FR2994015A1 - Musical improvisation method for musical instrument e.g. piano, involves generating audio signal representing note or group of notes, and playing audio signal immediately upon receiving signal of beginning of note - Google Patents

Abstract

The method involves receiving a clock signal generated by a clock generating circuit, and generating a variable tempo as a function of the clock signal. A signal of beginning of a note or beginning of group of notes is received, and a note or a group of notes to be played is defined by an automatic music generator according to a set of composition rules. An audio signal representing the note or group of notes is generated and immediately played upon receiving the signal of beginning of a note. The play of the audio signal is stopped upon receiving a signal of end of note. Independent claims are also included for the following: (1) a computer program for musical improvisation (2) an information storage unit storing a program for musical improvisation (3) a musical instrument in the form of an information processing device.

Description

The present invention relates to the field of musical instruments and more particularly to an instrument allowing a user, musician or not, to improvise a melody that is always in harmony with a musical background broadcast by the instrument.

Preliminary definition: the term of the musical vocabulary "measure" is used interchangeably in French to express two notions: - the grouping of rhythmic values in small sets, delimited visually by bars of measures, which are vertical bars occupying all the range (Claude Abromont: The theory of music). This is the French equivalent of the English word "bar"; - Each type of grouping has its own particular number, which indicates a fundamental dimension of rhythm: musical support. The complete duration of a measurement is broken down into several shorter periods called time. The type of a measure will depend on the number of times that this measure has and the duration of these different times (Claude Abromont: The theory of music). This is the French equivalent of the English expression "time signature". In the following, we will use: - the word "measure" to express the first notion; - the expression "measurement (encryption)" to express the second notion. There are a large number of musical instruments, which allow you to play in one or more different styles, but none of them allows you to play without having acquired a minimum knowledge of musical theory, and other on the other hand, to have followed an important apprenticeship in the instrumental technique related to this instrument. Moreover, the practice of improvisation requires a much deeper knowledge of musical theory, as well as a long apprenticeship on the instrument aiming on the one hand to acquire an important dexterity allowing to develop a varied musical discourse, and the other hand, the knowledge and practice of a large number of musical motifs related to the instrument and often used during such improvisations. The term improvisation here means "to play a music that one composes on the spot and without preparation". By definition, the playing of an instrument made by reading a score does not fit within the framework of improvisation. The improvisation can be done by the instrumentalist in collaboration with other musicians, or through the use of a soundtrack, ie recorded accompanying music (eg a beach of any compact disc of any artist). In both cases, the instrumentalist relies on the rhythmic and harmonic discourse delivered by his or her collaborators to build his musical discourse. There are so-called "diatonic" instruments, for example the diatonic harmonica, which are simpler to use because they are linked to a tone, but their use is in fact limited to this tone and does not make it possible to play, especially in the context of improvisation, on harmonies using several tones, or even simply on an agreement outside this tone as it often exists in music. There are electronic instruments called keyboards arrangers, in the form of a piano keyboard, and incorporating into the form of musical sequences different musical styles, which allow a single musician to play both the accompaniment and a melody or an improvisation. After having selected a style and a tempo, possibly chosen the instruments he wants to use, he starts the accompaniment, his left hand generating the harmonic frame, that is to say the sequence of chords, most often following a particular and simplified fingering, different from that a pianite would use to tackle the same chords. The computer program of the keyboard-arranger analyzes the fingerings thus plated, and modifies a pre-recorded musical sequence so that it is played in the harmony desired by the instrumentalist. The accompaniment is thus realized, and the instrumentalist can play or improvise a melody with the right hand, using the piano technique. But here again, the instrumentalist must know both musical theory and piano technique to provide a coherent and adapted musical discourse. There are musical video games, in which the player using an interface most often in the form of a musical instrument, must execute the orders that are indicated on the screen, along with a piece of music is executed by the game. The orders are of the type "press on this key" and generally connected to the melody played by one of the instruments of the musical background. Pressing more or less synchronously to the melody that is played brings a number of points more or less important. It is not a question here of music or musical interpretation, but reflex. The player does not play music: nothing in his behavior is close to that of an instrumentalist. He just executes the instructions displayed on the screen. On the other hand, there are electronic systems for automatic music composition capable of generating complete music pieces, with or without the prior intervention of a user. Some of these automatic composers allow the user to select a particular style, tempo, particular musical structure, and as they work together with a MIDI Musical Instrument Digital Interface (MIDI) synthesizer-what they have composed is played by the synthesizer-, they allow the user to select an instrumentation. Sometimes, using the commands of the MIDI language, the user can modify the mix in real time, for example by varying the volume, the position in the stereo space or the equalization of one of the instruments. a sound engineer's profile, but at no point can he interact with the machine in a way that makes him feel like playing music, since he's just editing sounds rather than playing notes. Finally, there are electronic interfaces, connected to a computer, which allow through the execution of a computer program on this computer, to superpose sounds or recorded sequences, or even to superimpose these sounds or sequences to a piece of music played by said computer program. The user can sometimes decide which sequence or which sound he will trigger and when, but he does not really play music as an instrumentalist could do. For example, it does not choose the length of the notes in the sequences, or the height of these sequences, nor even the orientation (towards the bass or the treble) of these sequences, since they are pre-recorded and can not to be modified. In addition, these devices generally require a long work of construction and preparation of sounds and sequences that require extensive musical knowledge and sound, especially so that the so-called sounds and sequences sound just between them when they are played simultaneously. It appears that these instruments or systems require from the user a certain instrumental technique and / or musical knowledge if he wants to be able to play one of these instruments or systems in order to produce a coherent musical discourse (that is to say at least: just musically), especially as part of an improvisation, and especially if it wishes to derive pleasure. Or otherwise, these systems are not really geared towards the improvisation game in which the musician, generally, chooses what he does and how he does it. In other words, it is not possible for a novice to practice music, especially in the context of improvisation. On the other hand, the one who already practices an instrument must restart a complete learning if he wishes to practice a different instrument. There is, however, a very large number of people who do not practice an instrument (or even do not practice music at all) who would like to practice it, usually for lack of time or personal investment, and who would probably be relevant in this area. practical because having one of the qualities necessary to achieve it, namely the sense of rhythm, the rhythm being, from the point of view of some theoreticians of music, its founding element.

