EP0156334B1 - Method and device for simulating (electronic artificial head) the free-field transmission characteristics of the ear - Google Patents

Abstract

An electroacoustic simulation method and apparatus for imitating electroacoustically the human outer ear having transmission properties corresponding to those of the human outer ear when exposed to sound in the free field is proposed, wherein the physical acoustical causes of the outer ear, the head, the upper part of the body, the pinna rim, and the like, are represented mathematically by simple partial models and approximated in the form of electric circuits containing circuit elements such as high-pass filters, low-pass filters, all-pass filters . . . and the like, with the possibility to vary certain properties of the circuit elements continuously by varying the parameters so that arbitrary directions of sound incidence can be adjusted infinitely in the horizontal and median planes.

Description

The invention is based on an electroacoustic simulation method according to the preamble of claim 1 and a corresponding device according to the preamble of claim 2. In a known device of this type (US-A-3 970 787) an attempt is made to electronic or electrical in the manner of a directional mixer To provide switching means that represent a suitable mean outer ear transfer function for assumed sound incidence directions.

However, as with conventional directional consoles, the actual mechanisms of formation of external ear transmission properties are not discussed, for example because they are unknown or a determination is simply inaccessible due to their complexity, but the transfer functions are simply measured empirically with a microphone and then electrical circuits and their determinations are determined Parameter.

Such a method then also makes it necessary to create a transfer function by measurement for each desired transfer direction, and it is additionally proposed to switch the characteristics of the individual elements, in particular filters, used for the simulation circuits by appropriately selecting the filter sizes, so that, for example, at Rotation of the head of a test subject by a predetermined twist angle of, for example, 10 ° by appropriate changes in the filter properties of the filters involved can then be worked with the transfer function measured for this angular direction. However, intermediate values cannot be set, despite the considerable outlay on electrical circuits, since, for understandable reasons, only a finite number of measured transfer functions can be used.

• 1. bislang kein Mittelungsverfahren für Außenohrübertragungsfunktionen existiert, welches mit Sicherheit zu mittleren Übertragungseigenschaften führt, die dem Nachrichtenempfänger "menschliches Gehör" tatsächlich adäquat sind;
• 2. daß eben nur die Einstellung einer endlichen Anzahl unterschiedlicher Schalleinfallsrichtungen möglich ist; und daß
• 3. der sich aus einer solchen Arbeit ergebende Aufwand proportional mit der Anzahl der einstellbaren Schalleinfallsrichtungen, die man wünscht, ansteigt.

It should also be added that the known device from US Pat. No. 3,970,787 does not reproduce the human outer ear in the case of free field sound systems Corresponding transmission properties is used, but the goal is a room-acoustic reproduction of hearing events via headphones, which means that the measured transmission functions also contain additional reflection effects that go back to the room acoustics. In this respect, the publication mentioned does not deal with the phenomena, reflection events, hearing events and other influences that actually occur in the case of acoustic transmission, which arise, for example, with an artificial head, but rather determines results by simple empirical measurement, so that even with any number placed outer ear transmission functions a switchover of such a "directional mixer" is nevertheless only possible for discrete directions, with the remaining problems that

• 1. So far there is no averaging method for outer ear transmission functions which certainly leads to average transmission properties which are actually adequate for the message recipient "human hearing";
• 2. that only a finite number of different directions of sound incidence is possible; and that
• 3. The effort resulting from such work increases proportionally with the number of adjustable sound incidence directions that one desires.

Another possibility for the electronic simulation of a so-called outer ear simulator for simulation Average outer ear transmission functions in the time domain would consist, for example, in that shock responses measured on test subjects are appropriately averaged and stored for all directions of sound incidence, which initially requires an extremely high storage space requirement, depending on the desired screening. In such a case, the output signal would be the so-called convolution of the input signal with the two shock responses valid for the respective direction of sound incidence (for the left and right ear). However, such signal processing in real time is practically impossible, since at least the signal processors currently available are only able to carry out such processing with considerable effort. For the same reason, the possibility of performing the so-called Fourier transformation of the input signal with subsequent multiplication of the corresponding transfer functions and reverse transformation is also ruled out.

• 1. die Hörereignisse nur innerhalb des von der Lautsprecheraufstellung vorgegebenen Raumwinkelbereichs liegen,
• 2. die Höhe des Hörereignisses häufig unterhalb der Verbindungslinie zwischen den Lautsprechern wahrgenommen wird und abhängig von der Position des Zuhörers zu den Lautsprechern ist, und daß
• 3. bei Kopfhörerwiedergabe in der Regel das Hörereignis im oder am Kopf des Zuhörers auftritt, da dem Gehör ungewohnte bzw. unnatürliche Ohrsignale angeboten werden.

