Virtual acoustic reflectors
Patent Information
- Application Number
- PCT/IB2025/052085
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-02-26
- Publication Date
- 2026-09-03
Smart Images
Figure IB2025052085_03092026_PF_FP_ABST
Abstract
Description
VIRTUAL ACOUSTIC REFLECTORS
[0001] The room acoustics in theatres, concert halls or in front of other stages is one of the most important factors for the success of any event. It is mainly determined by the temporal and spatial staggering of the direct wavefront of the sound sources, their first sound reflections and by the duration and the timbre of the reverberation.
[0002] There is the desire for balanced sound reproduction from the foremost seats to the last rows is often countered by commercial considerations. With the maximum number of spectators, the differences in sound pressure distribution increase. In large buildings there are long detours for the first sound reflections, which then reach the listener too late and lead to poor speech intelligibility, a subjectively unsatisfactory impression, and even undesirable echoes.
[0003] Until the last century, the size of acoustically high-quality concert halls was therefore limited. Architects had to make sure that the first sound reflections on the audience seats arrived at the listener within a time window of about 50 milliseconds after the direct wave front. This corresponds to a maximum detour of seventeen meters. Only reflections arriving at the listener within this time window are still assigned to the original sound. They not only increase the subjectively perceived loudness, but they are the core of the spatial perception of the sound event.
[0004] In particular, if they arrive at the listener from an angle of about 55 degrees next to the median axis, they reduce the interaural cross-correlation. This determines our spatial perception because a signal is only perceived as spatial if the signals between the right and left ear differ as much as possible in their phase position. If both signals are equal, i.e. completely correlated, we perceive a frontal mono sound source without spatial effect. The superposition of the direct wavefront with the first reflections leads to significant frequency response changes. Strong comb filter effects in the frequency response occur especially when the amplitude of the reflection is not much smaller than that of the direct wavefront. Especially then, learned stimulus patterns in interaction with our Head Related Transfer Function (HRTF) give us an idea of the room size. Unconscious head movements, whose effect we associate with learned stimulus patterns, thereby confirm the perceived position of the sound source in space.
[0005] The large number of reflections that later hit us from all directions as reverberation cannot cause such significant changes in the frequency response because their multiplicity leads to a statistical superposition without narrowband frequency response changes. They change the overall spectrum and mainly carry information regarding the surface characteristics of the room, because they have been reflected from its surfaces many times before they reach the audience. In each case, the frequency response of the reflecting surface is imposed on them, so that they determine the "timbre" of the room. The ratio between direct sound to the relatively constant reverberation level, which is throughout the room, is therefore one of the most important indicators of the perception of distance from the sound source, but does not contribute to their directional determination.Therefore, reverberation determines the acoustic quality of the event space only to a limited extent. The decisive factor for the perception of good room acoustics is the temporal and spatial arrival of the first sound reflections.
[0006] Only if the architect has succeeded in providing the rear audience seats with sufficiently strong early reflections as well, can sound sources with largely omnidirectional radiation still be perceived loudly and clearly there. They increase the volume without obscuring the modulation depth of the signal and thus speech intelligibility and clarity of the audio information. Although reverberation also contributes to increasing the sound pressure level in the back rows, it masks the source signal up to the point of making speech unintelligible.
[0007] The acoustics of some concert halls are famous mainly for the fact that architects reach by means of targeted arrangement of reflection surfaces have succeeded in creating a balanced distribution of direct sound and first reflections over the entire audience area. This causing a balanced distribution of the perceived loudness of the signal. In the area of the central seats, this can often be achieved retrospectively with the help of appropriately arranged reflection surfaces, if the room itself has not been shaped accordingly. In order to supply more distant audience areas with sufficient energy, the reflectors would have to be very large. Their position and curvature would be determined by the geometry of the room. In very large venues, this is usually not possible in practice because either the room height is insufficient or the reflectors for the back rows would hang in view of the stage.
[0008] It is therefore the task of the invention to find a solution with which early strong sound reflections can be generated from virtual reflectors also for those audience areas for which this would not be feasible with real reflectors in the given room geometry.Electronically generated early sonic reflections are to be radiated to the audience seats from a suitable location and with targeted alignment. The aim is to extend the range of acoustically high-quality seats and to enhance the subjectively perceived perception of the sound event over a wide audience area in a large hall or possibly also in the open air.
[0009] The prerequisite for this is that an arbitrarily shaped audience area can be specifically hit with the sound waves without neighbouring areas being undesirably taken by the generated sound waves. With the distance from the sound source of such a virtual reflection to the listener, the propagation time changes, so that audience areas outside the targeted area would not hit inside the suitable time window of about 10 to about 50 milliseconds.
[0010] A suitable solution is to radiate the corresponding sound waves from a sound transducer array constructed as a surface, whose transducers are individually controlled according to the principle of wave field synthesis or beamforming methods. The control must offer the possibility to shape the wave front freely so that neighbouring areas are hardly hit unintentionally. In addition, the latency of the system must remain very low.
