IT201800004636A1 - GENERATION OF COMPOSITE AUDIO SIGNALS, OBTAINED WITH SYNTHESIS METHODS, WITH STRICTLY LIMITED SPECTRUM AND DURATION, SPECIFIC FOR MEASUREMENTS IN PEDIATRIC AUDIOLOGY, FOR ADULTS OR FOR TESTING EQUIPMENT ACOUSTICS. THE SIGNALS GENERATED CAN BE ASSOCIATED WITH SIMPLE OBJECTS, IMAGES OR UNIVOCALLY DETERMINABLE CONCEPTS. - Google Patents

Description

TITLE: Generation of composite audio signals, obtained with synthesis methods, with strictly limited spectrum and duration, specific for measurements in pediatric audiology, for adults or for testing electroacoustic equipment. The signals generated can be associated with simple objects, images or concepts that can be uniquely determined.

FIELD OF TECHNIQUE TO WHICH THE FOUND REFERENCE

As regards the field of manufacturing techniques, the invention belongs to the category of devices / methods suitable for generating composite audio signals, obtained with synthesis methods, with a strictly limited spectrum and duration.

As regards the field of application, the invention belongs to the category of measurement methods in pediatric or adult audiology and / or measurements of the functionality of electroacoustic apparatuses.

SALIENT FEATURES OF THE FOUND

A method for the realization of synthesis signals is shown below. This method uses different functional blocks, each of which can be realized:

- In hardware mode, with suitable devices, such as for example oscillators, function generators, modulators

- With methods where hardware devices, such as DACs, are interfaced with numerical controllers

- According to a mixed modality, where some functions are obtained with numerically controlled devices and others are realized through hardware-only equipment. The signals thus obtained have the following fundamental characteristics: Strictly limited band; Limited time duration; Arbitrarily shaped temporal envelope; Perfectly controllable amplitude both as a peak value and as an effective value; Possibility to set the different frequency components to obtain a stimulation in SPL, HL, or with any frequency weighting curve, without the need for a post equalization; Possible association with objects, images or concepts immediately recognizable by subjects even in preschool age; Possibility of covering the entire audio spectrum with a limited different typology of sounds produced, with the consequence that the number of objects, images or concepts associated with the test is also limited; Possibility of frequently modifying a signal while remaining associated with the same object / concept.

REASONS THAT PROMOTED THE DEVELOPMENT OF THE FOUND

consequent to current situations, needs, problems

Necessary characteristics of test signals: premises

To carry out a test on a subject the apparatus it is necessary that the test signals have some characteristics, including: Being able to carry out a test completely covering the frequency range and dynamics required to carry out a meaningful and complete test; Contain an adequate amount of information, even if not entirely corresponding to real situations, which allows to carry out a meaningful test on the behavior of the subject / apparatus; Possess a complexity that allows to test the subject / apparatus in a meaningful way without causing excessive complications in the analysis of the response; Possibility of creating sounds that cannot be generated by a human subject; Ensure an accurate level of acoustic stimulation in both SPL and HL or other frequency weighting modes; Allowing a comparison between tests performed by several operators in different environments, which means repetitiveness as the subject making the measurement changes and repetitiveness as the environment in which the test is carried out varies; Allow a comparison of tests performed on the same patient at different times / places; Have a not too high complexity of the hardware necessary for the generation of signals and the eventual transformation into an acoustic signal.

When carrying out tests, on subjects or devices, there is certainly the need to respect a series of parameters so that the examination is reliable. The most significant are the following: Repetitiveness as the subject making the measurement varies; Repeatability when the environment in which the test is carried out changes; Possibility of comparison of the measurements made in different times and / or places.

With the signals obtained by means of the technique described by the invention, the characteristics required for the signals used to carry out the tests, both on human subjects and on electroacoustic apparatuses, are satisfied.

This document, in addition to describing the general methods for obtaining such test signals, also shows some examples of signals, produced with the methodology described by the invention, associated with the corresponding object and concept.

THE CURRENT TECHNIQUE OF THE SIGNALS USED FOR AUDIOLOGICAL TESTS

Vocal phonemes and acoustic signals in general, as a starting point for setting possible test signals

A possible starting point for identifying and analyzing the fundamental characteristics that the new synthesis signals must possess is constituted by the phonemes, which contain, as a whole, the significant and indispensable frequency components for a correct transmission and perception of only verbal language. The temporal envelope, different from phoneme to phoneme, has an amplitude dynamics around 40-50 dB.

The study began with an in-depth frequency analysis on some phonemes, pronounced by different subjects. This work was carried out in the form of a degree thesis in Audioprosthetic Techniques, entitled: Creation of a statistically significant sample of phonemes for the analysis of the temporal evolution of frequency components.

academic year 2013-2014

First of all, a series of phonemes were analyzed, recorded in high audio quality, with the aim of identifying their significant spectrum and the associated temporal trend. From this first analysis it emerged that the useful spectrum of a phoneme, for its easy and correct interpretation, not only semantic but also perceptive acoustic, as well as for its production with electroacoustic equipment, ranges from 100 Hz to 10 kHz. This spectrum can be considered a kind of subset of the overall audible acoustic spectrum. Sufficient information is obtained even with a frequency extension between 125 Hz and 8 kHz, provided that the response of the electro-acoustic device used for generating the acoustic signal in this frequency range is between -3 dB and 3 d B.

Sound in the world of children: comparison with the phonemes pronounced by a person Consider the set of sounds present in the environments in which children live: family, school, leisure or holiday places, outdoor situations, etc. These sounds, from a frequency point of view, cover a significantly wider spectrum than phonemes (both male and female), with values ranging from 60-80 Hz up to 12-14kHz. Furthermore, the amplitude dynamics can reach 70 - 80 dB.

Such a frequency extension, combined with such a high dynamics, is certainly not obtainable with sounds generated with the human voice.

These considerations led to the development of the invention, which allows, among other characteristics, to overcome the frequency and dynamic limits of the human voice.

Carrying out auditory diagnostics with a higher frequency range and dynamics than an acoustic signal corresponding to a phoneme, produced by the human voice, allows to obtain significantly valid results, both in terms of accuracy of the measurements and as a broader diversification of the measurements themselves. Furthermore, by associating the invention with an adequate system for converting electrical signals into acoustic signals, such as for example DAC Amplifiers Diffusers, it is possible to extend the diagnostics to the point of covering the entire frequency and dynamic range of a child's hearing, as well as than an adult.

A more accurate diagnosis, compared to the frequencies of speech alone, allows to fine-tune, at least from a principle point of view, optimal corrective devices with an extended frequency range, with an operation that covers most of the acoustic signals perceived by a child. . In this way, a better cognitive and cerebral development of the child is stimulated than in the case of using corrective devices with a frequency and dynamic range of functioning limited in the spectrum of speech only.

Generic speech sounds used as test signals

For the sounds emitted, produced by a human individual, some fundamental issues must be kept in mind, which certainly do not favor the use of such sounds as test signals:

Advantages: Relative simplicity to obtain them, as a certain training by the pronouncer is sufficient; No equipment is required, other than a sound level meter for a check, which however can only be carried out in an approximate manner, of the intensity of the sounds produced.

Disadvantages: Obvious non-repetition; Spectrum not fully controlled and often too broad for audiological measurement purposes; Differences between sounds of the same type but produced by different subjects / situations; Spectral differences between men and women; Spectral differences based on the age of who emits the sound / phoneme; Difficult control of the sound / phoneme characteristics emitted by the pronouncer; Differences in pronunciation deriving from the different territorial origin of the pronouncer; Formant components with minimum frequency values that hardly drop below 125 Hz; Formant components with maximum frequency values that hardly rise above 8 kHz - 10 kHz; Dynamics of the obtainable intensities limited to values of 40 - 50 dB.

The “6 sounds of Ling” (vocal sounds) used as test signals

The Ling-6 sounds correspond to six sounds used, specifically, to assess the detention threshold in children with hearing aids. They belong to the category of test sounds produced by human operators. These sounds are presented to the patient by an examiner (usually a woman), directly live. Only in the following years, precisely in 2012, these sounds were then recorded and calibrated on a CD, always pronounced by a woman. The methods of carrying out the test were the same as for traditional audiometry, i.e. in the cabin in a free field. Ling-6 sounds include the following sounds: / m /, / u /, / a /, / i /, / ∫ /, and / s /; these, according to the bibliography examined, should cover the range of the most important speech frequencies.