Conventional mixer systems can use the so-called panorama controls to distribute individual microphone signals to the two channels of a stereo transmission in such a way that when played over two loudspeakers in a typical stereo configuration, there is a spatial distribution of listening events between the two loudspeakers (total localization). This method has the disadvantages that

• 1. the hearing events only lie within the solid angle range specified by the loudspeaker set-up,
• 2. the level of the hearing event is often perceived below the connecting line between the speakers and depends on the position of the listener to the speakers, and that
• 3. When listening to headphones, the hearing event usually occurs in or on the head of the listener, since unusual or unnatural ear signals are offered to the hearing.

Because of these considerable problems and the great effort involved when the ear signals are to be reproduced for many directions of sound incidence, there has hitherto been no practical form of implementation of a so-called electronic artificial head or outer ear simulator.

The invention is therefore based on the object to enter new territory here and to create an outer ear simulator as a so-called electronic artificial head, which is capable of a stepless directional adjustment without great effort, taking into account the human outer ear transmission properties, by amount and phase for all frequencies deliver, and which works in spite of the simplified structure with particularly high accuracy, that is, carries out the simulation so that the transmission properties of the electroacoustic equivalent circuit of the outer ear correspond practically identically to those of the human outer ear in free-field sound.

Advantages of the invention

The invention solves this problem with the characterizing features of claim 1 and the characterizing features of claim 2 and has the advantage that corresponding ear signals can be generated with little effort in any direction of sound incidence in the free field, which can serve as headphone reproduction. This enables the realization of a natural sound image.

It is also advantageous that, although, as is known to the person skilled in the art, outer ear transmission functions in particular have particularly complicated structures, the circuit parts necessary for the practical implementation of the outer ear simulator according to the invention are limited to runtimes, simple filters, high and low passes and possibly resonance systems. The parameters required for setting the circuits during operation, such as the size of a transit time or the cut-off frequency of a low-pass filter, can be determined directly from physically predetermined geometric dimensions using a model for the analytical description of the outer ear transmission properties. For the various directions of incidence of the sound, only a few filter parameters of the model of the electronic artificial head represented by an electronic circuit change in addition to the transit times. It is therefore possible to simulate the transfer function for a sound incidence direction by determining only a few parameters, it also being possible to simulate average outer ear transfer functions to program the values of the averaged geometric parameters so that, in a further, practical simplification, the corresponding control parameters are calculated directly by an electronic logic circuit, microcomputer or microprocessor and transferred to the controllable circuit blocks, which simulate the individual elements of the electronic artificial head. This makes it possible to realize arbitrarily fine subdivisions of the angular range in the horizontal as well as in the median plane without requiring a large amount of storage space, so that the corresponding ear signals can be generated effectively for each direction of sound incidence, also and in particular in the vertical, in the free field.

The electronic artificial head according to the invention generates ear signals familiar to the ear, so that on the one hand the in-head localization is avoided and on the other hand the setting of any hearing event directions is possible. This opens up completely new perspectives not only for artificial head technology, but also for multi-track recording technology. In artificial head technology, it is possible with the electronic artificial head to spatially correctly mix signals from support microphones with the head-related recording. In multi-track recording technology, the use of the electronic artificial head enables the signals from individual sound sources to be recoded so that the entire area available to human hearing can be used to generate hearing events when the head is played back. The direction of the hearing event can also be changed during a recording, so that movements of a sound source can be simulated. The loudspeaker compatibility of the electronic artificial head is comparable to artificial head recording systems, since both systems have or approximate free field transmission properties that are equivalent to one another.

Further improvements and refinements of the invention are the subject of the subclaims and are set out in these. It is particularly advantageous here that it is also possible to change the permanently stored geometric parameters, as a result of which other outer ear transmission functions can then be simulated. The outer ear simulator according to the invention can also be coupled to an external computer via an interface, so that the individual transmission functions of test subjects or the effect of hearing aids (HDO devices, in-the-ear devices), anomaly of the ear cup or change in the eardrum impedance are electrically simulated can be.

drawing

Fig. 1

ein stark schematisiertes Blockschaltbild des erfindungsgemäßen Außenohrsimulators, aus dem auch die Unterteilung in richtungsabhängige Schaltungselemente und richtungsunabhängige Schaltungselemente hervorgeht,

die Fig. 2a und 2b

den Verlauf einer entsprechend vorliegender Erfindung simulierten Freifeld-Außenohrübertragungsfunktion (I) im Vergleich zu einer gemessenen (II) über der Frequenz und die Kurvenverläufe (1), (2), ( 3 ) die Simulation einzelner akustisch wirksamer Parameter, nämlich des Ohrkanals (Kurve 1), der Schulter- und des Ohrmuschelrandes (Kurve 2) und des cavum conchae (Kurve 3);