[0011] Such a technical possibility is described in WO 2023280982 Al. Here, the shape of several independently and simultaneously radiated, non-spherical wavefronts can be adapted in a coordinate system to the desired shape of the selected listener areas. It is also possible differences in the sound pressure level over the selected area are largely compensated by a vector-based system.
[0012] This creates the technical prerequisite for generating reflections of several virtual reflectors over a wide listener area in the appropriate time window. How such two- dimensional sound transducer surfaces, which are now available as modular units, can be used for this purpose in practice is described in figures 1 to 6 by way of example in a multi-purpose hall in which because of their large dimensions the wall surfaces cannot be used as a reflection surface and whose room height does not permit suitable positioning of real reflectors, at least in the rear audience area.
[0013] Figure 1 shows a scaled sketch of this exemplary multipurpose hall (1), which actually exists in Germany. It has a floor area of 74 x 74 meters, is 12 meters high and holds about 5,500 spectators without seats. Among other things, it is used for live concerts, where two line arrays (3), (4) suspended to the right and left of the stage (2) usually provide the sound reinforcement. The audience area (5) is only slightly affected by reflections from the fissured ceiling structure in its rear part, the side reflections areweak and counterproductive due to the runtime. In this area at the selected audience location (6), the opening angle to the line arrays is only 28 degrees, and spatial perception is correspondingly tight.
[0014] The possibilities to improve this by suitable acoustic reflectors are strongly limited due to the relatively low ceiling height. The acoustic reflector (7) shown in the sketch would have to be suspended 18 meters above the floor and would therefore not be feasible. However, it is one of the few positions from which the reflection from the right speaker (3) would arrive at the audience seat (6) in the correct time window. The temporally usable area (8) for an strong early reflection, where the wavefront would arrive about ten to fifty milliseconds after the direct wavefront of the right loudspeaker (3), then extend over a wide strip (8) through the audience area (5).
[0015] The acoustic reflector (7) would have to be curved accordingly. The curvature distributes the small amount of acoustic energy impinging on the reflector at the given distance from the sound source over an area that is large compared to its own surface area. The sound pressure level of the reflection is correspondingly low. The comb filter effects from the superposition with the direct wavefront then do not lead to the deep dips or exaggeration in the frequency response that would significantly enhance spatial perception at the listener position (6). The effect of the reflector would therefore remain limited unless it was designed to be very large.
[0016] Accordingly, such a multi-purpose hall (1), like many other poorly designed concert halls, can hardly be acoustically enhanced with real reflectors in retrospect. The reverberation can still be adjusted in duration and frequency response by changing the surface materials, but the spatial distribution of the first sonic reflections, the core of spatial perception, remains predetermined by the size and shape of the room. Electronic solutions, such as programmable line arrays, delay lines or delta stereophony, reduce the sound pressure drop in the back rows, but rarely contribute to an improved spatial perception of the sound event.
[0017] To simulate early sound reflections electronically failed so far because the radiation of the loudspeakers could not be limited sufficiently precisely to a selected audience area. In addition to the timing problems, the level of a broadly radiating loudspeaker dominates the nearby audience, while more distant areas are not supplied with sufficient level. Moreover, the delay time to the direct wavefront may then be too short in some places and too long in other areas. In contrast to real reflectors, it wouldalso be possible for electronically generated reflections to arrive in some areas before the direct wavefront of the sound source if the radiation is not sufficiently targeted. Then the sound source would be localized according to the law of the first wavefront in the direction of the reflector.
[0018] In the bass range, amplitudes of the same size would cancel or superimpose each other at 10 milliseconds delay to each other at 50 Hertz. Because no directional determination of the wavefronts is possible in this frequency range, the superpositions would not contribute to spatial perception here. They would lead to relatively broadband dips and exaggerations in the frequency spectrum of the sum signal, which would be irregularly distributed over the audience area (5). Therefore, the bass range in the frequency response of virtual reflectors, how they are represented in the next sketch, must be lowered significantly.
[0019] Figure 2 shows a basic solution for setting up a virtual reflector with the aid of an electronically controllable loudspeaker array (9) constructed as a sound transducer surface. Here, the virtually sound reflection generated by the input signal of the right loudspeaker (3) does not originate from the acoustic reflector (7). At the listener position (6), a wavefront of the right ceiling-mounted loudspeaker array (9) arrives from the same direction and also with the same time delay regarding the direct wavefront of the right loudspeaker (3) as the reflection of a real acoustic reflector (7). The loudspeaker array (9) thus virtualizes the real acoustic reflector (7) for listener position (6) from a different position.
[0020] For the listener at the listener position (6), the wavefront generated by the loudspeaker array (9) is indistinguishable from the reflection of a real reflector that would be mounted at the position of the acoustic reflector (7). When this listener looks in the direction of the stage, the reflection hits him from an angle of about 55 degrees next to the median plane, i.e. from the area in which a sound reflection has the best spatial effect. In addition, the electronic solution has the advantage that the level and frequency response of the reflection can be largely freely adjusted. This expands the psychoacoustic possibilities of sound field design compared to conventional reflector solutions.
