JP2011085731A - Musical signal processing device and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To create a lead tone and an additional tone which is stable without pitch fluctuation, when a musical signal in which many pitches are fluctuated is input in a short period of time. <P>SOLUTION: Two similar control means with each independent structure are separately provided, that is: a first pitch detecting means for detecting a note corresponding to a pitch name, and a first musical tone creating means for creating a first musical tone signal in which a pitch is controlled to the detected note; and a second pitch detecting means for detecting a note corresponding to a pitch name, and a second musical tone creating means for creating a second musical tone signal in which a pitch is controlled to an arbitrary note determined based on the detected note. Different control is performed on each of them, according to control information. Thereby, when a musical signal with changing note by fluctuating up and down is input, the lead tone in which a note is controlled based on the musical signal is created, and the additional tone following to note variation of the lead tone is created as a stable musical tone without pitch fluctuation. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention generates a lead sound that is pitch-controlled based on the pitch (pitch) of the input musical sound or voice, and generates an additional sound that is pitch-controlled following the pitch fluctuation of the generated lead sound. The present invention relates to a musical sound signal processing device and a program. In particular, the present invention relates to a technique for generating a stable added sound that is stable in terms of audibility and has no wobbling in the pitch when musical sounds or voices whose pitch changes in a short time are input.

  Conventionally, the pitch of a musical tone signal such as an input musical tone or voice is detected (a specific pitch corresponding to one of the musical pitch names is finally detected), and the musical tone of the lead sound at the detected pitch. A signal (first musical tone signal) is generated, and a new pitch (similarly corresponding to one of the musical pitch names) is newly created based on the detected pitch and chord information input from the keyboard or the like. A tone having a tone generation function for automatically generating a tone signal (second tone signal) of a harmony tone having the determined pitch as a separate independent additional tone having the generated lead tone as a main tone. Signal processing apparatuses and programs are known. As an example of such an apparatus, there is an apparatus described in Patent Document 1 shown below. In this specification, the term “musical sound signal” is not limited to a musical sound signal, but is used in the sense that it may include a sound or any other sound signal.

  Here, a conventionally known musical tone generation processing procedure in the apparatus described in Patent Document 1 will be described. FIG. 4 is a conceptual diagram for explaining a conventionally known musical sound generation processing procedure. The left diagram in FIG. 4 sequentially shows the flow of processing executed in the apparatus, and the right diagram in FIG. 4 shows changes in signal waveforms accompanying the execution of each processing. The vertical axis is frequency and the horizontal axis is time. FIG. 5 is a conceptual diagram showing a data structure of a conventionally known pitch determination table that is referred to when determining the pitch of a harmony sound as described later.

  First, an audio signal input via a microphone or the like is converted into a frequency signal by “frequency detection” processing. For this “frequency detection” process, any known technique such as the zero cross method, which is a well-known technique in the field of speech analysis, may be used, and detailed description thereof is omitted here. Next, the frequency signal is flattened (or also referred to as smoothing) by “flattening” the frequency signal. The flattened frequency signal is discretized into one of twelve-tone scale names (pitch names) every predetermined time by “class name detection” processing. Specifically, the flattened frequency signal is rounded to a predetermined pitch corresponding to one of a plurality of musical names defined in units of semitones (100 cents) (referred to as a name signal). In this way, the pitch of the input audio signal is detected. In the “convergence curve” processing, the floor signal is a signal that changes continuously in time. Then, the output signal of the lead sound in which the pitch of the input audio signal is modulated is generated by performing “output modulation” processing on the signal that continuously changes over time. In this example, however, the pitch of the detected audio signal is determined as the pitch of the lead sound.