Fig.3

ein detaillierteres Ausführungsbeispiel, noch immer in vereinfachter Form und lediglich der richtungsabhängigen Elemente für einen Kanal (einkanaliges Blockschaltbild des Außenohrsimulators - richtungsbestimmender Teil),

Fig.4

das Prinzipschaltbild einer praxisgerechten Ausführungsform unter Steuerung der Parameter der jeweiligen Schaltungselemente durch ein Mikroprozessorsystem und unter Beachtung gespeicherter gemittelter geometrischer Kenngrößen,

Fig.5

in größerem Detail eine mögliche Ausführungsform eines spannungsgesteuerten Tiefpaß-/Hochpaß-Filters, wie er zur Realisierung des erfindungsgemäßen elektronischen Kunstkopfs Verwendung finden kann,

Fig.6

das Blockschaltbild einer Interface-Schaltung zur Erzeugung von Steuerspannungen für die einzelnen Schaltungsblöcke zur Veränderung von deren Parameter, unter der Führung eines Mikroprozessors und

die Fig. 7 und 8

jeweils für zwei verschiedene Schalleinfallswinkel in Form von Diagrammen den Verlauf von Freifeld-Außenohrübertragungsfunktionen (hier des linken Ohres) einer individuellen Versuchsperson im Freifeld (Horizontalebene), wobei der dick durchgezogene Kurvenverlauf die Rechnung anhand des durch die Erfindung realisierten Modells und die gestrichelten Funktionen die Standardabweichung einer vorgegebenen Anzahl von Messungen an derselben Versuchsperson darstellen.

An embodiment of the invention is shown in the drawing and is explained in more detail in the following description. Show it:

Fig. 1

2 shows a highly schematic block diagram of the outer ear simulator according to the invention, from which the division into direction-dependent circuit elements and direction-independent circuit elements also results,

2a and 2b

the course of a free-field outer ear transmission function (I) simulated in accordance with the present invention in comparison to a measured (II) over the frequency and the waveforms (1), (2), (3) the simulation of individual acoustically effective parameters, namely the ear canal (curve 1), the shoulder and auricle rim (curve 2) and the cavum conchae (curve 3);

Fig. 3

a more detailed embodiment, still in a simplified form and only the direction-dependent elements for a channel (single-channel block diagram of the outer ear simulator - direction-determining part),

Fig. 4

the basic circuit diagram of a practical embodiment under control of the parameters of the respective circuit elements by a microprocessor system and taking into account stored averaged geometric parameters,

Fig. 5

in greater detail a possible embodiment of a voltage-controlled low-pass / high-pass filter, as can be used to implement the electronic artificial head according to the invention,

Fig. 6

the block diagram of an interface circuit for generating control voltages for the individual circuit blocks for changing their parameters, under the guidance of a microprocessor and

7 and 8

For two different sound incidence angles in the form of diagrams, the course of free-field outer ear transmission functions (here of the left ear) of an individual test subject in the free field (horizontal plane), with the thick curve showing the calculation using the The model implemented by the invention and the functions shown in dashed lines represent the standard deviation of a predetermined number of measurements on the same test subject.

Description of the embodiments

The basic idea of the present invention is to divide the physical causes of the outer ear transmission properties by differentiation and tracing back to predetermined, then simplified acoustic elements, for example the upper body, shoulder, head, auricle with cavum conchae cavity, ear canal and eardrum; All of these bodies have different influences on the outer ear transmission properties, depending on the frequency, depending on their geometric dimensions, the resulting transfer function of the outer ear then being composed of the complex superimpositions of the resonances, reflections and diffraction waves caused by all partial bodies.

Direction-dependent features are essentially determined by the elements, upper body, shoulder and auricular rim. The basic calculation of such dependencies is possible (using the KIRCHHOFF diffraction integral, derived from the GREEN theorem), but it is unsuitable for a clear representation, but it is used to describe the mean outer ear transmission function and its simulation in a model, as by aimed at the invention is needed.

It is therefore important that the invention separate itself from the calculation of complicated and complex diffraction integrals and diffraction and reflection at one Bodies using means of system theory describe what is only possible with the technical implementation of an electronic outer ear simulator, but then with comparatively simple means.