[0021] Because the time that elapses before the wavefront of the sound source (A) hits a real reflector is eliminated in the electronic solution, there is also a high degree of freedom in positioning the loudspeaker array (9). It can be further away from the listener by the distance between the sound source (3) (A) and the loudspeaker array (9) (B)operating as a virtual reflector than a real reflector for the same coverage area. Small changes of the angle of incidence of the wavefront at the listener have only little influence on the spatial perception. However, the temporal staggering of the wave fronts, which is substantial for the perception, can be freely designed within wide limits.
[0022] The calculation of the transit time between the direct wavefront of the sound source (A) and the selected listener position (C) can be done on the basis of the coordinate system (10). The origin of the coordinates can be chosen freely. In interaction with the method described in [7] it is recommended to take the coordinate origin and also the temperature dependent value of the sound velocity from this vector-based system, if necessary, also from other suitable systems.
[0023] For the calculation of the delay time between the direct wave front, which in the sketch starts from the right loudspeaker (3) (A) and the early reflection, which is radiated from the loudspeaker array (9) to the listener place (6), the listener place (6) is chosen as an example in the sketch. Seen from him, the stage area has only a small base width. At this place, hardly any real reflections are to be expected, which arrive in the favourable time range of approx 10 to 50 milliseconds after the direct wave front. Accordingly, only an insufficient spatial impression is given here.
[0024] When planning the mounting positions of loudspeaker arrays that are to work as virtual reflectors, it is recommended to start at the rear of the audience area. This involves finding a suitable place where the loudspeaker array (9) can be mechanically mounted as an electronically controllable virtual reflector. Ideally, he should be about 55 degrees off the median plane of the audience member looking toward the stage, at an elevation angle of 30 to 45 degrees. At a lower elevation, it becomes more difficult to limit the radiation of the transducer area to a selected area of the listener. Above 45 degrees elevation angle, the reflection increasingly hits the averted ear of the listener, again increasing the correlation of the ear signals. In the sketch Fig.2 the position of the loudspeaker array (9) is perfect for the listener position (6).
[0025] For this reference point, the time delay is now calculated with which a directional radiation of the source signal (A) should arrive as a virtual reflection in the area of the listener location (6) (C). Because the usable area of the relevant beam should not only hit this listener place (6), but should be usable in as large an area around it as possible, the desired delay for this area is set here to the middle between the upper- and lower-time limits, i.e., exemplarily (10 ms + 50ms) / 2 = 30 ms.
[0026] In the coordinate system (10), the travel time of the direct wave front is then calculated from the distance of the sound source (A) to the listener (C) using the formula which can also be used to determine the diagonal of a cuboid from its edge lengths. This vector results for the sound source (A) and the listener place (C) from the coordinate points
[0027] It does not matter whether the source position is entered first or the listener position is entered first. The square root of the sum of squares of the coordinates always gives the amount of distance between the source and the listener. In the exemplary of the sketch 2 a distance of 52.738 mm results from the values referred to the coordinate origin. The corresponding sound travel times are then obtained by dividing the distances by the speed of sound. At a sound velocity of 344 m / 2, the travel time of the direct wave front from the right loudspeaker (2) A to the listener position (6) C is then 153.3 milliseconds.
[0028] Following the same procedure, the next step is to calculate the transit time of the sound from the loudspeaker array (9) (B) to the listener location (6) (C).
[0029] If the corresponding beam of the loudspeaker array (9) were driven by the signal of the right loudspeaker (3) without delay, the virtual reflection would arrive at the listener position (6) after the calculated sound propagation time (A to C) minus (B to C) before the direct wavefront of the right loudspeaker (3). Therefore, the input signal for the beam in question must be delayed by this delay time difference and additionally by the desired delay time of the reflection with respect to the direct wavefront so that it arrives at the listener position (6) in a suitable time window.tBEAM = - (tBC-tAc) + tDELC (3)
[0030] So that the selected listener point lies approximately in the middle of the area which can be supplied with the selected sound beam, exemplarily for this beam of the loudspeaker array (9) the input signal in the sound transducer array is additionally delayed with IDELC of 30 milliseconds. This delay is then fixed for this selected beam and is not changed anymore. Even if the position of the right loudspeaker (3) varies slightly at the next event, this has little effect on the virtual reflection reaching the listener position (6) in the correct time range and from an appropriate direction. For a delay of the reflection of 30 milliseconds compared to the direct wavefront, the delay time of the inputsignal in the exemplary sketch for the selected beam is - (71.21ms-153.31ms) + 30 ms = 112.1 ms. Latencies of the mixing console, the line array or the corresponding beam in the transducer array (9) have to be compensated in the course of this.
