EP1175812A1

EP1175812A1 - Method for the reproduction of sound waves using ultrasound loudspeakers

Info

Publication number: EP1175812A1
Application number: EP00925256A
Authority: EP
Inventors: Wolfgang Niehoff; Vladimir Gorelik; Oliver Gelhard
Original assignee: Sennheiser Electronic GmbH and Co KG
Current assignee: Sennheiser Electronic GmbH and Co KG
Priority date: 1999-04-30
Filing date: 2000-05-02
Publication date: 2002-01-30
Anticipated expiration: 2020-05-02
Also published as: DE50007789D1; US20050207590A1; EP1175812B1; WO2001008449A1; EP1484944A2; EP1484944A3; ATE276636T1; AU4403600A

Abstract

Method involves using ultrasonic generator, whereby audio signal to be reproduced is combined with carrier signal in ultrasonic frequency range by sideband amplitude modulation. There is arrangement for subjecting modulated ultrasonic signal to dynamic error compensation and the compensated signal can be subjected to frequency path linearization if appropriate and then fed to a loudspeaker. There is a carrier reduction arrangement. An independent claim is also included for a use of an ultrasonic reproduction device.

Description

Method of reproducing audio sound with ultrasound speakers

The invention relates to a method for reproducing audio sound with ultrasound speakers and a construction of the ultrasound speakers and their application.

From J. Acoust. Soc. Am., Vol. 73, No. May 5, 1983 "The audio spotlight: An application of nonlinear interaction of sound waves to a new type of loudspeaker design" is already known to build a loudspeaker from several ultrasound emitters. Using such ultrasound emitters, audio sound can be emitted in a frequency range in that the audio sound itself can no longer be perceived by the human ear. Nonlinear effects in the air cause a high sound pressure and the superposition of two ultrasonic waves audible sound is generated. The high frequency of the ultrasound in comparison to conventional audio signals means that the sound is emitted in a highly spatially directed manner because of its small wavelength and the transducer dimensions of the ultrasound radiator, which are large in comparison. The frequency dependence of the directional characteristics of conventional loudspeakers - spherical emitters at low frequencies, directional emitters at high frequencies - hardly occurs with an ultrasound loudspeaker.

There is also a shifting effect in the conference volume AES, 26.-29. September 1998, San Francisco, California, "The Use of Airborne Ultrasonics for Generating Audible Sound Beams". From this, too, considerations for generating audible sound based on the radiation of the audio sound by means of ultrasound are known.

Furthermore, the phenomenon of the generation of sound waves by means of ultrasound emitters is also known from the magazine Audio, issue 8, 1997, pages 7-8. It is described here that a loudspeaker system emits a first signal of 200 kHz and the loudspeaker system emits a second signal with the same frequency of 200 kHz, the second signal being modulated with the audio sound signal (20 Hz to 20 kHz). The non-linear behavior of the air produces a mixing result when the two signals are superimposed, so that the difference between the two signals can be heard as acoustic sound.

As further prior art, reference is made to documents US-A-4,872,148, US-A-4,439,642, US-A-4,439,641, US-A-4,409,441, US-A-4,280,204, US-A-4, 199,246, WO-A- 85/02748, EP-A-0 164 342, EP-A-0 154 256, CA 1 274 619, CA 1 215 164, CA 1 195 420, CA 1 120 578, AU-A-28287/77, AU- A-510193, WO98 / 39209, WO98 / 02976, WO98 / 02977, WO98 / 02978, WO98 / 26405, GB-A-2 225 426, DE-A-27 39 748, US-A-5,375,099, CA 1 274 619 , DE-A-196 28 849, US-1, 616,639, US A-1, 951, 669, US-A-2, 461, 344, US-A-3,398,810. Further features of the ultrasound loudspeakers are described in the aforementioned literature references.

Although there have been various approaches to ultrasonic loudspeakers, such a product has not yet been able to establish itself on the market. This is also due to the fact that despite the special properties of ultrasound loudspeakers, some problems arise that are partly related to the nature of the ultrasound propagation, but also with the ultrasound emitter itself.

The invention has for its object to improve a method for reproducing audio sound and an ultrasound loudspeaker compared to the previous approaches, so that high-quality sound reproduction is possible. The object is achieved with a method according to claim 1 and an ultrasonic loudspeaker according to claim 2. Advantageous further developments are described in the subclaims and in the description below.

The method according to the invention combines low-frequency audio sound with the strong directional characteristic of ultrasound. The directional characteristic of the loudspeaker is therefore almost independent of the signal frequency. To understand the invention and its nature, the following should be pointed out: Mathematically, with formulas of nonlinear acoustics, it can be shown that at high sound pressure levels (p> 110 dB at 40 kHz), new waves are generated due to the nonlinearity of the medium, when several waves interact to stand by each other. The frequencies of these waves correspond to the sum and difference frequencies of the original waves and multiples thereof (n • © ι ± m • ω ₂ where ωi and ω _∑ frequencies of the initiated sound waves (tones) and n, m are integers). The sum and difference frequencies occur in every frequency range. Clear advantages Compared to conventional loudspeakers, this results in the ultrasound range in that a very strong directional characteristic of the transducer can be realized and which is outside the human hearing range. The initiating signals - i.e. the ultrasonic waves - are inaudible.

If e.g. A first tone with a frequency of 200 kHz and a second tone with a frequency of 201 kHz is emitted into the air at high sound pressure, so there are sum and difference tones in the overlapping zone of the two tones. The first sum tone (f = 200kHz + 201kHz = 401kHz) is not audible. The first difference tone is used to generate audible sound

(f = 200kHz-201 Hz = 1 kHz) used (Figure 4). This difference tone is much louder than all other tones resulting from the interaction. Sum and difference tones only emerge in a nonlinear medium such as air as distortion products.