The invention therefore does not approach the solution of the problem empirically, but begins with the consideration and basis of the mathematically effectively determined, complex diffraction and reflection relationships and the resulting transfer functions, which are converted into a (simplified) model by analytical consideration, which can be represented by electrical circuits, whereby, for example, the auricle or head is then represented by the first mathematical consideration of the superposition of several diffraction bodies in the form of certain circuits, with a complex addition of the respective reflected and diffracted sound components simulated by the electrical circuit blocks for the overall outer ear transmission function the corresponding parts or areas of the body. A layer at different levels is taken into account by an additional term (superposition principle).

Without the description being overwhelmed by complicated mathematical relationships, which is also not necessary to understand the invention, the following example shows what this means. For example, to indicate the transfer function of the cavum conchae - this includes basic resonances and, like the influence of the ear canal and the eardrum impedance in this case, is independent of the direction - the cavum conchae can be understood as a system consisting of several nested openings are using the following transfer function:

This function includes so-called Rayleigh-Struve functions with K₁ and so-called Bessel functions with J₁, with l _n also information corresponding to the length and r _n the radius corresponding information of n openings, with k = wavenumber ( Ω / c).

wobei V_n der Verstärkung, Q_n der Güte und f_on der Resonanzfrequenz der Öffnung n entspricht und die Parameter des Resonanzsystems - Resonanzfrequenz, Güte, Verstärkung - in einem funktionalen Zusammenhang zu den geometrischen Abmessungen - Radius und Tiefe - der Ohrmuschelöffnungen stehen.With a good approximation, such a transfer function is approximated by a transit time in connection with a resonance system, as follows:

where V _n corresponds to the amplification, Q _n the quality and f _on the resonance frequency of the opening n and the parameters of the resonance system - resonance frequency, quality, amplification - are functionally related to the geometric dimensions - radius and depth - of the auricle openings.

The invention is therefore based on the knowledge that the external, acoustically effective geometry of a human being has a mathematically at least good approximation to the measured outer ear transfer function. Starting from average geometric dimensions, it is therefore possible in this way to determine an average outer ear transfer function for each direction of sound incidence without additional effort, which in a suitable manner is the one required for human hearing Transmission properties, since all features required for the signal analysis and pattern recognition processes in the ear are taken into account due to the physical connection of the outer ear and its transmission properties. The realization is then possible by the assumption that these mathematically describable physical causes of the outer ear transmission properties can be approximated with a model that is based on systems known from telecommunications (high and low passes, delay elements, resonance elements and the like). This model allows the outer ear transfer function to be approximated for all sound incidence directions directly on the basis of physical parameters by varying a few parameters.

The system outer ear with its direction-dependent transmission properties describes in the communications technology sense the frequency-dependent distortions that the sound signals experience depending on the direction of sound incidence during the recoding into ear signals for the message receiver "human hearing".

In the following, an example, namely the head, which can be approximated by an ellipsoid, will be used to demonstrate how diffraction and reflection can be approximated using a model.

Reference is first made here to the block diagram shown in FIG. 3 of the direction-determining part of an outer ear simulator 10 - only one channel - the approximation by circuit blocks for the head area at 10a, for the auricle area at 10b and for the area divided by dash-dotted lines The shoulder and upper body area is shown at 10c.