[0031] After the position of the loudspeaker array (9) and the delay time tBeam of the input signal for this beam have been determined, it has to be determined in which listener area the virtual reflection arrives within the limits of the suitable time window. For this purpose, grid lines (12) are laid over the audience area (5) in the coordinate system (10), at the intersection points of which it is determined whether the virtual reflection arrives within the given of the set limits, in the example between approx. 10 and 50 milliseconds, after the direct wavefront at the respective intersection point of the grid lines (12). The grid spacing should be a maximum of 2.5 meters, as shown in the sketch, because otherwise may significantly transit time differences between neighbouring points will already occur. If the procedure is coupled with the vector calculation of the software from [7] or comparable procedures, it is recommended to adopt the coordinate system used there and the grid lines with their intersection points.
[0032] In order to determine this usable time range (8), which is marked with dots in the sketch, the time delay with which the synthetic reflection emanating from the loudspeaker array (9) hits each individual intersection point of the grid lines (12) must be calculated for each intersection point of the grid lines (12). A possible way to calculate this time delay is to subtract the sound propagation time tAC from the sound source to the respective crossing point from the sound propagation time of the loudspeaker array (9) (B) operating as a virtual reflector and to add to the calculated value the delay time of the electrical signal tBeam defined for the beam.tl = (tBCl -tACl) + tBeam (4)
[0033] Thus, for each individual intersection of the grid lines (12), the time difference of the virtual reflection with respect to the arrival of the direct wavefront of the sound source (A) is obtained.
[0034] In the next step, it is then checked whether this time difference lies within the set limits in which the wavefront of the relevant beam of the loudspeaker array (9) is suitable to simulate an early sonic reflection of the sound source (A). In the sketch, these are all points (8) that lie in the range between 10 and 50 milliseconds after the direct wavefront. To avoid mislocalizations, the limit of 10 milliseconds should not be fallen short of. Themaximum delay time is less critical, only above about 70 milliseconds the reflection becomes really disturbing.
[0035] In the exemplary sketch, only a part of the temporally usable area (8) is selected as the reflection area (11) of the loudspeaker array (9) with the specified delay time tBeam. Reflection would increase the perceived loudness of the sound source in the time-usable area (8) extending beyond it, but would not extend the perceived base width of the stage. In the unused area, the reflection would come at the listener from a direction outside the base width of the loudspeakers. For an improved spatial perception, it would be counterproductive here. Accordingly, only the coordinates for the selected reflection area (11) are adopted in the software of the transducer modules. Thus, the beam can then be limited to this area using the method described in [3] or a comparable method.
[0036] In doing so, the sound pressure level can be kept largely constant by the software in [7] over the selected range, regardless of the distance between the loudspeaker array (9) and the intersection points of the grid lines. In principle, however, it would also be possible with the method described in [7] to program a fixed ratio of the levels of the virtual reflection to the direct wavefront or to design the level ratio as a function of the sound source.
[0037] A suitable ratio is, for example, when the reflection at the listener's position is about 10 dB quieter than the direct wavefront, which corresponds approximately to the level of first sound reflections in acoustically good concert halls. Then the comb filter effects are deep enough to address the stimulus patterns of the perception of the acoustics of a large room, but not deep enough to produce nulls in the frequency response.According to [4], comb filter effects are still perceptible in the frequency response when their level is up to 27 dB below that of the direct wavefront. Moreover, the reflections are then capable of equalizing the loudness of the signal over the entire audience area. As in good concert halls, the signal in the far distanced seats is then subjectively hardly perceived as quieter than in the front rows - in contrast to open-air events or environments with little reflection.
[0038] The loudspeaker array (9) can simultaneously generate additional beams in other directions if the individual transducers are driven separately according to the principle of wave field synthesis. Corresponding systems have been state of the art and available for several years. The signal content and thus also the delay time of the drive signal of the individual sound beams is independent of each other [7], Thus, the same loudspeakerarray (9) can be used to virtualize further reflectors at other room positions and with other orientations than the acoustic reflector (7).
[0039] Figure 3 illustrates this possibility. Thus, further virtual reflections from the direction of the loudspeaker array (9) can also be hit in the areas (13), (14) and (15) with other delay times tBeam of the input signal. As far as possible, there should be no gaps between the individual coverage areas. This process can be designed most effectively with appropriate graphic support of a calculation program. If the temporally usable area is marked in the 3D model, its position can be changed simply by changing the delay time tBeam of the respective beam. In the example, it moves towards the stage when the time offset is shortened.
[0040] If the value is found at which the new time usable area 2 (13) joins the previously defined usable area (8) without gaps, the corresponding delay time is fixed for this beam and the procedure is repeated for the next beam of the same speaker array (9). The limit of possibilities is reached where the delay times become negative or the number of available independent beams of the speaker array is reached. In the example this is the area (15) directly in front of the stage.
[0041] If the area should be extended to where the delay time of the beam would become negative, i.e. the direct wave of the sound source arrives before the wave front of the virtual reflector can reach the point, there would still be the possibility to delay the signal of the loudspeaker itself and to add this time to all defined delay times tBeam of the individual beams. However, this possibility could not be used for live events because of the additional latency of the loudspeaker signals.
