WO2017076955A1 - Device for generating musical sounds in a non-harmonic scale: the dynalophone - Google Patents

Abstract

The present invention relates to a device for generating musical sounds, the device comprising a keyboard (2) which includes keys for controlling a plurality of electronic circuits, each electronic circuit being configured to generate a vibrating wave representing a particular biological signal.

Description

 DEVICE FOR GENERATING MUSICAL SOUNDS IN A NON-HARMONIC RANGE: THE DYNALOPHONE

FIELD OF THE INVENTION

The present invention relates to the field of musical acoustics.

It relates more specifically to musical instruments, including keyboard, to generate a plurality of sounds that can be used for the production of music.

BACKGROUND OF THE INVENTION

The limbic system is composed of a group of brain structures playing an important role in behavior and in particular, in various emotions such as aggression, fear, pleasure.

The changes that occur in the body when listening to music depend on the music being listened to: a high rhythm and volume can be a bit of a stressful situation for the brain, while soft, calm music will create a soothing situation.

One of the goals of musical creation is to maximize the solicitation of the sensory areas of the cerebral cortex, in order to produce emotions in the listener.

The music consists of a succession of musical sounds. Each sound corresponds to a vibratory wave that propagates in the air between a transmitter generating the vibration and a receiver adapted to capture the wave stimulus. The musical sound is essentially distinguished from noise by the fact that it is organized. The theory of music defines musical sounds with four characteristics: the duration corresponding to an interval of time during which the vibrations of the sound (or vibratory wave) can be identified,

 - the height corresponding to the frequency of vibration of the sound (in Hertz),

 the intensity which is representative of the effective value of the sound pressure of the sound (in decibel), and

 - the timbre which groups together certain harmonic and sound envelope properties.

Today, the sounds produced by musical instruments belong to harmonic ranges, that is to say that the instruments generate vibratory waves of substantially sinusoidal form.

An object of the present invention is to provide a new type of musical instrument, for example to provide composers and musicians new ways to excite the cochlear and basal nuclei of a listener.

BRIEF DESCRIPTION OF THE INVENTION

For this purpose, the invention proposes a device for generating musical sounds, for example called a Dynalophone, the device comprising a keyboard including keys for controlling a plurality of electronic circuits, each electronic circuit being configured to generate a vibratory wave. representative of a respective biological signal.

The name Dynalophone is proposed in reference to Dynamophone T. Cahill, which is considered the first instrument of electro-acoustic music (Patent US 580036, filed April 6, 1897).

Thus, the present invention allows the generation of musical sounds from data from the human body (ECG, EEG, pulse, mass spectrometry or NMR nucleic acids - RNA or DNA - or proteins, ...), in view of their use in the musical field. Preferred but non-limiting aspects of the device for generating musical sounds are the following:

 each vibratory wave has rise times Tm and descent Td of different durations;

 - Each electronic circuit comprises at least first and second oscillatory circuits electrically connected to each other in series;

 each vibratory wave is composed of at least one fundamental oscillation and one harmonic oscillation:

 the first oscillatory circuit being configured to generate a fundamental sound corresponding to the fundamental oscillation,

 the second oscillatory circuit being configured to generate a harmonic sound corresponding to the harmonic oscillation;

- each fundamental or harmonic sound generated by an oscillatory circuit is a solution of the following Van der Pol equation:

 with "μ" a non-zero anharmonic parameter;

 each oscillatory circuit is an RLC type circuit comprising:

 an electric arc of resistance R,

 a coil of inductance L, and

 a capacitor of capacitance C;

 for each oscillatory circuit, the resistance R of the electric arc, the inductance L of the coil and the capacitance C of the capacitor are chosen so that the electric arc generates a fundamental or harmonic sound, the shape of which coincides with the form of the fundamental or harmonic oscillation associated with it;

at least one of the electronic circuits is configured to generate a representative vibratory wave:

an amino acid chosen from Alanine, Arginine, Asparagine, Aspartate, Cysteine, Glutamate, Glutamine, Glycine, Histidine, Nesoleucine, Leucine, Lysine, Methionine , Phenylalanine, Proline, Pyrrolysine, Selenocysteine, Serine, Threonine, Tryptophan, Tyrosine, Valine, or

 o a nitrogen base chosen from adenine, guanine, thymine, cytosine.

