DE10019984A1 - Three=dimensional virtual acoustic reality calculation method for simulation of acoustical events in signal processing system by placing sound sources and receivers in virtual space - Google Patents

Abstract

The method involves using a VAR-3D technique, especially sound ray calculations for simulating or representing sound wave propagation and acoustics. A virtual space is created in an information processing system, and geometric points arranged on a global coordinate system. Surfaces are defined on the geometric points. Sound source objects and sound receiving objects (virtual microphones) are placed in the virtual space. Sequential sound rays are generated for the samples (pulses) resulting from the signal to be processed, and the propagation of the rays is calculated. The propagation calculation makes use of the technique mentioned in a specification of VAR3D, to calculate the acoustic conditions. The generated signal is obtained from the interface of the sound receivers. Independent claims are included for the use of a frequency selection and band allocation, the holding and organisation of data within a reality information object (RIO), and the simulation of resonance or natural modes in objects and bodies.

Description

Humans absorb about 50 percent of their spatial orientation via the sense of hearing.

This happens due to several auditory psychological procedures.

The 3-dimensional location of sound sources is about the detection of phase shifts possible, which result from our head shape and the fit of our ears.

The size of the room is largely recognized by the sound reflection times that are in develop this space (Hall).

These two "procedures" largely contribute to our orientation.

The size of a room for the sense of hearing is determined by the "reverberation time". That is the Time that passes until all of the energy generated by the sound source passes through Has "discharged" absorption. During this time, the sound wave keeps coming off the walls this space reflects, and thus, in ever weakened form, reaches the sense of hearing. Since man has a "learning" sense of hearing, and only in one 3-dimensional spatial environment is the hearing of a "sonic dead" spatial sound signal extremely uncomfortable.

For this reason, systems for artificial production, of reflections.

Digital hall systems that work with mathematical algorithms have become widespread and enforced.  

The basic function of conventional Hall devices:
The signal to be processed (sound material) is first digitized.

In abstract terms, it is now a sequence of impulses of different strengths. For each pulse to be processed, a mathematical algorithm time-delayed, continuously weakening repetitions are calculated (compression algorithm). The pulse repetitions are then added to the signal to be processed.

The algorithm is chosen so that the resulting signal of the reflection characteristic, corresponds as closely as possible to a room chosen from reality.

Disadvantages of this technique of Hall generation:
Since only the reflection times are added up, no actual spatial calculation takes place. The sound source is therefore inevitably static in the center of the room. Furthermore, the sound source does not differentiate in terms of the radiation characteristics of the sound source. It is emitted, so to speak, spherically for all frequencies.

In the natural environment, however, the higher the frequency, the more it is directed emitted.

The addition of reflection pulses to the "dry" signal, in which the reflections are mimicked using a "compression equation" (mathematical algorithm), contains the following problems and weak points:

• 1. serious disadvantage: no mobility of the sound source or the sound receiver (Microphone) and the room.
• 2. No location of the sound source over the reflection times.  
• 3. Unclear and unrealistic reverberation sound.
• 4. No spatial optical reference for the user. (Doesn't see what he's hearing!)
• 5. The sound space is always an empty cube that only comes from the side walls, Floor and ceiling exist.
• 6. The sound space inevitably contains only a "wall material" which is only imperfect can be replicated.
• 7. The reflection is "logically" only on one level, so it always stays 2- Dimensional.
• 8. Natural acoustic elements such as diffraction and resonance cannot be inserted. 9. The user has no options for the creative design of the room Available.
• 9. Working with this type of Hall generation is cumbersome and not realistic, abstract.

3-dimensional virtual acoustic calculation technical description

VAR3D is a technology that digitally simulates realistically all acoustic phenomena within a virtual-realistic environment, makes possible.

The most important idea is a 3-dimensional virtual space based on one 3-dimensional simulation of acoustic laws, using the method the sound beam calculation to give acoustic properties.

The virtual space consists of geometric points in an XYZ "global" Coordinate system that is generated via an electronic information system (e.g. computer) and being represented. Areas are defined using these geometric points.

The user can create his "world" from any surface construct geometrically and visually (furnishings, rooms, houses etc.). Now there is a geometric model of surfaces and the associated points.

The user can now assign a previously created material to each surface. The material contains the information of the absorption coefficient (as well as other relevant ones Information). That means an information curve, how much energy of the sound wave at one certain frequency and a certain material thickness is withdrawn.

The next step is to insert one or more sound sources into the 3-dimensional one virtual "sound space".  

These are logical objects, represented by geometric points, or non-spherical sound sources such as surfaces, or bodies composed of surfaces,
where the body composed of surfaces is the closest to reality. For example, a loudspeaker can also be constructed in this way.

Furthermore, one or more sound collectors have to be defined (a virtual microphone) and in to insert the virtual "sound space".

