flowchart showing the procedure of the volume parameter change giving process 16. In the volume parameter change giving process 16, the melody line is raised in the case of performance data consisting of multiparts. Make a facial expression.

  In the figure, first, a melody part is detected (step S171).

  Next, the volume of the melody part is compared with the volume of the other part (accompaniment part), and the amount of change is determined so that the volume of the melody part is relatively higher than the volume of the other part (step S172).

  Then, the volume parameter (expression data) of the melody part is corrected based on the determined change amount (step S173).

  If the volume of the melody part cannot be increased beyond the output range of the tone generator circuit 7 even if it is attempted to increase the volume of the melody part, the volume of the other parts is reduced uniformly to relatively increase the melody part. You can increase the volume.

  Also, in order to make the melody line stand out for performance data that is in charge of melody to accompaniment with a single tone like a piano, the top note included in the performance data is extracted and used as a melody part. Then, the melody line and the accompaniment are separated by using the rest as the accompaniment part, and thereafter, the processes of steps S172 and S173 of the volume parameter change applying process 16 may be performed.

  According to one embodiment of the present invention, the technique of “volume parameter change giving process 16” described with reference to FIG. 78 can be applied to the pitch, and the basic pitch is changed according to the melody part or accompaniment part. It can be performed. In other words, one way to make the melody stand out in the multipart is to bring the perspective of the melody part closer and away from the accompaniment part, but detect the melody part from the performance data (MIDI data, etc.) S171) By calculating the pitch change (step S172) and applying the pitch change (step S173), this method is realized by the pitch change.

  Here, when the pitch change is realized by PitchBendEvent, even if it is attempted to increase the pitch of the melody part, it cannot be increased if the maximum value of PitchBend is exceeded. In that case, it is preferable that the pitch of the melody is relatively increased by uniformly decreasing the pitch bend of all other parts.

  Also, for a score that is in charge of melody and accompaniment with a single tone like a piano score, when the score is input, the top note is extracted as a melody part and the rest as an accompaniment part. After separation, when the pitch of the melody part is slightly increased and the melody is emphasized and played, an automatic arrangement performance device in which the melody part is raised can be configured.

  Furthermore, according to one embodiment of the present invention, the method of “volume parameter change applying process 16” described with reference to FIG. 78 can be applied not only to the volume and pitch but also to other performance parameters. It is possible to add a facial expression by changing the performance parameters according to the melody part and the accompaniment part. In other words, as one method of raising the melody in the multipart, the melody part is detected from the performance data (MIDI data, etc.) as compared with the method in which the perspective of the melody part is brought closer and the perspective of the accompaniment part is kept away (step) S171) This can be realized by calculating a sense of perspective (step S172) and assigning a change in performance parameter accordingly (step S173).

  Here, various perspective setting methods other than volume and pitch are conceivable. For example, there is a method of changing the depth of reverb. The deeper the reverb, the farther it feels. Moreover, if LPF and HPF can be set, if such a filter is used to emphasize the low range, a close feeling can be obtained, and conversely, if the low range is weakened, the feeling becomes far. This can be similarly realized by using an equalizer (equalizer) instead of the LPF or HPF. Moreover, the effect becomes even greater when set in combination with reverb or the like.

(S) Parameter Processing According to Fingering FIG. 79 is a flowchart showing the procedure of the volume parameter change giving process 17. In the volume parameter change giving process 17, a live performance is reproduced by the volume change according to the fingering. Make a facial expression.

  In the figure, first, in the selected performance data, a pitch that is considered difficult to perform is detected (step S181). Specifically, the pitch that is considered difficult to perform refers to a pitch that is played with a little finger in, for example, piano performance or guitar performance, or a pitch that corresponds to a particularly fast arpeggio in piano performance. Of course, other criteria may be used.

  Next, a change value is determined such that the detected volume of the pitch is relatively smaller than the volume of the other pitches (step S182).

  Then, the volume parameter (specifically, the velocity in the note event corresponding to the detected pitch) is corrected according to the determined change value (step S183).

  In one embodiment of the present invention, the volume parameter change applying process 17 can be given a facial expression that changes the randomness of the pitch according to the finger, considering fingering. For example, when a live performance of a stringed instrument without frets is imitated, there are sounds that tend to be higher and lower depending on the fingering, and this is particularly likely to occur during fast fingering. Therefore, a more vivid performance can be reproduced by giving a change in pitch in consideration of fingering. In addition, even woodwind instruments have sounds that tend to be higher and lower, depending on the combination of holes to be pressed. Since brass instruments are the same depending on the position of the length of the tube, such expression is not limited to stringed instruments.

  In order to change the randomness of the pitch in consideration of such fingering, the fingering is judged from the performance data (MIDI data or the like) (corresponding to step S181), and the pitch change corresponding to this fingering is calculated (step (Corresponding to S182), by assigning a pitch parameter based on the calculated pitch change, a process of automatically changing the pitch according to the fingering is performed.

  For example, in the case of a cello, if fingering called the first position is performed on the score as shown in FIG. 80, “mi” → 1 (index finger), “fa” → 2 (middle finger), “so ”→ 4 (little finger). In this case, due to the structure of the human finger, the index finger, middle finger, ring finger, and little finger do not open at equal intervals, and the space between the middle finger and ring finger is narrower than the others. As a result, the musical score of FIG. 80 waits for a tendency that the pitch of “Fa” tends to shift to a higher level. A skilled performer knows such a phenomenon and can correct it by training. However, this pitch shift becomes significant depending on the skill level. On the contrary, in the case of the ring finger, the pitch tends to decrease. Therefore, by registering such a tendency, it is possible to reproduce the poor performance of the player.