An analysis of the Ling-6 sounds was carried out at I.R.C.C.S. Materno Infantile "Burlo Garofalo" of Trieste, in the summer of 2016, in an extremely accurate way, with equipment consisting of Rode model NT2 studio microphones, a very low noise microphone preamplifier with 2Hz-250 kHz bandwidth and a digital recorder Tascam DV-RA1000HD, set at 192 ks / s and 24 bit. As an example, in the TAV, n ° 26 and n ° 27 and n ° 28 the spectra of three Ling sounds are reported, relating to the sounds "a", "i", "se", pronounced by a female subject, professional of verbal communication , of Italian nationality. From this spectra we can see the considerable frequency extension and the limited dynamic extension.

The “6 sounds of Ling”: frequency analysis

The 6 Ling sounds were analyzed from a frequency point of view and the results are reported in the TAV. No. 29. In each spectrum analyzed, the analysis was carried out for two different dynamic ranges: 24 dB and 18 dB below the component of greater amplitude. By observing the spectral content of the various signals, it can be seen that the spectrum, in the dynamic range of 24 dB, is sometimes broad up to a decade, corresponding to about three octaves, while in the dynamic range of 18 dB the spectrum can be smaller. In any case, the number of harmonic components is high. Each sound contains a large number of components at different frequencies, but of little different amplitude, in order to be considered frequency selective for the purposes of an audiometric test and to allow to obtain truly significant results. As an example, the sounds / a / and / i / have an extremely broad spectrum, with many components of significant amplitude; the spectra of / ∫ / (sc) and / s / (ss) can be comparable to filtered white noises, each with occupancy greater than one octave.

The results reported are the result of two degree theses in Audioprosthetic Techniques - University of Padua:

Synthesis of signals in order to produce sounds to perform auditory tests in the child Graduating: academic year 2015 - 2016 Ling's 6 sounds test: proposal for standardization and validation, also with alternative synthesis signals.

Graduating: academic year 2015-2016

The "6 sounds of Ling": sound-image association the concept

The associations usually used are shown below.

/ mmmmm / associated with the expression used when a food is very good / iiiiiiii / associated with horse neighing

/ uuuuuu / associated with the sound of a cartoon ghost

/ aaaaaa / associated with an expression or gesture of danger

/ ssshh / associated with the gesture and the verse of silence

/ ssssss / associated snake hiss

These are somewhat imaginative associations, which require, on the part of the administrator, some modulation of the voice, with the aim of a certain facilitation towards the child, diversified for each object / situation, which makes the sound itself more understandable to the child. child, but which at the same time creates a temporal evolution of the spectral content that is even less controllable by the pronouncer, with the consequence of adding further annonics, with consequent reduction of the quality of the test.

Ambient sounds used as test signals

For sounds produced in nature or in a residential and / or production environment, some fundamental issues must be kept in mind, which certainly do not favor the use of such sounds as test signals:

Advantages: Remarkable simplicity to obtain them, it is in fact sufficient to record them, as long as they are of adequate quality

Disadvantages: Obvious non-repetition; Spectrum not entirely controlled and often too large for a single sound; Spectral differences often significant even for sounds that seem similar; Spectral differences also for events l situations that are the same but in different contexts; Difficult to identify a sound that can be considered as a reference; Highly variable amplitude dynamics.

Using these sounds for audiological tests, even if recorded with high quality equipment, does not allow to obtain significant and reliable results, precisely because of the considerable frequency extension and the variable and not very controllable dynamics of the signals themselves. To reduce the spectrum extension it would be necessary to resort to frequency filtering operations, as well as to dynamics alteration operations, but this would alter, even significantly, the information contained in the sounds themselves, with a consequent difficulty in interpreting ranging from part of the very young subjects subjected to the test.

The pure stationary tones used as test signals

For pure tones, ie sinusoidal signals, some fundamental issues must be kept in mind, both in favor of and against these types of signal for use in pediatric audiology. Advantages: Remarkable simplicity in implementation, both hardware and numerical; Extreme accuracy obtainable both in intensity and in frequency during the test; Relative simplicity of the calibration of the equipment used; Absolute repetition; Remarkable dynamics obtainable.

Disadvantages: Absence of any logical association between sound and object, image and / or concept; Limited active participation of the subject at a young age, due to a monotony of the test; Need for a large number of signals at different frequencies, but all with a sinusoidal trend, to carry out a meaningful test.

The pure vobulated tones used as test signals

Basically, these are pure tones modulated by m frequency, so as to cover a determined "frequency slice" dependent on the modulation depth and the center band frequency.

Also for these signals some fundamental issues must be kept in mind, both for and against these types of signals for use in pediatric audiology.

Advantages: Discreet simplicity in the realization (both hardware and numerical); Accuracy obtainable both in intensity and in frequency during the test; Relative simplicity of the calibration of the equipment used; Absolute repetition; Remarkable dynamics obtainable Disadvantages: Absence of any logical association between sound and object, image and / or concept; Limited active participation of the subject at a young age, due to a monotony of the test; Need for a large number of signals with different central frequency, unless you do not perform frequency modulations with an octave extension or about.

Observations relating to pure, stationary or vobulated tones

Since there is no association between sound and any concept object, tests made with such sounds are usually boring for children, who quickly get tired. The attention of the child, especially if in preschool age, hardly lasts more than a few minutes. Even when a game or similar is used during the performance of the audiological test, there is no direct association between the type of sound and the object used in the game, with many different sounds and unique game or image. This leads to a decrease in participation in the test by the child being given the test as different sounds are heard. Only the initial part of the test is reliable, as it is linked to an active participation by the child, but this reliability decreases as the test continues.

THE SYNTHESIS SIGNS: FUNDAMENTAL ASPECTS

What are the synthesis signals:

Signals obtained with exclusively synthetic methods; Absolutely controllable and repeatable signals; Signals with a well-defined spectral content belonging to an area of the audio spectrum whose extension can be determined a priori; Signals that can be immediately associated with objects and concepts even by small children, as early as three - four years of age; Signals which, while representing the same object and concept, can occupy different parts of the audio spectrum; Signals that can be "shaped" in frequency according to diagnostic needs. Signals that allow to obtain a very high dynamic intensity, being able to reach even 80 - 90 dB (the only constraints are the environmental noise, the signal-to-noise ratio and the power of the electroacoustic system used for the generation of sounds); Signals that can be easily realized with a controlled intensity in SPL or in HL or other frequency weighting; Signals that allow you to check "what" the child has heard and not just "if he has heard"; Signals that can be administered simultaneously from two or more sources, where a different signal can come out from each source, both in terms of spectral content and intensity.

What synthesis signals ARE NOT:

Real signals modified with filtering operations; Signals obtained directly by hitting real objects or by recording sounds and / or noises produced by objects in particular situations; Signals obtained as a synthesis of Ling's sounds; Signals obtained by adding together real signals, possibly processed; Signals that come from phonemes produced by a human operator; Signals present in nature, without the need for manipulation of the harmonic content; Result of real sound recordings.

SYNTHESIS SIGNS: FUNCTIONAL DESCRIPTION OF THE FOUND

LIST OF TABLES

TAV n ° 1: General functional block diagram of the invention, configured in such a way as to show one of the possible ways of generating the sounds in question, also showing the association between sound and corresponding object, image or concept.

TAV n ° 2: Mode for obtaining a stationary composite signal, without modulation, by adding sinusoidal signals, without noise.

TAV n ° 3: Mode for obtaining a composite signal modulated in AM (without noise). TAV n ° 4: Mode for obtaining a composite signal with distinct AM modulation for each single sinusoidal component, without noise.

TAV n ° 5: Mode for obtaining a composite signal with distinct FM modulation for each single sinusoidal component, without noise.

TAV n ° 6: Association between signal and object the concept.

TAV n ° 7: Association between signal and corresponding frequency content.