  On the other hand, when adding a harmony sound, the pitch detection result (or the pitch of the lead sound determined according to the pitch detection result) of the audio signal obtained in the “class name detection” process and the keyboard Based on the input chord information, one of 12 scale names (pitch names) is determined every predetermined time according to a pitch determination table as shown in FIG. 5 prepared in advance. That is, the pitch determination table shown in FIG. 5 is stored in advance in a ROM, RAM, or the like, one table for each chord, and one corresponding table is specified according to the chord information. Here, as an example, only the table for the “C major” code is shown. Then, as soon as the pitch detection of the input voice signal is triggered (trigger), the specified table is referred to based on the pitch detection result, and the pitch of the harmony sound is set to one of the musical pitch names. A corresponding specific pitch is determined, and a harmony sound pitch-controlled to the pitch is generated. Here, in the pitch determination table shown in FIG. 5, “E (0)” indicates an “E” sound having the same octave as the pitch of the detected lead sound, and “C (+1)” is detected. This indicates that the pitch is “C”, one octave above the pitch of the lead sound. Therefore, for example, when the pitch of the lead sound is “E3”, the pitch of the first harmony sound is determined as “G3”, and the pitch of the second harmony sound is determined as “C4”. .

  Thus, based on the floor name signal composed of the pitch corresponding to one of the floor names of 12 scales determined according to the pitch determination table as shown in FIG. By performing the “convergence curve” process and the “output modulation” process in sequence, an output signal of one or more harmony sounds is generated. Note that the note-on timing of the lead sound and the harmony sound generated as described above is the time when the pitch of the input audio signal is detected, while the note-off timing is the pitch of the input audio signal. This is the point when it is no longer detected.

Japanese Patent Laid-Open No. 11-133954

  As described above, the conventional apparatus determines the pitch of the harmony sound based on the pitch detection result (that is, the pitch of the lead sound) of the input audio signal. It can be understood that depends on the pitch of the lead sound. If so, when a voice signal is input that fluctuates up and down like a vibrato within a short period of time between the detection of the vowel of the input voice signal and the detection of the next vowel. There is a risk that a harmony sound whose pitch is continuously fluctuating larger than the fluctuation of the lead sound may be generated, and such a harmony sound is inconvenient and difficult to listen to.

  For example, according to the pitch determination table shown in FIG. 5, when the input audio signal (and thus the lead sound) is a vibrato that fluctuates between the pitch “E3” and the pitch “F3”, The harmony sound of 1 becomes an output signal that continuously changes so as to fluctuate between the pitch “G3” and the pitch “C4”. This means that even though the input audio signal fluctuates for only a semitone, the added harmony sound repeats a jump with a large pitch fluctuation of 4 semitones within a short period of time. This kind of harmony sound is not very useful as an expression for vibrato.

  In order to avoid the inconvenience, as another method, it is conceivable to reduce the frequency of pitch detection of the input audio signal itself. However, since both the lead sound and the harmony sound are generated according to the pitch detection of the sound signal, in such a case, not only the harmony sound but also the frequency of the lead sound itself is reduced, and the input sound signal Adopting this method is very inconvenient because there is a risk of losing the musical personality and expressiveness that it had.

  The present invention has been made in view of the above points, and even if the pitch variation of the input musical sound signal repeatedly occurs within a short time, the musical signal of the signal is based on the input musical sound signal. A musical tone signal processing apparatus and program that can generate a lead sound that reflects individuality and expressiveness, and that can generate a stable additional sound that is calm and audible without fluctuation in pitch It is.

  The musical tone signal processing apparatus according to the present invention generates a first musical tone signal whose pitch is controlled based on an inputted musical tone signal, and the pitch follows the fluctuation in pitch of the first musical tone signal. In a musical tone signal processing apparatus for generating a second musical tone signal to be controlled, input means for inputting a musical tone signal, analyzing means for analyzing the frequency of the inputted musical tone signal, and a frequency signal obtained in accordance with the frequency analysis The first pitch detection means for sequentially detecting any of the pitches corresponding to the pitch names according to the flattened frequency signal, and the pitch control to the pitch detected by the first pitch detection means. A first musical tone generation means for generating the first musical tone signal, a designation means for designating control information, and a change in the frequency signal obtained in accordance with the frequency analysis is flattened, and according to the flattened frequency signal Second pitch detecting means for sequentially detecting any one of pitches corresponding to names, the second pitch detecting means being controlled based on the control information, and being detected by the second pitch detecting means. Second musical tone generating means for generating a second musical tone signal pitch-controlled to an arbitrary pitch determined based on the pitch, and an output for outputting at least one of the generated first and second musical tone signals. And the second pitch detection means sequentially detects pitches with pitch changes that are temporally different from the pitch changes according to the pitches detected by the first pitch detection means. Further, the control is performed independently of the first pitch detection means based on the control information.

