Method for calibrating a vehicle-integrated binaural 3D audio system, and vehicle

Abstract

The invention relates to a method for calibrating a vehicle-integrated binaural 3D audio system, wherein the 3D audio system outputs sound via at least two loudspeakers (1) in order to simulate a virtual sound source (2) for a user (3) at a location (4) different from the loudspeakers (1), and wherein the following calibration steps are performed by a controller (5): a) placing the virtual sound source (2) inside or outside a vehicle (6) comprising the 3D audio system; b) outputting sound via the virtual sound source (2); c) acquiring a location indication (7) output by the user (3), wherein the location indication (7) describes the position (8) of the virtual sound source (2) as assumed by the user (3); d) determining a location difference (∆x) between the actual (4) position (8) of the virtual sound source (2) and the position (8) of the virtual sound source (2) described by the location indication (7); e) on the basis of the location difference (∆x), changing a head-related transfer function (HRTF) used by the controller (5) to adapt the audio signal output via the at least two loudspeakers (1) to the position (4) of the virtual sound source (2); and f) performing steps a) to d) again until the location difference (∆x) falls below a defined threshold value. The method according to the invention is characterized in that vehicle-integrated loudspeakers (1) are used and the location indication (7) is acquired with the aid of vehicle passenger compartment sensors (9.1, 9.2, 9.3).

Description

[0001] Mercedes-Benz Group AG

[0002] Method for calibrating a vehicle-integrated binaural 3D audio system and vehicle

[0003] The invention relates to a method for calibrating a vehicle-integrated binaural 3D audio system according to the type defined in more detail in the preamble of claim 1 and to a vehicle with such a binaural 3D audio system.

[0004] The so-called head transfer function, also known as head-related or outer ear-related transfer function, is abbreviated by the letters HRTF, which are based on the English term Head-Related Transfer Function. The HRTF describes the transfer function of the outer ear and, along with time differences between a person's two ears, is crucial for localizing sound sources. Due to the different anatomy of each person, the HRTF is individual for each person. Of particular relevance are the shape of the head, which influences sound diffraction at the head and the distance between the ears, as well as the individual shape of the auricles and the auditory entrance. If a person's HRTF is known, a virtual positioning of sound sources becomes possible in order to create a localization of the sound source at a virtually predetermined position in a room.This functionality is often associated with the term 3D audio.

[0005] In many applications, however, a large number of loudspeakers are necessary. If a person's HRTF is known, such 3D audio can be virtually generated using just two loudspeakers aligned to the right and left of the person. In the field of general audio technology, complex measurements of test signals using miniature microphones in a person's individual ear canals are well known for recording the HRTF. This can now be made simpler and more efficient if optical recording of the anatomy or geometry of the head, the auricles, and their spatial relationship to one another is enabled. In this context, DE 102020 108449 A1 describes a method for providing at least one user-specific binaural sound signal for vehicle occupants, in which the user-specific HRTF is calculated based on the anatomy of the respective person.The anatomy of each person is recorded by interior cameras arranged in the vehicle in order to easily and conveniently enable an optimized audio experience using a larger number of speakers.

[0006] US 2013 / 0194107 A1 also describes capturing and storing an HRTF to generate virtual 3D audio via a signal processor. Here, too, cameras are used to capture the anatomy underlying the calculations.

[0007] Capturing the entire relevant anatomy using cameras installed in a vehicle, for example, in the front area in front of a person using the vehicle, is extremely difficult in practice, as these cameras often cannot reliably and accurately capture all relevant areas of the person's head. While the person can be asked to move and rotate their head accordingly in front of the cameras, this is comparatively time-consuming in practice and is perceived as a loss of comfort if it is necessary every time the 3D audio system is started.

