Wave field synthesis

from Wikipedia, the free encyclopedia

The wave field synthesis (WFS) is a spatial audio reproduction process with the aim of virtual to provide acoustic environments. The synthesis creates wavefronts that emanate from a virtual point. Its acoustic localization does not depend on the position of the listener and also not, as with conventional multi-channel methods, on psychoacoustic effects such as phantom sound sources.

WFS reconstruction of the wave front, as animation for web links

Physical basics

The synthesis is based on the Huygens principle , which describes that every wave front can also be viewed as a superposition of elementary waves . Therefore any wave front can be synthesized from such elementary waves.

In order to generate sound waves , a computer synthesis controls each of the individual loudspeakers - arranged in rows around the listener - at the exact moment at which a virtual wave front would pass through its point in space. Professor Berkhout invented this process in 1988 at Delft University of Technology .

The mathematical basis is the Kirchhoff-Helmholtz integral, which describes that the sound pressure is determined at every point within a source-free volume if the sound pressure and sound velocity are determined at all points on its surface.

Therefore, any sound field can be reconstructed if the sound pressure and sound velocity are reconstructed at all points on the surface of the volume. Such a spatial approach describes the principle of holophony. The entire surface of the volume, i.e. all the boundary surfaces of the reproduction room, would have to be fitted closely together with monopoles and dipoles for sound generation, all individually controlled with their respective signal. In addition, the room would have to be reflection-free in order to meet the requirement of a source-free volume. This is a practically impossible approach, which is why simplified, workable procedures had to be found. According to Rayleigh II, the sound pressure is determined at every point in a half-space if the sound pressure at every point in a plane is known. Because our acoustic perception is most accurate in the azimuth plane, the process has been generally reduced to a single row of loudspeakers around the listener.

The starting point of the synthesized wavefront can then be any point within the horizontal plane of the loudspeaker. It represents the virtual acoustic source, which hardly differs in its behavior from a real acoustic source at its position. It no longer seems to move when the listener moves in the listening room. Convex or concave wave fronts can be generated, the virtual sound source can also be located within the loudspeaker arrangement and it can even be "bypassed".

Advantages of the process

When reproducing an acoustic scene, stereophony , 5.1 and other conventional reproduction techniques achieve the optimum hearing impression in a narrowly limited area, the sweet spot . If a listener is outside the sweet spot, the perceived quality of the spatial reproduction is significantly reduced. In contrast to these conventional methods, WFS achieves at least an approximate reproduction of the wave field of an acoustic scene within an extended listening area. The optimal reproduction is not restricted to a single point and a high reproduction quality can be achieved for entire groups of listeners. In addition, the listeners have the opportunity to move freely within the listening area, which on the one hand intensifies the listening experience and on the other hand also makes the distance between a reproduced source and the listener directly perceptible. Assuming a suitable spatial recording, it is possible by means of WFS to physically restore the sound field of a recording room to a large extent, so that a change in the listener's position in the playback room would have the same acoustic effect as a corresponding change in location in the recording room.

It is also possible to use WFS to display the acoustics of a real or virtual room without directly recording an acoustic scene in this room. To simulate a real-world space, a non-echoed signal recorded correspondingly measured can be emulated with a plurality of room impulse responses folded be. These impulse responses represent the reverberation of the viewed room from different directions. For the acoustic representation of a virtual room, these impulse responses can be determined by a mirror sound source model.

The object-oriented transmission standard MPEG-4 was standardized by the Moving Picture Expert Group , which enables separate transmission of content (the dry recorded audio signal) and form (the impulse response or the mirror sound source model). Each virtual sound source needs its own (mono) audio channel.

The spatial sound field in the recording room consists of the direct wave of the sound source and a spatially distributed pattern of mirror sound sources. They arise from the fact that the direct wave is reflected from all surfaces of the recording space. Reducing their spatial distribution to a few speaker positions, as is the case with all conventional, channel-based methods, must inevitably result in a significant loss of spatial information.