  According to the present invention, the first pitch detection means for detecting the pitch corresponding to the pitch name and the first musical tone generation for generating the first musical tone signal pitch-controlled to the pitch detected by the first pitch detection means. Means, a second pitch detecting means for detecting a pitch corresponding to the pitch name, and a second musical tone signal pitch-controlled to an arbitrary pitch determined based on the pitch detected by the second pitch detecting means. The second musical sound generating means for generating the sound is separately configured. Then, the second pitch detection means is caused to perform control different from that of the first pitch detection means based on the control information. That is, the first pitch detecting means and the first musical sound generating means for generating the first musical sound signal, and the second pitch detecting means and the second musical sound generating means for generating the second musical sound signal. Are controlled in the same manner, but are provided separately in independent configurations, and different controls are performed according to the control information. In particular, the second pitch detection means sequentially detects pitches with pitch changes temporally different from the pitch changes according to the pitches detected by the first pitch detection means based on the control information. For example, even when a musical tone signal such as a vibrato that changes while the pitch fluctuates up and down is input, the first musical tone whose pitch is controlled based on the inputted musical tone signal. In addition to generating a signal, the second musical sound signal whose pitch is controlled following the pitch fluctuation of the first musical sound signal can be separately generated as a calm and stable musical sound. become.

  The present invention can be constructed and implemented not only as a device invention but also as a method invention. Further, the present invention can be implemented in the form of a program of a processor such as a computer or a DSP, or can be implemented in the form of a storage medium storing such a program.

According to the present invention, the first musical tone signal generating means and the second musical tone signal generating means are configured separately, and the first musical tone signal and the second musical tone signal are generated based on the inputted musical tone signal. Thus, even when an audio signal such as vibrato that changes while the pitch fluctuates up and down is input, the first that reflects the fluctuation of the pitch is used. Independent of the musical tone signal, the second musical tone signal having a calm and stable sense of hearing can be generated.
In addition, since the frequency of the first musical sound signal does not have to be different from the conventional frequency, it is not necessary to reduce the pitch detection frequency of the inputted musical sound signal. There is also an advantage that the expression power is not lost.

1 is a block diagram of a hardware configuration showing an embodiment of an overall configuration of a musical tone signal processing apparatus according to the present invention. It is a functional block diagram for demonstrating the musical tone production | generation function of the musical tone signal processing apparatus which concerns on this invention. It is a conceptual diagram which shows the frequency change of an audio | voice signal in order to demonstrate a floor name detection. It is a conceptual diagram for demonstrating the conventionally known musical tone production | generation process procedure. It is a conceptual diagram which shows the data structure of the conventionally known pitch determination table.

  Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

  FIG. 1 is a hardware configuration block diagram showing an embodiment of the overall configuration of a musical tone signal processing apparatus according to the present invention. The musical tone signal processing apparatus shown in this embodiment is controlled by a microcomputer comprising a microprocessor unit (CPU) 1, a read only memory (ROM) 2, and a random access memory (RAM) 3. The CPU 1 controls the operation of the entire apparatus. ROM 2, RAM 3, input operation unit 4, display unit 5, sound source 6, communication interface (I / F) 7, and storage device 8 are connected to CPU 1 via communication bus 1D (for example, data and address bus). Has been.

  The ROM 2 stores various control programs executed or referred to by the CPU 1 and various data such as a pitch determination table shown in FIG. The RAM 3 is a working memory that temporarily stores various data generated when the CPU 1 executes a predetermined control program, or a memory that temporarily stores a control program currently being executed and related data. used. A predetermined address area of the RAM 3 is assigned to each function and used as a register, flag, table, memory, or the like.

  The input operation unit 4 is, for example, an input device such as a microphone for inputting a voice signal such as a voice uttered by a person, a start / stop button for instructing automatic generation start / stop of a harmony sound, and various parameters ( In addition to various operators such as a switch for setting a switch to be described in detail later, a numeric keypad for inputting numeric data, a keyboard for inputting character data, a mouse, or the like may be used. The input device is not limited to a microphone, and a musical tone operator such as a keyboard that generates a musical tone signal such as a chord tone required for generating a harmony sound in response to a user operation, or a musical tone previously stored in the ROM 2 or the like. It may be an input device such as a sequencer that supplies signals in the order of performance.