  Since this process is a poor reproduction, if the degree is too strong, a favorable result cannot be obtained. However, since moderate skill is expressed as a natural performance form, it is preferable that this skill can be adjusted by parameters.

  In addition, when the fingering position moves, a sound with a pitch change corresponding to the position movement may be added, so after judging the fingering, evaluate the magnitude of this fingering position movement and portamento (Slide) is added (corresponding to step S182), and a pitch parameter based on this is added, and a process of automatically changing the pitch according to the evaluation of the position movement is preferably performed. When generating a pitch change according to the amount of position movement by evaluating the position movement in this way, for example, in the case of a stringed instrument without frets, it is preferable to add a continuous pitch change in a portamento manner. In the case of a stringed instrument having a fret, it is preferable to add a pitch change in a stepwise manner corresponding to the pitch of the fret.

  In addition, when it is evaluated as an open string by fingering judgment, a more natural expression can be reproduced by prohibiting vibrato. In the case of wind instruments, smooth slur performance becomes difficult when the register tube is opened and closed by fingering. Therefore, if the smooth slur is permitted or not permitted by evaluation of fingering, a realistic performance can be reproduced.

  Furthermore, according to one embodiment of the present invention, the method of the volume parameter change giving process 17 is applied to consider fingering, and when the velocity is large at a low position (Velocity), You can add a “chopper” expression. For example, fingering is determined from performance data (MIDI data or the like) in the first step (corresponding to step S181), and in the second step, a large velocity is detected at a low position and noise corresponding to this is calculated (step (Corresponding to S182), by adding this to the performance data in the third step (corresponding to step S183), a system for automatically adding noise according to fingering can be configured.

  If you follow the live performance of a stringed instrument, regardless of the presence or absence of frets, it will be easier to hear the string hitting the fingerboard if you repel it at a low position. By adding noise in consideration of such fingering, a more vivid performance can be reproduced. Also, even with woodwind instruments, there are sounds that tend to be higher and lower depending on the combination of holes to be pressed, and with brass instruments, the same is true depending on the position of the tube length. It is not limited to stringed instruments.

  When the tone is a bass guitar, it is also effective to switch the tone to a chopper bass instead of adding noise. In this case, a fingering response automatic tone color changing system that temporarily changes the tone color in response to detection of a low position and a large velocity in the second step (corresponding to step S182) may be configured. In addition, when it is evaluated as an open string by fingering judgment, it is also effective to change to a tone for an open string.

(T) Processing Accompanying Return of Bow FIG. 81 is a flowchart showing the procedure of the volume parameter change giving process 18. In the volume parameter change giving process 18, in the case of performance data of a stringed instrument played using a bow, The expression of the bow is reproduced by changing the volume according to the bow of the stringed instrument. Thus, in the present volume parameter change applying process 18, in order to give a volume change according to the bow return, the selected performance data must be for a stringed instrument played using a bow. Therefore, before entering the volume parameter change giving process, it is checked whether or not the selected performance data is of this type, and only when the selected performance data is of this type, the process proceeds to the following process.

  In the figure, first, the moving speed of the bow is calculated from the original volume value (velocity value), and the return position of the bow is detected using the calculated value (step S191). Specifically, a bow trajectory is calculated based on the calculated moving speed, and the bow return position is detected by comparing the trajectory with a parameter indicating the bow length, and this time is obtained. Actually, it is necessary to consider the pressure of the bow, but in this embodiment, it is omitted for simplification. Of course, this bow pressure may be taken into account.

  Next, for each detection position, performance data in the vicinity of the bow folding position is read, and the amount of change is determined using the change imparting pattern (step S192). FIG. 82 is a diagram showing an example of this change imparting pattern (table data), where the vertical axis indicates the volume change value and the horizontal axis indicates time. As shown in the figure, the curve indicating the volume change value has the minimum volume change value at the bow folding position (center position in the figure). In the case of a live performance, there is a case where it is better to make the bow not noticeable. Therefore, when using the table data of FIG. 82, the vertical axis is scaled to express the expression. It is desirable to adjust the effect.

  Then, the value of the volume parameter (velocity) is corrected according to the determined change amount (step S193).

  Note that if the bow speed is simply calculated from the volume, bow folding may occur at an undesirable time. Therefore, it is preferable that the bow speed can be adjusted independently of the volume.

(U) Parameter Processing Associated with Corresponding Lyric Information FIG. 83 is a flowchart showing the procedure of the volume parameter change giving process 19. In this volume parameter change giving process 19, the contents of the lyrics are changed in the case of performance data with lyrics. Appropriate expression to change the intonation accordingly.

  In the figure, first, it is determined whether or not there is lyric information in the analysis performance data (selected performance data) (step S201). If there is no lyric information, the volume parameter change giving process 19 is immediately terminated. When there is lyric information, the process proceeds to the next step S202.

  In step S202, the word which should give a volume change is detected from the word in lyrics information. As this detection method, for example, a list of words to which volume change is preferably given is created in advance, and a word to which volume change should be given is detected by comparing the list with words in the lyrics information. Is mentioned.