The difference tones generated have the property that the propagation of the difference tones (secondary sound) takes place in the direction of the ultrasound to be generated (initiating tones, primary sound). Furthermore, the difference tones are only audible in the area of ultrasound, i.e. the directional characteristic of the difference tones corresponds to that of ultrasound. Finally, the sound pressure of the differential tones increases with the frequency of the ultrasound.

In the technical implementation of an ultrasound loudspeaker according to the invention, the still low-frequency audio signal to be reproduced is first subjected to frequency response linearization (FIG. 1, FIG. 2). This signal is then linked to a carrier signal in the ultrasound frequency range by means of a double sideband amplitude modulation. This ultrasound signal is then subjected to dynamics (error compensation (compression)), the compressed signal is subjected to a second frequency response linearization and this signal is in turn fed to the ultrasound loudspeaker. As an alternative to the above-described formation of the ultrasound signal, instead of double-sideband amplitude modulation, single-sideband amplitude modulation can be provided, the ultrasound carrier preferably being suppressed by a few dB, for example 12 dB (FIG. 2).

The ideal center frequency, i.e. the mean value between the ultrasound carrier frequency and the sideband frequency (range) of the emitted ultrasound signal results from the intended application. Two groups can be specified here: A. Use in the close range up to approx. 50cm; B. Use at a distance of more than 50cm up to remote sound.

Different requirements for the center frequency can be derived from this division of the range. The level of the audible sound pressure depends to a large extent on the sound pressure of the ultrasound signal, the non-linearity parameter of the medium, the frequency of the resulting audio signal and the distance to the source and the attenuation of the medium. The differential frequency wave DFW - the audible sound - builds up with increasing distance from the source. Due to the damping of the ultrasonic wave in the air, the greatest sound pressure is reached at a certain distance until the level drops again as the distance increases due to damping. The damping of the ultrasound in the air in turn depends on the ultrasound frequency. The higher the frequency, the higher the absorption of ultrasound in air.

For practical applications, this means that an ideal frequency range of approximately 40 kHz to 500 kHz (or more) can be specified for applications at a distance of greater than 50 cm to a few meters. On the one hand, the frequency range is chosen high enough to produce a DFW as effectively as possible and to ensure a sufficient frequency distance from the audible sound, but on the other hand low enough that the attenuation by the air does not have too great an influence on the audio sound. Another criterion is the directional characteristics of the Ultrasonic emitter. The higher the radiated frequency, the more directional the radiation.

A higher frequency makes sense for the close range (smaller than 50 cm), because the absorption of air in the near range is negligible, while the dimensions of the ultrasonic transducer are so small depending on the application that a stronger directivity is not achieved by the shape of the transducer, but can only be realized by increasing the ultrasound frequency.

The frequency shift of the low-frequency signal (speech, music, noises, sounds) in the ultrasound range takes place by means of amplitude modulation. This creates a carrier signal and an upper and a lower sideband, which contain the modulated information. At high sound pressure, the carrier signal, e.g. 200kHz and the lower sideband emitted via a converter and superimposed in the air. The non-linear behavior of the air creates a signal whose frequency corresponds to the difference between the carrier and sideband frequencies. The higher the frequencies of the emitted tones with constant amplitude, the louder the resulting difference tones. The sound pressure of the difference tones increases quadratically with the difference frequency of the emitted ultrasound tones. A high ultrasound frequency can maximize the directivity that can be achieved and increase the frequency distance of the emitted ultrasound from the human hearing range.

The sound pressure of the differential frequencies results, among other things, from the product of the signals to be mixed. When an amplitude-modulated signal is emitted, the carrier is emitted in full even in the event of a modulation pause, ie when there is no signal at the modulator. The amplitude of the wearer means constant noise pollution for the ears and permanent electrical stress on the transducers. With a normal amplitude modulation, the amplitude of a sideband is mxA _τ / 2 (with m = modulation index and A _τ = carrier amplitude). The Carrier is continuously emitted and has a greater amplitude than the sideband that is modulated in time with the low frequency. These above-mentioned problems can be sensibly eliminated with the measures described below. A noise reduction can be achieved if the amplitude of the carrier is reduced, for example by a filter or already in the modulator by partial carrier suppression, and at the same time the amplitude of the upper sideband is increased. This reduces the continuous level and increases the relative, level-related change in the level due to the modulation. In the event of carrier suppression, the lower sideband must be strongly suppressed to prevent the two sidebands from mixing with one another, which would cause severe distortion. The measure described above can also be generally referred to as "carrier reduction".

If the carrier amplitude is modulated with the amplitude of the signal to be transmitted, no signal is emitted in the event of a pause in modulation. An additional controlled compressor stage is then required, which compensates for amplitude errors resulting from the modulation of the carrier. To eliminate the problem described above, the carrier amplitude can be modulated in time with the signal to be modulated.

Furthermore, a problem described above can be countered by compressing the signal to be modulated so that the signal's dynamics are reduced and, in particular, the volume of the quiet signal passages is increased. This allows the modulator to be optimally controlled. After the modulation, the compression must be compensated for by an expansion in order to maintain the original dynamics. With the described compression of the modulation signal before the modulation, very good results could be achieved.

Another measure to eliminate the above problem is in Modulation pauses to suppress control of the converter with the carrier signal (muting), so that the modulator output signal is faded out when no input signal is present.