  The display unit 5 includes, for example, a liquid crystal display panel (LCD), a CRT, and the like. The generated lead sound and / or harmony sound score, parameter setting states set by various operators, or various kinds of pre-stored information Various information such as a list of data and a control state of the CPU 1 is displayed.

  The sound source 6 can simultaneously generate musical sound signals on a plurality of channels, and for example, a lead sound (first musical sound signal) or harmony based on an audio signal input via a communication bus 1D, for example, via a microphone. A musical tone signal such as a sound (second musical tone signal) is generated, and a musical tone is generated based on the generated musical tone signal. The musical sound generated from the sound source 6 is generated from a sound system 6A including an amplifier and a speaker. When the sound source 6 generates a lead sound, a harmony sound, etc., for example, gender (type and depth of voice quality such as male voice and female voice), vibrato (depth and period change rate, delay time until vibrato start) Various effects such as tremolo, volume, pan (stereolocation), detune, and reverb (reverberation) can be applied. Note that any conventional configuration may be used for the configuration of the sound source 6 and the sound system 6A. For example, the tone generator 6 may employ any of various tone synthesis methods such as FM, PCM, physical model, formant synthesis, etc., may be configured with dedicated hardware, or configured with software processing by the CPU 1 or DSP. May be.

  The communication interface (I / F) 7 is an interface for transmitting and receiving various information such as a tone signal, a pitch determination table, and a control program between the apparatus and an external device (not shown). The communication interface 7 may be, for example, a MIDI interface, a LAN, the Internet, a telephone line, or the like, and may include both wired and wireless ones.

  The storage device 8 stores various information such as a pitch determination table prepared in advance and various control programs executed by the CPU 1. Or you may enable it to memorize | store musical sound signals, such as the produced | generated lead sound and a harmony sound.

  If no control program is stored in the ROM 2, the control program is stored in the storage device 8 (for example, a hard disk) and read into the RAM 3 to store the control program in the ROM 2. It is possible to cause the CPU 1 to execute the same operation as in FIG. In this way, control programs can be easily added and upgraded. The storage device 8 is not limited to a hard disk (HD), but can be attached in various forms such as a flexible disk (FD), a compact disk (CD), a magneto-optical disk (MO), or a DVD (Digital Versatile Disk). A storage device using a recording medium may be used. Alternatively, a semiconductor memory or the like may be used.

In the above-described musical tone signal processing apparatus, the input operation unit 4, the display unit 5, the sound source 6 and the like are not limited to being built in one apparatus body, but each is configured separately and communicates via a MIDI interface or various networks. Needless to say, each device may be configured to be connected using an interface.
The musical tone signal processing apparatus and program according to the present invention can be applied to any type of apparatus / equipment such as a karaoke apparatus, an electronic musical instrument, a personal computer, a portable communication terminal such as a mobile phone, or a game apparatus. Good. When applied to a portable communication terminal, not only when a predetermined function is completed with only the terminal, but also a part of the function is provided on the server side, and the predetermined function is realized as a whole system composed of the terminal and the server. You may make it do.

  Also in the musical tone signal processing apparatus according to the present invention, for example, the pitch of an audio signal input via a microphone or the like is detected (finally a specific pitch (pitch) corresponding to one of the musical pitch names). Detecting), generating a musical tone signal of the lead sound of the detected pitch, and separately generating a new pitch (similarly, any of the musical pitch names) based on the detected pitch and chord information inputted from the keyboard or the like. And a tone generation function for automatically generating a tone signal of a harmony tone of the determined pitch. However, in the musical tone signal processing apparatus according to the present invention, unlike the conventional one, each of the functions for generating the lead sound and the harmony sound is configured independently. Therefore, the tone generation function of the tone signal processing apparatus according to the present invention will be described with reference to FIG. FIG. 2 is a functional block diagram for explaining the tone generation function of the tone signal processing apparatus according to the present invention. In FIG. 2, the arrows in the figure represent the flow of signals.