  Next, a volume change pattern to be used for each detected word is read and a volume change value is determined (step S203). Specifically, a volume change pattern (table) describing what volume change should be applied to each word in the list is prepared, and in step S202, the volume change pattern is read and the volume change is read out. Determine the value. In the volume change pattern, for example, a volume change curve while a word is pronounced is recorded, and the time axis of this curve (the vertical axis represents the volume change value and the horizontal axis represents time) is expanded and contracted. By doing so, the change value of the volume is determined. It should be noted that in determining the volume change value, it is desirable to consider the original volume of the note corresponding to the word position.

  Then, the volume parameter (expression data) is corrected according to the determined change value (step S204).

  According to one embodiment of the present invention, the volume parameter change adding process 19 described with reference to FIG. 83 is applied to adding a change in pitch, and when a poem is input in a composition system, word inflection data, syllables are input. From the data, it can be configured to change the pitch slightly. In the case of a song with a song, even if the melody is the same, it may be more natural to change the inflection depending on the contents of the lyrics, and it is also a good method to realize the inflection by a subtle change in pitch. Therefore, a certain word is registered together with the pitch coefficient in syllable units, and the pitch is changed (subtly) when the word appears in the performance data. Here, the pitch change according to the word means a pitch change within the time of the note corresponding to the syllable of the word, instead of changing the pitch in units of words.

(V) Various Embodiments Note that if the selected performance data is subjected to an expression process accompanied by various volume changes by the above-described volume parameter change applying processes 1 to 19, a very large volume is produced over the entire selected performance data. There may be a case where the sound volume is applied or, on the contrary, a very small sound volume is applied. In order to prevent this, an average value of the volume of the entire selected performance data is calculated, and an offset is added to the volume of the entire selected performance data so that the average value becomes a desired value. . At that time, a certain volume may exceed a maximum value or a minimum value that the sound source circuit 7 can output, so when calculating the average value of the volume of the entire selected performance data, Find the maximum and minimum values. When adding an offset, there is no problem if the result of adding the offset to the maximum value does not exceed the maximum value of the tone generator circuit 7. When exceeding, the maximum value not exceeding is added to the whole as an offset value. Just do it. In addition, when the maximum value exceeds the maximum value of the sound source circuit 7 only for a moment, it may be preferable to clip only that value with the maximum value of the sound source circuit 7, so that the time exceeding the maximum value is An allowable time (threshold value) for what percentage of the time when reproducing the entire selected performance data is allowed may be determined in advance. The same applies to the minimum value.

  Further, the type of expression applicable to the selected performance data may be defined according to the type of musical instrument assumed for the selected performance data. For example, in a piano tone, it is unrealistic to crescendo the note after it is once pronounced. If you are aiming for the most unrealistic effect, you may do it. However, when simulating a piano sound, it is preferable to prohibit changing the volume of a certain note after its pronunciation. For this purpose, a volume control flag may be provided for each tone color. With this flag, for example, in the case of the above-mentioned decay sound, the volume control after sounding is prohibited, and in the case of a continuous sound, the sound volume control after sounding is permitted. decide. FIG. 84 is a diagram showing an example of table data for defining volume control after sound generation, and a value read from this table data is set in the volume control flag. Furthermore, using this table data, if such data has already been set for a tone whose volume control is not permitted after sound generation, it is included in the range from the note-on to the note-off. It is also possible to provide a cleaning function for erasing all sound volume control data to automatically eliminate unnaturalness.

  Further, in the present embodiment (volume parameter change providing processes 1 to 19), the present invention is applied only to the control of the volume parameter. However, the present invention is not limited to this, and the present invention includes parameters other than the volume parameter, For example, it is also effective for controlling parameters relating to pitch, effect (reverb, chorus, panning, etc.), timbre, etc.

  Further, when the orchestra score is used as performance data and is reproduced by the tone generator circuit 7, if the number of parts becomes large, it may not be very effective to perform individual volume control for each part. In such a case, conventionally, volume control data (particularly, expression data and pitch bend data) created for one part is copied and applied to other parts. However, this causes a waste of duplicating the same data. Therefore, it is better to group a plurality of parts and control the plurality of parts with a single volume control data.

[Embodiment 3]
Further examples (a) to (e) according to the present invention will be described below as "Embodiment 3".

(A) “Accent sounds subtly make sounds”:
In order to express an accent, the main focus is on enhancing the volume, but if a facial expression with a change in pitch is made following the live performance expression, a more natural expression is realized. can do. In addition, not only accents but also performance expressions that are generally said to change the volume often actually involve a change in pitch. For this purpose, it is desirable to make the performance symbol correspond to the volume change process and at the same time correspond to the pitch change process. Therefore, in one embodiment of the present invention, when an accent is detected from note data, temporal changes in volume and pitch are calculated, and expression is applied based on these changes.

  FIG. 85 is a flowchart showing an example of processing for slightly swaying an accented sound according to one embodiment of the present invention. First, in step Kk1, a section to which volume change is to be given is specified from the selected performance data, and in step Kk2, a portion to which volume change (accent) is given is detected from the specified section. As shown at 86, the time change of the volume and pitch of this part is calculated. In addition, it is good also considering the performance data whole as an analysis area, without designating an area in this way.