BRIEF DESCRIPTION OF THE DRAWINGS

Other advantages and characteristics of the device for generating musical sounds according to the invention will emerge more clearly from the following description of several variant embodiments, given by way of non-limiting examples, from the appended drawings in which:

 FIG. 1 schematically illustrates an example of a biological relaxation signal having a rise time of different duration than a descent time,

- Figure 2 schematically illustrates an example of a device for generating musical sounds.

DETAILED DESCRIPTION OF THE INVENTION

1. General principle

The device for generating musical sounds according to the invention will now be described with reference to the figures. In these different figures, the equivalent elements are designated by the same reference numeral. This device allows the production of specific vibratory waves not belonging to the conventional chromatic range, these vibratory waves each including information representative of a biological signal.

In the following, the invention will be presented with reference to vibratory waves representative of protein and / or nucleic spectrometry peaks, it being understood that the invention is not limited to this particular type of biological signals. In particular, the sound generation device can be parameterized to generate vibratory waves representative of any type of non-symmetric relaxation biological signal - that is to say a signal having an asymmetry between its rise and fall times ( see Figure 1).

In the context of the present invention, the term "rise time" is intended to mean a duration Tm that the relaxation signal sets to go from a minimum of amplitude to a maximum of amplitude.

In the context of the present invention, the term "descent time" is intended to mean the duration Td that the relaxation signal makes to go from a maximum of amplitude to a minimum of amplitude.

The conversion of the biological signal into sound is based on the use of a so-called "Dynalet" transform, making it possible to break down the biological signal by:

 - a fundamental and

 a harmonic component,

satisfying a differential equation formulated by physicist Balthasar van der Pol:

- - x - - ~ f- - 0

 tli? 11.1 ·

 With "μ" a non-zero anharmonic parameter indicating non-linearity and the force of damping.

The fundamental and harmonic components of the biological signal thus determined are then used to size the oscillatory circuit elements adapted to generate oscillations corresponding to said fundamental and harmonic components. More precisely, each fundamental and harmonic component is associated with a respective oscillator (RLC circuit) capable of generating a shape oscillation corresponding to said component. In particular, the fundamental component is, for example, associated with a first Van der Pol oscillator, whose values of the electrical elements (resistance, inductance, capacitance) are chosen so that said first oscillator generates a vibratory wave (or "fundamental sound" Corresponding to said fundamental component. The harmonic component is associated with a second Van der Pol oscillator, sized so as to generate a vibratory wave (or "harmonic") corresponding to said harmonic component.

2. Lutherie Referring to Figure 2, there is illustrated an example of a device for generating musical sounds.

The musical game device comprises a control unit 1, a keyboard 2 and a sound generator 3. These different elements 1, 2, 3 can be integrated with each other or be distant from each other.

2.1. Control unit

The control unit 1 allows the control of the keyboard 2 and the sound generator 3. In particular, the control unit 1 includes a program allowing the matching of each key of the keyboard 2 with a respective note of the anharmonic range adopted for the musical composition. The control unit 1 is also programmed to control the sound generator 3, so as to synchronize the generation of each sound with the actuation of a key of the keyboard 2.

The control unit 1 can be composed of one (or more) station (s) of work, and / or one (or more) computer (s) or may be of any other type known to those skilled in the art. The processing unit is for example a mobile phone, an electronic tablet (such as an IPAD®), a personal assistant (or "PDA", acronym for the term "Personal Digital Assistant"), etc.