[0008] Furthermore, US 2018 / 0249271 A1 discloses a binaural audio calibration method. The method disclosed in the document is used to calibrate a virtual reality device for experiencing a virtual three-dimensional sound environment. The virtual reality device also takes into account a user-specific head transmission function in order to position a virtual sound source in a virtual reality environment in such a way that it is located in the virtual reality environment at the exact location where the user of the virtual reality device also hears it. For calibration, a virtual sound source is placed in the virtual reality environment, and a sound is output via the virtual sound source. The user of the virtual reality device indicates where in the virtual reality environment they perceive the virtual sound source.The difference is then determined between the location where the virtual reality device has placed the virtual sound source and the position assumed by the user. The head transmission function is then adjusted, and a new measurement is performed. The goal is to minimize the difference between the position where the user assumes the virtual sound source to be and the position where the virtual reality device has placed the virtual sound source in the virtual reality environment. The process steps are repeated until this difference falls below a specified threshold. The head transmission function, thus adapted to the user, can be assigned to a user profile and thus reused for subsequent virtual reality experiences. The use of a virtual reality device is required to experience the three-dimensional sound experience.

[0009] The present invention is based on the object of providing an improved method for calibrating a vehicle-integrated binaural 3D audio system, with the aid of which the effort for calibrating the binaural 3D audio system can be reduced, but at the same time a higher degree of correspondence is achieved between the location at which the 3D audio system places the virtual sound source and the location at which a user locates the virtual sound source.

[0010] According to the invention, this object is achieved by a method for calibrating a vehicle-integrated binaural 3D audio system having the features of claim 1. Advantageous embodiments and further developments as well as a vehicle for carrying out the method emerge from the dependent claims.

[0011] A generic method for calibrating a vehicle-integrated binaural 3D audio system, wherein the 3D audio system emits sound via at least two loudspeakers in order to simulate a virtual sound source for a user at a location different from the loudspeakers, and wherein the following calibration steps are carried out by a control unit: a) placing the virtual sound source inside or outside a vehicle comprising the 3D audio system; b) emitting sound via the virtual sound source; c) detecting a location indication output by the user, wherein the location indication describes the location of the virtual sound source presumed by the user; d) determining a location difference between the actual location of the virtual sound source and the location described by the location indication;e) changing, depending on the location difference, a head transfer function used by the control unit to adapt the audio signal emitted via the at least two loudspeakers to the location of the virtual sound source; and f) repeating steps a) to d) until the location difference falls below a predetermined threshold value; this is further developed according to the invention by using vehicle-integrated loudspeakers and detecting the location indication with the aid of vehicle interior sensors.

[0012] The method according to the invention allows for quick and easy calibration of a vehicle-integrated binaural 3D audio system used in a vehicle context. The method minimizes the difference between the assumed location and the actual location of the virtual sound source, so that a high degree of agreement is achieved once the location difference falls below the specified threshold. The complex measurement of the user's body is thus no longer necessary. Furthermore, an adaptation method for head transfer functions based solely on the measurement of a user's physical characteristics can still lead to deviations, since a head transfer function determined in this way may not correspond to reality. A remaining location difference cannot therefore be compensated for.However, the method according to the invention allows an evaluation of the head transfer function used by the control unit and allows the head transfer function to be changed accordingly until a desired quality with regard to the local correspondence between the presumed sound source position and the actual sound source position has been found.

[0013] To minimize the spatial difference, suitable optimization algorithms can be used, particularly based on machine learning. Such an optimization algorithm then determines which alternative head transfer function should be used or which parameter of the used head transfer function should be varied and to what extent.

[0014] By using vehicle-integrated speakers, the use of other devices such as headphones, virtual reality glasses, or the like can be dispensed with. The use of vehicle interior sensors enables comparatively particularly precise detection of the location indication. Vehicle interior sensors can be used that can obtain depth information, such as radar sensors, LiDAR sensors, ultrasonic sensors, and the like. Using one or more cameras, the location indication can also be recorded using machine vision. The camera images generated by the camera(s) can be evaluated using suitable image recognition algorithms. Depth information can also be generated using a stereo camera. Motion sensors can also be used as vehicle interior sensors.This also includes radio-based sensors, which detect movements by detecting changes in the radio signal strength. A Wi-Fi signal, for example, can be used as a radio technology. The Wi-Fi signal is generated by multiple antennas, and changes in signal strength are detected by multiple antennas. Such vehicle interior sensors can then detect, for example, the user's gestures, their posture, their direction of gaze, and the positioning of body parts in a specific location.