  Further, in step Kk3, a pitch change tendency corresponding to the volume change is determined for each detected portion. The determination of the pitch change tendency includes a method of determining by calculation using the volume change tendency, and a method of reading and determining a change change pattern of a pitch change prepared in advance corresponding to the volume change tendency. In step Kk4, the pitch parameter of the detected portion is changed based on the determined content, and the process proceeds to step Kk5. When it is determined that an end operation has been performed, the process ends. Otherwise, step Kk2 Returning to step S3, the processes in steps Kk2 to Kk5 are repeated.

  It should be noted that it is more natural not to use this function for processing such accented sounds for those that have few elements that are out of pitch, such as keyboard instruments. It is also effective to automatically perform an accent-like process on the note at the beginning of the time signature and the note at the beginning of the tuplet, not just the accent. Furthermore, for large numbers such as quintuples and quintuplets, they are broken down into small tuplets such as doublet + triplet or doublet + doublet + triplet. Thus, it is effective to apply normal accent processing to the original head notes and to apply somewhat weak accent processing to the decomposed head notes, since the beat feels easily.

  When changing the pitch, the volume is also changed at the same time. In vibrato and portamento, when there is movement in the pitch, lowering the volume slightly allows a more natural expression. In order to achieve this, processing is performed to search for expression symbols that change the pitch, such as vibrato, from the note data, calculate the magnitude of the pitch change of the note with the searched expression, and reduce the volume based on this. You can do it. Here, it is preferable to use extraction by HPF for the calculation of the pitch change. However, when the change of the output of the HPF is too severe, it is desirable to arrange the LPF after the HPF.

(B) “With double choking, do not make pitch changes in parallel / divide channels arbitrarily”:
When playing two guitar strings at the same time and performing choking at the same time, it is impossible for a live musical instrument to change the interval between two pitches in parallel. In one embodiment of the present invention, in accordance with such a fact, a natural performance expression can be reproduced by intentionally shifting the timing of the pitch change during double choking. For example, when there is double choking, the time change timing of the volume is shifted between the upper and lower sounds of double choking as shown in FIG. As described above, the “method of shifting the timing” may be the same time change but the start time of the change may be shifted, or the shape of the time change itself may be different as shown in FIG.

  FIG. 88 is a flowchart showing an example of processing that is not parallel in double choking according to an embodiment of the present invention. When this process starts, a double choking part is detected from the selected performance data in step Ll1. Here, when a plurality of double choking parts are detected, a facial expression imparting process in step L12 and subsequent steps is performed for each detected part. Moreover, you may enable it to select the part which wants to perform a facial expression provision process from the detected several part.

  In the next step L12, first, in order to give different pitch (or volume) changes to the upper and lower two sounds in double choking, the upper and lower sounds are separated and stored in separate parts, and then each part is stored. Determine the trend of changing the pitch (volume). As for this change tendency, the time change timing is shifted as shown in FIG. 87, but the change tendency as shown in FIG. 87 may be stored in advance as a change tendency pattern, or may be obtained by calculation. Good. When the change tendency pattern is determined in this way, in step Ll3, the pitch (volume) parameter of the double choking part is changed based on the determined change tendency, and this process ends.

  Note that setting a parameter called “string properties” and associating this parameter with a choking curve makes it practical. In addition, it is more natural to shift the target pitch itself from the original pitch difference. That is, in the example of FIG. 87, the pitch interval between the upper tone and the lower tone is 5 degrees, but will deviate from 5 degrees while the pitch is changing, but even when the pitch change ends. If the state is shifted from 5 degrees, a more natural feeling can be obtained.

  Further, in order to automatically realize the above-described “method for shifting the timing”, it is necessary to independently change the pitch corresponding to the two strings, so it is necessary to automatically separate the two MIDI tracks. is there. For this reason, in one embodiment of the present invention, the channel is divided arbitrarily for double choking. That is, when double choking is detected from performance data (MIDI data or the like), performance data of a plurality of parts with expression is generated, and the plurality of double choking is automatically allocated to the plurality of parts (see step L12).

(C) “Sequential choking makes pitch change non-uniform”:
For example, in the example of continuous choking shown in FIG. 89, choking should have been repeated four times, and the same pitch change should be performed. However, following the actual performance, the pitch change was not necessarily the same. Rather, there are many cases where the pitch is slightly different each time, which sounds more natural. Therefore, in one embodiment of the present invention, in accordance with such a fact, a portion where choking designations appear continuously is searched, and the pitch variation is not uniform so that the pitch variation of each choking does not become the same. By adding, you can reproduce natural performance expression. In the example of FIG. 89, such expression is realized by slightly increasing or decreasing the pitch of the centered chorded sound “Le”.

  FIG. 90 is a flowchart showing an example of processing for making pitch changes non-uniform in continuous choking according to an embodiment of the present invention. In this process, first, in step Mm1, a portion where choking is continuously performed from the selected performance data is detected. Here, when a plurality of continuous choking portions are detected, a facial expression imparting process of step Mm2 or less is performed for each detected portion. Moreover, you may enable it to select the part which wants to perform a facial expression provision process from the detected several part.

  In the next step Mm2, the change tendency of the pitch to be applied is determined by selecting a change applying template or performing a predetermined calculation based on the number of times choking is continued and the change tendency of the pitch. In step Mm3, the detected pitch parameter of the choking continuation part is changed based on the determined change tendency, and this process is terminated.