TAV n ° 8: Association between signal and object the concept, which the child realizes, and between signal and frequency content.

TAV n ° 9: Example of synthesis sound that simulates a bell

TAV n ° 10: Example of synthesis sound that simulates a train

TAV n ° 11: Example of synthesis sound that simulates the field cricket and corresponding realization using hardware circuits

TAV n ° 12: Example of signal corresponding to a musical movement

TAV n ° 13: Frequency shift of the generated signals and corresponding vanazwm del! ' object the corresponding concept.

TAV n ° 14: Frequency shift of the generated signals, with example of frequency shift to center the frequencies associated with the electrodes of a cochlear implant.

TAV n ° 15: Working group for the identification of possible synthesized sounds.

TAV n ° 16: First synthesis work, by technical staff

TAV n ° 17: First listening of the synthesized sounds, by the working group

TAV n ° 18: Listening to the synthesized signals, by a group of children with normal hearing TAV n ° 19: Execution of the test with the synthesis signals (general aspects)

TAV n ° 20: Execution of the test, with synthesis signals, on a single normoacusic child TAV n ° 21: Execution of the test, with synthesis signals, on a single hearing impaired child TAV n ° 22: Execution of the test, with the synthesis signals, on a group of children

TAV n ° 23: Execution of the test, with the synthesis signals, with an automatic method, on a single child

TAV n ° 24: Execution of the test, with the synthesis signals, with an automatic method, on a group of children

TAV n ° 25: Execution of the test, with synthesis signals, with direct and competing signals TAV n ° 26: Example of frequency analysis of a sound of "Ling", vowel "a"

TAV n ° 27: Example of frequency analysis of a sound of "Ling", vowel "i"

TAV n ° 28: Example of frequency analysis of a "Ling" sound, "if" sound

TAV n ° 29: Summary table of the spectra of "Ling" sounds

The overall structure of the functional blocks, shown in TAV n ° 1, describes the method for generating synthesis signals.

The blocks are of different types: signal generators obtained with one or more sinusoidal signals, white noise generators or similar, band pass filters, ramp generators or arbitrary functions to achieve the modulation of signals, AM or FM modulators, adders of the various signals , selectors (switches) to set the interconnection between the various functional blocks. Some blocks can be made in different ways, described below.

(SIGNAL1) Composite signal generator: it is obtained by adding together several sinusoidal signals, each with its own amplitude and frequency. By acting on the various pure tones generated, the spectrum of the signal produced is practically shaped: by choosing the appropriate frequency components and dosing their specific amplitude, the spectra of sounds similar to the concepts l objects l situations chosen for the test are obtained. To obtain bandwidth-limited signals, the frequency shift can stay within one octave or drop to 1⁄2 octave. As many sinusoidal oscillators are used as there are sinusoids which, when added together, create the composite signal. For each oscillator it is possible to set the frequency and the amplitude of the generated sinusoid. Each sinusoidal signal, in the frequency domain, is represented by a single line. The composite signal is represented by the set of lines corresponding to the single sinusoidal signals, each produced by the corresponding oscillator. Note: some sounds can also consist of single pure tones, in which case the spectrum consists of only one line.

The composite signal obtained can be stationary or modulated: see Table n ° 2, Table n ° 3, Table n ° 4 and Table n ° 5.

(DBl) Variable attenuator, possibly calibrated in dB, which allows you to set the amplitude of the composite signal. The action of this attenuator maintains unchanged the ratio between the amplitudes of the different sinusoidal components that form the composite signal, as set in the block (SIGNAL1) that creates the composite signal. By being able to set the level of the sinusoidal spectral components independently from that of any noise, it is possible to obtain an excellent realism of the sounds produced.

(MOD1) AM modulator: allows you to create a time envelope of the composite signal to obtain greater realism of the generated signals. Since, in reality, no sound has a null attack time and decay time, again to increase the correlation of sounds with real concepts, it is possible to introduce a simple AM amplitude modulation with 100% modulation coefficients, obtained with a trapezoidal modulating signal: the rise ramp corresponds to the attack time (Tr), the horizontal part to the duration of the sound at constant intensity (Ts) and the descent ramp to the decay time (Tf).

(GEN1) Generator of the modulating signal. Normally configured as a trapezoidal signal or the like, this signal can take any shape to obtain an arbitrary evolution of the time envelope of the signal produced by SIGNAL1.

(CLK1) Clock generator. To obtain a signal repeated several times, it can be envisaged that at each new clock pulse the envelope generator (GEN1) provides the same wave form, or a different wave form to obtain more complex signals.

(SW2) (SW3) (SW4) (SWS) Switches with which the amplitude modulation of the composite signal, achieved by means of (GEN1) and (MOD1), is activated. There are two modes of operation:

Modulation activated: SW2, SW3, SWS closed and SW4 open.

Modulation NOT activated: SW2, SW3, SWS open and SW4 closed.

(NS1) White noise generator. The choice of starting from a white noise allows to obtain different noise spectra through the action of filter banks. Furthermore, white noise is easily obtainable both by hardware and numerically.

(FLT1) By suitably filtering a white noise it is possible to obtain a noise with a delimited spectrum, with a particular distribution of the spectral energy. The introduction of the noise has the purpose to increase the realism of the obtained sounds or to carry out tests with masking superimposed noise.

(SW1) Diverter with which the filtering action against white noise is activated. There are two modes of operation:

Filtering activated: SW 1 in position 1.

Filtering NOT activated; SW1 in position 2 and in this case the filter is bypassed.

(DB2) Variable attenuator, possibly calibrated in dB, which allows you to set the amplitude of the noise. The action of this attenuator allows you to set the amplitude of the noise, regardless of the amplitude of the composite signal (SIGNAL1) and its possible modulation.

(MOD2) AM Modulator: Creates a time envelope of noise. The modulation of the noise can follow that of the composite signal or be independent. The amplitude modulation of the noise has, as in the case of the amplitude modulation of the composite signal, the purpose of making the overall signal obtained more realistic, making it associable with an object / concept.

(GEN2) Generator of the modulating signal towards noise. Normally configured as a trapezoidal signal, it can take any shape to obtain an absolutely arbitrary course of the noise envelope. It can have the same time course as GEN1: in this case the same time envelope is obtained for the composite signal and for the noise.

Note: it can be expected that with each new clock pulse the GEN2 envelope generator will provide a different waveform to obtain more complex signals.

(SW6) (SW7) (SW8) (SW9) Switches with which the amplitude modulation of the noise signal, achieved by means of GEN2 and MOD2, is activated / deactivated. There are two modes of operation:

Modulation activated: SWó, SW7, SW9 closed and SW8 open.

Modulation NOT activated: SWó, SW7, SW9 open and SW8 closed.

(SUM1) Adder block between the composite signal and the noise.

(OUT) This is physically the output signal which constitutes the actual SYNTHESIS SIGNAL. This signal is applied to the power amplification audio equipment which, in turn, sends the signal to the loudspeakers.

(OUT-T) This is the output signal, described in the time domain. The visualization can be obtained with an oscilloscope or similar instrument.

(OUT-F) This is the output signal, described in the frequency domain. The visualization can be obtained by means of a spectrum analyzer or similar instrument.

(CONCET) This is the concept / object that the subject (child) associates with the perceived sound. The audiological test is preceded by a short “training” during which preliminary instructions are given to the subject (child) through a series of associations between object / concept and corresponding sound.

(SOGG) This is the subject (child) who is being tested.

The way to obtain a stationary sign obtained from the sum of pure tones is shown in TAV n ° 2

This is the way in which, by means of the invention, a stationary composite signal is generated. As many sinusoidal oscillators are used as there are sinusoids which, when added together, create the composite signal. For each oscillator the frequency and amplitude of the generated sinusoid are set. Each sinusoidal signal, in the frequency domain, is represented by a single line. The composite signal, in the frequency domain, is represented by the set of lines corresponding to the individual sinusoidal signals, each produced by the corresponding oscillator. N sinusoidal generators are used, indicated with (SIN -1, SIN-2, SIN-N) and the adder block (SUM), and thus the composite signal (SIGNAL1) is obtained. The amplitude and frequency of each single sinusoid remain constant and the amplitude (RMS) of the resulting composite signal remains constant over time.