[0015] Accordingly, in the simplest case, the user's gaze direction or the direction of movement or pointing of a gesture can be used as a location indicator. The direction can be defined in a two-dimensional plane, for example, as an angle measured around the vehicle's vertical axis, or as an angle measured freely in three-dimensional space. In this case, the location difference is defined by the solid angle at which the user essentially looks past or points past the virtual sound source. However, this only represents a first approximation, since the distance of the virtual sound source from the user is not taken into account – i.e., how far the virtual sound source perceived by the user is in front of or behind the actual location. However, the first approximation can already be sufficient, especially since in the case of a slightly different location (position and direction), the direction should be preferred as an indicator for HRTF adaptation.

[0016] The binaural 3D audio system can have a plurality of speakers, for example, five, seven, ten, or even more speakers, but at least two speakers. The speakers can be arranged anywhere in the vehicle, with at least one speaker located in the user's left hemisphere and one speaker located in the user's right hemisphere.

[0017] The sound emitted by the virtual sound source can be diverse. For example, it could be birdsong, the whistling of a whistle, bubbling boiling water, the whistling of a kettle, the ringing of a bell, or similar. The user can specify which type of sound to use via a human-machine interface such as a touch-sensitive display for the control unit.

[0018] The virtual sound source can be placed anywhere within the vehicle interior. For example, the virtual sound source can be located on a vehicle seat, in the trunk, on the dashboard, under the headliner, or similar. It is also possible to place the virtual sound source outside the vehicle. This allows ambient noise to be simulated in a direction-dependent manner. The 3D audio system or control unit can take the influence of the vehicle structure into account, so that the corresponding sound is perceived by the user as acoustically muffled, as if it were actually coming from outside the vehicle.

[0019] The specified threshold for the spatial difference can be freely selected depending on the application scenario. For example, if the binaural 3D audio system is used for entertainment purposes, a higher threshold and thus a larger permissible spatial difference can be used. However, if, for example, location-specific warning tones are emitted via the 3D audio system, a lower specified threshold and thus a smaller permissible spatial difference can be used.

[0020] An advantageous development of the method provides that, in addition to the head transmission function, the control unit considers an individual room transmission function for each loudspeaker to adapt the audio signal emitted via the respective loudspeaker to the location of the virtual sound source. The geometric structure of the vehicle, the materials installed in the vehicle, and the location of the loudspeakers in the vehicle also influence the location of the virtual sound source. The previously listed variables are incorporated into a corresponding room transmission function. The room transmission function can, for example, be determined experimentally by the vehicle manufacturer by conducting acoustic measurements. The room transmission function can also be approximated using calculation or simulation methods.Due to the different mounting positions of the individual speakers in the vehicle, the spatial transmission function is unique for each speaker. If the control unit takes into account a specific spatial transmission function for each speaker, it is possible to further reduce the spatial difference between the assumed location of the virtual sound source and the actual location of the virtual sound source.

[0021] According to a further advantageous embodiment of the method according to the invention, the vehicle detects external noise and / or background noise within the vehicle with the aid of acoustic detection means and varies the amplitude of individual frequencies of the audio signal emitted via the loudspeakers depending on the interference and / or background noise. The acoustic detection means can be, for example, one or more microphones. The microphones can be mounted externally or internally, i.e. in the vehicle interior. The interference noise can be, for example, wind, construction noise, traffic noise or the like. The background noise can be, for example, a conversation between several vehicle occupants, music played via the vehicle's infotainment system, a telephone call conducted in the vehicle interior or the like.The frequencies of the audio signal delivered via the loudspeakers, whose amplitude is to be compensated, depend accordingly on the type of noise or background noise. For example, the amplitude of frequencies that are also contained in the corresponding noise and / or background noise can be increased. This prevents the corresponding noise or background noise from negatively affecting the user's ability to locate the virtual sound source. The audio signal is adjusted individually for each loudspeaker. This also takes into account the directionality of the loudspeakers relative to the user's location.