[0042] The selected reflection areas (11) (13), (14) and (15) are then also determined according to whether the direction from which the respective reflection from the loudspeaker array (9) arrives at the listener increases the perceived base width of the stage and thus improves the spatial perception of the sound event. As far as it is important to use the audience area (16) to be used from the time domain, although it is counterproductive for the directional perception of the sound source, but suitable to increase the perceived loudness, the planner may decide to use this option.
[0043] As far as only single loudspeaker arrays are installed, there will also remain partial areas of the audience area (5), in which a supply of virtual sound reflections is not possible in the usable time range and from a suitable direction. With several appropriatelydesigned loudspeaker arrays, however, it is possible for these to complement each other in their selected reflection areas.
[0044] Figure 4 shows such a combination of several loudspeaker arrays (9), (17), (18), (19) working together in a modular system. Distributed modules can better supply smaller parts of the audience area (5) from the area to be preferred for spatial perception about 55 degrees next to the median plane of the listener. In contrast, the advantage of better concentration of sound waves in the bass range, which would be associated with combining multiple loudspeaker arrays into a single unit, is less important for reflections.
[0045] From its position, the loudspeaker array (17) can reach the reflection areas 5 (20) and 6 (21) from a direction in which the angle of incidence of the wave fronts of the loudspeaker array (9) increases the perceived stage width and thus improves the spatial perception. This would not have been possible from the position of the loudspeaker array (9). This would not have been possible from the position of the loudspeaker array (9).
[0046] Thus, the coverage areas of the individual loudspeaker arrays complement each other in such a way that the overlapping areas of the individual beams are as small as possible. It is true that even in good concert halls several sound reflections arrive simultaneously at the individual listener positions. However, their number should remain as single-digit as possible, because otherwise the position-dependent comb filter effects in the frequency response add up to a statistical compensation of the effects, comparable to the superposition of many reflections to the reverberation.
[0047] In the depicted coverage, almost over the entire audience area (5), each listener is hit by an early sound reflection, which hits him from the right in the time window suitable for the spatial perception of the sound event. If the described system is mirrored in the depicted room, then each listener is also hit from the left side by another reflection whose sound source is the left loudspeaker (4). As far as the room is not symmetrical, this side is calculated separately in the described way. Such a design of the sound field is suitable to extend the base width in the phantom sound source reproduction significantly.
[0048] Insofar as the technical limits of the loudspeaker arrays are not exhausted by this is possible to divide the audience area (5) into a larger number of smaller reflection areas. Then the usable time window can be tightened, for example limited to 25 to 55 milliseconds, so that a relatively wide impression is created at all audience locations. If, in another setup, a time window of, for example, 10 to 35 milliseconds is preferred in order to represent a smaller space, modified coordinates of the selected reflection areasmust defined in the software of the loudspeaker arrays and, if necessary, become stored as a preset.
[0049] In any case, the sound pressure level of the virtual reflections should at least below approx 100 Hz be strongly reduced in the bass range. In this range, there are no narrowband comb filter effects as with shorter wavelengths. We also cannot spatially assign the direction of incidence of the reflections in that range, which is why they do not contribute to the improvement of spatial perception. The superposition with the direct wavefront in the frequency range only leads to spatially dependent dips or overshoots in the frequency response instead of improving the spatial perception. It is therefore important the frequency response of the loudspeaker arrays (9), (17), (18), (19) below 100 Hertz to lower significantly, better already from 200 Hertz. Accordingly, relatively light loudspeaker arrays with low volume can be used for the chambers of the individual sound transducers.
[0050] Superficially, the virtually reflections above about 5 kHz contribute less to the spatial perception of the sound event, because in this range several wavelengths fit into the ear distance. Therefore, here the determination of direction via the interaural time differences becomes ambivalent. However, localization based on Head Related Transfer Function HRTF works in this range. Small head movements can help here to confirm the position of the perceived sound source.
[0051] Not only for this reason, the loudspeaker arrays used should be so tight equipped with transducers that a directional radiation of their wavefront is possible up into the high frequency range. This also has the advantage that in very large venues, even the back rows can be supplied with harmonics by the virtual reflectors. Especially in dry room air, a significant level drop of the direct wavefront of the sound source cannot otherwise be avoided in the frequency range due to the large distance to the listener.
[0052] The described method can be used to significantly improve spatial reproduction at events in which phantom sound sources are generated between two loudspeakers.However, the described method also can be used for real sound sources, for example voices or instruments of an orchestra, which are not directly amplified electronically. Also, virtual sound sources of the wave field synthesis must have the reflections related to the source itself, not to the loudspeakers.
[0053] Figure 5 shows how a real sound source or a virtual sound source of wave field synthesis should be supported in the localization of the sound source by sound reflectionsfrom both directions, i.e. by reflections from the left and from the right. The direct wavefront here emanates from the sound source itself (22) or also from loudspeakers in its immediate vicinity on the stage artists. The electrical signal needed to create virtual reflections of the sound source can come from microphones or from pickups on the instruments.
