EP3178184B1 - Separating component for influencing sound transmission - Google Patents

Description

Technical area

The invention relates to a separating component for influencing the transmission of sound. It should serve to avoid disturbances and to ward off eavesdropping attacks.

State of the art

There are essentially two motivations for wanting to reduce the intelligibility of a speech signal. It is known that speech signals distract a person more than signals with no discernible information content, even if the volume is higher. This also applies to a certain extent when the signals are louder than the voice signal with no discernible information content. Distraction by voice signals occurs in open-plan offices, for example. Precisely because a low volume level can be achieved in open-plan offices through sound insulation and sound absorption, the distraction by a voice signal is relevant. One approach is to superimpose a noise signal on the speech signal, for example, so that an incomprehensible superimposed sound is created. The disadvantage here is that the superimposed sound has a comparatively high noise level.

The intelligibility of speech signals should often also be reduced in order to make eavesdropping more difficult. When it comes to security against eavesdropping on conversations, communications technology eavesdropping on radio or cable transmissions is undoubtedly in the foreground. From a technological point of view, these information technology systems can be protected by means of special technical measures and behavior, even if this temporarily leads to limited efficiency and convenience. It is different with the spoken word in rooms of buildings, even if Faradaysche Cages are in place, data lines are cut and mobile communication devices are switched off. Even through thick concrete walls, simple technical aids such as structure-borne sound sensors can be used to record the word spoken in the room and make it understandable. In the case of the building envelope - whether opaque or transparent - modified laser vibrometers enable invisible and unnoticed eavesdropping even from a great distance. In "normal buildings" sensitive microphones or simply listening in an otherwise quiet environment are sufficient for this. Finally, the various technical installations in the building, such as heating pipes or ventilation ducts, are also sound transmitters through the entire building. Structural and technical recommendations according to the state of the art to minimize eavesdropping include, for example, the abandonment of windows, a seamless solid construction, protection against structure-borne noise and absolutely unadorned rooms. Such rooms, however, completely contradict the ideas of the users.

A viable approach must therefore be able to be integrated into typical building structures and separating components, e.g. facades, walls, ceilings, etc., in order to achieve rooms that are as completely bug-proof as possible. This applies to both new buildings and the retrofitting of existing rooms. The starting point is the structural and technical substance and the contemporary components of office buildings and their available soundproofing potential. Since the facades, walls, ceilings, etc., are mostly multi-layered and multi-layered, their realistic sound insulation values are between approx. 25 and 55 dB, and slightly higher for heavy floor ceilings. These values can of course be increased, but only through more mass or more expensive building technologies. These expensive and complex technical approaches to increasing sound insulation also include active systems consisting of (vibro) acoustic sensors, actuators and signal processing to influence the vibration behavior of separating components (" Active Control of Vibration "Elsevier Ltd. 1996 , " Active control of noise and vibration "CRC Press 1996 ), e.g. plates or discs ( DE 102005044448 , DE 19826171 ). The influencing can take place as vibration damping or superposition. It proves to be particularly valuable at low frequencies, since in this area the sound insulation of all walls, ceilings, etc. is low overall or in part, in the manner of resonance. The frequency range of Understandable language is, however, at medium and high frequencies, where an increased sound insulation with active systems is not or hardly achievable.

In contrast to this, one approach to reducing audibility is to superimpose masking signals on speech so that a noise can be heard or detected, but its content is incomprehensible. On the one hand, these masking systems must alienate the voice content that is penetrating to the outside in such a way that it cannot be deciphered. Both word and sentence intelligibility must therefore be reduced to a level that can no longer be semantically reconstructed. On the other hand, there must be no noticeable acoustic disturbance (annoyance) in or outside the room, as user acceptance depends on this. Noise that drowns out is therefore not an acceptable solution as it would have to significantly exceed the speech signal level. Conventional systems for the noise of confidential voice information, such as. B. the device described in "http://www.global-security.org/de/produkte/lauschabwehr/76-akustisches-verrauschungssystem" are therefore subject to severe restrictions with regard to their practical application. In addition, the coherent speech signal can be at least partially extracted from the incoherent noise with appropriate effort and suitable mathematical or electronic methods.