[0022] A further advantageous embodiment of the method according to the invention further provides that, if multiple users are present in the vehicle, user-specific virtual sound sources are simulated for at least two users using user-specific head transmission functions. In other words, it is possible to simulate a separate, individual virtual sound source for different vehicle occupants. A head transmission function specifically tailored to the respective user is determined for each user. The procedure for locating these head transmission functions can be as described above. In particular, the position of the user within the vehicle interior, for example, the vehicle seat used by the user, plays a role.

[0023] Head transmission functions can be assigned to users through personal identification. For example, corresponding individuals can log in to the vehicle using a username or be recognized through biometric features such as a facial scan, an iris scan, or voice analysis. Users can also be recognized through a device carried on board, for example, the individual MAC address or Bluetooth address of a smartphone. Two user-specific audio signals can be output via the same pair of loudspeakers. However, it is preferable to use an individual pair of loudspeakers for each user to output the user-specific audio signal. This reduces or even eliminates the negative influence of acoustic interference on the localization of the virtual sound sources.

[0024] According to a further advantageous embodiment of the method according to the invention, the control unit maintains a database in which a plurality of HRTF profiles are stored, each HRTF profile being assigned a differently configured head transfer function, and the control unit selects a different HRTF profile and / or adapts parameters of the head transfer function used to change the head transfer function in method step e). There are therefore several options for changing the head transfer function. The HRTF profiles contained in the database or the head transfer functions underlying these profiles can be so-called standard or default functions. These standard head transfer functions can, for example, have been defined by the vehicle manufacturer for the most common application scenarios.By changing the HRTF profile, a particularly quick and easy adjustment of the head transfer function is possible. However, the corresponding head transfer function may not match the user's actual head transfer function sufficiently. This also makes it possible to further adjust the head transfer function used. Proven methods can be used to determine how the head transfer function parameters should be changed during each calibration step a) to e). For example, proven optimization algorithms or algorithms based on artificial intelligence or machine learning can be used. With the help of appropriate algorithms, it is possible to recognize patterns and thus change the parameters of the head transfer function used in such a way that the spatial difference is reduced in the next iteration.

[0025] A further advantageous embodiment of the method according to the invention further provides that the location difference comprises the Euclidean difference between the actual location of the virtual sound source and the location presumed based on the location indication. This enables an even more comprehensive and thus more precise determination of the accuracy of the 3D audio system when positioning virtual sound sources compared to solely considering, for example, an angular difference. For example, the 3D audio system can emit sound via the virtual sound source, and the user must then reach for the virtual sound source with their hand. The location and direction in the vehicle interior to which and in which the user reaches is then assumed to be the presumed location.The control unit can then determine the Euclidean distance between this presumed position and the actual position where the control unit has placed the virtual sound source. Since more degrees of freedom regarding the location information are thus taken into account, a faster and more efficient adjustment of the parameters of the head transfer function used is also possible, so that a correspondingly suitable head transfer function can be found even faster, i.e., using or performing a fewer iterations of the calibration steps a) to e).

[0026] According to a further advantageous embodiment of the method according to the invention, the user specifies where the virtual sound source should be placed to the control unit for method step a) via a human-machine interface. The user can thus influence the calibration process. For example, in certain regions of the vehicle interior, there may be a greater discrepancy between the presumed position of the virtual sound source and the actual location of the virtual sound source than in other regions. The user can then preferentially place the virtual sound source in this region of space, so that the error present in this region of space can be reduced even more quickly by calibration. This increases the efficiency of the calibration process. This procedure can also be carried out analogously by the control unit.This allows the control unit to divide the vehicle interior into different zones and gradually place virtual sound sources in each zone. The control unit then determines which zones have the greatest spatial difference and subsequently performs an increased number of measurements in the corresponding zones.

[0027] A further advantageous embodiment of the method according to the invention further provides that the control unit outputs an acoustic, visual and / or haptic indication to the user via output means if the location of the virtual sound source presumed by the user coincides with the actual location within a tolerated deviation or, in the case of a larger deviation, the control unit outputs the indication and supplements it with a location tip, wherein the location tip describes where the actual location of the virtual sound source is compared to the presumed location of the virtual sound source.

[0028] Acoustic cues can be provided using the vehicle's speakers. If the user touches the virtual sound source, the audio signal can be stopped, for example. Additionally or alternatively, a corresponding warning tone, such as a beep or jingle, can be played.