The control unit 1 comprises:

 processing means such as a processor),

 input means such as a touch screen,

 display means such as a screen, and

 transmission / reception means for exchanging data with the keyboard 2 and / or the sound generator 3, such as one (or more) antenna (s) and / or one (or more) cable (s) of connection.

2.2. Keyboard

The keyboard 2 is one of the input means, to allow the user to "play" a melody. In the embodiment illustrated in Figure 2, the keyboard is composed of 24 keys emitting the 24 notes of anharmonic range previously defined to compose melodies. Each note results from the choice of a characteristic NMR spectrometry peak, from among a set of peaks obtained during the compression of biological signals using a compression algorithm called "Dynalet", which will be described in more detail in FIG. the following.

Among these twenty-four notes, twenty notes correspond to amino acid NMR spectrometry peaks and four notes correspond to NMR spectroscopic peaks of nucleic bases.

For example, each note can be representative:

 an amino acid chosen from Alanine, Arginine, Asparagine, Aspartate, Cysteine, Glutamate, Glutamine, Glycine, Histidine,

Nesoleucine, Leucine, Lysine, Methionine, Phenylalanine, Proline, Pyrrolysine, Selenocysteine, Serine, Threonine, Tryptophan, Tyrosine, Valine, or

 o a nitrogen base chosen from adenine, guanine, thymine, cytosine.

2.3. Sound generator

The sound generator 3 makes it possible to generate sounds from electrical signals. More precisely, the sound generator 3 makes it possible to generate one (or more) sound waves according to the electric signal (s) resulting from the actuation of one (or more) key (s) on the keyboard 2.

Electroacoustic transduction is the transformation of an electrical signal into a musical signal consisting of acoustic pressures. Pressure variations resulting from electroacoustic transduction vibrate the eardrum, which is responsible for harvesting the sound wave.

The operating principle of the sound generator 3 is to vary the pressure of the air through an electric arc whose intensity is modulated. Unlike other types of transduction, there is no mechanical part in motion here, which is the main advantage of this technology, facing the "traditional loudspeakers".

The sound generator 3 is, for example, an electronic organ including a plurality of sound pipes and an electronic board to which are connected essentially a plurality of RLC circuits, and possibly a microcontroller and a memory.

Each RLC circuit is composed of a resistance electric arc R - such as an arc lamp, an L coil (inductor) and a capacitor C (capacitance) connected in series. RLC circuits are associated in pairs. Each pair of RLC circuits excites the sound pipes to generate a respective note of the anharmonic range. The RLC circuits of a pair are connected in series and tuned to satisfy the following Van der Pol equation:

 With "μ" a non-zero anharmonic parameter indicating non-linearity and the force of damping. Preferably, the anharmonic parameter "μ" is chosen between 0.05 and 0.5. This makes it possible to generate sounds in the range audible by the human ear. This equation can be formulated in the form of the following very general Van der Pol system:

 dx / dt = y, and

dy / dt = μ (1 -Ι < ² χ ² ) ν - ω ² χ Each pair comprises a first RLC circuit and a second RLC circuit connected in series:

 The values of the components - i.e. resistance, inductance and capacitance - of the first RLC circuit are chosen so as to correspond to the parameters of the van der Pol systems having as a solution the fundamental of the Dynalet approximation of a keyboard note,

 The values of the components - i.e. resistance, inductance and capacitance - of the second RLC circuit are chosen so as to correspond to the parameters of the van der Pol systems having as solution the first harmonic of the Dynalet approximation of a keyboard note.

In the embodiment illustrated in FIG. 1, the sound generator comprises twenty-four pairs of RLC circuits corresponding to the twenty-four notes of the anharmonic range selected. 3. Theoretical aspects

3.1. Theory of Dynalet Compression The main principles associated with Dynalet compression will now be described in greater detail. 3.1.1. Introduction

We propose a strategy of compression and retrieval of information related to spectral peaks, from mass spectrograms or magnetic resonance, protein or nucleic, or physiological relaxation signals (ECG, EEG, pulse, ...) .