The amplitude-modulated low-frequency vibration is emitted with a transducer at high sound pressure. In the air, the interaction between the carrier oscillation and the modulated sideband creates a difference frequency spectrum that corresponds to the spectrum of the low frequency. In order to achieve a low distortion factor, single sideband modulation is particularly preferred. If the carrier is partially suppressed in a conventional two-sideband amplitude modulation, then suppression of the lower sideband is essential because the mixing of the two sidebands with one another brings about additional difference frequencies which are undesirably noticeable in the form of a distortion factor.

With piezoelectric transducers, however, the radiation of the modulated signal is so narrow-band that the lower sideband is reproduced only very quietly. The mixing of the side bands with each other is therefore negligible in terms of sound pressure. However, this presupposes that the carrier is so loud that the mixture of carrier and sideband gives a much louder signal than mixing the sidebands with each other. Accordingly, the modulation is implemented either as a conventional double-sideband amplitude modulation or as a single-sideband amplitude modulation, in which the carrier is suppressed by, for example, 12 dB for further function optimization.

The relationship between the electrical input signal of the piezoelectric wander and the sound pressure level of the differential tones is not linear. A linear transmission can be achieved with a compensation circuit (dynamic compression). Frequency response linearization, which is required in particular in the case of piezoelectric transducers with a strongly nonlinear frequency response, compensates for frequency-dependent amplitude errors in the transmission system. The equalization can take place before the modulation in the low frequency range or after the modulation in the ultrasound range. Equalization after modulation has the advantage that the modulation reserve of the modulator is not restricted when a frequency range is raised.

The difference sound wave arises in the emitted ultrasound cone. The cross-section of the cone has an influence on the resulting audio frequency response. The audible signal is generated at an interface that is held into the sound beam. The lower limit frequency depends on the cross-sectional area of the object placed in the beam. In order to achieve a linear frequency response for a reflector on a wall, an equalization matched to the surface of the reflector is necessary (area-related equalization).

The maximum of the sound pressure results at a certain distance from the ultrasound source. It occurs at different intervals for different audio frequencies. A linear frequency response can therefore only be achieved for a certain distance by special distance-related equalization. The signal processing must therefore include a special distance-dependent frequency response equalization for a linear frequency response.

To generate a high level of ultrasound, a larger number of transducers are connected in parallel. It was found that the arrangement of the transducers plays a major role. So transducers are arranged on a plate as close as possible, so that the depth reproduction of the loudspeaker is quieter than in an arrangement in which the same number of transducers is mounted in a ring. The described analog amplitude modulation can also be implemented digitally. The multiplication of a sinusoidal oscillation (carrier) with a low frequency signal, partial suppression of the carrier as well as the suppression of the lower sideband with a digital signal processor module is possible. Frequency response contours can also be carried out relatively easily when using a digital signal processor.

The level of the audio sound pressure also depends, among other things, on the non-linearity parameter of the acoustically permeable medium. For air, the parameter is _ = 1, 2. For the medium water is _ = 3.5. It has now been found that an extreme value of _ of over 5000 can be given for a water-air mixture, which means that the sound pressure can be increased by a factor of 4000 compared to the medium air with a water / air mixture. In this way it is possible, for example, to implement a water / air mixture in a headphone cup, so that the water / air mixture medium is arranged between the ultrasound emitter and the receiver and increases the sound pressure of the audio signal.

The audio sound pressure can also be increased further by other measures. Due to the increasing division of the wavefront in the course of the propagation, which is synonymous with the creation of harmonics. After an energy balance, the energy contained in the harmonics is not available for the differential sound wave. In a way, there is an energy flow from the fundamental to the harmonics. If this energy flow can be slowed down, the audio sound pressure could be increased. A proposal for this is as follows:

A sound-permeable medium contains small cavities, which together with the material result in a large number of Helmholz resonators. The resonators are tuned to the first harmonic of the signal and thereby slow down the energy flow higher harmonics. If the cavities are filled with a non-linear medium, for example a liquid, this measure can achieve a higher value for the non-linearity parameters, which would increase the sound pressure of the differential tones.

With this technology, reflectors can be built that passively amplify the sound pressure of the differential tones.

For an ultrasound loudspeaker built into the headrest of a car, the described "damping plate" enables a higher level of audio sound to be achieved with simultaneously reduced ultrasound. For wireless headphones, it would be conceivable to transmit inaudible ultrasound wirelessly and to bring the differential tones to a sufficient level via the absorber described above.

Mathematically, with formulas of nonlinear acoustics, it can be shown that at high sound pressure levels (p> 110dB at 40kHz) new waves arise due to the nonlinearity of the medium air, if several woes are mutually interrelated.

The frequencies of these waves correspond to the sum and difference frequencies of the original waves and multiples thereof.

(n ^* ω, + m (ö! with ωι, ω∑: frequencies of the initiated tones and n, m: gaäne numbers).

The sum and difference frequencies occur in every frequency range. There are clear advantages over conventional loudspeakers in the ultrasound range, in which a very strong directional characteristic of the transducers can be realized and which is outside the human hearing range; the initiating signals are inaudible.

Example: If a tone with a frequency of 200 kHz and a second tone with a frequency of 201 kHz are emitted into the air at high sound pressure, sum and difference tones are produced in the overlapping zone of the two tones. The first sum tone (f = 200kHz + 201kHz = 401 kHz) is not audible. The first differential tone (f = 200kHz-201kHz = 1kHz) is used to generate audible sound. It is also much louder than all the other tones generated by the interaction. It is only in a nonlinear medium such as air that distortion products emerge that produce sum and difference tones.