From the U.S. 3,213,199 A a method is known in which sound is recorded in a room with a microphone. The room is separated from the neighboring room by a partition. In order to prevent conversations in the room from being overheard in the neighboring room, a loudspeaker is provided which, depending on the frequency and the amplitude of the sound recorded in the room, generates a masking sound in the neighboring room at any time.

From the CA 2,461,148 a special method for generating masking sound using a frequency filter is known. The sound to be masked or encrypted is broken down into time blocks. The frequency spectrum and the power level are determined and broadband sound is continuously generated with a suitable level spectrum for masking. Here is the amplitude of the masking sound comparatively small when it comes to pure masking applications and high when encryption is to be achieved.

From the US 2012/031689 A1 an electronic device for generating masking sound with a plurality of microphones and a loudspeaker is known. Voice signals and ambient signals are recorded. Depending on this, a masking sound is generated.

From the DE 10 2010 014 819 A1 a structure for active noise suppression in a vehicle interior is known. The active noise suppression takes place depending on the level of the sound to be suppressed. It is taken into account that after the sound to be suppressed has been attenuated, the active noise suppression can be weaker.

Also from the WO2004 / 107806 A1 a system is known with which sound can be masked.

From the US 2005/0065778 A1 an apparatus and a method for masking speech sound are known. The amplitude and the frequency of the masking signal can be varied in certain time segments. The time segments can have a fixed, variable or random length. Using a so-called "babble generator", a speech-like signal, an artificial voice, is generated as a masking signal. Then the artificial speech signal generated in the "babble generator", the masking signal, is superimposed with the real speech signal.

From the US 2003/0002687 A1 a structure and an associated method for the acoustic improvement of an environment are known. In one embodiment, a partition element in the manner of a curtain is provided in order to provide a discontinuity in a sound-conducting medium, such as air, and in order to absorb sound. One or more microphones pick up acoustic energy on one side of the separator and convert the sound into electrical signals. These are fed to digital signal processing. This uses an algorithm to analyze the electrical signals and to provide a control signal based on the analysis. From the Control signal, an electrical output signal is generated that is converted into sound. Overall, it is - at least largely - prevented that the sound from an acoustically loaded side of the separating element reaches the other side of the separating element. At the same time, a sound is provided on the acoustically unloaded side of the separating element, which, however, should be more pleasant than on the acoustically loaded side.

From the US 2009/0175848 A1 is known to convert systems without acoustic function in an office environment into a system of distributed audio sources for active sound control.

From the U.S. 5,834,647 A an arrangement for damping sound from moving sources is known. A set of electroacoustic sources is provided for this purpose, which emit anti-sound waves in response to an incident sound wave. The electroacoustic sources are preferably designed to emit a synchronous anti-sound wave with a radius of curvature approximately as large as the incident sound wave.

The requirements for largely tap-proof separating components with masking systems on top of a room therefore consist of minimizing speech intelligibility and preventing the outwardly penetrating speech signal from being reconstructed, as well as minimizing the potential for acoustic interference for the users of the adjoining rooms.

The object of the invention is therefore to achieve improved security against eavesdropping in rooms in which confidential conversations are conducted. The solution is given in the independent claim. The dependent claims teach advantageous configurations. The description and the drawings provide further explanations.