Note; a signal of this type can be used for the calibration of the acoustic stimulation system, being easily measured by means of a sound level meter in slow mode, or similar instrument, without the need for the peak measurement function.

The way to obtain a composite signal obtained from the sum of pure tones, repeated several times by means of an amplitude modulation, is shown in TAV n ° 3

The time envelope is unique for all the sinusoidal components that make up the signal (SIGNAL-1), and is repeated N times, with N ≥ 1. This time envelope is obtained by means of the amplitude modulator (MOD1) added to the structure shown in TAV n ° 2. The repetition of the envelope, controlled by the clock generator (CLK1), allows to obtain a signal that can be more easily interpreted by the subject subjected to the test, compared to a signal that is presented only once. The generator (GEN1) realizes the modulating for the AM modulator (MOD1) and is synchronized by the clock generator (CLK1). With the structure shown, a single amplitude modulation applied to the composite signal is obtained.

Note; if the repetition of the signal is not requested, a single modulation cycle can be carried out by setting the number N of the repetition clock equal to one.

The way to obtain an amplitude modulated composite signal, with a distinct envelope for each single sinusoidal component, is shown in TAV n ° 4

The AM modulation and the consequent time envelope are distinct for each sinusoidal component that participates in the realization of the signal. The distinct temporal envelope for each sinusoidal component makes it possible to obtain sounds of a different type than those obtained with a single temporal envelope. N sinusoidal generators are still used, indicated in the with (SIN-1, SIN-2, SIN-N), but each sinusoidal signal enters an amplitude modulator, indicated with (MOD-A1, MOD-A2, MOD-AN) . The shape of the modulating signal is decided by the generators (GEN-A1, GEN-A2, GEN-AN). Even if the waveform can be arbitrary, the most common case is that of a waveform of the trapezoidal type, or similar, which applied to the amplitude modulator allows to realize an envelope of each signal with an attack time, a pre-established sustenance time and decay time. The different amplitude modulations can be synchronized with each other by means of the repetition clock generator, indicated with (CLK-A). The individual modulated signals are applied to the inputs of an adder, indicated by (SUM-A). Note: the repetition clock has a finite number of cycles between 1 and N.

Note: it can be expected that at each new clock pulse the envelope generators (GEN-A1, GEN-A2, GEN-AN) each provide a different waveform to obtain signals with more complex temporal evolution.

The way to obtain a frequency modulated composite signal, with a distinct modulation for each single sinusoidal component, is shown in TAV n ° 5

The FM frequency modulation is distinct for each sinusoidal component that participates in the realization of the signal. This allows to obtain particular signals, different from those obtained with amplitude modulation. N sinusoidal generators are still used, indicated with (SIN-1, SIN-2, SIN-N), but each sinusoidal signal enters a frequency modulator, indicated with (MOD-F1, MOD-F2, MOD-FN). The shape of the modulating signal is decided by the generators (GEN-F1, GEN-F2, GEN-FN). As in the case of amplitude modulation, the waveform can be arbitrary. The different modulators can be synchronized with each other by means of the repetition clock generator (CLK-B). The individual modulated signals are applied to the inputs of an adder (SUM-B).

Note: the repetition clock has a finite number of cycles between 1 and N.

Note: it can be expected that at each new clock pulse the envelope generators (GEN-B1, GEN-B2, GEN-BN) each provide a different waveform to obtain more complex signals, with frequency characteristics aimed at particular tests.

Note: an AM amplitude modulation can be associated with the FM frequency modulation, realized by means of the MOD1 modulator of Table n ° 1.

A fundamental aspect of synthesis sounds is the association of each sound to a particular object / concept, as shown in TAV n ° 6.

The concept / object (CONCET1-A) is associated with the signal (SIGNAL1-A), the concept / object (CONCET1-B) is associated with the signal (SIGNAL1-B), the concept / object is associated with the signal (SIGNAL1-C) object (CONCET1-C) and so on. The objects / concepts must be sought among those that children consider as attractive things, games or that they see naturally during their daily life, In this way, such objects can be immediately identified by a child with a minimum of preliminary training, ensuring a good participation to the audiological test, with a consequent good reliability of the test itself.

For a diagnostic / analytical use of the synthesis signals, a perfect delimitation of the spectrum is essential, as shown in TAV n ° 7

A well determined and finite frequency content is associated with each signal. in this way a significant audiological test is obtained, thanks to the extreme accuracy both in terms of frequency content and in terms of the time course of the signal amplitude. Being able to narrow the spectrum of each signal by reducing it down to about 1⁄2 octave, or even a pure tone modulated AM or FM, there is the possibility of carrying out some tests rather targeted both towards patients, even those with hearing aids, and against any electroacoustic devices. In the example, the spectrum (SPECTR1-A) is associated with the signal (SIGNAL1-A), the spectrum (SPECTR1-B) is associated with the signal (SIGNAL1-B), the spectrum (SPECTR1) is associated with the signal (SIGNAL1-C) -C). The strictly limited and perfectly controllable frequency band, for each signal generated, constitutes the fundamental measurement aspect that makes the audiological test accurate and meaningful. This characteristic is guaranteed by the method with which the signals are generated by means of the invention.

For an audiological test, in the pediatric field, the double Association is fundamental: the first between Signal and Spectrum; the second between Signal and Concept / Object, as shown in TAV n ° 8

What is shown in Tables n ° 6 and Table n ° 7 is summarized. The child (SOGG) easily associates the concept / object (CONCET1-A) to the acoustic signal (SIGNAL1-A), which corresponds to a spectrum (SPECTR1 -A) perfectly controlled.

From an audiological and measurement point of view, the test is rigorous and with significant results, as it is linked to perfectly defined and controllable signals.

From a child's point of view it is more of a kind of play with sounds, rather than a hearing test.

This certainly leads to an active participation of one or more children, with a consequent greater reliability of the test, also thanks to a longer duration of the same before the test becomes boring, as, usually, occurs with tests carried out with pure tones. or similar.

A fundamental characteristic of the signals obtained with the invention lies in the possibility of shifting in frequency, while maintaining the same associated object / concept. TAV n ° 13 By varying the frequency values of the forming components, indicated with SIN-1, SIN-2, SIN-N in TAV n ° 2, it is possible to associate a different spectrum to the same object / concept, allocated in adjacent frequency portions. This allows you to use a few objects / concepts, to cover the entire acoustic spectrum of interest, making the test easier for the child, also simplifying the introductory part of training. This feature constitutes one of the strengths of the invention. Associated with the indicated spectrum (SPECTR-M) there is an object (in the example a medium-sized bell), indicated with (CONCET-M). Associated with the spectrum allocated on higher frequencies, indicated (SPECTR-S), there is an object (in the example a small bell), indicated with (CONCET-S), but of the same type as the previous one. Associated with the spectrum allocated on lower frequencies, indicated (SPECTR-L), there is an object (in the example a larger bell), but of the same type as the previous one, indicated with (CONCET-L).

As an example we can assume a shift in frequency, equal to 1⁄2 octave, of a composite signal. To move up by 1⁄2 octave, multiply the frequency value of each component of the original signal by 1.41 (square root of 2). To shift down by 1⁄2 octave, divide the frequency value of each component of the original signal by 1.41 (square root of 2). As a further example, to shift the spectrum by 1/3 of an octave, the coefficient to be entered is 1.26 (cube root of 2). The coefficient k is arbitrary.

Frequency shift of the composite signal, with respect to the original signal, to center some specific frequencies of the electrodes of a cochlear implant. TAV n ° 14 With the technique described above (Table n ° 13) it is possible to realize a frequency shift according to the needs, applying a specific multiplicative coefficient of the frequency, for each of the sinusoidal components that form the overall signal. to make the frequencies of the components forming the composite signal coincide with those (fx, fy, fw, fz) associated with the electrodes of the cochlear implants. Such devices, as known, realize a sort of Fourier transform of the acoustic signal incident in the microphone of the system, distributing the energy, in the form of an electrical signal, to various electrodes inserted inside the snail of the inner ear, each used to a very specific part of the acoustic spectrum.