The present invention addresses these problems and limitations and proposes a method and devices having the advantage of allowing any person, with or without musical qualifications, to improvise on a musical background, and this in harmony with the musical background, as can the to make qualified instrumentalists on their instrument. The invention relates, according to a first aspect, to a method of musical improvisation, characterized in that it comprises, by an information processing device, a step of receiving a clock generated by a generating circuit. clock built into the device or by an external device; a step of defining a measurement (encryption); a step of generating a varying tempo according to said clock signal and said measurement (encryption); a step of receiving a start signal of note or start of group of notes; a step of defining by an automatic music generator a note or a group of notes played sequentially according to a set of composition rules according to the moment of reception of said start signal relative to said tempo and said measure (encryption); a step of generating an audio signal representative of the respective note pitch of said note or notes of said group of notes; an immediate play step of the audio signal generated upon receipt of said start signal; a step of receiving an end of note signal or end of group of notes stopping the play of said generated audio signal. According to a particular embodiment of the invention, the method further comprises a step of receiving at least one other beginning signal of note or beginning of group of notes, each signal being assigned to a part of the range and triggering the definition and then playing a note or a group of notes in this reduced range. According to a particular embodiment of the invention, the method further comprises a step of receiving at least one other beginning signal of note or beginning of group of notes which triggers the generation of a note or a group of notes at a higher pitch than the previous note; or at the same height as the previous note; or at a lower height than the previous note. Thanks to these provisions, the user has the means to act on the one hand on the height of the notes he improvises and on the other hand on the sequence of these heights when he improvises a sequence of notes, it that is, the relative height of a note relative to the one preceding it. According to a second aspect, the invention relates to a method of musical improvisation, characterized in that it comprises, by an information processing device, a step of transmitting an audible signal of musical type representative of the tempo and its variations, the measurement (quantification) and its changes. According to a particular embodiment of the invention, the method further comprises a step of reading in a memory space of said device of said musical signal. According to a particular embodiment of the invention, the method further comprises a step of generating in real time said musical signal by the automatic music generator according to the tempo and the measurement (encryption). According to a third aspect, the invention relates to a method of musical improvisation, characterized in that it comprises, by an information processing device, a step of defining at least one sequence of chords compatible harmonically with said musical signal, each chord being associated with a scale, which makes it possible to share the pitch of the notes of each scale into three families of notes, the first family being made up of the notes common to the scale and the associated chord, the second family being composed not only of the other notes of the scale, the third family consisting of the notes of the chromatic scale amputated notes of the range associated with said agreement; an association step according to said tempo, the measurement (encryption), the position and the duration of the notes played previously, the membership of a potential moment of note to play to one of the first two families of notes, the position of said potential note instant to play being quantized and rounded according to the temporal resolution of the associated automatic music generator. According to a particular embodiment of the invention, the method further comprises a step of determining the maximum difference between the pitch of the note to be played and the height of the preceding note played according to the family to which belongs the moment to play and the family to which belongs the previous note played. Thanks to these arrangements, the melody improvised by the user is fair, coherent and melodious with respect to the musical background serving as harmonic and rhythmic support to said user. It is also of a great variety. According to a particular embodiment of the invention, the method further comprises a step of replacing a note played, if the note played is held during a change of agreement, and if this note does not belong to the first family of this new agreement, by the nearest note being part of the first family of the said 35 new agreement. According to a particular embodiment of the invention, the method further comprises a step of automatically replacing a played note which does not belong to the first family of the current chord and which is held too long by the note. the closest being part of the first family of the current agreement. Thanks to these arrangements, the melody improvised by the user remains fair and pleasant to the ear compared to the harmony of the background music.

According to a particular embodiment of the invention, the method further comprises a step of receiving a signal which, if it is received in conjunction with one of the start of note signals, triggers the generation of a second note, this second note being generated a sixteenth note after the note generated following the reception of said start signal and after stopping the game of this previous note.

According to a particular embodiment of the invention, the new note created is not separated from said previous note by more than two degrees, a degree designating any one of the notes of the range associated with the current chord. Thanks to these provisions, the user can improvise a very dense melody in the number of notes played without having to acquire a great dexterity and while obtaining a melody of a great fluidity. According to a particular embodiment of the invention, the method further comprises a step of choosing, when said group of notes is a pattern extracted from a library of patterns, the height of the first note so that at least half of the notes of said pattern are notes of the first family.

Thanks to these provisions, the user can introduce into his improvisation patterns usually played by musicians who will always be fair and pleasant to the ear. According to a fourth aspect, the invention relates to a method of musical improvisation, characterized in that the generation of the audio signal representative of the respective note pitch of said note or notes of said group of notes comprises, by a processing device of information, a step of reading and processing representative audio samples of a simulated instrument. According to a particular embodiment of the invention, the generation of the audio signal representative of the respective note pitch of said note or notes of said group of notes comprises a step of mathematical modeling of a simulated instrument. According to a particular embodiment of the invention, the generation of the audio signal representative of the respective note pitch of said note or notes of said group of notes comprises an additive or subtractive synthesis step. According to a fifth aspect, the invention relates to a computer program comprising instructions adapted to implement each of the steps of the method of the invention as succinctly set forth above when said program is run on a computer.

According to a sixth aspect, the invention relates to an information storage means, removable or not, partially or completely readable by a computer or a microprocessor comprising code instructions of a computer program for the execution of each of the steps of the process of the invention as succinctly set forth above. According to a seventh aspect, the invention relates to a musical instrument in the form of an information processing device, characterized in that it comprises means for receiving a clock generated by a generating circuit. clock built into the device or by an external device; means for defining a measurement (encryption); means for generating a tempo that can vary according to said clock signal and said measurement (encryption); means for receiving a start of note signal or start of group of notes; means for defining by an automatic music generator a note or a group of notes played sequentially according to a set of composition rules as a function of the moment of reception of said start signal relative to said tempo and said measure (encryption); means for generating an audio signal representative of the respective note pitch of said note or notes of said group of notes; means for immediate play of the audio signal generated on reception of said start signal; means for receiving an end of note signal or end of group of notes stopping the play of said generated audio signal. The characteristics of the invention mentioned above, as well as others, will emerge more clearly on reading the following description of an exemplary embodiment, said description being given in relation to the attached drawings, among which: FIG. Fig.1 shows an advantageous embodiment of the musical improvisation device; FIG. 2 shows the diagram of a musical improvisation device according to the invention in which the musical backgrounds are stored in the form of encoded musical data; FIG. 3 shows the diagram of a musical improvisation device according to the invention in which the musical backgrounds are stored in the form of either encoded musical data, or in the form of encoded or unencrypted digital audio data; Fig. 4 shows how synchronization is possible through the use of temporal and harmonic data; FIG. 5 shows how the notes are represented in FIGS. 6 and 7; Fig. 6 shows how, in the process of musical improvisation, the musical background and the improvising software are synchronized but out of phase, the resolution being black; Figure 7 shows how, in the process of musical improvisation, the musical background and the improvising software are synchronized but out of phase, the resolution being the sixteenth note; FIG. 8 shows how the method classifies notes into 3 families; Fig. 9 shows how the process associates tessitura with game keys; FIG. 10 shows an example of associating the tessitaries with the game keys FIG. 11 shows how the process attributes its family to a new note; Fig. 12 shows how the process distributes the intervals between the 1 and 2 families; Fig. 13 shows how the method evaluates the possible note pitches - Fig. 14 shows how the process generates sixteenth notes automatically; Fig. 15 shows how the process generates a binary pattern of four sixteenth notes at the correct height; Fig. 16 shows how the process can change the pitch of a binary pattern by four sixteenth notes in a chord change. A first embodiment of the invention consists of a musical instrument containing a set of musical backgrounds. Each musical background is associated with a temporal description stating: - its measure (encipherment) and its tempo as well as any changes of measure (encryption) and / or tempo; a harmonic description stating the possible chord (s), their associated range (s) for improvising, and their temporal position in the musical background (commonly referred to as a grid) of agreements) expressed in measures and divisions of the measurement according to one or more measures (encryption) specified. The user chooses a musical background, the instrument having the technical means necessary to broadcast this musical background. The user can then play notes in harmony with the musical background by means of an interface that changes the state of at least one switch. It should be noted that the interface may be the switch itself. The musical instrument has a synthesizer of sounds, allowing to play in real time the notes that have been composed. The user decides when he plays his notes and the duration of these notes, as can an improvisational musician on a musical instrument. He thus has the possibility of constructing a melodic musical discourse. A note is composed and played by the instrument as soon as a user action on the interface changes the state of the switch, and lasts as long as another action of the user on the interface does not bring back switch to its original state, unless the sound used by the synthesizer to play the notes played by the user, referred to as the solo sound, has a limited duration and stops before the user change the state of the switch, depending on the type of synthesis and the type of sound used. To compose the notes, the instrument has an improvisers software perfectly synchronized on the musical background, synchronized meaning here that the software has the full knowledge of the precise position where it is in the score of the musical background at the moment when the The switch changes state, and therefore indicates to him that he must improvise, and this thanks to the temporal description of the musical signal, this position being expressed in terms of measurement (encryption) of measurement, time and division of time. It also knows the position and duration of the previous notes played, since these notes were played using the same principle. He deduces from his current position what chords and associated scales he can use. Based on all this information and musical rules, he can compose a new note and build a melody that will appear stable and pleasant to the ear. The stability is ensured by the fact that the improvising software declares, for a given musical background, that the notes played in certain particular positions of the measure are notes exclusively from the current chord.