  As shown in FIG. 2, the tone generator 6 includes a signal input unit I, a frequency detection unit F, a lead tone tone generator M1, a harmony tone tone generator M2, an effect applying unit E, and a signal output control unit O. It has a function. The signal input unit I acquires an audio signal input via a microphone or the like with the start of the musical sound generation function, and sequentially supplies the acquired audio signal to the frequency detection unit F. When receiving the audio signal, the frequency detection unit F performs “frequency detection” processing on the input audio signal and converts it into a frequency signal. The frequency signal output from the frequency detector F is supplied to the lead tone musical sound generator M1 and the harmony musical tone generator M2.

  The lead tone music generator M1 and the harmony tone generator M2 are separately generated tone generators for generating the lead tone signal and the harmony tone signal, respectively. Part H1 (H2), pitch conversion part C1 (C2), lead sound generation part A1, or harmony sound generation part A2. The flattening unit H1 (H2) performs a “flattening” process on the frequency signal output from the frequency detection unit F, thereby flattening (smoothing) the change in the frequency signal. However, as will be described later, in the flattening unit H1 and the flattening unit H2, the “flattened frequency signal” obtained as a result of the “flattening” process according to the content of the designated parameter (flattening time constant described later). "Can be different. The parameters are instructed as lead sound parameter information from the lead sound parameter setting unit T1 to the flattening unit H1, and as harmony sound parameter information from the harmony sound parameter setting unit T2 to the flattening unit H2. The

  The flattened frequency signal is supplied to the pitch converter C1 or the pitch converter C2. The pitch conversion unit C1 discretizes the flattened frequency signal into any one of 12 scale names (pitch names) at predetermined time intervals by performing “class name detection” processing. That is, a specific pitch corresponding to one of the musical pitch names is detected based on the pitch of the input audio signal, and the specific pitch corresponding to one of the musical pitch names thus obtained is used as the lead sound. The pitch is determined and used. Of course, the pitch detection result of the input sound signal itself is not limited to the pitch of the lead sound, and the pitch detection result of the input sound signal is raised or lowered by a predetermined pitch such as one octave or three semitones. Then, the pitch converted may be determined as the pitch of the lead sound.

  On the other hand, the pitch conversion unit C2 performs a “class name detection” process on the flattened frequency signal to discretize it into one of 12 scale names (pitch names) at predetermined time intervals. FIG. 5 prepared in advance shows the pitch detection result of the input voice signal obtained by the “floor name detection” process based on the pitch detection result and the chord information input from the keyboard or the like. According to a pitch determination table, the pitch is converted into one of the twelve scale names (sound names) and discretized. A specific pitch (which may be one or more) corresponding to any of the pitch names of the music thus obtained is determined and used as the pitch of the harmony sound. However, as will be described later, the pitch conversion unit C1 and the pitch conversion unit C2 input data obtained as a result of the “floor name detection” process according to the instructed parameter contents (floor name change threshold and offset value described later). The pitch detection results of the audio signals thus made can be different. The parameters are as lead sound parameter information from the lead sound parameter setting unit T1 for the pitch conversion unit C1, and as harmony sound parameter information from the harmony sound parameter setting unit T2 to the pitch conversion unit C2. Instructed.

  A specific pitch (name signal) corresponding to one of the pitch names of music detected by the pitch converter C1 (C2) is supplied to the lead sound generator A1 or the harmony sound generator A2. When the lead sound generation unit A1 or the harmony sound generation unit A2 receives the floor name signal, the lead sound generation part A1 or the harmony sound generation part A2 separately generates a lead sound or a harmony sound based on the supplied floor name signal. That is, the lead sound generation unit A1 uses the “convergence curve” process to convert the rank signal into a signal that changes continuously in time, and the input voice signal by performing the “output modulation” process on the rank signal that changes in time continuously. An output signal of a lead sound in which the pitch of the signal is modulated is generated. The harmony sound generator A2 uses the “convergence curve” process to convert the rank signal into a signal that continuously changes in time, and performs an “output modulation” process on the rank signal that changes in time continuously. The output signal of the harmony sound with modulated pitch is generated independently independently of the generation of the lead sound by the lead sound generation unit A1.