[0029] Visual cues can be provided using lamps or display devices such as displays. For example, a corresponding lamp can light up briefly when the virtual sound source has been correctly located. A lamp can also be used to shine a spotlight on the location in the vehicle interior where the virtual sound source has been placed. More comprehensive information can be shown on a display. For example, the location tip can be provided in the form of a sectional view of the vehicle interior from a bird's eye view, showing the position where the control unit has placed the virtual sound source. The location tip can also be provided acoustically. For example, the following message can be spoken: "The bell is ringing on the passenger seat" or "The bell is ringing ten centimeters to the left".

[0030] Furthermore, haptic cues can be output, for example via actuators integrated into structural elements touched by the user. For example, the vehicle's steering wheel can be stimulated to vibrate if the user correctly locates the virtual sound source. By outputting the location tip, the user can be trained to compensate for remaining discrepancies between the presumed location of the virtual sound source and the actual location of the virtual sound source. In this way, the user can gradually learn that, for example, the virtual sound source is always slightly to the right below the position where the user would otherwise expect the virtual sound source to be. This correspondingly increases the reliability with which the user correctly locates the virtual sound source.

[0031] Preferably, the control unit adjusts the location tip depending on the magnitude of the location difference. In other words, a different location tip is issued if the user reaches far past the virtual sound source than if they reach directly past the virtual sound source or only slightly past it.

[0032] If, for example, the user reaches further past the virtual sound source, the light pulses emitted by the lighting devices can appear brighter or last longer. Acoustic alert messages can be played more loudly in this case. Other output modalities can also be used. For example, if the user hits the virtual sound source, a simple alert tone can be heard. However, if the user reaches past the virtual sound source, points past it, or looks past it, the actual location of the virtual sound source can be illuminated, and an alert message can be read out, indicating the actual location of the virtual sound source.

[0033] In a vehicle comprising a binaural 3D audio system and vehicle interior sensors, the 3D audio system and the vehicle interior sensors are configured according to the invention to carry out a method described above. The vehicle can be any vehicle, such as a car, truck, van, bus, or the like. It can also be a rail vehicle, watercraft, or aircraft. The 3D audio system comprises at least two loudspeakers for outputting the corresponding audio signal and a control unit for controlling the loudspeakers.

[0034] The control unit can obtain information from other processing units in the vehicle, such as an ABS control unit, and use it to control the loudspeakers. This allows driver assistance functions to be provided. If, for example, hazards are detected in the vehicle's surroundings, such as static or dynamic road users with whom a collision is imminent, corresponding warning tones can be emitted. The warning tones are positioned relative to the vehicle in such a way that the driver can perceive the hazard when looking in the direction of the warning tone.

[0035] Further advantageous embodiments of the method according to the invention for calibrating a vehicle-integrated binaural 3D audio system and a corresponding vehicle also emerge from the exemplary embodiments which are described in more detail below with reference to figures.

[0036] Showing:

[0037] Fig. 1 is a schematic plan view of a vehicle according to the invention, comprising a binaural 3D audio system; and

[0038] Fig. 2 is a flowchart of a method according to the invention for calibrating the 3D audio system of the vehicle shown in Fig. 1.

[0039] Figure 1 shows a vehicle 6 according to the invention, which has a binaural 3D audio system. With the help of the 3D audio system, it is possible to create a three-dimensional virtual soundscape for a user 3 present in the vehicle 6. The 3D audio system comprises at least two loudspeakers 1, which are arranged offset from one another in the vehicle 6. An audio signal is emitted via the loudspeakers 1, which simulates the presence of a virtual sound source 2 in the vehicle 6. In the exemplary embodiments shown, the virtual sound source 2 is symbolized by a bell. The virtual sound source 2 sounds to the user 3 at a location different from the loudspeakers 1. This simulation of sound sources can be used, for example, for entertainment purposes or to inform or warn the user 3.