The sound field can be synthesized much more precisely on the playback side. For this purpose, the individual sound sources are recorded dry. The reflections of the recording space are then synthesized on the playback side. Problems on the recording side, which cannot be solved with conventional methods, do not even arise with such an object-oriented method. In addition, not all signal components are inseparably mixed during playback; direct wave, first reflections and reverberation can be manipulated separately on the playback side.

The process also offers a clear advantage for conventional recordings: virtual sound sources called “virtual panning spots”, which reproduce the signal from the associated channels, can be positioned far outside the real reproduction area. This reduces the influence of the listener's position, because the relative changes in the angle of incidence and level are significantly lower than with the real loudspeaker boxes nearby. This extends the sweet spot considerably, it can now extend over almost the entire playback space.

Remaining problems

The most clearly perceptible difference to the original sound field is the reduction of the sound field to the horizontal level of the loudspeaker rows. It is particularly noticeable because the necessary acoustic damping of the playback room means that there are hardly any mirror sound sources outside this level. However, without this acoustic damping, the condition of the source-free volume from the mathematical approach would be violated. The consequence would be double-space reproduction.

The "truncation effect" is also disturbing . Because the wave front is not formed by a single sound source, but by the superposition of all wave fronts of the individual radiators, there is a sudden change in pressure when no further radiators make their contribution at the end of the loudspeaker arrangement. This “shadow wave” can be weakened if the level of the external loudspeakers is reduced. For virtual sound sources within the loudspeaker arrangement, however, this change in pressure leads the actual wavefront, making it clearly audible.

Since the WFS tries to simulate a different room than the existing one, the acoustics of the playback room must be suppressed. One way of doing this is to make the walls appropriately absorptive. The second possibility is the reproduction in the near field. The loudspeakers are located very close to the listening area or the membrane surface must be very large.

Another problem to this day is the high cost. A large number of individual sound transducers must be set up very close together. Otherwise, spatial aliasing effects will be audible. They arise because an unlimited number of elementary waves cannot be generated, as described in the mathematical approach. As a result of the discretization, position-dependent narrow-band notches in the frequency response arise within the playback area. Their frequency depends on the angle of the virtual sound source and the angle of the listener against the loudspeaker front:

For a largely aliasing-free reproduction in the entire listening area, the distance between the individual radiators would then be less than 2 cm. Fortunately, our ears are not particularly sensitive to this effect, so that it is hardly annoying at a distance of 10 to 15 cm. On the other hand, the size of the radiator field limits the display area; outside of its limits, no virtual sound sources can be generated. Therefore, the reduction of the procedure to the horizontal level seems justified to this day.

Research and market maturity

The newer methods for WFS were first developed at the TU Delft from 1988 onwards. As part of the EU-funded CARROUSO project (January 2001 to June 2003), ten institutes across Europe carried out research in this area. The participating research institutes were IRCAM, IRT and Fraunhofer IDMT , the participating universities were TU Delft, University of Erlangen, AU Thessaloniki and EPFL Lausanne (Integrated System Laboratory), participating companies Studer , France Telecom and Thales .

For the horizontal rows of loudspeakers around the listener, however, also because of the need to acoustically dampen the playback room, the acceptance problems are so great that the method has not been able to establish itself in the home. A breakthrough in this market segment is hardly to be expected for such a two-dimensional process, also because Ambisonics is a comparable three-dimensional solution for the home sector. Additional WFS converters on the ceiling or the inclusion of a 2.5-D synthesis operator will hardly change this situation. A three-dimensional proposal for a solution that includes the acoustics of the playback room in the synthesis is known, but due to the high level of effort, it has hardly been possible to this day. Other efforts are aimed at integrating the sound transducers invisibly into the playback room, for example through flat-panel speakers . An alternative approach is the psychoacoustic wave field synthesis: It does not aim at a perfect physical copy of the sound field, but synthesizes the parameters of the sound field that humans use for the localization of the source and the perception of timbre, pitch and source extension. It is sufficient to only generate the parameters with the same level of accuracy as can be achieved by the human ear. In this way, the required number of loudspeakers and the computing effort can be reduced enormously.