This can again be a geometric point, or a non-spherical sound absorber like a surface, or a body composed of surfaces.

In contrast to the sound source, the "geometric point" must have a certain value Have a radius to be touched by sound rays at all.

Calculation: (first step, calculation of an impulse (a single sound wave))

The laws of sound propagation are very similar to those of optics regarding reflection even the same.

Sound reflection on a flat surface: "alpha" = "alpha 1" where "alpha" = more incident Beam, and "alpha 1" = reflected beam.

The incident beam and the reflected beam lie on one level.

With VAR3D technology, the point of view is the basis that a spreading Sound wave can also be represented in simplified form as a sound beam. Because there is a sound wave "Basically" spreads out spherically, the first example can be represented as follows.

Step 1 Consider an impulse

A cube-shaped space is specified that consists of the eight geometric corner points and the areas defined on the corner points, and a sound source (here a simple ball) and a sound absorber.  

All sides of the cube are provided with a uniform degree of absorption (one certain frequency would correspond).

The sound source is now (evenly distributed over the entire surface of the sphere (Spherical) at 90 ° to the "surface") a quantity of sound beams to be determined "sent out".

The rays have a uniform "amount of energy" in the form of a variable E, which the "Energy quantity" (volume) of the pulse to be calculated corresponds.

A collision calculation is used to calculate where the beam touches a surface. If such a touch has come about, then the law of reflection the further path (angle of reflection) of the beam is calculated.

With each "contact" with one of the side walls of the cube (of the room) the variable E the value of the degree of absorption is subtracted.

After a touch-dependent amount of touch, the "energy" (Variable E) equals zero, which ends the reflection calculation.

Inside the room is the virtual microphone, in the form of a sound absorber. The sound beam travels a path determined by the room (cube). This is in the Variable W saved and calculated at every point of contact (any object).

When the "catchment area" of the sound absorber is touched by an incoming sound beam the sound beam passes the information, its variable E, the time of its start at the sound source (variable S), as well as the distance covered (variable W) within one certain protocol to be defined.

From the distance covered (variable W), and the duration of the sound, which must be specified (approx. 340 M / s in normal air), the duration is calculated that the sound beam was "on the way".  

This "space" results from this duration within the time course of the one to be formed Signal.

2nd step Frequency dependent absorption (multiple impulses, sound with different Frequencies)

Absorption is a crucial acoustic property in the natural sound space. The absorption strength is frequency-dependent (which means that your sense of hearing is oriented towards you), therefore frequency selection technology is used in the VAR3D.

The signal to be processed is before the actual spatial calculation analyzed in its frequencies. Frequency bands are of a size to be defined selected and provide the respective samples (impulses) with a variable Frequency membership contains.

Example: a sound lasting one second is digitized.

According to this, the digitization rate is 48,000 Hz (48,000 numerical values per second) 48,000 different numerical values = samples. These are now viewed as impulses.

In our example, the sound is broken down into 15 different frequency bands.

Ascending according to the respective frequency band, each sample / numerical value now gets within a variable F its frequency band membership.  

When calculating the sound beam, the absorption is no longer calculated as generalized in step 1, but for each sound beam as follows:
Each sound beam contains the variable F, which indicates which frequency band it belongs to. From the material that has been assigned to the surface touched, it can now be read from the absorption coefficient curve by how much "energy" the variable E of the sound beam is reduced within the frequency range that the variable F of the sound beam indicates.

3rd step Sound diffraction

Another important acoustic property in connection with bodies is sound diffraction. Is a surface (of a body) in relation to the wavelength of the incident on it Small sound wave, the sound wave "bends" around it. If the area is large enough, the sound wave reflects.

For the lowest frequency still reflected applies: "Calculation for sound diffraction"

FU = lowest frequency still reflected
c = speed of sound
d = diameter of the reflector
alpha = angle of incidence of the sound beam
a1 = distance reflector → sound source
a2 = distance reflector sound absorber (virtual microphone)

Sound diffraction is calculated within the 3D VAR using frequency selection.

In a precalculation of the entire "geometric scene", (entire virtual Sound space), a selection is made before the actual "reverb generation".

All objects that are small enough within a relevant frequency range to be determined to lead to sound diffraction become special marked (program technical "FLAG").

In connection with the variable F of the incoming sound beam, with these specially marked objects, the "calculation for sound diffraction" made. The result of this calculation is either sound diffraction or reflection. In the case of sound diffraction, the incoming sound beam keeps its direction.

The surface of the object to be "overcome", the path the sound wave takes on it, If you bend "around", it is included in the "travel time calculation".

Step 4 Resonance and non-spherical sound sources, as well as sound propagation in others Media as air

Other sound-determining elements of acoustics are the natural resonance of bodies, and non-spherical sound sources, as well as the propagation of sound waves in others Media as air.