EXAMPLES OF REALIZATION OF SYNTHESIS SOUNDS

Example of realization of a composite signal, AM modulated, without noise, with fundamental at 4 kHz. Corresponding object: TAV bell n ° 9

The example of a signal is shown from which a sound is obtained that is very reminiscent of a bronze bell or similar, with a pleasant sound as it corresponds to a major tonal chord; the chosen repetition frequency can be varied as desired in order to pass from an almost continuous sound to clearly distinct dings according to the diagnostic needs. The components formed are distributed exactly on an octave, starting from 4000 Hz and arriving at 8000 Hz, creating a chord in a major key. There is no superimposed noise. For the realization of this signal, the SIGNAL1, DB1, GEN1, MOD1 and CLK1 blocks of the general diagram of Table n ° 1 are activated. For the generation of the SIGNAL1 composite signal, refer to TAV n ° 2,

The frequencies of the sinusoidal components that make up the composite signal, all of the same amplitude, are indicated below:

F1 = 4000 Hz: F2 = F1 * 1.26 = 5040 Hz; F3 = F1 * 1.58 = 6320 Hz; F4 = 2 * F1 = 8000 Hz corresponding to a "major key" chord. Amplitude normalized to a level equal to - 1 dB. The envelope of the amplitude, obtained by means of an AM modulation, with a depth of 100%, has the following parameters:

Tr = 5 ms; Ts = 25 ms; Tf = 150 ms; Repetition period = 200 ms for a total of 5 shots with the total duration equal to ls. The repetition frequency and the duration of each single chosen sound can be varied, so as to pass from an almost continuous sound to clearly distinct dings according to the diagnostic needs

Noise: absent.

Example of realization of a composite signal, modulated AM, with noise, with fundamental at 1 kHz. Corresponding object: Old Far West train TAV n ° 10 An example of a signal is shown which gives a sound that is very reminiscent of a Far West train, with a pleasant sound as it corresponds to a major tonal chord; the signal is repeated only twice, with a different duration. The components formed are distributed exactly on an octave starting from 1000 Hz and arriving at 2000 Hz. There is superimposed noise, obtained from white noise with the combined action of a low pass filter and a high pass filter of the second order of Bessel (obtaining a band pass response). This noise makes the corresponding sound extremely realistic. For the realization of this signal, the SIGNAL1, DB1, GEN1, MOD1, CLK1, NS1, FLT1, DB2, MOD2, GEN2 blocks of the general scheme of TAV are activated. n ° 1. For the generation of SIGNAL1 refer to TAV n ° 2. The waveform of GEN2 is the same as that of GEN1 The frequencies of the sinusoidal components that make up the composite signal, all of the same amplitude, equal both for the first and for the second are mentioned below:

F1 = 1000 Hz; F2 = F1 * 1.26 = 1260 Hz; F3 = F1 * 1.58 = 1580 Hz; F4 = 2 * F1 = 2000 Hz corresponding to a "major key" chord. Amplitude normalized to a level equal to - 1 dB. The envelope of the amplitude of the first whistle, obtained by means of an AM modulation, with 100% depth, has the following parameters: Tr = 25 ms; Ts = 250 ms; Tf = 75 ms. The envelope of the amplitude of the second whistle, obtained by means of an AM modulation, with 100% depth, has the following parameters: Tr = 25 ms; Ts = 500 ms; Tf = 75 ms. Pause between the two whistles = 200 ms. Total duration = 1 .15 s

White noise filtered by “Bessel second order band pass filter” F1 = 2000 Hz Fh = 8 000 Hz, with max amplitude -40 db with respect to the sinusoidal components, with the same AM modulation of the composite signal.

Example of signal obtained with hardware devices: pure tone at 5 kHz, AM modulated, without noise. Corresponding object: TAV n ° 11 country cricket

The example of a signal obtained by hardware is shown, consisting of a pure tone on which an amplitude modulation (AM) acts to obtain a sound very similar to the call of the country cricket (the typical cricket cricket). ).

For the realization of this signal the functional blocks SIGNAL1, DB1, GEN1, MOD1 and CLK1 of the general diagram of TAV are activated. n ° 1. Since there is only one component formed, which can be with a frequency between 4000 Hz and 6000 Hz, the SIGNAL1 block is made up of a single sinusoidal oscillator. There is no superimposed noise. The modulating GEN1 is obtained as the sum of two sinusoids with periods of 40 ms and 5 ms respectively, with the amplitude of the sinusoid with a higher frequency (the one with a period of 5 ms) which is approximately equal to 1/9 of the amplitude of the sinusoid with a lower frequency (the one with a period equal to 40 ms). To obtain the characteristic sound of the "cricket" of the cricket, the envelope is repeated three to four times. The action of the MOD1 modulator is such as to obtain a modulation depth equal to 100%. In the hardware diagram shown, U1, U2 and U3 realize the composite modulating signal (previously indicated as arbitrary), indicated with GEN1 of the TAV. n ° 1. U4 applies the SIGNAL1 signal from the TAV. n ° 1. U5 realizes the actual modulator. U6 forms an impedance decoupler between the AM and SUC modulator. U5 and U6 realize the MOD1 block of the TAV. n ° 1.

Example of signal consisting of a succession of pure tones, with a frequency range of one octave, which creates a short melody: musical motif composed of TAV n ° 12

It is a melodic sequence of the notes of the central octave between 523 Hz and 1046 Hz. The measuring aspect lies in the fact that the components formed are distributed exactly over an octave, even if this value is not mandatory. There is no superimposed noise. Each note has an attack time of 10 ms and a decay time of 50 ms. The duration depends on the musical score. Being a musical melody, the signal is represented with the musical notation on the staff. The frequency spectrum evolves over time and for each note, in the corresponding frequency representation, there is a line corresponding to the frequency of the note itself. Note; the same melodic motif can be transposed in frequency and, practically, it can range across the entire audible acoustic spectrum, occupying in any case the frequency range of one octave. The sinusoidal components, all of the same amplitude, present one at a time, have a frequency from do (523 Hz) to do (1046 Hz). By means of transposition it is possible to always occupy an octave, but positioned in an arbitrary frequency range. The envelope of the amplitude of each note, obtained by means of an AM modulation, with 100% depth, has the parameters relating to the rise and fall time, indicated below;

Tr = 10 ms; Tf = 50 ms; Total number of notes = 21; Total duration = 3 s.

GUIDELINES FOR OBTAINING ADEQUATE SUMMARY SIGNALS

The first step, shown in TAV n ° 15, for the realization of the synthesis sounds consists of the meeting of Audio Technicians, Healthcare Professionals, Speech Therapists, Engineers, Operators in Primary and Infant Schools (GROUP-WORK), subjects who they have to do with children. This meeting aims to identify a first series of objects / concepts (CONCET-1, CONCET-2, CONCET-4, CONCET-5), quite familiar to children, with the prospect of being able to generate corresponding sounds that possess both characteristics of "analytical sounds suitable for audiometric measurements and clinical tests" and those of "sounds immediately recognizable by children". Having to merge the aspect of the measures and the clinical aspect, the meeting must be shared between different professional figures. The presence of an operator in the Kindergarten is important, who, thanks to direct interaction with children for a long time, is well aware of all those sounds and objects that children hear and see more often. It may be useful to set different groups of signals and corresponding sounds, depending on the geographical area of origin of the children, depending on the environment in which they live, depending on previous experiences, or depending on other factors. It can also be useful to identify different types of sounds, belonging to different contexts: school, family environment, games, nature, etc.