Solution

First of all, a method will be described which is particularly suitable for the device according to the invention described later. It was recognized that this was a

1. a) Erfassung von Amplitude, Zeit und Frequenz des Sprachsignals; wobei die Zeit als Parameter dabei nicht der Zeitpunkt der Signalerfassung, sondern wie aus Schneider-Stickler, B.; Bigenzahl, W.: Stimmdiagnostik. Springer-Verlag Wien, 2013 (ab S. 153 ) bekannt und vom Fachmann angenommen, die zeitliche Fluktuation der Sprache, sprich die zeitliche Änderung des Frequenz-Amplitudenspektrums ist.
2. b) Erzeugung eines Maskierschalls, der von Amplitude, Zeit und Frequenz des Sprachsignals so abhängig ist, dass durch Überlagerung des Sprachsignals mit dem Maskierschall ein unverständlicher Überlagerungsschall entsteht, wobei der Überlagerungsschall keine wesentlich höhere Lautstärke hat als das Sprachsignal.

Method for lowering the intelligibility of a speech signal is to be provided with the following steps.

1. a) Detection of the amplitude, time and frequency of the speech signal; whereby the time as a parameter is not the time of the signal acquisition, but how from Schneider-Stickler, B .; Bigenzahl, W .: voice diagnostics. Springer-Verlag Vienna, 2013 (from p. 153 ) is known and accepted by those skilled in the art, the temporal fluctuation of speech, i.e. the temporal change in the frequency-amplitude spectrum.
2. b) Generation of a masking sound which is so dependent on the amplitude, time and frequency of the speech signal that an incomprehensible overlapping sound is created by superimposing the speech signal with the masking sound, the superimposing sound not having a significantly higher volume than the speech signal.

The core idea is therefore that the masking sound depends in a differentiated manner on the speech signal, the intelligibility of which is to be reduced, and is specifically matched to the speech signal. The prior art is that a masking sound is provided with a higher volume when the speech signal is louder. The present method, on the other hand, uses the fact that a masking sound matched to the speech signal, which takes into account not only the amplitude but also the time and frequency of the speech signal, can already be superimposed into an incomprehensible superimposed sound with a volume in the range of the volume of the speech signal. So it is possible that the superimposed sound does not have a significantly higher volume than the speech signal. The masking sound itself can be a more or less intelligible speech sound. In the prior art, the speech signal is simply drowned out by unspecific noise, which is louder than the speech signal, so that the superimposed sound cannot be understood. In the present case, the masking sound could also be quieter than the voice signal. The superimposed sound is of course somewhat louder than the speech signal, but not significantly louder.

It should be mentioned that the masking sound must be generated in real time, since the masking sound has to be present at the same time as the speech signal for the superimposition. Particularly complicated, computationally expensive methods for generating the masking sound are therefore ruled out. As shown later using an example, the speech signal is recorded, the masking sound is generated and the resulting overlay in different places. The speech signal thus has a certain transit time until it comes to superimposition with the masking sound. This time can be used to generate the masking sound.

In the present case, the speech signal is broken down into intervals. The masking sound is formed by inversing the short intervals over time. To put it clearly: While the speech signal usually contains forward spoken text, the masking sound is, as it were, backward "spoken" sequences. The superimposed sound thus results from the superimposition of the speech signal with temporally inverted sequences of the signal. Long sequences cannot be inverted due to the requirement mentioned above to provide the masking sound in real time. Rather, it is a question of short intervals that have a length of 20 ms to 200 ms, in particular 50 ms to 100 ms, i.e. the length of the intervals is arbitrarily variable within these interval limits.

At this point it should be mentioned for the sake of understanding that the masking sound is formed directly from the speech signal. That is, in the present method, only the speech signal is used and changed in order to then superimpose the speech signal in a changed form as masking sound with itself. In the U.S. 3,213,199 A on the other hand, a new signal is generated from the amplitude and the frequency of the primary signal, i.e. the recorded speech signal.