Properties of the generated differential tones

- The propagation of the secondary sound (the differential tones) takes place in the direction of the primary sound (the initiating tones),

The secondary sound can only be heard in the area of the primary sound, i.e. the directional characteristic of the secondary sound corresponds to that of the primary sound,

- The sound pressure of the differential tones increases with the frequency of the initiating tones.

Technical implementation (exemplary embodiment of the invention):

FIG. 1 and FIG. 2 show block diagrams of an ultrasound loudspeaker, FIG. 2 representing an improved circuit compared to FIG. 1.

As can be seen in FIG. 1, the low-frequency audio signal is first subjected to frequency response linearization and then subjected to double-sideband amplitude modulation (and / or frequency and / or phase modulation), the carrier frequency being in the ultrasound range. Thereafter, dynamic compression or dynamic error compensation (signal-dependent) is carried out, if necessary. Another frequency response linearization then takes place and the signal that is then output is fed to the ultrasound transducer. The circuit according to FIG. 2 differs from FIG. 1 essentially in that instead of the double sideband amplitude modulation, a single sideband amplitude modulation is carried out, the carrier being suppressed by approximately 12 dB in the ultrasound range.

The ideal center frequency, i.e. the mean value between carrier frequency and sideband frequency (range) of the emitted ultrasound signal results from the intended application. Two groups can be specified:

1. Applications in the close range up to approx. 50cm

2. Applications at a distance> 50cm and remote sound

Different requirements for the center frequency can be derived from this division of the range. The level of the audible sound pressure depends on the sound pressure of the ultrasound signal, the non-linearity parameter of the medium, the frequency of the audio signal that is generated, and the distance from the source and the attenuation of the medium. The differential frequency wave builds up with increasing distance from the source. Due to the damping of the ultrasonic wave in the air, the greatest sound pressure is reached at a certain distance until the level drops again as the distance increases due to damping. The attenuation of ultrasound in the air depends on the frequency. The higher the frequency, the higher the absorption of the sound in air.

For practical applications, this means that an ideal frequency range of approx. 80 kHz to 180 kHz can be specified for applications at a distance of> 50 cm to a few meters. On the one hand, the frequency range is chosen high enough to produce a DFW as effectively as possible and to ensure a sufficient frequency distance from the audible sound, but on the other hand low enough that the attenuation by the air does not have too great an influence on the audio sound. Another criterion is the directional characteristic of the emitter. The higher the radiated frequency, the more directional the radiation.

A higher frequency makes sense for the near range, because the absorption of air in the near range is of negligible size, while the dimensions of the transducer are so small, depending on the application, that a stronger directivity is not achieved by shaping the transducer, but only by increasing it Ultrasonic frequency can be realized.

Frequency shift of the low frequency signal

The frequency shift of the low-frequency signal (speech, music, noises, sounds) into the ultrasound range takes place by means of amplitude modulation. This creates a carrier signal and an upper and a lower sideband, which contain the modulated information.

At high sound pressure, the carrier signal, e.g. 200kHz, and the upper sideband emitted via a converter and superimposed in the air. The non-linear behavior of the air creates a signal whose frequency corresponds to the difference between the carrier and the sideband frequency. The higher the frequencies of the emitted tones with constant amplitude, the louder the resulting difference tones. The sound pressure of the difference tones increases quadratically with the difference frequency of the emitted ultrasound tones. A high ultrasound frequency allows the directivity that can be achieved to be maximized and the frequency distance of the emitted ultrasound from the human hearing range to be increased.

Inadequacy in amplitude modulation: permanent carrier amplitude The sound pressure of the differential frequencies results, among other things, from the product of the signals to be mixed. When an amplitude-modulated signal is emitted, the carrier is emitted in full even in the event of a modulation pause, ie when there is no signal at the modulator. The high amplitude of the wearer means constant noise pollution for the ears and permanent electrical stress on the transducers. With a normal amplitude modulation, this is

Amplitude of a sideband m ^* - (with m = modulation index and

A _τ : carrier amplitude). The carrier is continuously emitted and has a greater amplitude than the sideband, which is modulated in time with the low frequency. The following measures therefore make sense:

carrier reduction

Noise reduction can be achieved if the amplitude of the carrier is reduced, e.g. by a filter or already in the modulator by partial carrier suppression, and at the same time the amplitude of the upper sideband is increased. This reduces the continuous level and increases the relative, level-related change in the level due to the modulation. In the event of carrier suppression, the lower sideband must be strongly suppressed to prevent the two sidebands from mixing with one another, which would cause severe distortion.

Modulation of the carrier amplitude in time with the signal to be modulated

If the carrier amplitude is modulated with the amplitude of the signal to be transmitted, no signal is emitted in the event of a pause in modulation. An additional controlled compressor stage is then required, which compensates for amplitude errors resulting from the modulation of the carrier. Compression of the modulation signal before the modulation

Compression of the signal to be modulated means that the dynamics of the signal are reduced and, in particular, the volume of the quiet signal passages is increased. This allows the modulator to be optimally controlled. After the modulation, the compression must be compensated for by an expansion in order to maintain the original dynamics.

mute

In order to suppress activation of the converters with the carrier signal during modulation pauses, the modulator output signal is faded out when no input signal is present.

Practical design of the modulator

The amplitude-modulated low-frequency vibration is emitted with a transducer at high sound pressure. In the air, the interaction between the carrier oscillation and the modulated sideband creates a difference frequency spectrum that corresponds to the spectrum of the low frequency. In order to achieve a low distortion factor, single sideband modulation is optimal. If the carrier is partially suppressed in the case of an ordinary double-sideband AM, suppression of the lower sideband is essential because the mixing of the two sidebands with one another brings about additional differential frequencies, which are noticeable in the form of distortion.