The second step, for the realization of the synthesis sounds, consists of the analysis of real signals and the subsequent synthesis that approximates these signals. TAV n ° 16

the technical staff (TECHNICAL), possibly assisted by medical and / or school staff, carries out a first frequency and temporal analysis of real sounds corresponding to the objects / concepts identified by the WORK-GROUP with the aim of verifying whether the sound in question can be reconstructed , by synthesis, with an acceptable approximation with only a few frequency components allocated on an extension not exceeding one octave or at most one octave and a half. If this is not possible, this sound is discarded. If, on the other hand, this is possible, we move on to the choice of the minimum formant components to obtain a sound, which corresponds as closely as possible to the chosen concept / object (CONCET-1, CONCET-2, CONCET-3, CONCET.4). Once the optimal formant components have been identified, a temporal evolution is set that makes the sound as realistic as possible (SIGNAL-1, SIGNAL-2, SIGNAL-3, S1GNAL.4) and uniquely associated by a child to the corresponding concept I object. To do all this, the "Sound Engineer" requires knowledge of some fundamental concepts: objects, situations, sounds, equations, Fourier's theorem, oscillators, ramps, time, envelope, white noise, filters, modulation, as well as that the meaning of each block of those shown in Table n ° 1,

THE! third step, for the realization of the synthesis sounds, consists of a first listening, part of the! working group (GROUP-LA VORO), synthesized sounds. TAV n ° 17

This listening is necessary to verify the effective association between sound (SIGNAL-1, SIGNAL-2, SIGNAL-3, SIGNAL.4) and the corresponding object / concept (CONCET-1, CONCET-2, CONCET-3, CONCET- 4). Some sounds could be modified and improved, thanks to indications provided to the sound TECHNICIAN by the WORKING GROUP. It is a joint action between WORKING GROUP and SOUND TECHNICIAN, to obtain sounds that are well suited to children.

It may happen that some sounds are not immediately interpreted or even become unclear for a child. In this case, already at the working group level, this sound and the corresponding concept are discarded (CONCET-2 and SIGNAL-2).

Note: Table n ° 17 does not show the acoustic diffuser obviously necessary for the transformation of electrical signals into acoustic signals. This is in order not to complicate the drawing unnecessarily.

The fourth step consists in listening to the synthesized sounds by a group of normal hearing children (GROUP-TEST). AV n ° 18

Also in this case, the purpose is to verify the effective and immediate association between sound (SIGNAL-1, SIGNAL-3, SIGNAL-4) and object / concept (CONCET-1, CONCET-3, CONCET-4). It may happen that some sound is not immediately interpreted or is interpreted incorrectly by children, or it corresponds to an object / concept not entirely familiar to children (SIGNAL-4 and CONCET-4). Some associations, which are not entirely intuitive, may require more effort on the part of the children and in this case the corresponding sound is discarded. A very important aspect to check is the participation of children, certainly linked to the sounds produced. Some sounds can be associated with playful situations or, in any case, pleasant: these sounds and the corresponding objects / concepts are to be preferred over other sounds associated with objects not familiar to children.

Implementation of the test with the synthesis signals: execution modality. TAV n ° 19

Once the signals suitable for performing the test have been identified, the test is carried out on patients to determine the hearing threshold.

The test is administered by medical personnel, possibly assisted by school personnel (GROUP-ANALYSIS). The various sounds, corresponding to the synthesis signals (SIGNAL-1, SIGNAL-3), which cover the frequency range suitable for the purpose of the test, at different intensities, to identify the minimum level of intensity at which each sound is still perceived by the subject, which indicates the corresponding concept / object (CONCET-1, CONCET-3). In this way the perceptual threshold in the different frequency zones is determined. Based on the value of this threshold, it can be deduced whether the subject can be considered normal or hearing impaired.

Note; the actual test is preceded by a short training to validate, by the subjects participating in the test, the correct association between sound and corresponding object / concept.

Note: Table n ° 19 also shows the acoustic diffuser, obviously necessary for the transformation of electrical signals into acoustic signals.

Execution modality of the test with synthesis signals: case of normoacusico subject. TAV n ° 20 Different images (CONCET) are made available to the child (SOGG), each corresponding to one of the sounds (SIGNAL) proposed by the GROUP-ANALYSIS during the test. The images, corresponding to the different signals, could be printed on a board and the child indicates the correct image, or they could be printed on different supports, for example cards, and the child lifts the card (ANSWER) corresponding to the correct corresponding image to the perceived sound. If the child correctly perceives all sounds, he is able to correctly associate each sound (SIGNAL) with the corresponding object / concept (CONCET).

Execution modality of the test with synthesis signals, in the case of a hypoac tisic subject. TAV n ° 21 The test method is the same as that applied for the normal-hearing child: on the other hand, the condition of any hearing loss is identified only at the end of the test. If the child perceives all sounds correctly, he is able to correctly associate each sound with the corresponding object / concept. If the child (SOGG) has some difficulty in perceiving some sounds, he will show some indecision, or even some error, (ANSWER) in determining the exact object / concept.

Execution modality of the test with synthesis signals on a group of children. TAV n ° 22 Different images (CONCET) are made available to the group of children (SOGG-1, SOGG-2, SOGG-N), each corresponding to one of the sounds (SIGNAL) that are proposed by the GROUP-ANALYSIS during the test. In the case of groups of several children, the images corresponding to the different concepts / objects (CONCET) can be printed on different supports, for example cards. When perceiving the sound, each child lifts the card (ANSWER) corresponding to the correct image corresponding to the perceived sound. This modality has already been successfully tested on two groups of children in class before a primary school. If the child perceives all sounds correctly, he is able to correctly associate each sound (SIGNAL) with the corresponding object / concept (CONCET). In the case of hearing-impaired children (SOGG-N) the answers provided may be incorrect or given after a time significantly longer than the response time of other normal-hearing children. The staff (GROUP-ANALYSIS) that manages the test observes and takes note of the different situations. A possible problem with this method lies in the mutual proximity of the children, who can copy the action of the neighboring child.

Execution modality of the test, with automatic method, with the synthesis signals, on a single child. TAV n ° 23

A master PC manages the entire test and, through communication with the notebook, or similar device, knows the children's response, both in terms of correctness and reaction time. The child has his own notebook. Different images are made available to the child (SOGG), each corresponding to one of the sounds presented during the test. The images can be inserted in a synoptic displayed on the screen of the (NOTEBOOK) equipped with a touch screen, or similar equipment. The child's answer is given by touching the correct image on the screen. The Notebook is connected to the computer (PCMASTER) which also manages the generation of synthesis signals. In this way, by developing an appropriate software application, it is possible, by the master PC, to store the child's answers, verifying their correctness and response times. The staff (GROUP-ANALYSIS) verifies the correct execution of the! test even without the need to intervene directly on the sequence and the generated signals. In this way, at the end of the test, you already have the information relating to the child's acoustic perception, with the knowledge of the association between the minimum level of perception and the corresponding part of the spectrum. Furthermore, the test can be developed completely automatically, thanks to appropriate algorithms that set the sequence of signals to be applied to the subject: these algorithms, based on the child's responses, modify the sequence of signals, varying their intensity and spectral content. , in order to obtain an accurate test precisely in the frequency zones where the child shows uncertainty in the response, The medical and school staff observe the progress of the test. The evolution of the test takes place, essentially, in the same way as it would if checked manually by an audiometrist technician.

Execution modality of the test, with automatic method, with the synthesis signals, on a group of children. TAV n ° 24

The execution of the test is very similar to the test performed on a single child, with the only difference that the 1 (PC MASTER) is connected with several notebooks, or similar devices, (NOTEBOOK1, NOTEROOK2), one for each child. Also in this case the test can be developed in a completely automatic way, thanks to suitable algorithms that set the succession of signals to be applied to the subject, as indicated in the previous situation shown in TAV n ° 23

Warnings for an accurate test execution

Since the sound corresponding to the different synthesis signals must be perceived at the same intensity by all the children of the group, an accurate calibration of the acoustic diffusion system in the environment in which the test is carried out is necessary, as well as an accurate distribution of children in the environment.