Accordingly, the generation of the masking sound is subject to a random change. It goes without saying that this random change must remain within a framework that ensures that the superimposed sound remains incomprehensible. The lowering of the intelligibility of a speech signal can also serve to prevent eavesdropping on a conversation. In this case, generating the masking sound according to a method that is always unchanged, in this case the change in the masking sound does not result from the change in the masking rule but from the variation in the length of the interval, could result in the method being recognized by the eavesdropper and thus a reconstruction of the Speech signal would be possible. Such a reconstruction is made considerably more difficult by the random change in the generation of the masking sound. The way to change randomly is to change the length of the short time intervals described above according to a random function. It should be noted that the in US 2005/0065778 A1 procedure described differs from it. In the US 2005/0065778 A1 the way in which a masking sound is generated within the time intervals depending on the recorded speech signal is to be changed. This variation should not be ruled out in the present case. What is decisive, however, is that in the present case a random change in the time intervals has been recognized as sufficient to encrypt the masking rule as a whole.

On the basis of the above considerations, a large number of methods are conceivable with which the masking sound can be generated.

In one embodiment, the time intervals are changed randomly in a range from 20 ms to 200 ms, in particular in a range from 50 ms to 100 ms.

In one embodiment, an attenuation and / or attenuation of the speech signal is taken into account when generating the masking sound, wherein in particular a masking sound with a low volume is provided when the superimposition sound occurs from the superimposition of the masking sound with the attenuated and / or attenuated speech signal. In many cases it will be possible to attenuate or attenuate the speech signal before it reaches an area in which there is interference or eavesdropping is to be fended off. A masking sound with a lower volume is sufficient. Correspondingly, the superimposed sound is also less loud, so that possible disturbances are further reduced. If a comparison is made above between the volume of the masking sound, the superimposed sound and the speech signal, it is sensible to focus on the speech signal after an attenuation or attenuation that may have taken place.

As mentioned above, it is particularly important to increase the security of rooms from being eavesdropped. Accordingly, the invention relates to a partition component for separating a room from a room environment. The room environment is usually the outside area of a building. But it is also conceivable that it is a neighboring room in a building. The separating component contains sensors, actuators and a signal processing unit for influencing the sound transmission. The separating component is characterized by the components described below:
There are sensors for detecting the voice signal coming from the room. There are also actuators for generating a masking sound. The masking sound depends on the amplitude, time and frequency of the speech signal. As already explained above, it is not enough to take into account the volume alone, rather the time and frequency of the speech signal must also be taken into account. It is essential that a signal processing unit is available which can process the signals recorded by the sensors in such a way that the actuators can generate such a masking sound that by superimposing the masking sound on the speech signal in the room environment, an incomprehensible superimposed sound is created without the superimposed sound must have a much higher volume than the voice signal. The masking sound itself has roughly the same volume as the voice signal in the room environment, but it can also be quieter.

The sensors and actuators are spaced apart in a defined manner so that the speech signal has a certain transit time until it is superimposed with the masking sound. In the U.S. 3,213,199 A the sensors and actuators are also spaced apart. As has been recognized within the scope of the present invention, however, the distance cannot be taken into account or even used. With one, as in Figure 1 the U.S. 3,213,199 A The microphone shown as a sensor in the middle of the room, the speech signal from a speaker near the partition arrives at the partition more quickly than at the microphone. In the case of a speaker seated facing away from the partition, the speech signal is more likely to be at the microphone than at the partition. Appropriate consideration of the runtime of the speech signal from the sensors to the actuators is therefore in the U.S. 3,213,199 A not possible. This is necessary for the execution of the in U.S. 3,213,199 A masking method described is also not necessary because the masking sound is not generated and superimposed in real time using the primary speech signal itself, but only from its specific parameters frequency and amplitude. In the case of the present separating component, on the other hand, this is possible as described below and is also necessary to fulfill the task according to the invention. Correspondingly, the signal processing unit is designed to delay the masking sound in terms of time with respect to the speech signal in order to take into account transit times of the speech signal in the separating component. It should be noted here that the transit time of the electrical signal from the sensors and the signal processing unit to the actuators is generally much shorter than the transit time of the sound in the separating component. If the sensors are now in the area of the side of the separating component that faces the room and the actuators are in the area of the side of the separating component that is facing the room environment, this transit time of the sound from the sensors to the actuators must be taken into account. This transit time is advantageously used according to the invention in that the time delay resulting from the signal transit time is designed to be sufficient for the signal processing unit to generate the masking sound directly, ie in real time, and to make it available for the superposition.