The corresponding signal consists of a series of unsymmetrical peaks, which are currently processed by Fourier transform, wavelet transform or functional estimation, according to the following equations and solutions: i) Fourier Transform (1807)

Simple pendulum: dx / dt = y, dy / dt = - ω ² χ

 Solution x (t) = kcoscot ii) Harley Wavelet Transform (1997)

Cumbersome pendulum: dx / dt = y, dy / dt = - (ω ² + τ ² ) x - 2xy

Solution x (t) = kcoscot e ^{~ Tt} , y (t) = -kco sincot e ^{~ Tt} -kx coscot e ^{~ Tt} ,

The Dynalet transformation is based on the Liénard equation, where P and Q are arbitrary polynomials:

 - Liénard system: dx / dt = y, dy / dt = - P (x) x + Q (x) y

- Van der Pol system: P (x) = co ² and Q (x) = μ (1 - ^ χ ² )

3.1.2. Description Principle (with ω = 1)

When μ »1, we observe that the period of the limit cycle of the van der Pol equation is mainly determined by the time during which the system remains around the cubic function, when x and y are both of the order of magnitude of 1 / μ, ie 0.

Thus, the period T of the oscillation is estimated approximately by T ^¾ μ.

The amplitude a _x (i) of the abscissa x on the limit cycle of the van der Pol equation is estimated by 4 / k <a _x (i) <4.048 / k, for each μ> 0.

The amplitude at _y {i) = 2b / k of the ordinate on the limit cycle of the van der Pol equation is estimated from the equations:

x + μ (1 -χ ² ^ = 0 and

(x ² + b ²⁾ / 2 ^ 2 + -3μχ 3μbx ^3/8 -μχ ^ / 24 to 2.023 = 0 if μ is small.

More generally, a _x (i) is estimated by:

8 _χ (μ) "(4 + μ ^2/3) / (1 + 7μ ^2/96)

T, in the vicinity of the bifurcation value μ = 0, is equal to

2π / Ιηιβ, where β is the eigenvalue of the Jacobian matrix J of the van der Pol equation given by:

The characteristic polynomial of J is equal to:

β ² - μβ + 1 = 0,

So β = (μ ± (μ ² -4) ^1/2 ) / 2 and we have:

T "+ 2π πμ ^2/8. Let us now give, as polynomial of order 2 + j, an approximate solution S (kj, / ^2j ) (t) of the differential system:

dx / dt = y and dy / dt = - x + μ (1 -k ² x ² ) y / 2j

We search for an approximate solution x (t) = S (1, 1) (t), like a polynomial of order 3, in the case where μ = 1:

x (t) = c ₀ + Cit + c ₂ t ² + c ₃ t ³ , y (t) = ci + 2c ₂ t + 3c ₃ t ² , where the coefficients Ci are obtained by identification in the following equation:

(x ² + y ²⁾ / 2 -3μχν / 2 + 3μνχ ^3/8 -μχγ ³ / 24-2023 = 0,

We obtain as follows:

c ₀ ^2/2 + Ci ^2/2 -3c ₀ Ci / 3c + 2 ₀ ³ Ci / 8 - c ₀ The ^3/24 = 2, c ₂ c ₃ - 9c ₃ ^2/2 - 9c ₀ c ₂ ³ + 9c c _{0 2} ^3/4 ₀ ² + 27c _c3 ^2/8 -3c ₀ c _{2 c3} ^2/4 - c ₂ ^4/24 = 0

=> C ₂ c ₃ - ³ 27c _2/2 + 9c -3c ₃ ² ₂ c ₃ ^2/2 - c ₂ ^4/24 = 0,

Which implies: c ₀ = 2, d = 0, c ₂ «0.46 and c ₃ « 0.04. Due to the symmetry of the limit cycle, all the solutions {S (k _j ^ / 2 ^j )} _i | _N are orthogonal and one can decompose any continuous function f on this basis, thanks to Weierstrass's theorem. We will call this decomposition, decomposition "Dynalet".