  The lead sound or the harmony sound generated by the lead sound generation unit A1 or the harmony sound generation unit A2 is supplied to the effect applying unit E, and the effect applying unit E provides gender, vibrato, tremolo, volume, pan, detune, reverb, etc. Various effects can be imparted. The signal output control unit O outputs the lead sound and / or the harmony sound supplied from the effect applying unit E to the sound system 6A. In this case, it is possible to appropriately select a musical sound signal to be output such as only a lead sound, only a harmony sound, a lead sound, and a harmony sound.

  As described above, in the musical tone signal processing apparatus according to the present invention, each of the functions for generating the lead sound and the harmony sound is configured independently. Therefore, each of the lead tone musical sound generation unit M1 and the harmony sound musical tone generation unit M2 can be separately controlled, so that the generation timing of the lead sound and the generation timing of the harmony sound are not necessarily matched. I have to. Specifically, by instructing the parameters with different contents (parameter values) to the flattening unit H1 and the flattening unit H2, and the pitch converting unit C1 and the pitch converting unit C2, respectively, independent time for both of them. It is possible to control the signal by realizing the proper behavior. This will be described below.

  First, control of the flattening part H1 and the flattening part H2 will be described. In the flattening process, when the input audio signal changes from silence to voiced sound, there is often a period during which the frequency cannot be detected well, so the frequency detection is performed after detecting the voiced sound (vowel). By indicating the grace period up to (this is called a flattening time constant) as a parameter, it is possible to eliminate unnecessary frequency information that can be detected when the frequency cannot be detected well. The time constant of flattening is the time from the time when it is detected as a vowel to the start of flattening, and is set, for example, between 0 and 50 ms (milliseconds). For example, when the flattening time constant is set to 0 ms, the flattening is started from the time point detected as a vowel. On the other hand, when the flattening time constant is set to 10 ms, the detected pitch is output as it is from the time when it is detected as a vowel until 10 ms, and flattening is performed after 10 ms.

  By the way, in general, if the time constant of flattening is slowed, the possibility of misrecognizing a wide pitch change close to a semitone as a name change is lowered, but the followability to a pitch change exceeding a semitone is deteriorated. On the other hand, if the flattening time constant is increased, the followability to a pitch change exceeding a semitone is improved, but the possibility of misrecognizing a wide pitch change close to a semitone as a name change is increased. Conventionally, however, the lead sound and the harmony sound are not independent of the flattening process, and the lead sound and the harmony sound can only be controlled using the same flattening time constant. That is, in the conventional case, the flattening process can be controlled only by one certain time constant. Therefore, when the time constant of flattening is delayed, the lead sound is not convenient and the harmony sound is convenient. When the flattening time constant was increased, the lead sound was favorable and the harmony sound was inconvenient. Therefore, in the present invention, the flattening process is made independent for the lead sound and the harmony sound, and the time constant of the flattening is increased for the lead sound that needs to follow the pitch change exceeding the semitone. On the other hand, for a harmony sound that does not require the ability to follow a pitch change exceeding a semitone, the time constant of flattening is set to be slow, and the flattening part H1 and the flattening part H2 are controlled separately. To be able to.

  Next, control of the pitch converter C1 and the pitch converter C2 will be described. When discretizing frequency information into floor names at the time of “floor name detection”, conventionally, the floor name is determined by determining which one is closer to the center frequency of the upper and lower floor names. In the embodiment, once a floor name is determined, a new floor name is not adopted unless the floor frequency is separated from a center frequency equal to or higher than a specified threshold value. As a parameter for that purpose, a range (referred to as a floor detection threshold) in which the transition of the target pitch for pitch correction is suppressed is designated. For example, if the floor detection threshold is set to 75 cents, and the detected pitch is within ± 75 cents from the current target pitch, the current target pitch is maintained (no new floor name is adopted). I am doing so. Thereby, a hysteresis effect can be brought about in transition of the target pitch of pitch correction.