In one embodiment, the separating component is designed such that the generation of the masking sound is subject to a random change. This means that the generation of the masking sound can only be reproduced with great difficulty, if at all. This prevents the speech signal from being reconstructed. As already shown in the description of the process, the change may only be within limits that allow the creation of incomprehensible superimposed sound.

In an important embodiment, the separating component is designed such that the method described above can be carried out with all its advantages and in the various embodiments of the method.

In one embodiment, the separating component is designed with multiple layers or multiple shells. Such structures are common and readily available. In addition, gaps are created so that the signal processing unit, for example, can be easily accommodated there.

In one embodiment, the separating component is set up in such a way that the masking sound is superimposed on the speech signal within the separating component. It can thus be achieved that in an area of the separating component facing the room environment, structure-borne noise is formed, which radiates incomprehensible superimposed sound into the room environment.

Alternatively, under certain conditions also in addition, the separating component is set up in such a way that the portion of the speech signal passing through the separating component is superimposed with the masking sound in the room environment. This is particularly the case when the actuators for generating the masking sound are located on the side of the separating component facing the room environment. In this case, a portion of the speech signal that is greatly reduced by the separating component passes through the separating component. This then becomes, so to speak, only in the room environment through an overlay with the masking sound to an incomprehensible overlay sound.

Both approaches have advantages. A corresponding superimposition within the separating component so that, as in the first described alternative, structure-borne noise is formed in the area of the separating component facing the room environment, which radiates incomprehensible overlaying sound into the room environment, eavesdropping should be even more difficult. It is no longer possible to tap the structure-borne noise on the side facing the room environment.

In one embodiment, the sensors are designed as vibroacoustic sensors. Vibroacoustic sensors are suitable for the precise detection of a speech signal.

In one embodiment of the invention, the sensors for detecting the speech signal are attached to the side of the partition component facing the room or in this area. This is where the voice signal has the highest intensity, so that sensors there can easily detect the voice signal. As an alternative or in addition, the actuators for generating the masking sound are attached to the side of the separating component facing the room environment or in this area. The masking sound is required in the room environment, so it makes sense to attach the actuators in this area. Of course, it is also conceivable, for example in the case of a multilayer separating component, to mount the actuators in a position close to the room.

It goes without saying that all of the above measures cannot make an eavesdropping attack entirely impossible. The invention can, however, make an important contribution to making such an eavesdropping attack significantly more difficult. In many cases this is very desirable.

Embodiment

Also based on the Figure 1 which shows a schematic representation of a double-shell partition wall with a vibroacoustic system integrated according to the invention for masking the speech sound transmission between two rooms, the invention is to be described in more detail below.

The starting point is, as in Figure 1 to recognize a multi-shell partition component 1, for example a double-shell wall or ceiling construction or a single-shell wall with facing shell, with adequate sound insulation in the information-containing voice frequency range. This acoustic transmission path between the tap-protected room 2 and its external environment 3, which forms the room environment, is manipulated non-linearly by means of a vibroacoustic speech signal masking system. It comprises (vibro) acoustic sensors 4 and actuators 5 as well as a signal processing module 6, which are integrated into the multi-shell separating component 1. In this method for speech masking, a modified speech signal 9 serving as masking sound is generated in real time from the primary speech signal 8 detected on the room side 7 of the separating component 1, which is fed into the multi-shell separating component 1 by means of the actuators 5 or preferably its outside 10 for sound emission stimulates. The type, number and distribution of the sensors 4, actuators 5 etc. depends, among other things, on the construction of the separating component 1. Finally, the superimposed radiation of both signals, i.e. the damped primary sound signal 11 and the modified speech or masking signal 9, on the outside 10 to a quiet and at the same time incomprehensible superimposed sound 12. The above-mentioned generation of the modified speech signal 9 in real time should simply mean that the superimposition of the primary speech signal 8 and the modified speech signal 9 to the superimposed sound in the external environment 3 must take place simultaneously. The primary speech signal 8 requires a certain transit time from the vibroacoustic sensors 4 to the actuators 5. This is beneficial in that the generation of the modified speech signal 9 takes a certain amount of time.