Non-spherical sound sources:
A quantity of sound beams to be determined is evenly distributed over the body (surface) defined as the sound source. From the point determined in this way, each sound beam is calculated at a 90 ° angle to the surface ("sent").

It is thus achieved that the sound source is not from a geometric point to locations. These non-spherical sound sources are the most common in natural space occurring type of sound sources.

With non-spherical sound sources, the material properties of the respective Object used for calculation (absorption).

Resonance:
Almost every body in our real acoustic environment has the ability to vibrate in itself due to its nature. So it can be viewed as a sound source itself.

About the geometric and physical properties that apply to each object are defined, the resonance behavior of the object (body) is determined (calculated). The resonance behavior is shown in a frequency curve that the object over a Variable R is attached (resonance frequency characteristic).

Objects that should be "equipped" with resonance become special by the user characterized.

When a sound beam comes into contact with a specially marked one object capable of resonance is determined by the variable F of the sound beam, the absorption characteristic, as well as the resonance characteristic and the excitation characteristic, the value of the variable E (energy) of the Sound beam determined.

The excitation characteristic curve reflects the energy value that the body needs to be around to be able to assume states resulting from the resonance frequency curve (To in Natural vibration).

In connection with the variable E (energy) of the sound beam, the absorption characteristic, and the energy / excitation characteristic curve results in the value "EO" (energy of the object).  

Sound propagation in media other than air.

The speeds of sound, i.e. the speeds of propagation, vary from substance to substance different in our natural environment.

This only plays a subordinate role in a simulation, since humans are the largest Partially absorbs sound waves over the air.

The sound beam moving through an object, which is a medium other than air as such is not affected in its direction.

The distance covered, which is stored in the variable W, is, however, followed by the object's given rate of propagation is calculated while it is within of the object.

If the sound beam emerges from the object, the object is treated as a sound source with the properties defined for the sound source.

With the insertion of sound propagation in media other than air, in the simulation and their description, are all relevant phenomena that affect our sense of hearing, shown.

Such a "logical system" to simulate our (human) acoustic Environment is a virtual system.

That means it represents a technology based on information processing electronic Systems such as a computer, for example.

Software can be created, one on an operating system such as Windows Microsoft executable program, or an electronic hardware environment that uses this technology to generate spatial reverberation (reverberation device).

The VAR3D technology can also be used, for example, as "Logig" on a microchip realize.  

Step 5 Data organization

Since the generation of computer-generated images with the help of ray tracing a similar one Using technology as the basis is a combination of these (and similar) "logical" Information very important for the future.

Reality Information Objects "RIO"

So-called abstract "objects" contain optical and acoustic as well as others information accessible to our senses, within a defined protocol.

Such a protocol contains the following data technology information:
RIO = the following data is a "RIO object".
Size xxxx = contains the start and end address in the memory.
Type xxxx = contains the information in which order which information is.
Datatype xxxx = Contains the information which data follows.
Data xxxx = Contains the data.

Data is organized within the "Rio Objects" that is accessible to the human senses are. That means data that are used to generate sensory stimuli, as well Data that contain direct information for one of these senses. The senses to be addressed can be: sense of sight, sense of hearing, sense of smell and sense of touch. The corresponding data and Information can be optical or acoustic, for example.

The VAR3D technology will basically work with "RIO" objects when it comes to keeping Processing and storage of information is coming.  

VAR3D technology results in the following direct improvements over the problematic old procedure of generating Hall using mathematical algorithms:

• 1. Both sound sources and sound absorbers, as well as all geometric objects, are free flexible and unlimited use (limited by the one used Information system). Movements can be freely defined by the user for each object become.
• 2. The sound sources are with appropriate use of the virtual microphones (Sound absorbers) and listening systems, locatable, both via the reflections of the room (Hall) as well as via the sound beams that hit the sound absorber directly without detour (Direct sound).
• 3. The more complex the geometric model and all the others are Simulation parameters can be designed, a realistic and natural sounding reverb. The authenticity of the hall depends only on the accuracy with which the user designed the sound space.
• 4. Depending on the technical design of the resulting products, the user can be provided with a visualization. Through this he has the opportunity to see and design the "reverberation room" directly.
  It provides the user with a direct visual reference to the "sound" of the room that is experienced via the sense of hearing.
• 5. With VAR3D, the user has completely free possibilities of a "sound space" design acoustic environment. He can focus on it in reality to virtually mimic existing things - to design, or to completely create new acoustic environments.  
• 6. The materials are also free to design and are independent of those of the manufacturer given wall materials.
• 7. The VAR3D technology provides a sound propagation that is similar to reality. In nature, sound waves generally spread three-dimensionally.
  In the old method with mathematical algorithms, only a "logical" spread can be achieved in two dimensions. A location of a sound source, or of bodies in the sound field, was not possible with this method.
• 8. Important sound-determining elements such as diffraction of sound waves and resonance were not to be realized within the old procedure.
• 9. Realistic work is possible with VAR3D technology.
  The user has optically and acoustically objects that are known from his daily sensory experience, such as materials, 3-dimensional bodies, sound sources, and virtual microphones.
• 10. Because of the similarity to ray tracing for the generation of optically detectable images, both technologies can be linked.
  This is not possible with the old method of generating sound reflections based on mathematical algorithms. The commercial applications possible through VAR3D.