  Here, a case where the detected frequency changes as shown in FIG. 3 after it is first detected as a floor name (specific pitch) X will be described as an example. FIG. 3 is a conceptual diagram showing a frequency change of an audio signal in order to explain floor name detection. The vertical axis is frequency and the horizontal axis is time. In FIG. 3, the floor name x and the floor name y are adjacent floor names (100 cents difference), and the broken line in the figure is the frequency of 50 cents difference on the basis of the frequency of the floor name x. For convenience, the alternate long and short dash line represents a frequency with a difference of 75 cents, and the alternate long and short dashed line represents a frequency with a difference of 25 cents.

  When the frequency changes and exceeds the 50 cent difference for the first time at a time point, conventionally, the floor name (specific pitch) changes from x to y at this time point. After that, it falls below 50 cents again at time b, so the floor name changes from y to x. However, such a change often changes the pitch with a slightly deeper expression, and does not correspond to a change in name on the notation. Therefore, in the present invention, for example, 75 cents is designated as the threshold for changing the name. In this case, as the frequency changes, even if it exceeds the 50 cent difference at the time point a, it does not reach the 75 cent difference. Since this is the first time, the floor name is changed from x to y at this point. Once the floor name is changed from x to y, the floor name is not changed from y to another floor name unless the frequency is changed more than 75 cents from the frequency of the floor name y. The floor name changes from y to x for the first time at the time point f when the frequency changes. It should be noted that the floor name detection threshold of 75 cents may be set to an appropriate value depending on how the user sings.

  In addition, the values indicated by the broken lines and the alternate long and short dash lines in FIG. 3 can be offset up and down according to the offset values (parameter information) instructed from the lead sound parameter setting unit T1 or the harmony sound parameter setting unit T2, respectively. As a result, when determining the pitch from the pitch of the input audio signal, by adding different offsets to the pitch determination criterion, the point of change of the name of each rank is adjusted between the lead sound and the harmony sound. Can be directed. For example, for an input voice with an upward tendency, an offset value that offsets the floor change threshold (75 cents in the above example) up and down by about 5 cents as a whole may be instructed only to the pitch conversion unit C2. By giving the various parameters as described above to the pitch converter C1 and the pitch converter C2, the pitch converter C1 and the pitch converter C2 respectively receive wide vibrato audio signals. Even in such a case, that is, when the pitch change is large as a musical expression, an extra floor name change is not detected.

As described above, the musical tone signal processing apparatus according to the present invention sequentially performs the “frequency detection” process, the “flattening” process, the “floor name detection” process, the “convergence curve” process, and the “output modulation” process. In this way, the lead sound and the harmony sound are generated, and the lead sound tone generation unit M1 and the harmony sound music generation unit M2 having the same configuration including each of the processes performed after the “frequency detection” process are provided. A lead sound generation and a harmony sound generation are provided separately, and different parameter information is given to them so that different control can be performed. In other words, even when a lead sound is generated by immediately determining the pitch based on the pitch detection result in response to the pitch detection of the audio signal, the harmony tone musical sound generating unit M2 is different from the conventional one. In some cases, the pitch may not be detected when the pitch of the signal is detected (parameter information is given to the musical tone generator M2 for the harmony sound in this way). In other words, the harmony sound generation unit M2 sequentially detects the pitches accompanied by a change in pitch that is temporally different from the change in pitch according to the pitch detected by the lead tone generation unit M1. Sound parameter information is instructed from the harmony sound parameter setting unit T2. By doing so, even when a voice signal such as a vibrato whose pitch changes while moving up and down is input, it is possible to generate a harmony sound with a calm and stable sense of hearing. .
In addition, since the frequency of pitch detection of the input audio signal does not have to be reduced, the frequency of occurrence of lead sounds is the same as before, and the musical personality and expressive power of the input audio signal are lost. There is nothing.

In the above-described embodiments, the musical sound signal that is the basis for generating the lead sound and the harmony sound has been described by taking the sound input to the microphone as an example. However, for example, the musical instrument performance sound input to the microphone may be used. . In the case of a musical instrument performance sound, the additional sound may be an accompaniment sound.
The harmony sound is not limited to generating only one sound at a time, and may generate a plurality of sounds at the same time.