This system can also be used in conjunction with solid components. Here, for example, there is the possibility of using the sensors 4 and actuators 5, similar to

Connection boxes for electrical power or data networks to be integrated directly into the component. Alternatively, the component can also be provided with a facing shell on the room side, for ceilings e.g. also in the form of a floating screed, whereby the voice masking system is located in the gap between the facing shell and the solid component. The facing shell does not generally have to serve soundproofing purposes, but can also act as a sound absorber in room 2 to improve the room acoustics. In any case, the multi-shell or multi-layer structure offers the advantage that the masking sound 9 emitted from the outside 10 of the separating component 1 does not reach the room 2 or only reaches it weakly and is therefore not perceived as disturbing. From this point of view, an asymmetrical structure of the separating component 1 is of particular importance.

The algorithm for generating the modified speech signal 9 uses the detected primary speech signal 8. The acquisition is followed by filtering, time delay and time inversion of separate signal sections. The known sound transmission properties (level, phase and frequency behavior) of the separating component 1 are taken into account. In the end result, instead of quieter language on the outside 10 of the separating component 1, only a kind of “Babylonian language tangle” with a low level can be perceived. The speech and speaker recognition as well as the associated interception are made considerably more difficult or even completely prevented.

The filtering, time delay and time inversion of short signal sections are based on the knowledge, see e.g. Sound Masking Performance of Time-Reversed Masker Processed from the Target Speech. Acta Acustica United with Acustica. Vol. 98 (2012), pp. 135-141 that this manipulation achieves a significantly higher masking effect when the original speech signal is presented at the same time than the superposition with the known, mostly differently filtered noise signals. This knowledge arose in the context of language masking between workstations in open-plan offices in order to reduce the disruptive effect of understandable background language. Eavesdropping security or the like is not a relevant goal there.

The speech masking according to the invention is dependent on the language prevailing in the application, since the speech patterns differ. Although nothing changes in the basic manipulation steps, some details, such as filter cut-off frequencies and algorithms, can be optimized for a specific language. The language-dependent selection and adaptation of the system properties is therefore provided as an embodiment.

In this context, an additional algorithmic module for speech recognition can also be integrated in two ways. On the one hand, it distinguishes speech from other non-speech sounds. Such algorithms already exist. On the other hand, it recognizes the language family.

A supplementary feature is a type of control or feedback structure with which the success and the degree of speech incomprehension can be checked almost in real time in order to then adapt the parameters of the masking noise to be generated.

When implementing and adapting the vibroacoustic system according to the invention to mask the sound transmission of light partition walls, there are several design options. If there is a risk of mutual eavesdropping between two neighboring rooms, a bidirectional system is possible that works separately or in a coordinated manner in both directions. All of the features mentioned can be transferred to so-called flanking components, that is to say, for example, to walls which run perpendicular to the actual separating component 1 and enable longitudinal sound transmission between rooms. The flanking sound transmission also applies analogously to lines, ducts, pipes, etc. that transmit speech signals in the building stimulated by the sound in room 2. In this case, the separating component 1 consists of a multi-layer duct casing with an otherwise identical structure to the elements of the vibroacoustic system according to the invention for speech masking. Joints pose a particular challenge, especially the unavoidable functional joints, e.g. on doors. However, the vibroacoustic system according to the invention for speech masking can be used again both in the case of a basically advantageous covered design of door and frame and also in the case of a special element for covering the joint become. In this case, a design tailored to the actual partition component 1, that is to say, for example, the door, is provided.

The application of the vibroacoustic system according to the invention for speech masking is not only limited to buildings, but can also be used advantageously in vehicles and means of transport of various types.