Development

The Dynalet transformation consists in identifying a Lienard system that expresses interactions between its variables through its Jacobian matrix, analogous to that of the studied experimental system, whose empirical limit cycle is the closest (in the sense of the distance Δ between subsets, or the mean squared distance between points and set, or the Hausdorff distance between subsets) of the set of points of the signal in the phase plane (xOy), where y = dx / dt. In practice, if the Liénard system is a van der Pol system, it is necessary to perform the following transformations to obtain a Dynalet approximation of the original signal:

 i) estimate the parameter μ (possibly ω, if it is different from 1), so that the period of the van der Pol signal is equal to the empirical average period (calculated from the original signal) and

 ii) translate the abscissa of the origin of the axes in the phase plane, then

 iii) make a homothety of the abscissae, so as to match the van der Pol signal and the original signal.

In the case of the processing of an ECG signal, the whole of the approximation method involves the following steps: a. Delete the time intervals during which the signal is under the critical plateau value Ι_ (ε) of the Levy time, equal to the time that the signal passed between 0 and ε; this step yields the QRS complex of the experimental ECG, b. Set the value of the parameter μ (and possibly ω), so that the period of the Van der Pol signal is equal to the duration of the QRS complex, c. Perform a translation of the abscissa of the origin of the phase plane (xOy) and a scaling setting the value of the parameter b on the x-axis of the Van der Pol signal, so as to adjust them to the maximum X abscissae of the QRS complex, d. Finish the approximation with an optimization of the parameters (mainly the parameters μ and k), by matching the points of the QRS complex with the Van der Pol limit cycle (using the scaling on the x and y axes), to minimize the distance Δ between the interiors of the set of QRS points and all the points of the Van der Pol limit cycle (denoted respectively ECG and VDP, with ECGO and VDPO interiors) in the phase plan:

 A (ECG0, VDP0) = Area [(ECG0 \ VDP0) (VDP0 \ ECG0)], using a Monte Carlo method to estimate the area of the interiors of the linear approximation of empirical points of the experimental QRS complex and the limit cycle Van der Pol, calculated from a sample of points in the phase plane, respectively {Ei} i = 1, 100 and {Vi} i = 1, 100.

It is also possible to minimize the mean squared distance between Van der Pol boundary cycle points and empirical points with the same phase, or to minimize the Hausdorff distance between the inner subsets of the curves considered,

Repeat the process to obtain successive harmonics, so as to respect, for example, a fixed threshold of 20 dB for the SNR signal-to-noise ratio and 10% for the MSE mean square error,

Calculate a polynomial approximation of the signal from the quadratic estimate of the van der Pol limit cycle corresponding to step 4, for example, if

: H (x, y) = (x ² + y ² ) / 2 - μχν / 2 + μνχ ³ /8-μχν 3/8

Theory of the invention

The present invention aims to build a non-harmonic electronic violin synthesis, leading to the production of specific sounds not belonging to the classical chromatic range, and thus allowing a purely aesthetic use in the musical field. This violin, called Dynalophone, is based on the conversion of biological relaxation signals (that is to say non-symmetrical signals) from the human body (ECG, EEG, pulse, mass spectrometry or NMR of nucleic acids - RNA or DNA - or proteins, etc.) in sound, for their use in the musical field.

This conversion makes it possible to create new achromatic ranges of sounds, each achromatic range being able to be composed of a variable number of musical sounds, according to the desired use.