Note that the chord information input for generating the harmony sound was detected from the input information input in response to a user operation from a performance operator such as a keyboard connected to the apparatus or the apparatus as described above. It may be a thing which can be obtained in the form of inputting chord names sequentially.
In the above-described embodiment, the harmony sound is generated based on the chord information. However, the present invention is not limited to this, and other known methods for generating the harmony sound without using the chord information may be used. . For example, a method of generating a harmony sound with a pitch three times higher than the pitch of the lead sound may be adopted.

DESCRIPTION OF SYMBOLS 1 ... CPU, 2 ... ROM, 3 ... RAM, 4 ... Input operation part, 5 ... Display part, 6 ... Sound source, 6A ... Sound system, 7 ... Communication interface, 8 ... Memory | storage device, 1D ... Communication bus, A1 ... Read Sound generation unit, A2 ... Harmony sound generation unit, C1 (C2) ... Pitch conversion unit, E ... Effect imparting unit, F ... Frequency detection unit, H1 (H2) ... Flattening unit, I ... Signal input unit, M1 ... Musical tone generator for lead sound, M2 ... Musical tone generator for harmony sound, O ... Signal output control unit, T1 ... Parameter setting unit for lead sound, T2 ... Parameter setting unit for harmony sound

Claims (3)

A first musical tone signal whose pitch is controlled based on the input musical tone signal is generated, and a second musical tone signal whose pitch is controlled following the pitch fluctuation of the first musical tone signal is generated. In the musical sound signal processing device to
An input means for inputting a musical sound signal;
Analyzing means for frequency analysis of the input musical sound signal;
First pitch detection means for flattening a change in the frequency signal obtained according to the frequency analysis, and sequentially detecting any one of pitches corresponding to the pitch names according to the flattened frequency signal;
First musical sound generating means for generating a first musical sound signal pitch-controlled to the pitch detected by the first pitch detecting means;
A designation means for designating control information;
A second pitch detecting means for flattening a change in the frequency signal obtained in accordance with the frequency analysis and sequentially detecting any one of pitches corresponding to pitch names in accordance with the flattened frequency signal; The pitch detection means is controlled based on the control information,
Second musical sound generating means for generating a second musical sound signal pitch-controlled to an arbitrary pitch determined based on the pitch detected by the second pitch detecting means;
Output means for outputting at least one of the generated first and second musical sound signals,
The second pitch detecting means is based on the control information so as to sequentially detect a pitch accompanied by a pitch change temporally different from a pitch change according to the pitch detected by the first pitch detecting means. The musical tone signal processing apparatus is controlled independently of the first pitch detecting means. The designation means uses control information as time information from vowel detection to flattening start when flattening a change in frequency signal obtained according to the frequency analysis, and a sound corresponding to a pitch name according to the flattened frequency signal. Specify at least one of threshold information for determining which pitch is to be the upper or lower pitch when detecting any of the high, offset information for offsetting the threshold further up and down,
The second pitch detection means detects a pitch so as to suppress a temporal pitch change compared to a pitch change detected by the first pitch detection means based on at least one of the information. The musical tone signal processing apparatus according to claim 1. A first musical tone signal whose pitch is controlled based on the input musical tone signal is generated, and a second musical tone signal whose pitch is controlled following the pitch fluctuation of the first musical tone signal is generated. A computer-executable program, the program being stored in the computer,
The procedure for inputting musical sound signals,
A procedure for frequency analysis of the input musical sound signal;
A procedure for flattening a change in the frequency signal obtained according to the frequency analysis, and sequentially detecting any one of the pitches corresponding to the pitch names according to the flattened frequency signal;
Generating a first musical tone signal pitch-controlled to the detected pitch;
A procedure for specifying control information;
Based on the control information, the change in the frequency signal obtained according to the frequency analysis is flattened so as to be accompanied by a pitch change that is temporally different from the pitch change according to the detected pitch, and the flattening is performed. A detection procedure for sequentially detecting one of pitches corresponding to the pitch name according to the frequency signal, and a second musical tone signal pitch-controlled to an arbitrary pitch determined based on the pitch detected by the detection procedure The steps to generate
A program for executing a procedure for outputting at least one of the generated first and second musical tone signals.

Images