With piezoelectric transducers, however, the radiation of the modulated signal is so narrow-band that the lower sideband is reproduced only very quietly. The mixing of the side bands with each other is therefore negligible in terms of sound pressure. But that presupposes that the carrier is so loud that the mixture of carrier and sideband gives a much louder signal than mixing the sidebands with each other.

The modulation is therefore realized either as a conventional double-sideband AM or as a single-sideband AM, in which the carrier is suppressed by approximately 12 dB for further function optimization.

Dynamic compression (dynamic error compensation)

The relationship between the electrical input signal of the piezoelectric transducers and the sound pressure level of the differential tones is non-linear. A linear transmission can be achieved with a compensation circuit.

Linearization of the frequency response

Frequency response linearization, which is required in particular in the case of piezoelectric transducers with a strongly nonlinear frequency response, compensates for frequency-dependent amplitude errors in the transmission system. The equalization can take place before the modulation in the low frequency range or after the modulation in the ultrasound range. Equalization after modulation has the advantage that the modulation reserve of the modulator is not restricted when a frequency range is raised.

Area equalization

The difference sound wave arises in the emitted ultrasound cone. The cross-section of the cone has an influence on the resulting audio frequency response. The audible signal is generated at an interface that is held into the sound beam. The lower limit frequency depends on the cross-sectional area of the object placed in the beam. To create a linear for a reflector on a wall To achieve a frequency response, an equalization that is tailored to the surface of the reflector is necessary.

Distance-related equalization

The maximum of the sound pressure results at a certain distance from the source. It occurs at different intervals for different audio frequencies. A linear frequency response can therefore only be achieved for a certain distance by special distance-related equalization. The signal processing must therefore include a special distance-dependent frequency response equalization for a linear frequency response.

Increasing the sound pressure through a large number of transducers

To generate the high ultrasound level, a larger number of transducers are connected in parallel.

The arrangement of the transducers plays a role here: If the transducers are arranged as close as possible on a plate, the depth reproduction of the loudspeaker is quieter than in an arrangement in which the same number of transducers is attached in a ring.

Modulation through digital signal processing

The described analog amplitude modulation can also be implemented digitally. Multiplication of a sine wave (carrier) with a low-frequency signal, partial suppression of the carrier and suppression of the lower sideband are possible with a DSP module - FIG. 3 -. Frequency response corrections can also be carried out relatively easily. Nonlinearity parameter

The level of audio sound pressure depends, among other things. on the nonlinearity parameter of the medium. For air, the parameter is ε = 1, 2. For the medium water is ε = 3.5, for water with air bubbles an extreme value of ε = 5000 can be specified. Compared to the medium of air, a sound pressure that is 4000 times higher can theoretically be achieved.

A suitable medium between the ultrasound emitter and the listener can increase the sound pressure of the audio signal.

The audio sound pressure can be increased by another measure. Due to the increasing division of the wavefront in the course of the propagation, which is synonymous with the creation of harmonics. After an energy balance, the energy contained in the harmonics is not available for the differential sound wave. In a way, there is an energy flow from the fundamental to the harmonics. If this energy flow can be slowed down, the audio sound pressure could be increased.

A proposal for implementation looks like this:

A sound-permeable medium contains small cavities, which together with the material result in a large number of Heimholtz resonators. The resonators are tuned to the first harmonics of the signal and thereby brake the energy flow to higher harmonics. If the cavities are filled with a non-linear medium, e.g. a liquid, this measure allows a higher value for the nonlinearity parameter to be achieved, as a result of which the sound pressure of the differential tones has been increased.

This technology makes it possible to build reflectors that passively pass through Increase the sound pressure of the differential tones.

For an ultrasound loudspeaker built into the headrest of a car, the described “damping plate” enables a higher level of audio sound to be achieved with simultaneously reduced ultrasound.

For wireless headphones, it is conceivable to transmit inaudible ultrasound wirelessly and to amplify the differential tones to a high level using the absorber described above.

Practical applications

Since audible sound is only generated in the overlapping zone of the mixed ultrasound signals, the spatially separated radiation of carrier and sideband signals via own transducers enables an almost point-like "projection" of the sound. The radiation of both signals via a single transducer or a transducer array, however, changes the point-like into a linear characteristic along the direction of propagation of the ultrasound.

Practical applications of the ultrasound speaker are primarily those in which the strong directivity of the speaker is used. In applications a) - e), an absorbent material behind the area to be sonicated prevents backward reflection of the ultrasound.

a) Art objects that "speak" sonication of an art object in such a way that the sound can only be heard in the immediate vicinity of the object. The transducer can, for example, be arranged above the object and can only be heard within a small area around the object surrounding area does not occur. b) Active noise compensation for cars, planes, buses, trains: The ambient noise is recorded and analyzed with a microphone. With an electronic circuit, a signal with the opposite phase is generated and with the ultrasound transmission method it is emitted in a directional and seat-dependent manner. The superimposition of the sound with the counter-sound produced reduces the ambient noise.

c) Conference systems for spatially addressable sound in different languages: In conference rooms, the individual seats are selectively sounded without the respective neighbor being disturbed. Different languages can be transmitted simultaneously without headphones.

d) Loudspeakers in the airplane, bus, train as a replacement for headphones: The strong directional effect of the ultrasound loudspeaker enables sound to be delivered using loudspeakers instead of headphones. This is possible through the implementation of electrically or mechanically swiveling spotlights and allows "audio on demand".

e) Directed sound on stage (prompter)

f) In the car as an addressable loudspeaker (transducers installed in the headlining or headrest can be controlled via a control panel with a matrix display)

g) Sound from computer workstations on the monitor. Transducers are attached around the picture tube of the monitor. The sound is only audible in front of the monitor.