For example, in the case of a conversion of protein and nucleic spectrometry peaks, the achromatic range may consist of twenty-two sounds from:

 peaks of nuclear magnetic resonance spectrograms (NMR) of twenty amino acids,

 peaks of NMR spectrograms of the triplets "initiation" and "termination" of the genetic code. More generally, the sounds of the violin could correspond to a reproduction of any signal of molecular spectrometry (NMR, mass, Raman, ultrasound, fluorescent, ...) and to any source of non symmetrical relaxation signals (ie, having an asymmetry between the rise and fall of the relaxation signal). The conversion of biological signals into sound data can be implemented using a compression method known as the "Dynalets method" based on the anharmonic pendulum equation, described in US Patent Application No. 14 / 864,029 filed September 24, 2015. It is therefore:

 i) after compression of the signals by a new method of approximation of the relaxation peaks (of non-sinusoidal shape, even damped),

ii) to construct a musical instrument based on the electronic circuit of the van der Pol oscillator, whose parameters (values of capacitor capacitor C parameters, coil inductance L and resistance R of the circuit) are fixed by those obtained for the fundamental signals representing each peak, during the compression of the signal.

This specific violin, based on this electronic assembly, allows to play and hear a music based on the sounds of the new range, without going through a classical violin giving the sounds of the chromatic scale. The van der Pol equation corresponding to the variations, over time, of the potential of the electronic circuit is given by:

Cd ² U / dt ² + (- α + 3γΙΙ ² ) dU / dt + U / L = 0, (1) where the values of a and γ are related to those of C, L and R.

The sounds generated by the violin are then interpreted by the human sensory areas. This interpretation is based on human cochlear and post-cochlear treatments, which perform the segmentation and contrasting of sounds that charge the sinusoidal spectrum of Fourier over a homogeneous range of related frequencies, and provide, in cortical areas of auditory projection, the aesthetic feeling. The reader will have understood that many modifications can be made to the invention described above without physically going out of the new teachings and advantages described herein. Therefore, all such modifications are intended to be incorporated within the scope of the appended claims.

Claims

1. Device for generating musical sounds, for example called Dynalophone, the device comprising a keyboard (2) including keys for controlling a plurality of electronic circuits, each electronic circuit being configured to generate a vibration wave representative of a respective biological signal, each vibratory wave having rise times (Tm) and descent time (Td) of different durations.

2. Sound generation device according to claim 1, wherein each electronic circuit comprises at least first and second oscillatory circuits electrically connected to each other in series.

3. Device according to claim 2, wherein each vibratory wave is composed of at least one fundamental oscillation and one harmonic oscillation:

 the first oscillatory circuit being configured to generate a fundamental sound corresponding to the fundamental oscillation,

 the second oscillatory circuit being configured to generate a harmonic sound corresponding to the harmonic oscillation.

The sound generation device according to claim 3, wherein each fundamental or harmonic sound generated by an oscillatory circuit is a solution of the following Van der Pol equation:

 With "μ" a non-zero anharmonic parameter.

5. Sound generating device according to any one of claims 2 to 4, wherein each oscillatory circuit is an RLC-type circuit comprising: - a resistance electric arc R, a coil of inductance L, and

 a capacitor of capacitance C.

6. A sound generation device according to claims 3 and 4 taken in combination, wherein, for each oscillatory circuit, the resistance R of the electric arc, the inductance L of the coil and capacitor C of the capacitor are selected from so that the electric arc generates a fundamental or harmonic sound, the shape of which coincides with the form of the fundamental or harmonic oscillation associated with it.

7. A sound generation device according to any one of the preceding claims, wherein at least one of the electronic circuits is configured to generate a representative vibratory wave:

 an amino acid chosen from Alanine, Arginine, Asparagine, Aspartate, Cysteine, Glutamate, Glutamine, Glycine, Histidine, Nesoleucine, Leucine, Lysine, Methionine , Phenylalanine, Proline, Pyrrolysine, Selenocysteine, Serine, Threonine, Tryptophan, Tyrosine, Valine, or

 o a nitrogen base chosen from adenine, guanine, thymine, cytosine.