(D) “In Arpeggio, overtones are played in a separate channel”:
For example, if an arpeggio with a sound shape as shown in the example of the arpeggio shown in FIG. 91 (1) is played, in the case of a live musical instrument, “ The phenomenon that the sound of "Seo" resonates all the time. Therefore, in one embodiment of the present invention, a common arpeggio overtone is detected from performance data (MIDI data, etc.), and this overtone is assigned to another sound source part and played at a low volume in another part. Thus, it is possible to simulate a phenomenon in which overtones continue to sound.

  FIG. 92 is a flowchart showing an example of processing for producing overtones in another channel in the arpeggio according to one embodiment of the present invention. In this process, first, in step Nn1, an arpeggio part is detected from the selected performance data. Here, when a plurality of arpeggio parts are detected, a facial expression imparting process of step Nn2 or less is performed for each detected part. Moreover, you may enable it to select the part which wants to perform a facial expression provision process from the detected several part. After the common overtone of each sound is extracted in step Nn2 following step Nn1, the extracted overtone is stored in step Nn3 so as to be sounded as a new part of the selected performance data, and this process is terminated.

(E) “Select a timbre by the symbol of the score”:
Considering the correspondence between the symbols of the musical score and the timbre, for example, in the musical score of a violinstrument such as a violin, arco (alco performance method of rubbing a string with a bow) and pizz. In the case of a switching score example of (pizzicato: playing a pizzicato that repels a string), the first bar is instructed to perform by bowing (stringing), and the second bar is a performance by plucking a string (plucking a string) (pizzicato string) Then, the first measure is again instructed to perform with bowing. Here, when “pizz.” Is displayed on the score, it is convenient to automatically change the tone to the pizzicato string, and when “arco” is displayed, the tone is returned to the string tone again. It is convenient to do so. For example, in the timbre of the current GM system (MIDI sound source standardization standard), there is only one Pizzicato timbre, and many types of bowed timbre are prepared. Therefore, when changing to a pizzicato string, a means for storing which timbre will be restored when “arco” is displayed should be prepared.

  Therefore, according to one embodiment of the present invention, data indicating the Pizzicato method ("pizz." Symbol) is searched from performance data (MIDI data or the like), and when this data is detected, the current tone is retained. By setting the timbre for the pizzicato and expressing it, the timbre can be automatically changed according to the pizzicato symbol. When the pizzicato tone is not limited to 1 as in the GM system, it is preferable to register a tone corresponding to the pizzicato symbol separately.

  Also, the correspondence between the score symbol and the tone is not limited to the above-mentioned “pizz.”, And this mechanism is also used when the base tone is temporarily changed to the chopper base tone. Can be used as is. In addition, there are various modes, such as a case where the string spurt is temporarily made a solo tone, and a case where a special performance technique such as Corleño is used. Furthermore, as a similar function, it is also effective to automatically set the piano pedal symbol corresponding to the damper on of the control change. Accordingly, in one embodiment of the present invention, a system for selecting a predetermined tone corresponding to a symbol of a musical score is provided.

  FIG. 94 is a flowchart showing an example of processing for selecting a timbre in response to a musical score symbol according to one embodiment of the present invention. When this process starts, first, in step Pp1, a predetermined musical symbol (data corresponding thereto) A instructing a change in tone color is detected from the selected performance data. Here, when a plurality of parts are detected, a facial expression imparting process in step Pp2 and subsequent steps is performed for each detected part. Moreover, you may enable it to select the part which wants to perform a facial expression provision process from the detected several part.

  In the next step Pp2, the music symbol B for returning the timbre is detected corresponding to each of the music symbols for which the timbre change instruction is detected in step Pp1. Then, in the following step Pp3, corresponding timbre change data is inserted in accordance with the positions of the music symbols A and B detected in steps Pp1 and Pp2. For example, a timbre change event (program change data and bank select data) representing the timbre after the change from the position of the music symbol A instructing the change to the position of the music symbol B to which the timbre is restored correspondingly. Is inserted, and after the position of the music symbol B, the restored tone is inserted. Here, the timbre after the change uses a predetermined timbre for each musical symbol, and the restored timbre retains the timbre before the change and inserts it.

(F) “In the piano, the location of the sustain pedal is calculated without permission”:
In a piano score, if a pedal symbol is attached, the dumbbell control by the control changer may be performed at the position of the symbol, but there is also musical score data in which the pedal symbol is omitted. Therefore, according to one embodiment of the present invention, in order to deal with such a score, the range of expression can be expanded by interpreting the phrase and automatically calculating the damper control position. For example, by interpreting a phrase from performance data (MIDI data, etc.) and detecting a phrase break, the damper is turned on (sustain pedal on) at the start position of the phrase, and the damper is turned off (sustain pedal off) at the phrase end position. Damper control is automatically performed on performance data.

(G) “When a score is entered, an instrument is automatically assigned”:
In one embodiment of the present invention, the score editing apparatus is configured to automatically set the tone color of a sound source by only setting the number of score scores and comparing it with a typical score score. In addition, after assigning a musical instrument tone automatically according to score score input in this way, when a note is further input, the melody is surely compared with the learning result whether the melody seems to be the set tone or not. Configured to evaluate

[Various Embodiments]
A program in which a storage medium storing software program codes for realizing the functions of the above-described embodiments is supplied to a system or apparatus, and a computer (or CPU 1 or MPU) of the system or apparatus is stored in the storage medium. It goes without saying that the object of the present invention can also be achieved by reading and executing the code.