These particular positions may be related to the measure (ciphering) and the tempo of the musical background, and are identical for several measures that follow each other, so as to maintain a regularity of the presence of the notes of the agreement in these places which participates in the stability of the melody during this duration. For the same musical background, the improvising software can also modify these positions for a certain duration, the positions then being identical to all the measures of this duration to maintain this new regularity. Stability is also ensured by the fact that when a generated note does not belong to the chord, ie is a note of the range associated with the chord range except those of chord , or even a note external to this range, then the one that follows, except special cases due to specific algorithms (for example the addition of ornamentations or altered notes, succession sixteventh or seventh-sixth), belongs to this agreement .

On the other hand, when a note is generated which does not belong to the chord, and if this note lasts too long, this note can be replaced after a certain time by a note of the chord close to this note to replace. Finally, when a note is generated before a chord change, and if the note is not a note of the new chord, then the note is replaced at the time of chord change by a note of the new chord near that chord. note to be replaced. In a preferred embodiment, the interface of the instrument enables one of three additional switches to trigger the improvisation of a new note in a limited part of the range of the solo sound used for improvisation , each of the switches being linked to a part of the range, in this case the serious part of the range, the middle part of the range and the acute part of the range. It is understood that this arrangement is not essential, or can be reduced to sharing the range of the two-part solo sound, or on the contrary lead to a larger number of parts, for example four or five or more, each part then being connected to a switch. It can thus be seen that in this variant embodiment, the user has four different means for generating a new note and can direct his improvisation in different parts of the range of the solo sound. All of these means are called the game interface, and the sensors linked to the four switches the game keys, even if one can imagine that these sensors can generate different types of physical quantity (the sensor being the switch, or a pressure sensor, or a capacitive sensor, etc.) other than an on / off state characteristic of the switch. We call respectively the bass, medium and treble keys related respectively to improvisation in the bass, midrange, treble of the range. We call global touch the key related to improvisation over the entire range. In a preferred embodiment, the instrument has in addition to the game interface a so-called function interface for extending the playing possibilities of the instrument. This interface functions in the same way as the game interface, namely that a sensor causes the switch state to change. The sensors of the function interface are called function keys. The first variant of this embodiment requires the presence of the three keys bass, medium, acute. In this first variant, the prior activation of the so-called "direction" function key modifies the interpretation of the improvisational software when activating one of the low, medium and high keys. Thus, the activation of the bass key will produce the improvisation of a note below the previous one, unless the previous note played is already the lowest note that the solo sound can produce. The activation of the medium key will produce the improvisation of a note identical to the previous one, or else very close to the previous one if certain criteria of stability or accuracy must be fulfilled. Finally, the activation of the high key will produce the improvisation of a note located above the preceding one, unless the previous note played is already the highest note that can produce the solo sound. With this embodiment, the user has three additional means to generate a new note, and can master the melodic orientation that takes his improvisation. In a second variant of this preferred embodiment, the prior activation of the so-called "double note" function key modifies the operation of the improvising software when one of the game keys is activated. user to generate during his improvisation, twice as many notes as he actually plays, and this in rhythm, especially to avoid, when the musical background is of a fast tempo, that the melody generated is too choppy or rhythmically unstable, or simply to help it generate fast melodic strokes that it could not generate otherwise, unless a long and tedious training. For example, if the user wants to hear four notes per beat (that is, four sixteenth notes per beat if the song is 4/4), he can do this using this feature and playing two notes per time (ie two eighths per beat if the song is 4/4). The improvising software generates a first note following the same stability rules mentioned above, then a sixteenth note later a second note (after stopping the first one) that will last until the play key or the double note key is released. , following the same limitations on the length of sound mentioned above. This second note has the particularity that: - it is different from the first note; - the difference in pitch with the first note is not more than two degrees, a degree designating any of the notes in the range associated with the current chord (which in most cases consists of seven repeated notes of octave in octave) so as to maintain a melodic coherence when several of these double notes are linked together; - the "direction" function key has no influence in determining its height. However, it should be noted that, in the case of the joint use of the first variant of this embodiment, the "direction" function key retains its influence on the first of the two notes. Similarly, this first note is concerned with the stability criteria. In a third variant of this preferred embodiment, two additional function keys make it possible to generate repetitively patterns or stored in memory, or built from notes previously played by the user. For example, in a measurement (encryption) in x / 4 (for example 4/4), the "binary pattern" key is used to repeat a pattern of four sixteenth notes, while the "ternary pattern" key is used to repeat patterns. three eighths triplets or six sextolets. Thus, the prior activation of the so-called "binary pattern" function key modifies the operation of the improvising software when one of the game keys is activated. The improvising software, instead of composing a note, chooses a pattern in its pattern bank, and plays the first note of this pattern, then successively the following notes at the tempo rhythm and thanks to the synchronization, stopping each time the previous note. When he has played the fourth note, he plays the first, and so on. Different criteria of stability are taken into account when choosing the height of the first note of the pattern, the most important being that at least half of the notes of the pattern are part of the notes of the agreement. Thus, when the musical background changes agreement, this last criterion is reevaluated, and the improvising software possibly changes the pitch of the first note, as well as that of the following notes in proportion and so that each note of the pattern is just relatively to the range associated with the chord, the criterion then being filled again. The pattern is thus globally transposed. It should be noted that, in the case of the joint use of the first variant of this embodiment, the function key "direction" retains its influence in the choice of the first note of the pattern, and subsequently in the modification the overall height of the pattern, as will be explained in the detailed description. It should be noted that the principle is the same for ternary reasons. It would also be the same if the binary patterns were for example 8 semiquavers, the patterns then being repeated every other time. In the same way, the principle would be the same if the ternary motifs consisted for example of six eighth note triplets or 12 sextoles, the patterns then being repeated every two times. The musical instrument must necessarily have an interface to sensors to change the state of the switches. The sensors can be the switches themselves. Or else pressure sensors, capacitive sensors, inductive, etc ... the action on the sensor delivering a signal that can be quantified on different values, with two thresholds, one to trigger the beginnings and the other note-ends, these values can also be used to describe for example the velocity, or any other parameter having an influence on the improvisation or the texture of the solo sound played.

In addition to the interface, the musical instrument must also include a processor system associated with RAM and ROM or programmable. The musical backgrounds and their temporal and harmonic descriptions are stored in the read-only memory or programmable memory and can be in various forms: audio-digital data (for example AIFF format, WAV, etc.); - encoded audio-digital data (for example mp3 format ...). The musical instrument must then have a hardware or software decoder; - encoded musical data, using a format such as the MIDI (Musical Instrument Digital Interface) format. The musical instrument must then have a multi-channel sound synthesizer and multi-timbral, hardware or software type. Different types of synthesis are possible: based on samples, additive, subtractive, modeling, etc ... or a mixture of these different syntheses.

The solo sound also requires the use of a sound synthesizer, hardware or software, whether it is based on samples, additive synthesis, subtraction, modeling, etc ... The improvisation software can come modify the synthesis parameters in real time, so during the game of the notes, and in particular in rhythmic synchronization with the musical background, so as to make the solo sound more realistic (the solo sound can indeed simulate a known instrument, like the piano , guitar, flute, etc.) and / or more expressive. The processor must also have a random number generator, because many decisions internal to the algorithms depend on random numbers.

We can notice that the musical background is not essential for some applications. In this case, the user knows the tempo and the measurement (encryption) by the use of light signals generated for example by means of an electroluminescent diode, or else by the addition of an audio signal. metronomic at the signal of the solo sound.