Execution modality of the test with direct and competing synthesis signals. TAV n ° 25

Having extremely accurate signals available (such as frequency content, time envelope and amplitude) there is the possibility of carrying out more demanding tests on the subject, compared to tests with a single acoustic source. There is a need for a signal generator capable of generating two different signals separately but simultaneously (GEN-DUAL). Having two acoustic sources (SORG-1, SORG-2), each providing a different signal (SIGNAL-H, SIGNAL-L), even at different intensities, there is the possibility to test not only the acoustic perception, but also the interpretative faculty at the highest level. In fact, stimulation with multiple sounds, different from each other, requires a cognitive and elaborative commitment on the part of the subject (SOGG) subjected to! test, larger than the test case carried out with a single sound. The answer can be given verbally by the child, indicating the images on a board (or similar support) or on the touch screen of a notebook, where each concept / object is represented with two images of different sizes. If the perceived sounds have the same intensity, the child will point to two images of equal size. If the perceived sounds have different intensities, the child will indicate two images of different size depending on the intensity of the perceived sounds. A test of this type requires a greater mental effort on the part of the child than the simple test of perception of a single sound, but it also allows an evaluation of the child's processing skills to be associated with the perceptual test.

CHARACTERISTICS, APPLICATIONS AND RESULTS WITH THE SYNTHESIS SIGNS

General characteristics of synthesis sounds

Absolute repetition; Totally controlled spectrum distributed over a frequency range extended as desired, with the possibility of allocating it, with frequency components even beyond the spoken spectrum; Exclusion of variants related to any subject who utters the sound, as the sound is not produced by a subject. Absolutely controlled signal structure such as time envelope and amplitude; Ability to simply switch from SPL stimulation to HL stimulation, or according to any other frequency weighting curve; Possibility of applying two or more signals, different from each other, at the same time, even at different intensities.

Characteristics in frequency and amplitude of synthesis sounds

A fundamental characteristic of a synthesized signal is that it has a strictly limited spectrum. The frequency extension can be between 1/3 of an octave 1 octave and 1⁄2, with a total absence of spurious components outside the frequency band corresponding to the signal produced. The two values indicated as minimum and maximum frequency extension are not mandatory, as they can also go down to 1/6 of an octave to extend up to 2 octaves and beyond, for composite signals. It is also possible to create signals consisting of a pure tone modulated in AM or FM.

With synthesis signals also the amplitude characteristics are extremely accurate and controlled, both as amplitude of each single forming component, and as overall amplitude intended as the effective value of the composite signal obtained. Furthermore, the dynamics obtainable of the corresponding acoustic sounds is limited only by the background acoustic noise of the environment and by the electronic noise of the generation and amplification equipment. In practice it is possible to obtain dynamic values even higher than 80 - 90 dB.

Being able to generate preliminary calibration electrical signals, and corresponding stationary acoustic signals, it is possible to obtain an accurate intensity measurement using sound level meters or the like. These sounds, after having calibrated the stimulation system, can be modulated in amplitude to obtain the required synthesis sounds.

Furthermore, there is the possibility of generating two or more sounds at the same time, even at different intensities, but in any case perfectly controlled, to carry out tests with noise masking or for the simultaneous action of signals belonging to disjoint parts of the acoustic spectrum. Sounds can also come from different directions.

Temporal characteristics of synthesis signals

Thanks to the synthesis technique, the duration of each signal produced is absolutely controllable and with an extremely accurate value. This makes these signals particularly suitable for carrying out tests of comprehension of the acoustic message by a subject. It is possible to set the duration of the signal starting from a minimum time, gradually lengthening it until the exact identification of the sound by the subject. Such a test is a further step forward from the simple frequency response test. Being able to set the time envelope in an absolutely arbitrary way, there is the possibility of freeing oneself from a type of sound linked to an object / concept suitable for a child, to switch to a "more technical" sound suitable for testing both on human subjects both on electroacoustic devices, verifying their behavior towards transient signals.

Characteristics relating to perception by the subjects: sound - object association Although these are synthesized sounds, their sound characteristics are such that they can be immediately associated, even by a preschool child, with objects / concepts well known to the child himself . This is due to the fact that the generated sounds can very clearly resemble real sounds. With a careful choice of synthesis sounds and the corresponding association with an object / concept, the identification of the meaning of the sound by the child is often immediate and does not even require preliminary training,

Possible applications in pediatric audiology

Different functions can be obtained with the synthesis signals: Assessment of auditory perception by determining the audibility threshold in the different frequency zones of the audible spectrum in a manner more similar to a game (even collective) rather than to an acoustic visit; Verification of the correct interpretation of sounds in the absence of background noise; Verification of the correct interpretation of sounds even in the presence of competing signals or background noise; Verification of the correct functionality of the hearing aids (external prostheses or cochlear implants) in the event that the subject is a wearer of such devices, thanks to the possibility of setting the frequencies of the formants in correspondence with the frequencies of the electrodes of the cochlear implants; Assessment of the auditory perception level up to the extremes of the human audio band; Verification of an adequate level of processing in the case of complex signals or stimulation with multiple acoustic signals at the same time.

Possible applications in adult audiology

In addition to applications in pediatric audiology, further applications should be considered; Assessment of auditory perception in subjects with good auditory memory such as musicians or the like (using composite signals with particular acoustic / musical characteristics).

Possible applications in the tests of electroacoustic devices

Determination of the functionality in the frequency domain with respect to more complex signals than a pure tone, a burst or similar; Verification of the correct temporal response to transients referred to signals very similar to the real ones; Verification of the functioning of certain control algorithms active in hearing aid devices, used to improve the patient's auditory perception: Verification of the repetitiveness of the response to repeated complex signals; Verification of the behavior of implanted systems, such as cochlear implants.

The idea of synthesis signals was born as early as 2007-2008, with a degree thesis in "Audioprosthetic techniques" at the University of Padua, entitled: "Generation and analysis of electrical signals for electroacoustic tests" - Graduate student:

in which some foundations were laid for the generation of synthetic electrical signals in the acoustic spectrum. An interesting aspect was in the description of! generation method, which could be obtained with hardware equipment, with numerical methods or with mixed methods by combining numerical generation with hardware equipment.

Possible applications of synthesis signals in the speech spectrum

The various synthesis signals, which can be used to carry out tests aimed at determining the perception, by a subject, can be allocated frequently well beyond the 125 Hz -8 kHz band, typically used for speech perception tests. Any extension between 100 Hz and 10 kHz or, even more so, between 80 Hz and 12 kHz, allows for a more accurate analysis that covers the acoustic spectrum of speech, both male and female. Taking into account that the range between 125 Hz and 8 kHz is six octaves, and the range between 80 Hz and 12 kHz slightly exceeds seven octaves, in the hypothesis of generating synthesis sounds that cover an octave, or a little less, the need arises to generate from six to eight different signals, associated with as many corresponding concepts / objects. It is a limited typology of different signals, well suited to be associated, without difficulty, with corresponding objects / concepts by a child.

Possible applications of synthesis signals in the music spectrum

Unlike the verbal message, a musical information ranges frequently in a much wider range, which, to allow a good listening to the music, should start from no more than 30 - 40 Hz and reach at least up to 14 - 16 kHz. By means of a vocal type test, precisely because of a limited frequency content of the human voice, it is absolutely not possible to verify the correct behavior of a subject / device towards a musical message.

Synthetic sounds do not have this frequency limitation, being able to range over the entire audible range of a human subject and even beyond.

A further characteristic of a musical message lies in the greater dynamics compared to a verbal message. In the case of a verbal message, a realistic dynamic hardly exceeds 40 dB, while for a musical message this value can reach 80 - 90 dB. Consequently, the corresponding test signals should also reach these dynamics; this is possible only with signals obtained with synthesis methods and certainly not with signals obtained vocally.

It can therefore be concluded by saying that in order to carry out significant tests on subjects / devices, to evaluate the behavior towards an acoustic message of a musical type, the synthesis signals are certainly adequate.

The various test signals, for musical use, are frequently allocated in the 32 Hz - 16 kHz band (excursion that can also be extended outside this range), with the components of each signal that can be spread over a range of one octave or lower. Taking into account that the extension between 32 Hz and 16 kHz is 9 octaves, in the hypothesis of generating synthesis sounds that cover one octave, or a little less, the need arises to generate 9 different signals, which can be extended to 10 in the case of test up to 16 Hz. The applications of these sounds, as well as being suitable for children, to test hearing abilities on the entire acoustic spectrum, can be applied to all those patients who "work with sounds" such as musicians, audio technicians and all those subjects who have a good musical auditory memory.