  In this case, the program code itself read from the storage medium realizes the novel function of the present invention, and the storage medium storing the program code constitutes the present invention.

  As the storage medium for supplying the program code, for example, the flexible disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, magnetic tape, nonvolatile memory card, ROM 2 or the like can be used. . Further, the program code may be supplied from the server computer 20 via another MIDI device 17 or the communication network 19.

  Further, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also the OS running on the computer based on the instruction of the program code performs the actual processing. It goes without saying that a case where the functions of the above-described embodiment are realized by performing part or all of the above and the processing thereof is included.

  Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. It goes without saying that the case where the CPU 1 or the like provided in the board or the function expansion unit performs part or all of the actual processing and the functions of the above-described embodiments are realized by the processing.

[Summary of Features of Examples]
The present invention has been described with reference to various embodiments. The basic features of these embodiments are summarized as follows:
A correspondence relationship between predetermined feature information and musical tone control information for facial expression is set as a rule in the facial expression module (facial expression algorithm), and generation method information representing the facial expression rule is stored in the storage means in advance. When the characteristic information of the supplied performance data is acquired, the musical tone control information (time parameter, pitch parameter, volume parameter, etc.) according to the expression module (expression expression algorithm) based on the generation method information corresponding to the acquired characteristic information Various performance parameters) can be added to the performance data, and according to the acquired feature information, various expressions can be added to the song with simple operations even if the user is a beginner. Musical performance data can be created automatically. Furthermore, the performance data output with the musical tone control information added is evaluated, and the musical tone control information is adjusted according to the evaluation result, so that the expression with the optimum musical tone control information can be performed.

  Also, the performance data features note density and two-tone intervals as note time information, progress of performance, long tone trill, pitch bend, vibrato, etc. as minute vibration information, phrase delimiters (phrase end) , Pitch information, pitch change direction change information (pitch up / down change part), continuous or similar phrases of the same or similar pattern, registered tone form (phrase template), volume information, “tension” as atmosphere information , Note group sequence information (grouped notes, long tuplet), chord number information, timbre information, finger as fingering information, position movement, position, guitar pulling off, hammering on, piano as performance information There are trills, lyric information, performance symbols such as dynamics and staccato, etc. According to the characteristics of people, it is possible to obtain a performance output with versatile and diverse expressions.

  Further, if the relationship between the predetermined characteristic information of the performance data already supplied and the musical tone control information is stored and the characteristics of the newly supplied performance data are extracted, the musical tone is extracted according to the relationship stored in the storage means. By generating control information and adding it to newly supplied performance data, it is possible to add a facial expression for setting the tempo using the learning function. In addition, the relationship between predetermined characteristic information and musical tone control information regarding performance data is stored in a library, and when characteristic information of the supplied performance data is extracted, musical tone control information is generated by referring to the library, By adding it to the performance data, it is possible to add a facial expression for setting the tempo using the library.

  Further, musical tone control information is generated based on predetermined characteristic information of the supplied performance data, the generated musical tone control information and the musical tone control information of the supplied performance data are compared with the entire performance data, and based on this comparison result Thus, by correcting the generated musical sound control information, it is possible to review the entire performance data, set the optimal musical sound control information in a balanced manner while looking at the performance data.

  Furthermore, features such as pronunciation length, same tone part, melody part, volume change (accent), double choking, consecutive choking, arpeggio performance, tone change / restoration music symbol information from performance data supplied Based on these feature information, volume parameters, pitch parameters, harmonic overtones in different parts, timbre data, etc. can be edited to obtain performance outputs with a wide variety of expressions. .