On the other hand, it is easily understood that another embodiment of the invention consists of a musical instrument also having an interface for receiving a synchronization signal, for example by using the MIDI protocol (the midi protocol allows the sending of messages of clock, tempo, measurement (encryption), beginning of measurement). In this case, the instrument can easily be synchronized with a sequencer software running on a microprocessor system, for example a computer or a synthesizer incorporating such a sequencer, as it exists on the market. We also understand that such software can send either beforehand or in real time the harmonic description of a musical background that would be played by musicians, using the same interface and the same MIDI protocol, the musicians being themselves same "synchronized" on the sequencer, as is common today. In this case, on the one hand, the instrument no longer depends solely on its own musical backgrounds, and on the other hand the user, without being a musician, can play in a collective context with musicians. Finally, it is easily understood that another embodiment of the invention consists of two musical instruments having an interface for transmitting and receiving a synchronization signal, for example using the MIDI protocol. In this case, one of the two instruments is chosen as master, the other as a slave, and as in the paragraph above, the master instrument sends to the slave instrument, either beforehand or in real time , the harmonic description of the musical background that it diffuses, the slave diffusing in this case only the signal of its solo instrument. Such an embodiment allows the two users to improvise together each on their instrument, the slave instrument providing a musical discourse that is right and consistent with the background broadcast by the master instrument through synchronization. The present invention is described in the following in more detail with reference to advantageous devices for understanding how the invention works. Different variations or substitutions of the constituent elements of the invention are possible leading to different devices but nevertheless based on the same method. It should be noted that in the following, we sometimes use the word composition instead of the word improvisation. First of all, it must be understood that the invention deals with a particular musical instrument which makes it possible to improvise a melody on a musical background broadcast by the instrument, and which is coupled with the services of a computer software. algorithmic improvisation to create said melody, which means that to function, ie to produce the melody, a device using the method of the invention requires interaction with a user, in the same way as in a concerto for piano played live, the piano will not produce any sound as long as the pianist does not play (ie does not interact with the piano), even if the accompanying orchestra plays. It is also fundamental that, when the user decides to play using the interface described below, a note is generated immediately, that is to say in real time, as the piano produces a sound immediately when the pianist press one of the keys on the keyboard of that piano. It is necessary, moreover, for the device using the invention is pleasant to use, that this note is just harmonically and even pleasant to the ear compared to the musical background played by said device.

It is first of all useful to describe the interface between the user and the method of the invention used by the advantageous device. FIG. 1 shows that in this device the interface consists of eight switches 1 to 8, also called keys, whose state changes (released or pressed) will trigger an action of the improvisation process. These eight keys are of two types: - keys 1 to 4 are the game keys, which are always used to trigger the composition and play of a note or group of notes; keys 5 to 8 are the function keys, which are used prior to pressing one of the game keys, and which make it possible to use algorithms specific to the composition process. The operation of the game keys is as follows: the pressing of a key triggers the composition of a note and then its game. This note is played (that is to say we can hear it) as long as the key remains depressed unless the sound used for the melody instrument is a short one, or the sound synthesis used does not allow long sounds. As soon as button 1 is released, the sound stops, unless the sound used is a long-release type (ie the sound will stop slowly decreasing). To compose a series of notes in order to build a melody, you must successively press and release one or more of the four game keys 1 to 4 consecutively.

The 1 key is used to compose notes throughout the range defined for the sound of the melody. The keys 2, 3 and 4 have a behavior that varies depending on whether the key 7 is pressed or not. If the 7 key is not pressed, then: - The 2 key is used to compose notes at the top of the range defined for the sound of the melody (high notes); - The 3 key is used to compose notes in the middle of the range defined for the sound of the melody (medium notes); - The 4 key is used to compose notes at the bottom of the range defined for the sound of the melody (low notes). If the 7 key is pressed, then: - The 2 key is used to compose notes such that any composed note is sharper than the previous note played, as long as the composer does not exceed the highest note that the instrument of the melody can generate. If he has arrived at this note, then he repeats it, or modifies it slightly so that it remains just in harmony when an agreement changes, according to the harmonic constraints related to the new agreement; - The 3 key is used to compose notes such that any composed note is at the same height as the previous note played, or slightly different (higher or lower) so that it stays in harmony when a chord changes according to the harmonic constraints of the new agreement; - The 4 key is used to compose notes such that any composed note is more serious than the previous note played, as long as the composer has not reached the lowest note that the instrument of the melody can generate. If he has arrived at this note, then he repeats it, or modifies it slightly so that it stays in harmony just when a chord changes, according to the harmonic constraints of the new chord. A prior press on the key 5 allows, when pressing one of the four game keys 1 to 4, to generate a second note, which will be called automatic note, a sixteenth note after the note generated by the press on said game key, and this after having stopped this first note. The distance between the beginning of the two notes, a sixteenth note, is therefore dependent on the tempo. In other words, the first note still has a sixteenth note. In addition, the duration of the second note lasts as long as one of the two keys 5 or the game key 1 to 4 is pressed, but stops as soon as a new play key is pressed. Hold key 5 pressed, and press any one of the game keys 1 to 4 at the eighth note, to generate a melody composed of a series of sixteenth notes. When keys 5 and 7 are used together, the behavior of the key 7 described above only applies to the notes generated by the pressing of the game keys 1 to 4, and not to the automatic notes. The algorithmic operation of the method for the generation of these automatic notes will be described later. A prior press on the key 8 makes it possible, when pressing one of the four game keys 1 to 4, to generate and repeat a pattern of four sixteenth notes, a pattern that can: - be chosen from a library of process motifs ; 30 - Be built from notes previously played by the user. The pattern is repeated as long as key 8 is pressed. As long as key 8 is pressed, and in addition key 7 is pressed together: - at play key 2, then the pattern will be repeated while ascending; - to the game key 4, then the pattern will be repeated downhill. As will be explained later in the description of the algorithmic operation of the method for the generation of these patterns. When the key 8 is released, then the pattern stops, and ends on a note whose height depends on different conditions that will be explained later.

A prior press on the key 6 allows, when pressing one of the four game keys 1 to 4, to generate and repeat a pattern of three triplets eighth notes, a pattern that can: - Be chosen in a library of patterns of process ; - Be built from notes previously played by the user. The pattern is repeated as long as key 6 is pressed. As long as key 6 is pressed, and if in addition key 7 is pressed together: - at play key 2, then the pattern will be repeated while ascending; - to the game key 4, then the pattern will be repeated downhill. as will be explained later in the description of the algorithmic operation of the method for the generation of these patterns. When the key 6 is released, then the pattern stops, and ends on a note whose height depends on different conditions that will be explained later. It has been said above that, in the advantageous device relating to the invention, the musical instrument broadcasts a musical background on which the user can improvise using the interface which has just been described. Figures Fig.2 and Fig.3 show two possible block diagram representative of such a device according to the invention. In FIG. 2, the device generates musical backgrounds that are stored in the form of encoded musical data. The reference 9 comprises all the elements necessary for the development of the process when it is requested by the interface 19 consisting of a set of switches 20. The microprocessor 10 controlled by a clock 11 uses a ROM 12 in which are stored all the algorithms necessary for the proper functioning, all the musical data used to generate the musical backgrounds and to improvise the melody, as well as descriptors associated with the musical backgrounds. It also has a ram 13 for storing its temporary data, the whole being connected by a digital bus 14. The microprocessor 10 must internally have a timing circuit ("timer" in English language) controlled by the clock 11 which allows him, thanks to the information contained in the musical data of the musical background and associated descriptors, to generate and send at the right tempo the musical information of said musical background, generally in the form of midi messages (see standard Musical Instrument Digital Interface), the sound synthesizer 15, controlled by a clock 16. It should be noted that, according to the configurations, the clock 16 can be replaced by the clock 11 already mentioned. The sound synthesizer 15, upon receipt of the midi messages, and in accordance with the midi standard, generates and mixes said sounds in audio-digital form, the result of the mixing being sent to the digital-to-analog converter 17 controlled by the same clock 16. An amplifier 18 upgrades the signal from the converter 17, so as to be able to control a speaker, an audio headset, or any external device, and transform into an audible signal the analog electronic signal from the amplifier 17. Said circuit timer allows the improvisation process to position precisely in the "partition" of the musical background, a user intervention on the interface 19, thanks to the associated descriptors, then to generate a note of the melody which will also be sent by a midi message to the synthesizer 15, and then mixed by said synthesizer to the other midi data as explained above. It should be noted that often references 15, 17 and 18 are grouped in the same electronic component. There are even low cost electronic components grouping all the references 10, 12, 13, 14, 15, 17 and 18, the clock of the set then being the reference 11. It should be noted that, since the musical background is built solely from musical data and not from audio-digital data, there is no problem of synchronization between the audio-digital data generated by the synthesizer 15 and the temporal data manipulated by the micro-processor. processor 10 and this even if the clocks 11 and 16 are different because the information exchanges between the microprocessor 10 and the synthesizer 15 are made using high level messages (note message in the MIDI protocol) , not at the low digital audio level (ie, at the audio-digital sample scale).