Claims (14)

1. Separating component (1) for separating a room (2) from a room environment (3) comprising sensors (4), actuators (5) and a signal processing unit (6), for influencing sound transmission, characterized in that the separating component (1) has the following component parts:

   sensors (4) for detecting the speech signal (8) coming from the room (2);

   actuators (5) for generating a masking sound that is dependent on amplitude, time and frequency of the speech signal (8);

   a signal processing unit (6), which can process the signals detected by the sensors (4) in such a way that the actuators (5) can generate such a masking sound (9) that an incomprehensible superimposed sound (12) arises as a result of the superimposition of the masking sound (9) with the speech signal (11) in the room environment, without the superimposed sound (12) having to have a significantly higher loudness than the speech signal (11), wherein the sensors (4) and the actuators (5) are spaced apart, with the result that the speech signal (8) has a certain propagation time until superimposition with the masking sound (9) occurs, wherein the signal processing unit (6) is configured to delay the masking sound (9) temporally relative to the speech signal (8) in order to take account of propagation times of the speech signal (8) in the separating component (1).
2. Separating component according to Claim 1, characterized in that the separating component (1) is configured in such a way that the generation of the masking sound (9) is subject to a random variation.
3. Separating component according to Claim 1 or 2, characterized in that the separating component (1) is configured to carry out a method having the following steps:

   a) detecting amplitude, time and frequency of the speech signal (8);

   b) generating a masking sound (9) that is dependent on amplitude, time and frequency of the speech signal (8) in such a way that an incomprehensible superimposed sound (12) arises as a result of the superimposition of the speech signal (8) with the masking sound (9), wherein the superimposed sound (12) does not have a significantly higher loudness than the speech signal, wherein the generation of the masking sound (9) is subject to a random variation, and wherein the speech signal (8) is split into intervals, characterized in that the length of the intervals is varied randomly in accordance with a random function.
4. Separating component according to Claim 3, characterized in that the intervals are varied randomly in a range of 20 ms to 200 ms, in particular in a range of 50 ms to 100 ms.
5. Separating component according to either of Claims 3 and 4, characterized in that insulation and/or damping of the speech signal (8) are/is taken into account when generating the masking sound (9), wherein in particular a masking sound (9) having low loudness is provided if the superimposed sound (12) is effected from the superimposition of the masking sound (9) with the insulated and/or damped speech signal (11).
6. Separating component according to any of Claims 3 to 5, characterized in that the masking sound (9) is formed by temporal inversion of the short intervals.
7. Separating component according to any of Claims 1 to 6, characterized in that the separating component (1) is multi-layered or multi-shelled.
8. Separating component according to any of Claims 1 to 7, characterized in that the separating component is designed in such a way that the superimposition of the masking sound (9) with the speech signal (8) takes place within the separating component (1), with the result that in a region (10) of the separating component (1) facing the room environment (3) a structure-borne sound forms which emits incomprehensible superimposed sound (12) into the room environment (3).
9. Separating component according to any of Claims 1 to 8, characterized in that the separating component (1) is designed in such a way that a superimposition of that portion of the speech signal (11) which passes through the separating component (1) with the masking sound (9) takes place in the room environment (3).
10. Separating component according to any of Claims 1 to 9, characterized in that the sensors (4) are embodied as vibroacoustic sensors.
11. Separating component according to any of Claims 1 to 10, characterized in that the sensors (4) are fitted at the side (7) of the separating component (1) facing the room (2) or in this region (7) and/or the actuators (5) are fitted at the side (10) of the separating component (1) facing the room environment (3) or in this region (7).
12. Separating component according to any of Claims 1 to 11, characterized in that the separating component (1) is embodied as a door between room (2) and room environment (3).
13. Separating component according to any of Claims 1 to 12, characterized in that the separating component (1) is embodied as a solid component.
14. Separating component according to any of Claims 1 to 13, characterized in that the separating component (1) is formed by a wall-like component with an attachment shell.

Images