FIG. 1 is a block diagram showing a hardware configuration of a performance data creation system according to an embodiment of the present invention. FIG. 2 is a functional block diagram showing an outline of functions by the performance data creation system of the present invention. FIG. 3 is a diagram showing a musical score used in the embodiment (1) of the present invention. FIG. 4 is a diagram showing an example of a “note density-tempo coefficient” table used in the embodiment (1) of the present invention. FIG. 5 is a flowchart showing an example of the note density response process according to the embodiment (1) of the present invention. FIG. 6 is a flowchart showing another example of the note density response process according to the embodiment (1) of the present invention. FIG. 7 is a flowchart showing a process for responding to the progress of music according to the embodiment (2) of the present invention. FIG. 8 is a diagram showing an example of tempo change according to the progress of music in the embodiment (2) of the present invention. FIG. 9 is a diagram showing the relationship between the note time interval and the tempo in the embodiment (3) of the present invention. FIG. 10 is a flowchart showing processing for a long tone trill / vibrato according to the embodiment (4) of the present invention. FIG. 11 is a diagram showing a tempo change curve for a long tone trill / vibrato in the embodiment (4) of the present invention. FIG. 12 is a diagram showing an example of tempo change by phrase interpretation in the embodiment (5) of the present invention. FIG. 13 is a flowchart showing processing to the phrase end portion according to the embodiment (5) of the present invention. FIG. 14 is a flowchart showing an example of processing corresponding to the sound range according to the embodiment (6) of the present invention. FIG. 15 is a diagram showing an example of a “pitch-tempo coefficient” table used in the embodiment (6) of the present invention. FIG. 16 is a flowchart showing another example of processing corresponding to the sound range according to the embodiment (6) of the present invention. FIG. 17 is a flowchart showing the processing when the pitch shifts from rising to falling according to the embodiment (7) of the present invention. FIG. 18 is a diagram showing the relationship between pitch rise and fall and tempo change in the embodiment (7) of the present invention. FIG. 19 is a flowchart showing a process when the same pattern is continued according to the embodiment (8) of the present invention. FIG. 20 is a flowchart showing a process of assigning the same / similar tempo expression to the same / similar phrase according to the embodiment (9) of the present invention. FIG. 21 is a diagram showing an example of a phrase template in the embodiment (10) of the present invention. FIG. 22 is a flowchart showing the process using the registered sound shape according to the embodiment (10) of the present invention. FIG. 23 is a diagram for explaining slur processing in a keyboard instrument according to an embodiment of the present invention. FIG. 24 is a diagram for explaining slur processing in a wind instrument according to one embodiment of the present invention. FIG. 25 is a flowchart showing an example of processing for changing the tempo according to the volume according to the embodiment (11) of the present invention. FIG. 26 is a flowchart showing another example of the process for changing the tempo according to the volume according to the embodiment (11) of the present invention. FIG. 27 is a diagram showing an example of a “tension” knob used in the embodiment (12) of the present invention. FIG. 28 is a flowchart showing processing based on “tension” according to the embodiment (12) of the present invention. FIG. 29 is a flowchart showing processing for a group of musical notes according to the embodiment (13) of the present invention. FIG. 30 is a flowchart showing processing for a long tuplet according to the embodiment (14) of the present invention. FIG. 31 is a flowchart showing processing according to the number of chords according to the embodiment (15) of the present invention. FIG. 32 is a flow chart showing processing linked to a tone color according to the embodiment (16) of the present invention. FIG. 33 is a flowchart showing processing according to fingering fingers according to embodiment (17) of the present invention. FIG. 34 is a flowchart showing processing according to fingering position movement according to the embodiment (18) of the present invention. FIG. 35 is a flowchart showing processing according to the height of the fingering position according to the embodiment (19) of the present invention. FIG. 36 is a flow chart showing processing when pitch bend is applied according to the embodiment (20) of the present invention. FIG. 37 is a flowchart showing a process corresponding to guitar pull-off according to the embodiment (21) of the present invention. FIG. 38 is a flowchart showing a process corresponding to the piano sustain pedal operation according to the embodiment (22) of the present invention. FIG. 39 shows an example of a “vibrato depth-tempo coefficient” table in the embodiment (23) of the present invention. FIG. 40 is a diagram showing an example of a “vibrato speed-tempo coefficient” table in the embodiment (23) of the present invention. FIG. 41 is a flowchart showing processing according to the depth and speed of vibrato according to the embodiment (23) of the present invention. FIG. 42 is a view showing a reproduction example of the string trill in the embodiment (24) of the present invention. FIG. 43 is a flow chart showing processing in a plurality of parts for the string trill according to the embodiment (24) of the present invention. FIG. 44 is a flowchart showing processing based on the learning function in the embodiment (25) of the present invention. FIG. 45 is a flow chart showing processing by creating a library according to the embodiment (26) of the present invention. FIG. 46 is a flowchart showing processing based on lyrics according to the embodiment (27) of the present invention. FIG. 47 shows the relationship between the sound source output waveform and the correction timing in the embodiment (28) of the present invention. FIG. 48 is a flowchart showing processing according to the sound source output waveform according to the embodiment (28) of the present invention. FIG. 49 is a flowchart showing processing corresponding to the dynamic symbol string according to the embodiment (29) of the present invention. FIG. 50 is a flowchart showing processing corresponding to the staccato according to the embodiment (30) of the present invention. FIG. 51 is a flowchart showing an overall review process according to the embodiment (31) of the present invention. FIG. 52 is a flowchart showing the procedure of the main routine executed by the parameter automatic editing apparatus of FIG. 1, particularly the CPU. FIG. 53 is a flowchart showing the procedure of the parameter change giving process of FIG. FIG. 54 is a flowchart showing the procedure of the volume parameter change giving process 1 for performing the expression of changing the volume parameter of the performance data in accordance with the change in the pitch. FIG. 55 is a diagram illustrating an example of a note event sequence in which the pitch tends to increase. FIG. 56 is a diagram illustrating an example of time-series data after the filtering process is performed. FIG. 57 is a flowchart showing the procedure of the volume parameter change applying process 2 for performing facial expression of increasing the degree of excitement by gradually increasing the value of the volume parameter of the performance