[0054] As far as the microphone (23) is not positioned in the immediate vicinity of the sound source, the sound propagation time from the sound source to the microphone must be subtracted from all calculated delay times for the virtual reflections. Corresponding to it may be necessary to add the same delay time also to the via pickups or directly fed instruments in the mixing console, so that all signals are synchronous again.
[0055] The coverage areas do not change in the process because they are related to the position of the sound source. In all other respects, the procedure is similar to that described in Figs 1 to 4, with the difference that each sound source is supplemented with both a right and a left reflection.
[0056] In the sketch, only one beam of each of the loudspeaker arrays (9) and (17) is drawn, which hits the listener position (6) with a time delay of 30 milliseconds to the direct wavefront of the sound source. Of the respectively usable reflection areas 7 and 8 (24), (26), the selected reflection areas 7 and 8 (25), (27) are used.
[0057] The rest of the audience area is supplied by further beams of the same loudspeaker arrays, which are not shown. This raises the question of how the position of the sound source on the stage or the change in location of the sound source across the stage area affects the perception of the sound event.
[0058] Figure 6 shows such a change of location of the sound source (28), in the example by 7.50 meters from the centre of the stage to the right edge of the stage. At the listener position (6), the delay time in the example for the reflection arriving from the right side would increase from 30 to 32.6 milliseconds if the original delay time of the signal for the beam is maintained. The original coverage area would not change because neither the position of the speaker arrays (9), (17) nor the listener positions nor the propagation times would have changed.
[0059] As far as the values for the delay of the input signals of the beams would be corrected so that the time delay of the virtual reflection at the listener position (6) would remain constant, in the example the delay time of the corresponding beam on the right would have to be increased from 111.7 to 114.3 milliseconds and on the left from 123.0to 125.5 milliseconds. This would shift the area where the reflection arrives in the time window selected here from 10 to 50 milliseconds after the direct sound from the dotted areas to the areas marked with circles (29) and (30). The delay times for the beams in the adjacent audience areas would have to be adjusted accordingly. The slightly changed delay times would have no effect on perception.
[0060] Both methods would be technically feasible, although the second option would require more technical effort. From a psychoacoustic point of view, there is no advantage in changing the position of the coverage areas by time corrections if the sound source moves on stage or is positioned outside the center of the stage. For the perception of the concrete room dimensions, the optical reference to the reflection surfaces is missing, the virtual reflections can only convey the feeling of a large room. With very large stages, however, care must be taken to ensure that the reflections always reach the listener from the correct side. This limits the coverage areas if necessary, because this condition must be fulfilled for the outer right as well as for the outer left stage position.
[0061] Figure 7 shows the possibility of using the loudspeaker arrays (9), (17), (18), (19) additionally to further equalize the acoustic perception over the entire audience area. At many venues, the room acoustics are designed for the best possible perception in the center of the hall, where the highest ticket prices can be realized for the best seats. The rear seats are cheaper, because there the level of the direct wave drops significantly and the small opening angle to the stage area results in a high correlation of the ear signals with a correspondingly low spatiality of perception. Ceiling reflections only able for equalize the perceived loudness, the correlation of the ear signals is not reduced by reflections from the direction of the sound source.
[0062] Because the reverberation has approximately the same level everywhere around the room, it covers the direct wavefront in this rear area especially when the perceived loudness of the direct wavefront drops significantly due to large distance to the sound source. The Speech Intelligibility Index (STI), which is used to evaluate speech intelligibility, therefore drops sharply in this area, which also worsens the perception of musical performances.
[0063] Besides the improved spatial perception, the virtual reflections significantly increase the perceived loudness of the source signal in this audience area because they are still attributed to the direct wavefront within their time window. Therefore, they do not mask the signal content, but increase the level difference to the reverberation.Accordingly, the STI value also increases because it is formed by the ratio of the sound pressure levels in the first fifty milliseconds to all reflections arriving later.
[0064] Also, the front rows mostly offered at a lower price in concert halls, because here the high direct sound component masks the perception of the room acoustics. Early sound reflections can often be created in that area by reflective surfaces such as stage lamination, but a balanced reverberation component is also important for spatial perception. Its spectrum and duration give an impression of the nature of the surfaces and also of the size of the room, as we would expect from a hall with good acoustics. In the front rows, however, the reverberation is masked by the high direct sound level and, if necessary, the first strong reflections in this area.
[0065] Because the reverberation in a room hits us from all directions, the direction from which it reaches the listener is irrelevant for the localization of the sound source. Also its arrival in time is not subject to the strict rules we have to apply to the first sound reflections. For signals of mirror sound sources, whose starting points after umpteen reflections on the surfaces of the room are virtually several hundred meters away from the listener, are neither directionally nor temporally closely correlated with the source signal. Neither their level nor their direction of incidence change noticeably as the listener moves around the room. Therefore, we can direct synthetically generated convolution reverberation from all directions to the areas where a little more reverberation improves perception. In the sketch, area (31) is selected for this purpose in the audience area (5). However, it is also possible to define several reverberation level zones in which the reverberation component is defined separately. In the impulse response of a suitable concert room with which this convolution reverb is generated, the first reflections should be suppressed so that no strong wave fronts arriving too late are perceived as echoes. The reverberation level in the front speakers themselves should be reduced accordingly.