Based on the assumption that diffraction and reflection on a circular disk can be approximated very well, for example, with a simple model consisting of two delay elements and a low-pass filter (if one starts out from simple basic geometric figures), this is used to approximate the head model assumed that the sound pressure curve at a point on an ellipsoid as a head approximation depending on the direction of sound incidence can be approximated by the circuit blocks of a partial model indicated at 10a, the parameters of this model being functionally related to the geometric dimensions of the body under consideration; they also take into account special cases such as spherical heads, ellipses and circular disks. In this case, in all of the circuit blocks shown in FIG. 3, there is a coefficient element under K, a time element with T and any index, a low-pass filter with TP and a high-pass filter with HP. The physical reference of the model representative of the head at 10a is as follows. For the directly incident sound wave, depending on the sound angle, the sound field reflected by the ellipsoid (head) is added by the blocks K₁ and HP₁. The diffraction field emanating from the border is divided into two parts. The branch with the elements K₃, HP₃ and T₂ represents the part of the diffraction wave facing the sound source location, the branch with the elements K₂, HP₂, T₁ and TP the part facing away from the sound source location. The cut-off frequencies of the high and low passes, the amounts of the transit time and the coefficients are determined directly from the parameters head size, sound incidence direction and position of the ear canal entrance. As for the head, the influence of the diffraction body shoulder (with the elements K _S, HP _S, P _S and _S TP), upper body (K _O, HP _{O, O} T and TP _O) and can be table Geometric data of 6 male test subjects (m = mean, σ ≙ standard deviation) No. parameter m σ 1* Shoulder width 〈Mm〉 496 28 2 * Shoulder depth 〈Mm〉 269 29 3 * Shoulder slope 〈 ^O 〉 23.3 2.9 4 * BP over shoulder 〈Mm〉 160 11 5 * BP from above 〈Mm〉 156 11 6 * BP from below 〈Mm〉 105 5 7 * BP from the front 〈Mm〉 116 6 8th* BP from behind 〈Mm〉 102 5 9 * BP angle 〈 ^O 〉 11.9 4.6 10 * Head width 〈Mm〉 177 15 11 * Head height 〈Mm〉 261 10th 12 * Head depth 〈Mm〉 218 6 13 * Head radius above 〈Mm〉 86 9 14 * Middle over BP 〈Mm〉 70 11 15 * Head radius below 〈Mm〉 66 13 16 * Middle under BP 〈Mm〉 25th 7 17 * Head radius on the side 〈Mm〉 109 7 18 * Middle over BP 〈Mm〉 41 13 19 * Middle next to BP 〈Mm〉 14 10th 20 * Neck width 〈Mm〉 104 8th 21 * Neck depth 〈Mm〉 117 10th 22 * Neck angle 〈 ^O 〉 35.9 3.1 23 * Chin in front of BP 〈Mm〉 94 5 24 * Auricle height 〈Mm〉 70 6 25 * Auricle width 〈Mm〉 35 3rd 26 * Ear cup inclination 〈 ^O 〉 12.4 5.3 27 * Middle over BP 〈Mm〉 13 2nd 28 * Middle next to BP 〈Mm〉 5 1 29 * Cavum conchae height 〈Mm〉 30th 2nd 30 * Cavum conchae width 〈Mm〉 21st 1 31 * Cavum conchae depth 〈Mm〉 19th 2nd 32 * Middle over BP 〈Mm〉 4th 1 33 * Middle next to BP 〈Mm〉 8th 1 34 * Head radius above 〈Mm〉 81 11

Describe the auricle (with the elements K _a , T _a and TP _a as well as K _z , T _z ) with the model shown in FIG. 3 to the extent that it is sufficiently precise for the direction-dependent elements.

The table given on the previous page relates to geometric data of six test persons determined by measurement, this data can be used as geometric mean values when referring to the parameters of the individual approximation elements as shown in FIG. 3, and the calculation of the circuit elements. The bet of the averaged geometric parameters, as will be explained further below, can be permanently programmed to emulate mean outer ear transmission functions.

${\text{T}}_{\text{z}}{\text{(r}}_{\text{a}}{\text{,r}}_{\text{b}}{\text{,ϑ) = T}}_{\text{o}}{\text{K}}_{\text{e}}$

${\text{K}}_{\text{z}}{\text{(r}}_{\text{a}}{\text{,r}}_{\text{b}}\text{,ϑ) = 0.5 K (1 + |sin(ϑ)|)}$

${\text{K}}_{\text{a}}{\text{(r}}_{\text{a}}{\text{,r}}_{\text{b}}{\text{,ϑ) = 1 - K}}_{\text{z}}{\text{(r}}_{\text{a}}{\text{,r}}_{\text{b}}\text{,ϑ)}$

mit

ϑ:

Schalleinfallswinkel

r_a:

Hälfte des Parameters Nr. 24*

r_b:

Hälfte des Parameters Nr. 25*

Weiteres Beispiel Kopf:

${\text{K₂ = 0.5}}^{\text{.}}\text{K·[(1 - sin(ϑ)).(r' - R)/r' + + 0.5·(1 - cos(ϑ))·R/(R + r')]}$

$\text{K₃ = 0.5 K'·[(1 + sin(ϑ))·(r' - R)/r'+ + 0.5·(1 - cos(ϑ))·R/(R + r')]}$

${\text{ϑ'' = arccos(r'/r}}_{\text{a}}\text{cos(ϑ))}$

mit

c

Schallgeschwindigkeit in Luft

f_g

Grenzfrequenz des jeweiligen Hoch- bzw. Tiefpasses.