data. FIG. 58 is a diagram illustrating an example of a change imparting pattern. FIG. 59 shows an example of the relationship between note density and pitch instability in one embodiment of the present invention. FIG. 60 is a flowchart showing the procedure of the volume parameter change giving process 3 for performing facial expression of changing the volume according to the fineness of the continuous performance. FIG. 61 is a diagram showing an example of table data indicating the relationship between note density and volume. FIG. 62 is a diagram showing an example of table data indicating the relationship between note density and pitch instability in one embodiment of the present invention. FIG. 63 is a volume parameter change giving process 4 for adding a facial expression in which the volume parameters of the second and subsequent similar phrases are changed according to the similarity and appearance mode in the case of selected performance data in which similar phrases repeatedly appear. It is a flowchart which shows the procedure of. FIG. 64 is a flowchart showing the procedure of the volume parameter change applying process 5 for emphasizing the first beat when notes of the same note length appear continuously in the case of performance data of three beats. FIG. 65 is a flowchart showing the procedure of the volume parameter change applying process 6 for giving a facial expression of suppressing the volume at the end of the phrase. FIG. 66 is a flowchart showing the procedure of the volume parameter change adding process 7 for performing facial expression such that the volume is initially increased, decreased midway, and finally increased when trill or vibrato is applied to the long tone. is there. FIG. 67 is a diagram illustrating an example of a change imparting pattern. FIG. 68 is a diagram showing an example of the relationship between the trill / vibrato speed change with respect to the progress of music. FIG. 69 is a flowchart showing the procedure of the volume parameter change applying process 8 for performing natural expression when performing a roll performance of a trill or percussion instrument. FIG. 70 is a flowchart showing the procedure of the volume parameter change applying process 9 for adding expression to a chord that has its pronunciation pronounced in the case of performance data containing chords. FIG. 71 is a flowchart showing the procedure of the volume parameter change applying process 10 for performing facial expression imitating a raw staccato performance in the case of performance data including a staccato. FIG. 72 is a flowchart showing the procedure of the volume parameter change applying process 11 for giving a facial expression that the volume is fluctuated for a long tone. FIG. 73 is a diagram showing an example of table data for changing the amplitude of the random number generated by the random number counter in accordance with the count value of the random number counter. FIG. 74 is a flowchart showing the procedure of the volume parameter change applying process 12 for performing facial expression with volume expression according to the pitch average. FIG. 75 is a flowchart showing the procedure of the volume parameter change applying process 13 for performing facial expression such that the volume of each part is lowered during part division. FIG. 76 is a flowchart showing the procedure of the volume parameter change applying process 14 for performing expression with a volume expression according to a timbre change. FIG. 77 is a flowchart showing the procedure of the volume parameter change giving process 15 for giving a facial expression that the volume changes during pitch bend. FIG. 78 is a flowchart showing the procedure of the volume parameter change applying process 16 for adding a facial expression to make the melody line stand out in the case of performance data consisting of multiparts. FIG. 79 is a flowchart showing the procedure of the volume parameter change applying process 17 for performing facial expression that reproduces a lively performance by changing the volume according to fingering. FIG. 80 is a diagram illustrating an example of cello fingering. FIG. 81 shows the procedure of the volume parameter change giving process 18 for performing expression that reproduces a bow feeling by changing the volume according to the return of the bow of the stringed instrument in the case of performance data of the stringed instrument played using the bow. It is a flowchart. FIG. 82 is a diagram illustrating an example of a change imparting pattern. FIG. 83 is a flowchart showing the procedure of the volume parameter change giving process 19 for performing expression of changing inflection according to the contents of the lyrics in the case of performance data with lyrics. FIG. 84 is a diagram showing an example of table data for defining volume control after sound generation. FIG. 85 is a flowchart showing processing for an accent sound according to the embodiment (a) of the present invention. FIG. 86 is a diagram showing the relationship between the volume and pitch over time in the embodiment (a) of the present invention. FIG. 87 is a diagram showing the temporal change relationship of the pitch of double choking in the embodiment (b) of the present invention. FIG. 88 is a flowchart showing processing in double choking according to the embodiment (b) of the present invention. FIG. 89 is a diagram showing an example of continuous choking in the embodiment (c) of the present invention. FIG. 90 is a flowchart showing processing in continuous choking according to the embodiment (c) of the present invention. FIG. 91 is a diagram showing an example of harmonics in common with the arpeggio in the embodiment (d) of the present invention. FIG. 92 is a flowchart showing processing in the arpeggio according to the embodiment (d) of the present invention. FIG. 93 is a diagram showing an example of switching between “arco” and “pizz.” In the embodiment (e) of the present invention. FIG. 94 is a flowchart showing the timbre selection process using the symbols of the musical score according to the embodiment (e) of the present invention.

Explanation of symbols

1 CPU (analysis means, determination means, parameter editing means, extraction means, calculation means, detection means, setting means),
2 ROM (storage means),
9 External storage devices (supplying means, reading means) such as FDD, HDD, CD-ROMD for driving FD (storage means),
16 MIDI interface (supply means),
18 Communication interface (supply means),
OD original performance data,
ED performance data after facial expression,
Ko Initial tempo value,
to tempo change start point,
ta When reaching the lowest tempo value Kd,
tb When the target tempo value Kt is reached,
ΔK Tempo change width,
ΔKt Tempo target change amount,
n1, n2 Sound that gives a noisy pitch change,
13k “Tension” knob,
A sound source output waveform that reaches the maximum volume value at the rising time tm at the time ts ts,
tc The point at which the 80% volume value is reached to give correction timing.

Claims (1)

Performance data supply means for supplying performance data composed of a plurality of parts each including note information and musical tone control information representing volume;
Part detection means for detecting the same tone color part and the number of parts from the performance data;
Volume value calculating means for calculating the volume value of each part based on the number of parts, so that the volume of the part of the same tone color decreases.
A performance data creating apparatus comprising: a sound volume value assigning means for assigning the musical tone control information representing the sound volume value to the corresponding part.

Images