In FIG. 3, the device generates musical backgrounds that can be stored or in the form of encoded musical data, or in the form of encoded or unencrypted audio-digital data. The reference 21 comprises all the elements necessary for the development of the method when it is requested by the interface 31 consisting of a set of switches 32. The microprocessor 22 controlled by a clock 23 uses a ROM 24 in which are stored all the algorithms necessary for the proper functioning, all the musical or audio-digital data and their associated descriptors used to generate the musical backgrounds and to improvise the melody. It also has a ram 25 for storing its temporary data, the whole being connected by a digital bus 26. Compared with FIG. 2, the microprocessor 22 must also have a sound synthesis software ( which therefore replaces the synthesizer 15 of FIG. 2), possibly audio-digital data decoding software if the music backgrounds are stored as encoded audio-digital data, and also a software for mixing audio-digital data. to mix all these sounds. The microprocessor 22 must internally have a timing circuit ("timer" in English language) controlled by the clock 23 which allows it, thanks to the information contained in the descriptors associated with the musical backgrounds: - if the background music are stored in the form of musical data: generating at the right tempo audio-digital data from the musical data of the musical background; to precisely position in the "partition" of the musical background, thanks to the associated descriptor, a user intervention on the interface 31. As in FIG. 2, the improvisation process generates a note of the melody which is then transformed into audio-digital data by the synthesizer software, data that is mixed with the audio-digital data of the musical background. The resulting mix is sent to the digital-to-analog converter 27 controlled by the clock 28, then to the amplifier 29 whose role is the same as that of the amplifier 18 of FIG. It should be noted that, according to the configurations, the clock 28 can be replaced by the clock 23 already mentioned. As is customary to do with commercially available audio-to-digital (D / A) conversion components, the D / A converter 27 is connected to the micorprocessor 22 via a dedicated serial link, and is master of transfers on this bus. Those familiar with this field will readily understand that these data transfers are managed in the microprocessor 22 by a DMA (Direct Memory Access) associated with a dual buffering system. It should be noted that: - since the musical background is built solely from musical data and not from audio-digital data, there is no problem of synchronization between the audio-digital data generated by the microprocessor 22 and the time data manipulated by said microprocessor 22, and this even if the clocks 23 and 28 are different. Indeed, these low-level audio-digital data are completely decoupled from the high-level musical and temporal information, elaborated from the musical data of the musical background; when the musical background is constructed from digital data stored in memory, and if the clocks 23 and 28 are physically different, then there is a synchronization problem between the reading of the audio-digital data by the microprocessor 22 and the temporal data manipulated by said microprocessor 22. This is due to the fact that this reading is performed at the rate of the clock 28, while this clock drifts with respect to the clock 23 which controls the timing circuit (timer) allowing generate the tempo.

It is therefore necessary that the times managed by the timing circuit of the microprocessor 22 controlled by the clock 23 can be modified in real time by the microprocessor 22 according to this drift so that the audio signal generated by the converter N / A 27 remains in phase with the positioning of the actions of the user on the interface 31 in the "partition" of the musical background. The microprocessor 22 has knowledge of this drift by the variation of the reading address of the buffer mentioned above read at regular times, instants defined by the timing circuit.

In the two configurations of FIGS. 2 and 3, the clock signal necessary for the generation of the tempo is the signal coming from the timing circuit integrated in the processor, the processor being also responsible for the broadcasting of the background music. , and especially the start of this broadcast. The simultaneous start of the broadcast and timing circuit (or the process using the timing circuit to scan the temporal description of the background music) allows the improvisation process to position precisely in the "score" of the musical background, a user intervention on the interfaces 19 or 31. When the clock signal necessary for the generation of the tempo is generated by an external device, for example in the two cases mentioned above where the device is synchronized with another device by a serial link (using for example the MIDI protocol), it is necessary for this clockwise information to pass through this link and at regular intervals, as well as start-up information (which is the case with the midi protocol). ). It is interesting to note that in the two devices shown in Fig.2 and Fig.3, the microprocessor can also have automatic music composition algorithms to generate its own music backgrounds. Figure Fig.4 shows an example set of descriptors associated with a musical background. The descriptor 33 indicates: the musical background consists of two parts, the first part being associated with the descriptor 34 and the second with the descriptor 35; - how to chain the two parts to play the musical background: in this example, part 1, then part 2, then part 1, then part 2, etc .... It is thus understood that the number of descriptors depends on the number of parts, and that the sequence of the different parts can become complex if the number of parts is important. In the same way, the musical background can consist of a single part, repeated or not. The descriptor 34 linked to the first part gives two types of information: - a temporal description indicating the measurement (encryption) at the beginning of the game and the times of any change of measure (encryption) during the game, as well as the tempo at the beginning of the game and the moments of change of tempo during the game. The moments of change of measure (encryption) or tempo are described in musical terms: measurement number, black number in the measure, eighth note number in the measure, etc ... until the smallest subdivision accepted by the process in this musical background; A harmonic description giving the name of the chords and their temporal position in this part, a position described in the same musical terms as those defined in the temporal description, as well as the mode associated with each chord. In this description, there is, in this first part, variations around a starting tempo, indicated at 128. This happens if the part in question is a part in audio-digital data recorded by musicians not playing the metronome. On the contrary, a part in musical data, thus elaborated from a compositional computer tool, what is commonly called a Midi sequencer, will probably have no change of tempo, unless the composer of this part has wanted to simulate what happens when musicians do not play the metronome.

The chord names correspond to the English notation (A = la, B = si, C-do, D = re, E = mi, F = fa, G = sol) and use here the notation used in jazz or popular music. : C, Am, G7, etc .... The enumerated modes are 7-note scales which define the 6 intervals between each of the notes of this scale, and make it possible to identify the harmonic function of the chord with which it is associated . Thus, the ionic mode is followed by intervals of 1 tone, 1 tone, V2 tone, 1 tone, 1 tone, 1 tone, the interval between the seventh note and the first note of the same scale played an octave higher being a semitone. When this mode is associated for example with the chord of C, it means that we play the chord of C in the key of C major, and we obtain the following sequence of notes: Ionian mode on a chord C: do The following are the examples: (1) 1 1/2 1/2 1 1 1 1/2 It can be seen that the second part 35 has a measurement change (encryption) to the measure 7, then a new change to 8. Fig. 5 shows how notes, here blacks, will be represented in Figures 6 and 7. Thus, in a measurement in 4/4, the 3 'black 36 is represented by the box 37 identified n3, the length of the box being representative of the time dependent on the tempo of said black. FIG. 6 shows how the time spent is managed and how the moments of action (support or release) on the game keys of the interface 19 of FIG. 2 or 31 of FIG. 3 are discretized and quantified when the measure (encryption) is 4/4 and the resolution of the improviser is black. We assume here that the black, whatever the tempo, is divided into 480 equal parts called tics. The duration of the tick therefore depends on the tempo and corresponds to the duration programmed in the timing circuit of the microprocessor 10 or 22, relative to its clock 11 or 23 respectively. Those who are familiar with the timing circuits will understand that it generates an interruption at each tick, which allows the process, once it has received the start command related to the launch of the background music: - to count the blacks, then the measures, then the parts; - to reprogram the duration of the tick to a different value if there is a change of tempo in the descriptor of the current game. and therefore to be synchronized with said musical background.