In WFS, primary sound, early reflections and reverberation are usually processed separately. With a synthesis based solely on the impulse response, the method still reaches the limits of the available computing power for moving sound sources. Therefore, combinations of model-based methods for the direct wavefront and the early high-sound reflections with impulse response-based generation of the reverberation, which is less important for the localization of the sound source, are used. This enormously reduces the computing power required.

In the past few years, various WFS systems have been installed, especially in the public sector. For the Bregenz Festival , the Mörbisch Lake Festival , in the KinderMedienzentrum in Erfurt and at various universities, the rows of loudspeakers generate a sound field in which the sound source is incomparably more stable than with conventional sound systems. The WFS sound system IOSONO was developed by the Fraunhofer Institute for Digital Media Technology IDMT in the vicinity of the Technical University of Ilmenau for the research area and is sold under the name IOSONO by the company Barco . Internationally, wave field synthesis is currently the subject of research at several universities. The rapid development in the field of digital signal processors will make the method a possible alternative to conventional, channel-based methods, even in the home, especially since these, too, are confronted with increasing acceptance problems as the number of channels increases.

The TU Berlin and the WDR broadcast a concert from Cologne Cathedral on July 29, 2008 live in WFS in their lecture hall 104. The auditorium with 650 seats is equipped with the world's largest sound system to date based on the principle of wave field synthesis. For this purpose, a band consisting of a total of 2700 loudspeakers runs around the lecture hall at about the ear level of the audience; Added to this are front sound systems and ceiling speakers for height information. The loudspeakers are controlled via digital data lines with 832 independent signals, which in turn are generated by a cluster of 16 computers.

literature

  • Jens Blauert: Spatial hearing . S. Hirzel Verlag, Stuttgart 1974. ISBN 3-7776-0250-7
  • Jens Blauert: Spatial Hearing, 2nd Postscript - New Results and Trends since 1982 to 1997 . S. Hirzel Verlag, Stuttgart 1997. ISBN 3-7776-0738-X
  • Glen M. Ballou: Handbook for Sound Engineers, Third Edition . Elsevier / Focal Press, 2002. ISBN 0-240-80758-8
  • Jens Ahrens: Analytic Methods of Sound Field Synthesis, Springer Verlag, Berlin, Heidelberg 2012. ISBN 978-3-642-25743-8

Web links

Commons : Wave field synthesis  - collection of images, videos and audio files

Individual evidence

  1. ^ AJ Berkhout: A holographic approach to acoustic control. In: Journal of the Audio Engineering Society. Vol. 36, No. 12, 1988, pp. 977-995.
  2. ↑ Jens Blauert: Spatial hearing . S. Hirzel Verlag, Stuttgart 1974. ISBN 3-7776-0250-7
  3. ↑ Boone, Marinus M .: Multi-Actuator Panels (MAPs) as Loudspeaker Arrays for Wave Field Synthesis . In: Journal of the Audio Engineering Society . tape 52 , no. 7/8 , July 15, 2004 ( aes.org [accessed May 16, 2017]).
  4. ↑ van Dorp Schuitman, Jasper, de Vries, Diemer: Wave Field Synthesis using Multi-Actuator Panel: Further Steps to Optimal Performance . June 1, 2006 ( aes.org [accessed May 16, 2017]).
  5. ^ Ziemer, Tim: A Psychoacoustic Approach to Wave Field Synthesis . July 22, 2011 ( aes.org [accessed May 16, 2017]).
  6. ^ Ziemer, Tim, Bader, Rolf: Implementing the Radiation Characteristics of Musical Instruments in a Psychoacoustic Sound Field Synthesis System . October 23, 2015 ( aes.org [accessed May 16, 2017]).
  7. ^ William Francis Wolcott IV: Wave Field Synthesis with Real-time Control . Project Report, University of California, Santa Barbara 2007.
  8. ↑ http://www.pressestelle.tu-berlin.de/medieninformationen/2008/juli_nr_167-199/medieninformation_nr_1882008/
  9. ↑ Subject area audio communication: wave field synthesis. Retrieved May 10, 2020 .