[0040] The 3D audio system is controlled by a control unit 5. The control unit 5 can be distributed across multiple computing units or integrated into a single computing unit (not shown). In the exemplary embodiments shown, a computing unit 5.1 is used to evaluate vehicle interior sensors 9.1, 9.2, and 9.3. A computing unit 5.2 is used to calculate the audio signal. A computing unit 5.3 forms an optional communication interface to a central computing device 14 external to the vehicle. Additional information can be obtained from the central computing device 14, in particular wirelessly. To adapt the audio signal emitted via the loudspeakers 1 so that the virtual sound source 2 is heard at any location other than the loudspeakers 1, the control unit 5 takes into account a so-called head transfer function (HRTF).The head transmission function (HRTF) depends on the body dimensions of the respective user 3 and influences the directionality from which a user 3 hears a corresponding sound. If the head transmission function (HRTF) matches the actual head transmission function (HRTF) of user 3, the virtual sound source 2 will also sound for user 3 at the position (8) where the 3D audio system simulates the virtual sound source 2 (4). However, if the head transmission function (HRTF) deviates from the user-specific head transmission function, the virtual sound source 2 will also sound at a location other than the intended location.

[0041] With the aid of a method according to the invention, the 3D audio system can be calibrated quickly, easily and reliably in order to find a head transmission function HRTF that is suitable for the user and thus to improve the sound experience.

[0042] For this purpose, the control unit 5 controls the loudspeakers 1 in order to place the virtual sound source 2 at a location 4. The location 4 can be inside or outside the vehicle 6 and can also be referred to as the actual location or the desired location. At the beginning, the control unit 5 uses a standard head transmission function HRTF, which very likely differs from the actual head transmission function of the user 3. Accordingly, the user 3 assumes that the virtual sound source 2 is at a presumed location 8. The user 3 then outputs a location indication 7 to inform the vehicle 6 or the control unit 5 where they perceive the virtual sound source 2. As a location indication 7, the user 3 can, for example, point at the presumed location 8, look in this direction, or orient their body in this direction. The vehicle interior is monitored using vehicle interior sensors 9.1, 9.2, 9.3.These include, for example, an interior camera 9.1, a radar sensor 9.2 and a radio sensor system 9.3. For example, the interior camera 9.1, as shown in Figure 1, can be embedded in the dashboard of the vehicle 6 and aimed at the user 3. Other cameras not shown in detail can also be installed. With the help of image recognition algorithms, a differentiated and comprehensive analysis of user behavior is then possible. Depth information or distance values ​​can be determined with the help of the radar sensor 9.2 and the radio sensor system 9.3. This makes it possible to detect movements. The radio sensor system 9.3 can, for example, be an arrangement of several radio antennas, such as Wi-Fi antennas, whereby movements can be detected by analyzing the radio signal strength.

[0043] If the user 3 points or looks in the direction of the presumed location 8, the control unit 5 can determine an angle as the location difference between the location 4 and the presumed location 8. In this case, the virtual sound source could be located at any distance on a directional line 13. This therefore only allows a comparatively rough location determination. In a preferred embodiment, however, the user 3 accesses the exact location where they suspect the virtual sound source 2 to be. This enables the control unit 5 to determine the presumed location 8 more precisely. The control unit 5 can thus determine the Euclidean distance between the location 4 and the presumed location 8 as the location difference Ax. This means that an even more precise location difference is available, so that the head transfer function HRTF can be adjusted more precisely in the next calibration step.

[0044] The position difference Ax is compared by the control unit 5 with a specified threshold value, whereupon the head transfer function HRTF used by the control unit 5 is changed until the position difference Ax is less than the specified threshold value. Changing the head transfer function HRTF is possible in various ways. According to one embodiment, the control unit 5 comprises a database 11 in which a plurality of HRTF profiles 12 are stored. Each HRTF profile 12 is assigned an individual head transfer function HRTF. If the position difference Ax is greater than the specified threshold value, a different HRTF profile 12 can then, for example, simply be loaded, i.e. activated. Additionally or alternatively, it is possible to change the parameters of the respective head transfer functions HRTF stored in the individual HRTF profiles 12.This enables a targeted adaptation of the head transmission functions HRTF to the respective user s. An HRTF profile 12 found or adapted for the respective user 3 can be linked to the user 3 and reloaded whenever the user 3 is present in the vehicle 6. Thus, the calibration method according to the invention only needs to be performed once for each user 3.