Application of simulation of acoustics are conceivable in many areas, in four Main areas, however, are surrounded by high demand at the time of registration as a relevant implementation does not yet exist.

These four areas are: music production, film dubbing (especially multi-channel formats such as Dolby Digital), video game applications in the 3D area, virtual reality and thus the upcoming form of the Internet.
Possible commercial applications are not only limited to the generation of realistic reverberations, but also include a holistic simulation of acoustic phenomena.

In the area of film dubbing, for example, multichannel dubbing is for the medium DVD of great interest. Different systems are offered for realistic reverberation to be able to use in production. However, none of these systems is "non-static", or 3-dimensional. The ability to move sound sources, and more Possibilities that the VAR3D technology can implement are very much in demand. However, the "main area of application" will be the Vital Realty and thus the Internet.

Anyone who will move in virtual spaces will become the "acoustics" of that space perceive that is generated on the basis of VAR3D technology.  

Application example

An application of the VAR3D system would be conceivable as follows.
19 Hall device,
to create realistic-looking virtual acoustic environments, for a sound signal that is anechoic before processing with VAR3D.

In the classic 19 inch format, which is primarily used in recording studio environments, a metal housing is produced which is accessible from the front, the "user interface" in Contains form of controls and data screens. Still located on the front a device interface to one or more data entry devices.

The interfaces to the rest are on the back of the device Recording studio environment.

Audio signal inputs and outputs are used as interfaces in analog as well as digital form realized. Furthermore, data bus interfaces are implemented, which are used to control the device or the transfer of data can be used.

The electronic is not accessible to the user within the device Information system, in the form of a circuit board, on which electronic switching elements and Microprocessors are soldered on.

Controlled by an operating system, these microprocessors calculate the ones to be calculated Data.

The operating system coordinates the data input and output, maintenance and management within of memory. VAR3D technology is implemented within the operating system. The user has the option of displaying his geometric data via the data interface, and make minor changes.  

The device now calculates all from the anechoic signal coming from outside User defined acoustic laws and gives, in one by the user defined way, the calculated signal on its signal outputs.

The user can dynamically adjust his sound sources or sound absorbers with the help of the data input device connected to the front device interface, in virtual Move space.

Claims (4)

"VAR-3D" 3-Dimensional "Virtual Acoustic Calculation" is the simulation of acoustics and other physical processes that affect the area of acoustics on an information or signal processing system, characterized by the application of the methods described in VAR3D , especially the application of the method of (sound) radiation calculation for the simulation or representation of sound wave propagation, as well as acoustics. A virtual space is created on an information processing system by organizing geometric points via a global coordinate system. Areas are defined on these geometric points. Sound source objects and sound absorbing objects are then inserted within a virtual space to be constructed. Sound beams are now generated one after the other for the samples generated by the signal to be processed (considered as energy value → impulse) and their spread is calculated. During the propagation calculation, the procedures explained in the description of the VAR3D are used to calculate the acoustic laws. The generated signal is made available via the interface of the sound absorbers. 1. The application of the procedure of frequency selection and frequency band allocation; on impulses or sound beams to be processed, within a simulation of acoustics and other physical processes that affect the area of acoustics, - on an information or signal processing system - which mainly uses the method of radiation calculation to simulate or display Sound wave propagation and acoustics used, characterized by:
the integration of the frequency band information (e.g. VARIABLE F) in the information unit of the sound beam to be calculated, which is required for the calculation of diffraction, absorption and resonance. 2. Keeping and organizing data within a "Rio" "Reality Information Object" marked by, the simultaneous holding and storage of data and information, - The one for a simulation system, which for the representation of a data and information required for virtual reality can be used -, within a predefined format an information or signal processing system.   3. The simulation of resonance or natural vibrations of objects and bodies within a simulation of acoustics, on an information or signal processing system, characterized by

• - The use of energy value information in connection with an object - within a system for simulating acoustic laws on an information or signal processing system - as well as resonance frequency information and absorption frequency information.
• - Furthermore, the simulation of resonance is characterized in that an object within a simulation environment comes into a simulated oscillation by "energy supply" via properties defined for the object and thus emits energy again.