At 38, and for a 4/4 (4/4) measurement, the blacks n3 and n4 of the measurement n-1 M, 1_1 are represented at the tempo Tl, then the four blacks n1 to n4 of the measurement measure n Mn at tempo T2, and finally the 1st black n1 of the measure Mn +, again at the tempo Il. The tempo T2 is greater than the tempo T1, and therefore the duration in seconds of the negatives of the measures n-1 and n + 1 is greater than those of the measure n. All these blacks, however, have the same number of ticks, which shows that the duration of the tick depends on the tempo. The time diagram thus represented is that associated with the musical background. When a user activates one of the game keys, it is important to know when he or she did it in relation to that time diagram. In particular, if the user wants to play a black, for example the black n4 of the measure Mn_1, it is likely that it will not operate one of the game keys at the moment representing the separation between the black n3 and the black n4 of measure Mn_1, but rather a little before or a little later. It is therefore necessary to discretize these moments of support: the discretization is done at the separation between two blacks of the time diagram 38, and the quantification is related to the resolution of the improvising software, here the black one. Thus, at the user level, which is also that of the improvising software, will be considered as black n4 of the measure M, any time belonging to the zone located a forward eighth (ie half of a black) and a quaver after said separation. The time diagram 39 is obtained, according to the point of view of the improvising software, which is therefore one step ahead of that of the background music 38, that is to say half the resolution of the improvising software. . It should be noted that at tempo changes, the durations before and after the separation are different, even if they contain the same number of tics (the duration of the tick depends on the tempo, the number of tics per time - namely the black in 4/4 - being on the other hand constant). FIG. 7 is identical to FIG. 6, except that the resolution of the improvising software is sixteenth note. The measure (encipherment) is in 4/4, and the four semiquavers of the same black n, are denoted dl to d4, each sixteenth note being divided into 120 ticks. The discretization is done at the boundaries of the sixteenth notes of the time diagram 40. The time diagram 41, which is therefore the point of view of the improvisational software, is now three times ahead of that of the background music 40. here is interesting to note that the improvisation process is dependent by definition of the moment when the user chooses to play a new note. It is therefore not possible for him to browse sequentially melodies previously recorded and stored in memory, a principle widely used in the tools of automatic musical composition. On the contrary, it must be written note by note and in real time. He does not have the knowledge of the future to build his melodies, he only has the knowledge of what has already been played, so of the past, that is to say of the position and the height of the preceding notes relative to the temporal and harmonic descriptions of the playing part of the musical background, as well as the position of the new note to play. The musical theory and the listening of many improvisers 'recordings (jazz field mainly) also teach us that, on a given chord associated with a given mode, we can not play any note from the mode to n' any moment. It is interesting to separate the 12 available notes from the Western chromatic scale into three different parts: - the notes of the chord, which we will call the 1st family note; - the notes of the mode associated with the chord not included in the chord, which will be referred to as the notes of the 2nd family - the notes of the chromatic scale not included in the mode, which will be described as a note of the 3 "family or external notes. Figure Fig.8 shows for several agreements how are distributed the notes of different families. It should also be noted that each note of a mode is called degree, the 1st note of the mode being the degree, the 2nd note the 2nd degree, etc ... In the following, and for simplification, the degree will be the reference unit, which eliminates notions of tone or 1/2 tone. For example, in the figure Fig.8, the difference between the 3 'and the LI' degree of the agreement C is of 1/2 tone, and of 1 tone for the agreement Dm, whereas expressed in degree , the difference is 1 degree for both chords.

Thus, playing only notes of the first family on a chord poses no problem of stability. The disadvantage is that the melody may seem poor. On the other hand, playing only the notes of the second family on a chord will produce a feeling of imbalance, because the notes of the chord are never mentioned. Finally, playing only external notes on an agreement will seem false. Mixing the notes of the first and second families, to which we add some external but short notes, near notes of the other two families, will produce a rich and balanced melody if the notes of the second family are judiciously positioned in relation to that from the first family.

FIG. 9 shows, by way of example, how the method can evaluate the range associated with the four keys 1, 2, 3 and 4. In this example, the range of the solo sound ranges from la3 to 1a6. On a given current chord, these two extreme scores can be evaluated as the smallest possible degree deg min and the highest possible degree deg max. It should be noted that, according to the constraints imposed by the harmonic description exposed in figure Fig.4, deg min (respectively deg_max) can correspond to la3 # (respectively lab6) if la3 (respectively la6) does not correspond to a degree the mode associated with the current chord. In the figure Fig.9, the expression [0-3] means "a number between 0 and 3 drawn at random, each of the numbers having the same weight". Each time you press a play key, the tessitaries associated with the three keys 2, 3 and 4 are re-evaluated: thus, the boundaries between the three zones with reduced range (keys 2, 3 and 4) vary, the said zones can therefore be cover from one support to the other. This allows for greater diversity in the choice of notes during improvisation. By way of example, FIG. 10 shows concretely how the boundaries of the range associated with the key 3 change during three successive presses on the key 3. Here, the degrees do not go above deg max-3 nor below deg_min + 4, which allows, when the key 7 is activated, to exceed these borders to go deg min or deg_max depending on the direction chosen, so that the user, if it reaches these limits, has the sensation of going down or climbing higher than when it does not use the key 7. Figure Fig.11 shows by way of example how it is possible, depending on the moment of support on a touch of play then that of its relaxation, to attribute one of the first two families to a note to improvise, and this from the previous remarks on the notion of stability of a musical phrase. The whole process uses a random number generator which, depending on the value of one or more thresholds, chosen in advance and which can depend on musical parameters, decide to belong to one of these first two families, so that to ensure a variety in the choice of this membership. It is assumed here that the measurement (encryption) of the musical background is in 4/4 and that the resolution of the improvisation process is the sixteenth note, as it has been shown in Fig.7. Thus, each measure is divided into 16 positions of sixteenths dl to d16. The first step 42 specifies that the first improvised note in an agreement will definitively be from the 1st family, so a note of the chord suggesting this chord. The second step 43 introduces the notion of stability point: if a note is played in this position, then this note will necessarily and definitively of the 1st family, that is to say a note of the chord. This allows you to hear the notes of the chord at specific and regular times. For example, if the stability points are in: - dl, d5, d9 and d13; the four positions of black ni, n2, n3 and n4 (see figure Fig.5) of each measure are chosen in the 1st family when they are played; - dl and d9: the 2 white positions (n1 and n3) of each measurement are chosen in the 1st family; - d5 and d13: the 2 positions of black in the air n2 and n4 of each measurement are chosen in the 1st family when they are played; - d3 and dl 1: the eighth notes in the air of the blacks n1 and n3 of each measure are chosen in the family when they are played; - dl to d16: all the notes played are from the 1st family.