h) "Ultrasound wallpaper" or ultrasound ceiling for active noise compensation in the home, function see above.

i) Surround speakers: exploitation of wall reflections: "projection" of the Surround information on the room walls where virtual sound sources are to be located. The rear speakers do not necessarily have to be placed behind the listener.

j) PA systems for PA applications: Acoustic "illumination" of very specific zones. Thereby, delimitation of the surrounding areas (audio on demand).

k) Hands-free system (in the car for telephoning): Due to the strong directivity of the loudspeaker, the microphone can be attached appropriately so that there is no acoustic feedback between the loudspeaker sound and the recorded microphone sound. Combination of ultrasound loudspeaker and directional microphone to avoid acoustic feedback: The loudspeaker is arranged above the listener, for example, while the directional microphone is directed towards the speaker. The strongly directed sound of the ultrasound loudspeaker does not reach the microphone, so that there can be no acoustic feedback (e.g. in TV studios for viewer questions.

I) If an ultrasound loudspeaker is installed at each seat, one can

Forward the call to every seat without having to hand over the phone.

In the method for reproducing audio sound described here, only inaudible ultrasound is emitted into the air via a special transducer.

Non-linear effects in the air produce audible sound at high sound pressure and the superposition of two ultrasonic waves. The high frequency of ultrasound in comparison to conventional audio signals has the effect that the radiation of the sound is relative because of its small wavelength and in comparison to it large transducer dimensions are strongly spatially directed. The frequency dependence of the directional characteristics of conventional loudspeakers (spherical emitters at low frequencies, directional emitters at high frequencies) hardly occurs with this loudspeaker.

The process combines low-frequency audio sound with the strong directional characteristic of ultrasound. The directional characteristic of the loudspeaker is therefore almost independent of the signal frequency.

Reduction of distortion with amplitude modulation

The modulated signal is emitted using ultrasonic transducers. If the signal is a two-sided modulated AM signal, distortions caused by the principle can be reduced as follows:

1. through narrowband converters with high quality

2. with broadband converters through an upstream filter

The filter is omitted in the case of narrowband converters, since the transfer function of the converters is already equivalent to that of a narrowband filter.

The system is to be tuned so that the carrier frequency comes to lie approximately at the -6dB point of the filter edge. Cutting the lower sideband reduces the distortion.

Temperature-dependent drift of the filter flank of narrowband converters and filters must be compensated for by tracking the carrier frequency. The carrier frequency is updated as far as possible in signal pauses.

In the case of speech reproduction, the audio signal to be modulated should be filtered to increase speech intelligibility. The filter is to be designed in such a way that an attenuation of 3dB / oct. he follows. Reduction of distortion due to the converter geometry

If the transducer dimensions exceed the value of approx.% Of the lowest low-frequency wavelength to be emitted, distortions due to time differences of the signals increasingly occur in the near field of the transducer. The dimensions of the transducer should therefore be dimensioned smaller than the wavelength mentioned.

Supplement to the technical implementation of the modulation

An even more directed radiation of the audio tape can be achieved as follows:

The sound pressure of the audio band depends on the product of the sound pressure of the carrier signal and the sideband. By increasing the sound pressure - either the carrier or the sideband - the resulting sound pressure increases in the audio frequency range. The emission of a wide frequency range at high sound pressure poses certain difficulties.

The radiation from the carrier and sideband via a transducer or a transducer group places great demands on the transducers. Due to almost identical radiation conditions of the carrier and the sideband, the audio wave is generated in the entire overlapping area of the signals. This leads to a relatively broad radiation. An even sharper directional effect can be achieved by radiating the carrier and side band via separate transducers:

A special, very narrow-band, sensitive and very directional converter generates the carrier signal, while the sideband is overlaid with a broadband converter / converter array. Since the audio sound pressure results from the product of the two ultrasonic sound pressures to be superimposed, adjust the sound pressure of the wearer within wide limits the sound pressure of the audio wave and at the same time reduce the level of the ultrasound wearer at low volume levels. However, the superposition of the sound waves and the generation of mixed products takes place only in the area where both sound waves equally fill the room. The very strong possible directional characteristic of the carrier radiator also results in a very pronounced directional effect for the audio wave.

Absorption of the ultrasound signal by an ultrasound filter

To generate the audio signal from the modulated ultrasound signal, a certain path is required along which the wave demodulates in the air. If the ultrasound has covered the required distance, a filter that is permeable to audio frequencies but impermeable to ultrasound means that although the audio wave is clearly audible, the ultrasound signal is strongly attenuated. The filter has no significant effect on the directional characteristic of the converter.

The filter must be designed in such a way that it strongly attenuates frequencies above the listening area, while audio frequencies experience only a slight attenuation. It is sensibly arranged at the end of the generation zone.

Since a long generation zone is required for low audio frequencies, the lower limit frequency of the audio signal can be varied by varying the distance between the converter and the absorber.

Enrichment of the sound image through psychoacoustic effects The lower the frequency of the audio wave demodulated in the air, the lower the sound pressure of the wave, based on the constant sound pressure of the ultrasonic waves. For physical reasons, low frequencies can only be reproduced very quietly.

In order to create the subjective impression of reproducing deep tones that are not objectively present, a certain overtone spectrum can be generated by signal processing, which gives rise to this impression. Predistortion of the audio signal is required for this. The modulator contains a circuit that fulfills this function.

More applications

Virtual speaker

In order to make a sound object appear to move in the room, it is necessary to move the speaker in the room using conventional loudspeaker technology. This effect can be achieved more effectively with the ultrasonic loudspeaker.