In the following, only as an example and not to be understood as restricting the invention, there are also mathematical relationships which show how the parameters used in the model here of FIG. 3 (coefficient, transit time, high-pass, low-pass) in a mathematical that is now simplified to this extent There is a connection to the geometric dimensions (parameters *) of the table on page 12.
Auricle:

${\text{T}}_{\text{a}}{\text{(r}}_{\text{a}}{\text{, r}}_{\text{b}}{\text{, ϑ) = T}}_{\text{O}}{\text{· (2-K}}_{\text{e}}{\text{), T}}_{\text{O}}{\text{= r}}_{\text{a}}\text{/ c}$

${\text{T}}_{\text{e.g.}}{\text{(r}}_{\text{a}}{\text{, r}}_{\text{b}}{\text{, ϑ) = T}}_{\text{O}}{\text{K}}_{\text{e}}$

${\text{K}}_{\text{e.g.}}{\text{(r}}_{\text{a}}{\text{, r}}_{\text{b}}\text{, ϑ) = 0.5 K (1 + | sin (ϑ) |)}$

${\text{K}}_{\text{a}}{\text{(r}}_{\text{a}}{\text{, r}}_{\text{b}}{\text{, ϑ) = 1 - K}}_{\text{e.g.}}{\text{(r}}_{\text{a}}{\text{, r}}_{\text{b}}\text{, ϑ)}$

With

ϑ:

Sound incidence angle

r _a :

Half of parameter No. 24 *

r _b :

Half of parameter No. 25 *

Another example head:

${\text{K₂ = 0.5}}^{\text{.}}\text{K · [(1 - sin (ϑ)). (R '- R) / r' + + 0.5 · (1 - cos (ϑ)) · R / (R + r ')]}$

$\text{K₃ = 0.5 K '· [(1 + sin (ϑ)) · (r' - R) / r '+ + 0.5 · (1 - cos (ϑ)) · R / (R + r')]}$

${\text{ϑ '' = arccos (r '/ r}}_{\text{a}}\text{cos (ϑ))}$

With

c

Speed of sound in air

f _g

Limit frequency of the respective high or low pass.

For a complete description of the structure-defining features, the influence of the direction-independent elements of the ear canal and carvum conchae cavity must be determined. The resonance property of this cavity can be approximated very well by bandpass systems in the form of series resonant circuits. The parameters (resonance frequency, quality and amplification) are also functionally related to the geometric dimensions of the cavity. The ear canal can be understood as a tube with a complex firing impedance, the eardrum impedance. To a good approximation, this system is described by a model consisting of transit time, high pass and coefficient.

These considerations then lead to the highly simplified block diagram of an outer ear model shown in FIG. 1, only the left channel being indicated; The individual blocks of this model, in accordance with the refinement according to FIG. 3 already explained above, stand for the corresponding acoustic elements, which are also shown in FIG. 1 and which can be found in principle in all people, and which thus represent the super-individual structures of the Set outer ear transmission functions. It has proven to be useful to divide the model shown in FIG. 1 into a direction-dependent part 12, which serves to simulate the directional characteristic of the outer ear, and a direction-independent part 13, which simulates the free-field outer ear transmission function. When playing back via a free-field-equalized headphone, ear signals can be generated via the outputs for the free-field simulation, which correspond to the ear signals of a "middle" test person for the set sound incidence directions. The second output, designated 14b in FIG. 1, -14a is the free-field equalized output - is used to emulate the free-field outer ear transmission functions. The complete, schematically simplified model of FIG. 1 is limited in the necessary circuit parts to run times, simple filters, all-passports and adders - although the outer ear transmission functions, as explained further above, sometimes have extremely complicated structures -, the parameters of the circuit parts and - Blocks, such as the size of a transit time or cut-off frequency of a low-pass filter or the like, can be determined directly with the aid of a model for the analytical description of the outer ear transmission properties from physically predetermined geometric dimensions, namely from the table given above, with the further, very significant conclusion that by Variation or change of such or in any case predetermined parameters of the circuit blocks here, and as can be understood immediately, completely continuously opens up the possibility of the corresponding ear signals for each direction of sound incidence in the horizontal and median plane to generate, so that with such an electronic artificial head, a system is available which, for any sound incidence directions, corresponds to a free field sound system Headphone playback generated and the realization of a particularly natural, impressive sound. An improvement in transparency is also guaranteed for loudspeaker reproduction analogous to artificial head technology. There are therefore not only special areas of application in psychoacoustics that will be discussed, but this also opens up new possibilities for the artistic design of a recording, especially in recording studio technology.