These points of stability can vary from one musical background to another, or to each repetition of a musical background, or to each part of a musical background, or even within a part of a musical background. They can also be determined in real time by analyzing the game of the user by the method. If the note played is in the eighth-note position d2, d4, d6, ..., d14, d16, the third step 44 temporarily assigns it to the family, if not temporarily the first family. Here, 45 provides stability by modifying the family assigned by step 44 according to the family of the preceding note. Thus, if the previous note was not from the 1st family, then the new note is from the 1st family. If not, and according to the value of a threshold threshold, the process assigns the new note either the 1st family, or the 2nd family, and then step 46 places the new note in one of three categories: 1st category where the distance at the beginning of the preceding note is large and greater than a threshold threshold12, and provided that the preceding note is of the 1st family - the 2nd category where the distance at the beginning of the preceding note is small and less than a threshold threshold13, with threshold3 <threshold2 - the 3 'category for notes whose distance at the beginning of the previous note is between the two thresholds. Finally, step 47 definitively fixes the family of the new note: - if the grade is of the 1 category: the family is fixed according to the distance of the new note at the end of the previous note. The greater this distance, the more likely the new note is to be in the 1st family, and conversely, the smaller the distance, and the newer grade is likely to be from the 2nd family; - if the grade is in the second category: if the previous grade is from the 1st family, the new grade may, following a random draw, be equal to the previous one. Otherwise, if the previous note is a "me degree," then the new can, following a random draw, be a 7 "degree, and conversely, if the previous note is a 7" degree, then the new may, following a random draw, be a 6 "degree; - if the grade is from the 3" category: his family is not changed. The end of the previous note is known because the moment of release of each of the game keys is known. Fig. 12 shows how the process attributes to the new note a maximum interval Ai-j with the previous note, and this according to the families i and j of each of these two notes. It is recalled that these two notes are played via the game keys. The interval can be positive or negative, unless the function key 7 is activated. In this case, only the interval concerned is preserved. The maximum deviations, for example A2- 1 of value +/- 3 degrees for a passage from the 2 "family to the 1 st family, are given as an example and may change depending on the part of the musical background, the musical background In this example, this means that the new note is located within plus or minus three degrees of the previous note, as shown in Fig. 13, which illustrates this example. In degree 6, which is supposed to be a degree of the family, that of the new note will be of the first family. We suppose that the degrees of the first family are degl, deg3 and deg5. The possible degrees are deg3 and deg5 of the same octave as deg6 (octave 3), and deg 1 of the octave higher (octave 4), if we keep the same values as in the figure Fig.12. The choice of the note is made by a random draw. However, each of the degrees may have a weight different from that of the other degrees, depending on whether at this moment (or for the musical background as a whole), the method wishes to privilege one degree with respect to the others, or to privilege a direction relative to to the other. In the case where the function key 7 is activated, the mode "direction" of the process is then active, and if the game key activated is the key 2, then the only possible degree is the degree degl of the octave 4. Conversely, in the case where the function key 7 is activated, and if the activated play key is the key 4, then the two degrees possible are degrees deg3 and deg5 of the octave 3. A first remarkable characteristic of the process is that any degree of each mode can be classified as a degree to be avoided. Such a degree will usually be from the second family. For example, the musicians are well aware that the fa in a major C major C major (that is deg4), is a note to avoid.

Thus, the method, rather than prohibiting such a degree, makes it possible to play it, but replaces it by a degree close to the first family (deg3 and deg5 in the example of the fa) of two possible different ways: the duration of said degree is greater than a duration which can be parameterized, in particular making this duration inversely proportional to the duration of the time (for example, eighth note on a slow tempo or a white on a very fast tempo), c that is, proportional to the tempo; if the degree is still played at the position of the next black, or according to the tempo, at the next white which follows the play of the said degree. and this thanks to the timing circuit which makes it possible to memorize the durations and to know the positions of the notes with respect to the musical background. A second remarkable characteristic of the method, also due to the timing circuit associated with the temporal description making it possible to know the positions of the notes with respect to the musical background, is that any note during the game is evaluated during a change of chord, that is to say, it is assigned a degree in the new agreement. If this degree is from the 2nd family in the new chord, or if the grade is outside the range associated with the new chord, then he (she) is replaced by the nearest degree of the 1st family of the new chord . FIG. 14 shows the operation of the second variant of the preferred embodiment, linked to the prior activation of the "double note" function key. In this embodiment, the prior press on the key 5 makes it possible to generate an automatic note a sixteenth note after a note generated by one of the game keys, thanks to the timing circuit mentioned in FIGS. 2 and 3 using the temporal description described in Fig. 4, as well as the division of each sixteenth note into a large number of ticks (Fig. The use of this mode recurrently, that is to say when the user generates notes via the game keys at the eighth note, provides a list of sixteenth notes. This is clearly explained figure Fig.14 by showing a concrete example, when the measurement (encryption) is in 4/4. Time diagram 49 shows a sequence of 2 blacks n1 and n2, divided into four 16ths to d4 each. The user's point of view is shown here (which corresponds to the time diagram 41 of FIG. 7). We see that there is a change of agreement in nl dl, and that there is a point of stability in n2d3: the notes that will be played in these places will be of the 1st family. The timescale and the solid arrows in solid lines indicate the moments when the user presses a play key, and by dashed arrows the moments when the process generates the automatic sixteenth notes. It can be seen that the solid line arrows are not regularly spaced because of the irregularity of the user, whereas the dashed arrows representing the automatic notes are always spaced one eighth note from the solid line preceding them , thanks to the cutting of each sixteenth note in a large number of ticks (Figure Fig.7). Enumeration 59 indicates the degrees of the 1st and 2nd families of the current chord. Enumeration 58 shows how the height of an automatic double note is assigned: it is calculated in relation to the degree of the preceding note, by adding or subtracting 1 or 2 degrees. The choice is made by random draw. Percentages of chance of getting a particular interval can be changed from one musical background to another, or from one part of a musical background to another, or within the same part, following musical rules . We see that in 50, the note was given a degree of the first family (deg3) because there is a change of agreement. At 56, the degree (deg5) is also of the first family, since n2d3 is a point of stability. It can be seen that in 52 and 53, two notes of the second family follow each other, and that in 54, according to what has been explained in Fig. 11, the note is of the 'family. Finally, it is easily understood that the so-called "direction" function of the first variant of the preferred embodiment has no influence on the automatically generated notes, the height of which depends only on the previous note as seen in FIG. Fig.14. The notes affected by the "direction" function are the notes played by the user, and thus in figure Fig.14, the notes of the time diagram marked by the vertical arrows in full line. FIG. 15 shows the operation of the third variant of the embodiment to be preferred, allowing the start and then the repetition of a binary pattern of four sixteenth notes. In this embodiment, pressing the key 8 before pressing one of the game keys indicates to the method that it must choose a bit pattern in its bit pattern bank and play it according to the constraints related to the current agreement and the motive itself. It is understood in 60 that a pattern is defined as a sequence of intervals expressed in degrees: the first digit indicates the interval between the 1st and the 2nd degree, the 2nd digit the interval between the 2nd and 3rd degree , etc. The timing diagram 61 shows a sequence of 2 blacks n1 and n2, divided into four 16ths d1 to d4 each. The user's point of view is shown here (see time diagram 41 in Fig. 7). The time diagram 61 and the solid vertical arrow indicate the moment when the user presses a play key to start a pattern, and the dotted arrows the moments when the process automatically generates the duplicates. eighth notes calculated from the sequence of intervals of the pattern. The dashed arrows representing the automatic notes show that these notes, thanks to the timing circuit and the cutting of each sixteenth note in a large number of ticks (Figure Fig.7), are always spaced a sixteenth note exactly from the arrow which precedes them. To humanize the game of this binary pattern, it will suffice to shift a set of sixteenths of the pattern from a small number of ticks around the starting position, this shift being chosen randomly for each sixteenth note of the pattern. Line 62 shows how the sequence of intervals is applied following degrees, the degree of departure being chosen here as the 1st degree degl to better understand. Before choosing the starting degree of the pattern, if it is not imposed by the pattern itself, the method must ensure the stability of the pattern with respect to the current chord for each of the starting degrees possible: this stability is ensured for a given starting degree as long as the pattern contains at least as many degrees of the 1st family as the 2nd family. In Fig. 15, we give in 63 the distribution between the two families of the seven degrees of the mode associated with the chord, and in 64 the representation of the pattern for each of the seven degrees of the mode taken as the starting degree. The degrees of the family are enumerated in italics. We thus note that if the degree of the motive is a degree of the family, it can only be deg5, and that the process has the choice between three degrees of the second family as a starting point, which may be chosen. as explained in Fig. 11, unless the motive imposes its own degree of departure. If the "direction" mode is used when launching the pattern, and if for example, the user wants to generate a higher pitched pattern via the game key 2, then the process will choose a starting level that will necessarily be above the previously played, unless in this case one of the degrees of the pattern exceeds the maximum degree allowed in the current chord. Fig. 16 shows how the process re-evaluates the pattern during a chord change. The pattern is re-evaluated when the first degree turn of the pattern to be played comes. To understand it well, we show in 65 the chord on which the starting degree of the pattern was chosen. This is a major C accent associated with an Ionian mode. In 66, it is shown that the process has chosen deg2 as the starting degree, which gives the sequence of notes "mi mi mi" on the agreement defined in 65. The method has the particularity that when a pattern has started, it keeps this pattern at the same pitch as perceived by the listener when chord changes (ie if the pattern is do mi mi fa on a chord, it will still be d re mi fa on the following agreement), unless the stability criterion in the new agreement is not fulfilled, in which case it will seek to move the pattern up or down to make it stable. We show in 67 the new chord Dm, and in 68 how the notes of the pattern as built on the preceding chord 65 are translated into degrees of the new chord 67. The stability criterion is not fulfilled since the number of degrees of the chord The first family of the pattern in this new chord 68 is smaller than the number of degrees of the second family of that pattern. The method then determines, as in FIG. 15, which are the closest first degrees that fulfill the stability criterion, a possible result being given at 69. When a pattern has been launched, and the mode "direction" of the process is activated (key 7 pressed) at the same time as one of the two keys 2 (towards acute) or 4 (toward the bass), the motive is then transposed by one degree in the direction chosen and in its entirety with each new occurrence of the pattern, that is to say all the blacks in the case of a binary pattern of four sixteenth notes, without checking the criterion of stability. In fact, in this case, if the pattern is out of balance (less than the 1st family than the 2 'family), this imbalance will be temporary, since it will only last one black, and the next occurrence will be likely to meet the criterion. If, on the other hand, one leaves the direction mode, the motive being always played, the motive is then again evaluated, and will be able to be transposed towards the nearest stable degree if it is out of equilibrium.