By using the reflective properties for ultrasound hard surfaces, it can be achieved that the reflection of the ultrasound speaker on a wall or the like. similar to the light beam reflected in a mirror is perceived and a virtual source is created. For example, two implementations are possible:

1. Rotatable and swivel-mounted US loudspeaker

2. Fixed US loudspeaker that shines on a movably mounted reflector. Spatial signal transmission by a moving converter

When the listener moves, e.g. on a treadmill, escalator or similar can be carried by swiveling the transducer, so that only the moving listener is sonicated, but not the surrounding area.

Moving the audio sound can also take place with a connection of ultrasound emitters located above the listener, which are synchronized with the running speed of the treadmill / escalator, and which only sound at the areas of the room in which the listener is currently moving.

Spatial signal transmission through phased array

By specifically controlling individual transducer elements of an array, it is possible to carry a spatial signal (given the strong directional characteristic of the ultrasound loudspeaker) without moving the ultrasound emitter. The method is a combination of the "phased array" technique and the "ultrasound loudspeaker" described above.

FIGS. 4a and 4b show the propagation of an audio sound wave that is generated by an ultrasound transducer. In this case, the frequencies f, = 101 kHz and f ₂ = 100 kHz, for example, are emitted simultaneously by the ultrasound emitter (ultrasound transducer). Similar to a (non-linear) mixing stage of an AM medium-wave receiver, the mixed products f, + f ₂ = 201 Hz and f, -f ₂ = 1kHz and their multiples are created in the ultrasonic beam in the air. The sum frequency f, and f ₂ = 201kHz is not audible for humans, but the difference frequency f, - f ₂ = 1kHz. It is easy to imagine that one modulates f with the audio frequency range Δf = 100 ... 20kHz to f, = 100kHz + Δf. The mixture of the non-linearity of the air then creates exactly the audio frequency 100Hz ... 20kHz, among other things, whereby this has a similarly strong concentration as it is predetermined by the ultrasound beam.

In the mixing zone of the ultrasound beam, virtual audio sound sources (virtual loudspeakers) are created, which are added up in the direction of the continuous ultrasound, because ultrasound and audio sound propagate at the same speed of sound (340m / s). You can imagine this effect on a model. Small loudspeakers are mounted close to each other on a bar, all of which can emit audio sound as spherical emitters (FIG. 5) and which are controlled with the same audio signal with a time delay. The time delay t between two speakers is chosen so that it corresponds exactly to the time that the sound wave needs from one speaker to the next. It can be determined by the relationship t = c / l _L (c = speed of sound). The sound coming from the first loudspeaker is amplified by the second, etc. The large number of loudspeakers (an infinite number of virtual sound sources arise in the ultrasound beam), which are switched on depending on the location with the duration of the sound, results in a very strong bundling of the audio sound.

The audio sound in the ultrasound beam according to the invention is generated in the ultrasound beam itself. In contrast to the radiation from a conventional loudspeaker, it initially becomes louder with increasing distance until the ultrasound level has decreased to such an extent that the non-linear effect of the air no longer acts and therefore no more components are added to the audio sound generation become. The length of the active zone of audio sound generation in the ultrasound beam determines the lower limit frequency of the directional audio sound source. There must be at least as many virtual sound sources that the active zone is several wavelengths long at the lower cut-off frequency. For this reason, audio frequencies below 100 Hz require large distances between the listener and the ultrasound emitter (and therefore also high output powers). The use of psychoacoustic signal processing, as described above, offers a solution. It follows from the two effects described that the level and the lower reproduction frequency of the audio signal are location-dependent. The high ultrasound level, which is in principle necessary for generating the audio sound, only has to be present in the active zone of the ultrasound beam. Once the directed audio sound beam has been generated, the ultrasound component can be eliminated with an acoustic low-pass filter (ultrasound absorber that is permeable to audio sound).

FIGS. 6a and 6b show typical application examples of the ultrasound emitter, which is arranged under a ceiling and which directs ultrasound beams modulated with audio signals onto a wall, of which an ultrasound-absorbing coating (ultrasound reflection coating) is oriented so that ultrasound is absorbed. The then reflected audio signals are free of ultrasound and can be heard by people in front of the wall.

A conventional ultrasonic transducer can be used for the ultrasonic transducer itself. However, ultrasound foil converters are also particularly suitable which, in the manner of a capacitor (electret) transducer, have a foil and a correspondingly designed counter electrode (with grooves or holes).

The embodiment variant is also advantageous, in which a distance measuring device to an ultrasonic measuring device is used to determine where a listener to be exposed to is located. If this is in a critical area of the ultrasound beam that could be harmful to health, the ultrasound reproduction is switched off so that the person (or the animal) is not exposed to excessive ultrasound levels. If the ultrasound is to be directed to a specific area and if this area is also moving (this is the case, for example, with a single listener who is moving on a stage and is to be irradiated), it is advantageous for this if a device is designed by means of which the listener to be sonicated can currently be localized, so that the sonication then preferably only to the localized one Area. This can be achieved, for example, in that the listener to be sonicated carries a transmitting device with navigation (e.g. GPS) and thus constantly sends his own navigation data to a receiving device, which in turn is used to control the pivoting of the ultrasound beam. The listener to be sounded could also be equipped with a so-called TAG identifier, the exact position of which is determined by a corresponding interogator (interrogation unit for the TAG), which in turn controls the pivoting of the ultrasound beams. However, all other technical possibilities for localizing a single area or several areas can also be used to control the swiveling of an ultrasound beam, so that the audio reproduction can then only be heard in the desired narrow area, but not outside the desired area.