[0045] New standard HRTF profiles can, for example, be obtained from the central computing device 14. An HRTF profile 12 configured in the vehicle 6 can also be transmitted to the central computing device 14 and made available there for retrieval by other vehicles.

[0046] Furthermore, the vehicle 6 can have acoustic detection means 10, which can be embodied as an external microphone or an internal microphone. With the aid of the acoustic detection means 10, external noise and / or background noise within the vehicle can be detected. The audio signal output via the loudspeakers 1 can then be adapted to said background noise and / or background noise. For this purpose, the amplitude of the frequency components comprising the corresponding background noise or background noise can be specifically increased or decreased.

[0047] The vehicle 6 can also be configured to locate an emergency vehicle, such as an ambulance, relative to the vehicle 6. The emergency vehicle can, for example, emit a warning signal via a siren. In such a situation, the user 3 must locate the emergency vehicle in traffic and, if necessary, adjust their driving behavior accordingly. To make it easier for the user 3 to locate the emergency vehicle, a follow-up horn signal can, for example, be emitted as a virtual sound source 2 and guided around the head of the user 3 in the vehicle interior at a location that matches the position of the emergency vehicle relative to the vehicle 6. This makes it easier for the user 3 to recognize where the emergency vehicle is located.

[0048] Using the method according to the invention, it is also possible to define individual head transmission functions (HRTF) for several users 3 present in the vehicle 6 and, accordingly, to simulate a virtual sound source at a different location for each user 3. Thus, the method according to the invention can be implemented individually for each user 3. The presence of users 3 can also be detected by means of the vehicle interior sensors 9.1, 9.2, and 9.3.

[0049] Furthermore, the user 3 can be trained to locate the virtual sound source 2 even more accurately. If a remaining location difference Ax remains, information messages can be output to the user 3 describing the location 4 at which the control unit 5 has actually placed the virtual sound source 2, relative to the presumed location 8. The user 3 thus gradually learns this deviation, whereby the remaining location difference can be further reduced or even disappear completely. The sequence of the method according to the invention is illustrated again in Figure 2 using a flowchart. The method begins in method step 201.

[0050] In method step 202, control unit 5 checks whether an HRTF profile 12 adapted to the respective user 3 already exists. If this is the case, the method ends in method step 210 or runs in a loop. If this is not the case, method steps 203, 204, 205, and 206 are executed. In a distributed implementation of control unit 5, method steps 203 to 206 can be executed, for example, on processing unit 5.1.

[0051] In method step 203, the computing unit 5.1 evaluates sensor data generated by external and internal sensors of the vehicle 6. In this way, the computing unit 5.1 determines a vehicle usage context, an environmental context, and a perception context. The vehicle usage context describes, for example, whether the vehicle 6 is being used to carry out a journey from a starting point to a destination or is currently being cleaned or loaded. For example, the output of direction-dependent warning messages is not required during cleaning or loading, but is required during a journey with the vehicle 6. The environmental context describes, for example, whether and where static and / or dynamic traffic objects are located opposite the vehicle 6. The environmental context can also describe what type of environmental object is involved, such as an emergency vehicle.Directional cues tailored to the respective surrounding objects can then be output. The perceptual context describes, for example, how the 3D audio system is used. For example, only a reduced number of speakers 1 may be available to output the audio signal because, for example, music is already being played on some speakers 1.

[0052] In process step 204, the computing unit 5.1 evaluates sensor data generated using external microphones. Accordingly, external noise can be determined.

[0053] In method step 205, the computing unit 5.1 evaluates sensor data generated by the interior camera 9.1, the radar sensor 9.2, and / or the radio sensor system 9.3, enabling it to analyze user behavior and identify users. In step 206, the computing unit 5.1 checks the active HRTF profile 12, i.e., which HRTF head transmission function is currently being used.

[0054] The computing unit 5.2 used to control the 3D audio system, i.e., to generate the audio signal to be output via the loudspeakers 1, outputs the corresponding audio signal via the loudspeakers 1 in method step 207 to simulate the virtual sound source 2 in the vehicle interior. As indicated by a dashed arrow, method steps 208, 209, and 206 are subsequently executed.