In the model shown in FIG. 1, which comprises coefficients, low-pass, high-pass, all-pass, bandpass, adder, resonance elements and the like only for one channel combined by circuit blocks, only the runtimes for the different sound incidence directions only change few filter parameters. It is therefore also possible to simulate the transfer function for a sound incidence direction by determining these only a few parameters. The individual circuit blocks are designated in FIG. 1 for the head region with 10a ', for the ear cup and border with 10b' and for the shoulder and upper body with 10c '; an addition element which effects the additive superimposition of the respective complex partial transmission functions bears the reference number 15. The circuit block of the direction-independent part comprises the areas of the ear canal and cavum conchae and is designated by 16.

A further advantageous embodiment in the present invention is that all the runtimes occurring in the outer ear model are combined in a basic runtime circuit block 17 which is upstream of the circuit blocks 10a ', 10b' and 10c 'and which represents and implements the required signal delays and runtimes.

In this context, since the technical realization of the transit times using analog delay lines can lead to problems, for example, because in addition to an insufficient useful / interference power ratio, mixed products of the frequencies could occur in the audible frequency range, the present invention draws a digital realization of the high-quality construction Runtimes into consideration, which in principle take place in such a way that all the runtime elements assigned to the respective sub-models or circuit element chains are arranged as shown in FIG. 1, that is to say they are drawn in front of the individual other circuits, which makes it possible to use only one analog / Get by digital implementation. In particular, a 16-bit AD / converter is used for the basic runtime block 17, which operates at a sampling rate of 44 kHz, for example, which is sufficiently high. The quantized samples are read into a shift register after the conversion. The delay time is then determined by the time difference between writing and reading out different memory locations, which is controlled by a microprocessor, which is to be explained below and effects central control of the individual elements. Due to the short memory access times, it is possible to read out all memory locations required for runtime simulation (8 runtimes per channel - there are left and right channels) during a sampling period. With a fast DA / converter, the delayed samples can be output again in time-division multiplexing. Based on this concept, only one or two DA / converters (one converter for each channel) are then necessary. The filters and coefficients necessary for the simulation are then preferably implemented with the help of controllable operational amplifiers, which further will be explained below. A digital filter implementation - for example with fast signal processors - is also within the scope of the invention, but it is advisable, at least for the time being, in particular for cost and effort reasons.

It has already been mentioned above that, as the illustration in FIG. 1 shows, the electronic artificial head (outer ear simulator) forming the invention is preferably under a central control, which considerably simplifies practical handling; For this purpose, a microprocessor 18 is provided, in which, for example, the values of the averaged geometric parameters can also be permanently programmed, which are required for emulating mean outer ear transmission functions. In a corresponding manner, the corresponding control parameters can then be calculated by the processor 18 and transferred directly to the controllable circuit blocks. With this method, arbitrarily fine subdivisions of the angular range in the horizontal and median planes can be realized without large storage space requirements, so that the corresponding ear signals can be generated for each sound incidence direction in the free field. It is then also possible to change the permanently stored geometric parameters, which means that other outer ear transmission functions can also be simulated. Furthermore, it is possible to couple the outer ear simulator according to the invention to an external computer via an interface; this possibility is denoted by 19 in the more detailed representation of FIG. 4, where, assigned to the microprocessor 18 ', the keypad of an external computer, for example a personal computer, is shown.

The almost amazing simulation capability of the outer ear simulator according to the invention can be seen in the two diagrams of FIGS. 2a and 2b, FIG. 2a showing a free field outer ear transmission function (I) simulated according to the invention - here without upper body simulation - compared to an effective one measured, i.e. empirically determined transfer function, as shown in (II). In addition to this, FIG. 2b shows the simulation of individual acoustically effective parameters, to be referred to as free-field partial outer ear transmission functions, namely for the area of the ear canal at (1), the area of the shoulder and pinna border at (2) and the cavum conchae at (3); The outer ear transfer function (I) of Figure 2a is then composed of these two partial courses.

As already mentioned above, the individual circuit elements of FIG. 3 represent the head, auricle border and shoulder / upper body areas of the circuit blocks of FIG. 1 in more detail, they connect to the basic runtime block 17 realized with digital elements and each contain individual, not yet mentioned addition elements 15a, 15b, 15c, 15d with the end adder 15 'with the output connection 20 leading to the direction-independent elements. the high and low passes and the adders summarizing their output signals.