[0066] In principle, it would also be possible according to the described method to provide the selected reflection areas (11) of the audience area (5) with initial sound-strong reflections or additional reverberation by means other than the loudspeaker arrays described in [7] or comparable methods. Several audience areas can be sonicated simultaneously with several single beams from several virtual reflectors, generated by appropriately designed sound guides. These can be horn structures, acoustic lenses or surfaces of enclosures shaped according to the desired radiation characteristics, which areequipped with a large number of sound transducers and can thus form the desired wave front from their elementary waves.
[0067] In this case, only a single coverage area is illuminated at a time, so that a larger number of virtual reflectors is needed to cover the entire audience area. In return, however, it is also possible to dispense with separate control of the individual transducers in such an enclosure and to control all the loudspeakers of the respective virtual reflector with a common input signal delayed by the delay time tBeam. In addition, as in principle with the electronically controlled loudspeaker arrays, there is the possibility of first irradiating a reflection surface with the individual beam before the corresponding beam hits the supply area. The shape of the reflection surface can support the alignment of the wavefront. Thus, the design of the transducer units working as virtual reflectors could be adapted to the room and not be recognizable as transducers.
[0068] The field of application of the described method is not limited to improving the perceived acoustics of existing buildings. As far as the method is already included in the planning phase, the audience areas can be designed significantly larger, which can significantly reduce the relative construction costs per audience seat. There are also large open-air venues, where suitably positioned loudspeaker arrays acting as virtual reflectors could give the audience the impression that they are sitting in an acoustically high-quality hall. Finally, it would even be possible in the home to create virtual room acoustics that are clearly superior to the real playback room.
[0069] Even a textured surface of the virtual part of the concert hall can be simulated. For this purpose, only the input signal of the beams has to be structured accordingly. A smooth surface throws back a single wavefront. Structured surfaces produce multiple wavefronts with a few hundred milliseconds offset from each other, which are superimposed on the listener. If an appropriately processed signal is used as the source for the reflections, it is distributed in all the beams and creates the corresponding perception.
[0070] In all applications, it is also possible to customize the setup of the event with stored presets that can also be changed during the event. A speaker can remain easily understandable over the entire audience area with little reverberation and have about the same perceived volume, at a concert good acoustics can then be virtually created over a wide audience area and a choir can impress with somewhat stronger reverberation from all directions.
[0071] Reference signs1 Multipurpose hall2 Stage3 Right loudspeaker4 Left loudspeaker5 Audience area6 Listener Position7 Acoustic reflector8 Time usable area9 Loudspeaker array top right10 Coordinate system11 Selected reflection area12 Grid lines13 Selected reflection area 214 Selected reflection area 315 Selected reflection range 416 Timed but unused reflection area17 Loudspeaker array front right18 Loudspeaker array top left19 Loudspeaker array front left20 Selected reflection area 521 Selected reflection area 622 Sound source on stage center23 Microphone24 Usable reflection range 725 Selected reflection range 726 Usable reflection range 827 Selected reflection range 828 Sound source at right edge of stage29 Shifted reflection range right30 Shifted reflection range left31 Area supplied with additional reverberation
Claims
Claims1. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors,characterized inthat by means of an electronically controllable loudspeaker array constructed as a sound transducer surface or by means of a suitable other sound transducer construction, acoustic wave fronts are radiate targeted in at least one listener area, in which wave fronts arrive with a suitable delay time range relative to the direct wave , in which they subjectively become assign the direct wave front of the sound source, so that a reduced correlation of the ear signals and an increase in the perceived loudness of the sound event cause an enhancement of the spatial perception and the perceived loudness of the sound event, wherein can used the advantage of virtual acoustic reflectors over a real reflector, that the transit time of the direct wave front of the sound source to the real reflector is omitted in the electronic signal transmission to the virtual acoustic reflector, is used to have a higher degree of freedom in positioning and extended possibilities in the design of level and frequency response of the reflection compared to the position of a real reflector.
2. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat the sound transit times are determined by dividing the distances calculated in a vector calculation or by other suitable means with respect to the positions of sound source, whereby the calculation of the transit time difference between the arrival of the direct wave front of the sound source at a selected point in the audience area and the arrival of a virtual reflection emanating from a virtual acoustic reflector can be performed on the basis of a coordinate system whose coordinate origin can be freely selected and which, in cooperation with a method for the spatial control of virtual acoustic reflectors constructed on the basis of electronically controllable loudspeaker arrays, can also adopt the coordinate origin from this system.
3. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat the input signal for a selected sound beam of the virtual reflector becomes delayed by the negated time difference from virtual reflector to a selected listener Point minus the propagation time from the sound source to that listener point, additionally delayed by the desired delay time tBeam of the reflection with respect to the direct wavefront, so that it arrives at the selected listener position in a suitable time window after the direct wavefront of the sound source, in which it enhances the spatiality of perception, the perceived loudness and the speech intelligibility STI of the sound event.
4. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat only that part of the temporally usable reflection range of the loudspeaker array beam selected with the specified delay time tBeam of the selected beam is selected as coverage range in which the incidence direction of the virtual reflection would improve the spatial perception of the sound event, unless the reflections in the temporally usable range extending beyond this although not would extend the perceived base width of the stage, but the advantage that the perceived loudness of the sound source in this range could increase would outweigh.
5. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat further virtual reflections from the direction of the same virtual reflector with different delay times of the input signal hit other parts of the audience area, whereby no gaps should occur between the individual coverage areas as far as possible, which is most effectively achieved by carrying out this process with the aid of a calculation program that visualizes the usable time range of each virtual reflection with appropriate graphical support, whereby the usable time range is marked in the 3D modeland the usable range of each individual beam of the loudspeaker array can be shifted solely by changing the delay time tBeam of the respective beam, whereby the usable time range can also be extended to where the delay time tBeam of the relevant beam would be negative, because the direct wave of the sound source arrives there before the wavefront of the virtual reflector can reach this point, by delaying the signal of the loudspeaker itself and adding the corresponding time difference to all specified delay times tBeam of the individual beams, which however increases the latency of the loudspeaker signals, which is why this method can only be used for recorded signals.
6. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat the sound pressure level of the virtual acoustic reflectors in the bass range, at least below approx.100 Hertz, is strongly reduced, because the listeners in this range cannot spatially assign the direction of incidence of the reflections, which is why they do not contribute to the improvement of spatial perception, but the superposition of the virtual reflections with the direct wavefront in that frequency range leads to locationdependent dips or overshoots in the frequency response, instead of improving spatial perception, which is why the frequency response of the virtual acoustic reflectors should be slowly lowered already below 200 Hertz, which can also be done by the fact that loudspeaker arrays designed for the intended use can be relatively light and can be designed with a small volume for the individual, preferably closed chambers of the individual sound transducers.
7. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat the described method can be used to enhance the spatial reproduction of phantom sound sources or of real sound sources, for example voices or instruments of an orchestra, which are not electronically amplified but are available as an electrical signal via microphones or pickups, or of virtual sound sources of wave field synthesis, the transit times to microphones or also the latency times of mixing consoles and loudspeakersystems being included in the calculation of the sound transit times tBeam of the virtual acoustic reflectors.
8. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat the loudspeaker arrays installed as virtual acoustic reflectors can also be used to balance the perceived acoustic impression over the entire audience area in that synthetically generated convolution reverberation is directed from all directions to those areas where slightly more reverberation improves the perception, whereby several zones can be defined in which the reverberation component is defined separately, and whereby in the impulse response with which this convolution reverberation is generated, the first reflections are suppressed so that no strong wave fronts arriving too late are perceived as echoes, whereby the reverberation level in the front loudspeakers themselves should be reduced accordingly.
9. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthe perceived acoustics of planned buildings can be designed by incorporating virtual acoustic reflectors into the planning phase in that way, thereby becomes possible to make the audience areas of new objects significantly larger, or that in outdoor venues suitably positioned virtual acoustic reflectors can give the listener nearly the perception that he or she is sitting in an acoustically high-quality hall, or that the method can also be used in the home to create virtual room acoustics that are superior to the real acoustics of the playback room.
10. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1,characterized inthat the setup of the event can be stored in presets, which can also be changed during the event, whereby with the help of the virtual reflections, the acousticimpression of a structured surface of the reflection surfaces can also be simulated by structuring the input signal of the beams accordingly by means of emitting several wave fronts from the loudspeaker arrays, which overlap at the listener, with a time offset of several hundred milliseconds.
11. Method for enhancing the spatial perception of a sound event with the aid of virtual acoustic reflectors according to patent claim 1 to 14,characterized inthat the selected reflection areas (11) of the audience area (5) are provided with initial powerful reflections or additional reverberation by means of loudspeaker arrays other than those described under [7] or comparable methods with which several audience areas can be illuminated simultaneously with several beams in that way, that individual beams are generated by appropriately designed sound guides, which are designed as horn structures, acoustic lenses or as surfaces of housings shaped according to the desired radiation characteristics, which are equipped with a large number of sound transducers and thus form the desired wavefront from their elementary waves, whereby they rely on a separate control of the individual sound transducers can be dispensed with and all loudspeakers of the respective virtual reflector are controlled with a common input signal delayed by the delay time tBeam, wherein their beam also can being projected against a reflection surface which shape can support the desired alignment of the beam and the additional detour of the wavefront becomes added to the delay time tBeam.