Such applications are particularly advantageous in a theater (for the prompter) or in the television studio for a TV show, if the moderator moving across the stage is to receive instructions that should not be audible to the rest of the audience.

The ultrasound beam can be swiveled using the different techniques described in this application, that is to say by swiveling the ultrasound emitter or by a swiveling reflector or by the so-called “phased array” control, the ultrasound beams being determined electronically in the direction.

Claims

Expectations

1. A method and apparatus for reproducing audio sound by means of an ultrasound generating device, the reproduced audio signal being linked by a sideband amplitude modulation to a carrier signal in the ultrasound frequency range, means being provided for subjecting the modulated ultrasound signal to dynamic error compensation and the compensated ultrasound signal is optionally subjected to frequency response linearization and then fed to an ultrasound transducer (loudspeaker), means being provided to reduce the amplitude of the ultrasound carrier signal (carrier reduction).

2. The method and device according to claim 1, characterized in that the ultrasound signal is suppressed (muted) in modulation pauses, ie when no audio signal is to be reproduced.

3. The method and device according to any one of the preceding claims, characterized in that the (still low-frequency) audio signal to be reproduced is subjected to a frequency response linearization before the modulation.

4. The method and device according to one of the preceding claims, characterized in that the audio signal to be reproduced is subjected to a double sideband amplitude modulation or a single sideband amplitude modulation.

5. The method and device according to one of the preceding claims, characterized in that means are provided to suppress the ultrasound carrier by an amount of about 8 to 20 dB, preferably 12 dB.

6. The method and device according to one of the preceding claims, characterized in that the frequency of the ultrasound carrier signal is in the range of about 40 kHz to 500 kHz.

7. The method and device according to one of the preceding claims, characterized in that means are provided in a double sideband amplitude modulation to suppress the lower sideband.

8. The method and device according to any one of the preceding claims, characterized in that means are provided to carry out an equalization (frequency response linearization) after the amplitude modulation.

9. The method and device according to one of the preceding claims, characterized in that a plurality of ultrasonic transducers is provided, which are connected in parallel.

10. The method and device according to claim 9, characterized in that the transducers are arranged on a plate as close as possible.

11. The method and device according to one of the preceding claims, characterized in that the modulation is carried out by means of a digital signal processor.

12. The method and device according to any one of the preceding claims, characterized in that a water air bubble mixture is arranged in the ultrasonic propagation path.

13. The method and device according to claim 12, characterized in that the water bubble mixture is formed in a headphone cup.

14. The method and device according to any one of the preceding claims, characterized in that a sound-permeable medium is arranged in the propagation path of the ultrasound beams, which contains cavities, which together with the medium material have a plurality of Helmholz resonators, which preferably on the first harmonic of the ultrasound signal are coordinated.

15. The method and device according to claim 14, characterized in that the cavities are filled with a non-linear medium.

16. The method and device according to one of the preceding claims, characterized in that a plurality of ultrasonic transducers are arranged in a ring.

17. The method and device according to one of the preceding claims, characterized in that the ultrasound carrier signal and the sideband signal and separate transducers is supplied.

18. The method and device according to one of the preceding claims, that the opening angle of an ultrasonic transducer is approximately in the range of 0.5 to 10 °, preferably 1 °.

19. The method and device according to one of the preceding claims, characterized in that means are provided for subjecting the audio signal to predistortion.

20. The method and device according to one of the preceding claims, characterized in that means are designed to pivot the ultrasound beam in a desired direction.

21. The method and device according to claim 20, that means for pivoting the ultrasound beam consists of a mechanical pivoting device of the ultrasound emitter and / or of an electronic control of the ultrasound emitter in the manner of a so-called “phased array” and / or that a pivotable reflector is formed which reflects the ultrasound in a desired direction.

22. The method and device according to one of the preceding claims, characterized in that the ultrasound device forms an ultrasound wallpaper, so that when listening, the impression arises that the sound comes directly from the wall (or the wallpaper on the wall).

23. The method and device according to one of the preceding claims, characterized in that the carrier band of the ultrasound beam band and the ultrasound beam side band is generated with different transducers.

24. The method and device according to one of the preceding claims, characterized in that the audio NF signal is subjected to psychoacoustic preprocessing (in particular psychoacoustic predistortion) and appropriate means are designed for this.

25. The method and device according to any one of the preceding claims, characterized in that the device is designed as an acoustic treadmill, so that the movement of a listener of an ultrasonic transducer is only sounded the moving listener, but not the surrounding area.

26. The method and device according to one of the preceding claims, characterized in that at least one ultrasound transducer is provided, which serves exclusively or in addition to the ultrasound radiation as a transmitting and / or receiving device of an ultrasound-based distance measuring device.

27. The method and device according to any one of the preceding claims, characterized in that the properties of the audio signal to be reproduced, in particular its lower cut-off frequency, is determined by the size of the reflection surface, in order thus to preferably compensate for the frequency response linearization or the equalization of the audio signal.

28. The method and device according to one of the preceding claims, characterized in that the audio signal to be reproduced is subjected to frequency and / or phase modulation in a modulator.

29. Use of an ultrasound reproduction device according to one of the preceding claims in an art exhibition and / or in a museum or for active noise compensation and / or in conference systems and / or as a loudspeaker as a headphone replacement and / or for directional sound on a stage (souffleuse) and / or as an addressable loudspeaker and / or for the sound reinforcement of computer workstations and / or as a surround loudspeaker and / or for acoustic sound reinforcement in very specific zones and / or in a hands-free system.

30. Use according to one of the preceding claims, for sonication of an area through which the listener moves or through which the listener is moved, the reproduction level of the ultrasound signal always being directed at the moving listener.