[0055] In method step 206, the location indication 7 output by user 3 is then additionally recorded, and the location difference Ax is determined from this. If the location difference Ax is too large, the head transfer function HRTF is adjusted in method step 208.

[0056] In process step 209, a check is then made to determine whether the spatial difference Ax is now greater or smaller than the specified threshold. If necessary, process steps 206 and 208 are repeated until the spatial difference Ax is sufficiently small.

Claims

Mercedes-Benz Group AG Patent claims 1. A method for calibrating a vehicle-integrated binaural 3D audio system, wherein the 3D audio system emits sound via at least two loudspeakers (1) in order to simulate a virtual sound source (2) for a user (3) at a location (4) different from the loudspeakers (1), and wherein the following calibration steps are carried out by a control unit (5): a) placing the virtual sound source (2) inside or outside a vehicle (6) comprising the 3D audio system; b) emitting sound via the virtual sound source (2); c) detecting a location indication (7) output by the user (3), wherein the location indication (7) describes the location (8) of the virtual sound source (2) presumed by the user (3); d) determining a location difference (Ax) between the actual location (4) and the location (8) of the virtual sound source (2) described by the location indication (7);e) Changing, depending on the location difference (Ax), a head transfer function (HRTF) used by the control unit (5) to adapt the audio signal emitted via the at least two loudspeakers (1) to the location (4) of the virtual sound source (2); and f) Repeating steps a) to d) until the location difference (Ax) falls below a predetermined threshold value; characterized in that vehicle-integrated loudspeakers (1) are used and the location indication (7) is detected with the aid of vehicle interior sensors (9.1, 9.2, 9.3).

2. Method according to claim 1, characterized in that the control device (5) in addition to the head transmission function (HRTF) for each loudspeaker (1) has an individual room transmission function for adapting the via the respective loudspeaker (1) to the location (4) of the virtual sound source (2) is taken into account.

3. Method according to claim 1 or 2, characterized in that the vehicle (6) detects vehicle-external noises and / or vehicle-internal background noises with the aid of acoustic detection means (10) and the amplitude of individual frequencies of the audio signal emitted via the loudspeakers (1) varies depending on the noises and / or background noises.

4. Method according to one of claims 1 to 3, characterized in that when several users (3) are in the vehicle (6), user-specific virtual sound sources (2) are simulated for at least two users (3) using user-specific head transmission functions (HRTF).

5. Method according to one of claims 1 to 4, characterized in that the control unit (5) maintains a database (11) in which a plurality of HRTF profiles (12) are stored, each HRTF profile (12) being assigned a differently configured head transfer function (HRTF), and the control unit (5) selects a different HRTF profile (12) and / or adapts parameters of the head transfer function (HRTF) used in method step e) to change the head transfer function (HRTF).

6. Method according to one of claims 1 to 5, characterized in that the location difference (Ax) comprises the Euclidean distance between the actual location (4) of the virtual sound source (2) and the location (8) presumed by the location indication (7).

7. Method according to one of claims 1 to 6, characterized in that the user (3) specifies to the control unit (5) via a human-machine interface where the virtual sound source (2) is to be placed for method step a).

8. Method according to one of claims 1 to 7, characterized in that the control device (5) outputs an acoustic, visual and / or haptic indication to the user (3) via output means if the location (8) of the virtual sound source (2) assumed by the user (3) coincides with the actual location (4) within a tolerable deviation or, in the case of a larger deviation, the control device (5) outputs the indication and supplements it with a location tip, the location tip describing where the actual location (4) of the virtual sound source (2) is located compared to the assumed location (8) of the virtual sound source (2).

9. Method according to claim 8, characterized in that the control unit (5) adapts the location tip depending on the magnitude of the location difference (Ax).

10. Vehicle (6) comprising a binaural 3D audio system and vehicle interior sensors (9.1, 9.2, 9.3), characterized in that the 3D audio system and the vehicle interior sensors (9.1, 9.2, 9.3) are configured to carry out a method according to one of claims 1 to 9.

Images