The process, thanks to the timing circuit and the cutting of each sixteenth note in ticks, allows to chain patterns between them, the patterns can be chained only when they are repeated. If there is such a request for one of the two keys 6 or 8, proceed as in FIG. 15, the last degree of the preceding pattern serving as a basis for calculating the degree of the new pattern by following FIGS. , Fig.12 and Fig.13, unless the new motive imposes his family and his first degree. The pattern stops as soon as the key (6 or 8) that launched it is no longer active when it is looping, that is when it has to play the first note of the pattern again. In this case, the method calculates an end degree close to the last played degree of the pattern in the family. If at this time one of the three game keys 2,3 or 4 is pressed, then the process will seek a degree in the range concerned, provided that the "direction" mode is not active. If this mode is active, the process will generate: - if key 2 is active: a degree of the first family above the last degree of the pattern; if the key 3 is active: if the last degree of the motive is of the first family, that degree, if not a degree of the first family which is near; - if key 4 is active: a degree of the first family below the last degree of the reason.

Claims (19)

REVENDICATIONS1. A method of musical improvisation, characterized in that it comprises, by an information processing device: a step of receiving a clock generated by a clock generation circuit integrated in the device or by an external device; a step of defining a measurement (encryption); a step of generating a varying tempo according to said clock signal and said measurement (encryption); a step of receiving a start signal of note or start of group of notes; a step of defining by an automatic music generator a note or a group of notes played sequentially according to a set of composition rules according to the moment of reception of said start signal relative to said tempo and said measure (encryption); a step of generating an audio signal representative of the respective note pitch of said note or notes of said group of notes; an immediate play step of the audio signal generated upon receipt of said start signal; a step of receiving an end of note signal or end of group of notes stopping the play of said generated audio signal. 2. A method of musical improvisation, according to claim 1, characterized in that it further comprises: a step of receiving at least one other beginning signal of note or beginning of group of notes, each signal being assigned to a part of the range and triggering the definition then playing a note or a group of notes in this reduced range. 3. musical improvisation method according to one of claims 1 or 2, characterized in that it further comprises: a step of receiving at least one other signal of beginning of note or beginning of group of notes which triggers the generation of a note or group of notes: - at a higher pitch than the previous note; or - at the same height as the previous note; or - at a lower height than the previous note. 4. Method of musical improvisation, according to one of claims 1 to 3, characterized in that it further comprises: a step of transmitting an audible signal of musical type representative of the tempo and its variations, measurement (encryption) and its changes. 5. musical improvisation method according to claim 4, characterized in that it further comprises: a step of reading in a memory space of said device of said musical signal. 6. musical improvisation method according to claim 4, characterized in that it further comprises: a step of real-time generation of said musical signal by the automatic music generator in accordance with the tempo and measurement (encryption). 7. musical improvisation method according to one of claims 5 or 6 characterized in that it further comprises: a step of defining at least one sequence of chords compatible harmonically with said musical signal, each chord being associated with a scale, which makes it possible to share the pitches of the notes of each range in three families of notes, the first family being made up of the notes common to the range and the associated chord, the second family consisting not only of the other notes of the range, the third family consisting of the notes of the chromatic scale amputated notes of the range associated with said agreement; and an association step according to said tempo, the measurement (encryption), the position and duration of the notes played previously, the membership of a potential note moment to play at one of the first two families of notes, the position of said potential note instant to play being quantized and rounded according to the time resolution of the associated automatic music generator. 8. musical improvisation method according to claim 7, characterized in that it further comprises: a step of determining the maximum difference between the pitch of the note to play and the height of the previous note played according to of the family to which belongs the moment to play and the family to which the previous played note belongs. 9. A method of musical improvisation, according to claim 7, characterized in that it further comprises a step of replacing a note played, if the note played is held during a change of agreement, and if this note does not belong to the first family of this new agreement, by the nearest note being part of the first family of the new agreement. 10. A method of musical improvisation, according to claim 7, characterized in that it further comprises: a step of automatic replacement of a played note which does not belong to the first family of the current agreement and which is held too long by the nearest note in the first family of the current agreement. 11. musical improvisation method, according to claim 7, characterized in that it further comprises: a step of receiving a signal which, if it is received in conjunction with one of the start of note signals, triggers the generation of a second note, this second note being generated a sixteenth note after the note generated following the reception of said start signal and after stopping the game of this previous note. 12. A method of musical improvisation according to claim 11, characterized in that the height of the new note created is not separated from the height of said preceding note by more than two degrees, a degree designating any one of notes of the range associated with the current chord. 13. A method of musical improvisation, according to claim 7, characterized in that it further comprises: a choice step, when the said group of notes is a pattern extracted from a library of patterns, the height of the first note so that at least half of the notes of said pattern are notes of the first family. 14. A method of musical improvisation, according to claim 1, characterized in that the step of generating an audio signal representative of the respective pitch of said note or notes of said group of notes comprises: a step of reading and processing representative audio samples of a simulated instrument. 15. A method of musical improvisation according to claim 1, characterized in that the step of generating an audio signal representative of the respective pitch of said note or notes of said group of notes comprises: a step of mathematical modeling of a simulated instrument. 16. A method of musical improvisation, according to claim 1, characterized in that the step of generating an audio signal representative of the respective pitch of said note or notes of said group of notes comprises: a step of additive or subtractive synthesis. 17. Computer program comprising instructions adapted to the implementation of each of the steps of the method according to any one of claims 1 to 16 when said program is executed on a computer. 18. An information storage medium, removable or not, partially or completely readable by a computer or a microprocessor comprising code instructions of a computer program for the execution of each of the steps of the method according to any one Claims 1 to 16. 19. A musical instrument in the form of an information processing device, characterized in that it comprises: means for receiving a clock that can be generated by a clock generation circuit integrated in the device or by an external device; means for defining a measurement (encryption); means for generating a tempo that can vary according to said clock signal and said measurement (encryption); means for receiving a start of note signal or start of group of notes; means for defining by an automatic music generator a note or a group of notes played sequentially according to a set of composition rules as a function of the moment of reception of said start signal relative to said tempo and said measure (encryption); means for generating an audio signal representative of the respective note pitch of said note or notes of said group of notes; means for immediate play of the audio signal generated upon receipt of said start signal; means for receiving an end-of-note or end-of-note signal stopping the play of said generated audio signal.