Even the synthesis signals at the ends of the audio band have the same characteristics as the synthesis signals of the center of the band: spectral accuracy and well-determined amplitude and absolutely controlled temporal evolution.

With the signals allocated frequently at the extremes of the band it is also possible to carry out very accurate tests on electro-acoustic equipment or the like,

Translation in frequency with maintenance of the same object / concept The possibility of moving the different forming components on the frequency axis allows you to change the acoustic parameters of a sound, while maintaining the same concept associated with the sound. In this way the number of signal types, and corresponding sound, which can be associated with different concepts, can be extremely reduced, with a consequent simplification of the possible introductory part of the test consisting of training to recognize the various sounds and association with corresponding objects / concepts. ). As an example we can consider a sound that corresponds to a bell:

if the formant components are allocated in a lower part of the spectrum, the effect will be that of a larger bell than the bell corresponding to the formants allocated in a higher part of the spectrum. If the formant components are allocated in a higher part of the spectrum, the effect will be that of a bell of smaller dimensions than the bell corresponding to the formants allocated in a lower part of the spectrum, based on what has been said above it can be seen as a bell , which passes from a large bell (bell), to a normal bell (bell), to a small bell (bell), is always interpreted by the subject as the same object: the bell.

Good participation of very young subjects in the test carried out with synthesis sounds One of the most significant results lies precisely in the active participation of one or more children at the same time for the entire duration of the test.

A preliminary verification (without an actual acoustic calibration of the intensities) was carried out on two classes of first grade children (one class at a time) of the Comprehensive Institute The test was administered with six types of sounds, each made to be perceived at different intensity. Each sound corresponded to a representative figure printed on a tag, which each child had to lift when he recognized the perceived sound. For both classes, the test, including the first part of training, exceeded the duration of 18 minutes, during which the children maintained a fully active participation, just as if it were a collective game.

Further tests were carried out during the 2017-2018 school year, at the "Teatro delle Voci" of Treviso, on two groups of children, belonging to two School Complexes of Treviso: School Theater Laboratory "Teatriamo" aged between 6 and 14 years and two fourth grades of the “De Amicis” Primary School aged between 9 and 10 years. Also in this case the duration of the test exceeded 15 minutes. The test was integrated into a morning dedicated to acoustic physics experiments. After the tests were carried out, the participating children expressed their observations by actively contributing to the development of the method. Useful observations, which were found to be very useful for the optimization of the method, were also expressed by the school staff and some parents who participated in the morning of experiments at the Teatro delle Voci in Treviso.

During two of the four tests carried out, two children with mild hearing problems were identified, highlighted by a perception of the different sounds lower than that shown by the other children.

FINAL REMARKS

All the characteristics of the invention, indicated above as advantageous, convenient or the like, can also be missing or be replaced by equivalent solutions. The individual characteristics or solutions, set forth with reference to general teachings or particular embodiments, can be all present, or in part, in any other embodiments. They are also able to be replaced by solutions with different construction methods, but always inherent to the concepts expressed in the invention. The invention thus conceived is susceptible of numerous modifications and variations, all of which are within the scope of the inventive concept. In practice, the methodologies used, provided they are compatible with the specific use, as well as the mathematical rules may be any, according to requirements. Furthermore, all the details, solutions or elements of the invention can be replaced by other technically equivalent elements. It is understood that the invention is not limited to the particular configurations and / or manufacturing methods illustrated above, which constitute only non-limiting examples of the results resulting from the invention, but that numerous variants are possible, all within the reach of technicians operating in the signal sector, without thereby departing from the scope of the invention itself.

Claims (9)

CLAIMS TITLE: Generation of composite audio signals, obtained with synthesis methods, with strictly limited spectrum and duration, specific for measurements in pediatric audiology, for adults or for testing electroacoustic equipment. The signals generated can be associated with simple objects, images or concepts that can be uniquely determined. 1) Method for the generation of composite audio signals, specific for use in pediatric audiology, as they can be easily and immediately associated, even by a preschool subject, to very simple and uniquely determinable objects, images or concepts, from the spectrum and with a strictly limited duration, which method includes the steps of: - generation of at least one stationary or modulated signal, consisting of a single sinusoid or composite, of defined and adjustable amplitude, with a clock to control the repetition of the signal N times, with N greater than or equal to one - generation of at least one stationary or modulated noise, obtained from a white noise by means of a filtering operation, of defined and adjustable amplitude - sum of the stationary or modulated composite signal and the stationary or modulated noise. 2) Method according to claim 1, in which the generation phase of the stationary or modulated signal is carried out both with analog hardware devices, such as oscillators, function generators, modulators, adders, and other linear devices, and with devices controlled by numerical means such as DACs, digital attenuators, multipliers, adders, and other digital devices, both by combining the two solutions. 3) Method according to claims from n. 1 to n. 2, in which the generation phase of the stationary or modulated signal is achieved by adding together a finite number, with a value from 1 to N, of stationary sinusoidal signals, each with a frequency between fmin and fmax and with its own specific amplitude, obtaining thus a strictly limited band of the stationary or modulated signal. 4) Method according to claims from n. 1 to n. 3, in which the phase in which the amplitude modulation is implemented, is achieved by applying one at a time, or jointly, two different amplitude modulation modes: - unique AM modulation of the composite signal, obtained from the sum of sinusoidal signals with frequency between fmin and fmax - distinct AM modulation for each sinusoidal signal, as the waveform of the modulating agent is arbitrary, thus obtaining a strictly limited band of the modulated signal. 5) Method according to claims from n. 1 to n. 3, in which the phase in which the frequency modulation is implemented, is achieved by applying a distinct FM frequency modulation for each sinusoidal signal, the waveform of each modulating being arbitrary, this modulating obtainable by adding several sinusoidal components or other waveforms, thus obtaining a strictly limited band of the modulated signal. 6) Method according to claim n. 1, in which the phase in which a filtered noise is obtained, is realized starting from a white noise or similar, filtering this noise through the action of a high pass filter and a low pass filter, connected in cascade, with the frequency of cut of the high pass filter lower than the cutoff frequency of the low pass filter, obtaining a noise that remains in a limited band, except for the components at the extremes of the filtered noise band, of much smaller amplitude than the noise components in the center of the band. 7) Method according to claim n. 1 and n. 6, in which the amplitude modulation AM with respect to the filtered white noise, said noise added to the signal obtained according to claims from no. 1 to n. 5, does not alter the substance of the overall signal, which still remains in a limited band. 8) Method according to claims from n. 1 to n. 7, by which the result obtained by adding the composite signal, obtained as described in the claims from n. 1 to n. 5, with the filtered noise, obtained as described in the claims from n. 6 to n. 7, corresponds to a strictly limited band signal, with a frequency range between 1/3 octave and 2 octaves, extendable from a pure tone to 3 octaves and beyond, and limited time duration between a few fractions of a second and a few seconds. 9) Method according to claims from n. 1 to n. 8, through which the signal obtained has characteristics, in the time domain and in the frequency domain, suitable for diagnostic use in the audiological field, 10) Method according to claims from n. 1 to n, 8, by which, by changing the frequency domain of the signals generated around one octave, the acoustic frequency characteristics are changed, with consequent modification of the parameters of the audiological test, but the association with the same object / image / action, but with virtually different dimensions and / or types, thus being able to cover the entire auditory acoustic spectrum with a limited number of objects / concepts associated with the corresponding signals, with an association rule of the type "One to many" with an object that can be associated with several signals. 11) Method according to claims from n. 1 to n. 5, through which, by carrying out an adequate frequency positioning of the signal forming components, a test signal is obtained whose forming components coincide with the frequencies associated with the electrodes of the cochlear implants, or with the frequencies at which certain electroacoustic apparatuses perform specific functions. 12) Method according to claims from n. 1 to n. 11, through which, using: - a master personal computer, or similar, where appropriate control algorithms are implemented - one or more peripherals such as tablets with touch screens or similar, each associated with a subject under test, connected via network, via cable or wireless with the master personal computer, it is possible to carry out the audiological test, on a single subject or on a group of subjects, even automatically, where each subject subjected to the test interacts with the master personal computer through direct action on a synoptic or similar, displayed on the screen of the tablet or similar device.