The detailed exemplary embodiment in FIG. 4 represents the basic circuit diagram of a possible form of implementation of an outer ear simulator according to the present invention, with a block 21 containing operating and input elements and display elements, assigned to the microprocessor system 18 ', to which a central time control 22 is also assigned or contained in it. Via the multiple connecting lines 23a, 23b, the microprocessor influences the parameters of, for example, eight analog circuit channels 24 present here, which operate with their outputs on the summation element 15 ″. Depending on the type and structure of the model, the analog circuit channels 24 contain first and third order low and high pass filters 24a, 24b, bandpasses 24c and so-called coefficient elements 24d with a gain of - 1 ... + 1. The delay elements for the respective channels are realized as digital delay lines and arranged for this purpose so that only an AD / conversion to a digitization block 26 connected downstream of an input low-pass filter 25 is required. After the implementation, the quantized samples are read into a freely addressable memory 27 (delay memory RAM). The delay times that result between reading in and reading out the samples at different storage locations determine the time difference, the length of the register being determined by the maximum necessary delay time. Since a memory access is very short compared to the already mentioned sampling rate of preferably 44 kHz, all the sampling values necessary for the simulation of the different transit times can be read out in succession during a sampling period. It is therefore possible with the aid of a correspondingly fast DA / converter 26 to convert the sample values obtained in this way for the different transit times in time-division multiplex operation, for which purpose a time-division multiplex switch 29 controlled by the central time sequence control is shown schematically after a signal recovery block 28 is indicated, the outputs of which are connected to the inputs of the various channels 24. The combination of analog and digital circuit parts ensures problem-free implementation on the one hand and an extremely versatile, high-quality simulation system for the representation of outer ear transmission functions on the other. In FIG. 4 the right channel area, for example, is also designated 30a, and an associated left channel area is designated 30b; A low-pass filter 31a, 31b is also connected downstream of the adder elements 15 ', the right ear signal at the output 32a of the low-pass filter 31a and the left ear signal at the output 32b of the low-pass filter 31b.

Insgesamt ergibt sich durch eine solche Schaltung der Fig. 5 ein spannungsgesteuertes Tiefpaß/Hochpaß-Filterelement.A possible embodiment of a circuit element which can be designed as a low-pass filter or high-pass filter of the 1st order is shown in FIG. 5; the filter is constructed with the aid of a so-called "Operational Transconductance Amplifier - OTA" 33, which is connected as a controllable resistor, in which the forward steepness (transconductance) is the reciprocal of the gain and can be set with the aid of an externally fed direct current I _St. Depending on the feed-in point of this DC input signal, the transfer function of either a low pass or a high pass results for the overall arrangement. A normal operational amplifier 34 is still connected downstream of the OTA 33; the control current results from the lower circuit part, the control voltage U _{St being} fed to an operational amplifier 35 and reaching the OTA 33 via an FET transistor 36 at the output; only a capacitor C on the feedback branch and the resistors R3 and R4 in the input circuit for the inverting connection are essential to a feedback line 37. A limit frequency proportional to the control current I _St then results in the case of such a circuit, for example from the following formula

Overall, such a circuit of FIG. 5 results in a voltage-controlled low-pass / high-pass filter element.

The circuit in FIG. 6 is a block diagram of an interface circuit for generating the control voltages U _St , which can be taken off at the output 38 of the circuit and are required for the parameter setting of the filters and coefficient elements. The microprocessor 18 '(FIG. 4) writes the parameter data word via a data bus line 39 into a data register 40. At its outputs, the data word is converted into a voltage by a digital / analog converter 41 with a downstream current-voltage converter 42 for example 0 ... -10 V implemented. Via an address register 43 which is driven by the same data bus line 39, a channel of an analog multiplexer 44 connected downstream of the current / voltage converter 42 is addressed and thus the voltage generated is switched through to a corresponding output memory circuit (sample + hold), each for Such an S + H circuit 45 is provided for each filter element to be controlled. The sample + hold circuit 45 consists only of a storage capacitor C and a very high-resistance voltage follower 46 as an operational amplifier. If the capacitor C is charged, the channel is switched off again with an inhibit signal which is output by the address register 43. The whole process runs in the the same way cyclically for all other channels. In this way, the voltages across the holding capacitors C are refreshed in each case. An existing decoding logic 47 generates the loading pulses for the two registers 43 and 40 with the aid of address bus input lines 48 and control bus input lines 49 from the microprocessor system 18 '.

7 and 8 finally show in the form of diagrams for two different directions (0 ^o and 270 ^o ) the free field outer ear transmission functions of the left ear of an individual test subject in the free field (horizontal plane), the solid line being calculated by calculation according to the subject of the present Invention (model) and the two curve lines surrounding this solid line with dashed lines at the top and bottom represent the standard deviation in six measurements on the same test subject; one recognizes how highly precisely the aim of an electronic artificial head could be achieved by the present invention.

• 1. in der psychoakustischen Forschung zur leichten Realisation von speziellen Außenohrübertragungseigenschaften, z.B. zur Simulation einer Hörgeräteversorgung. Da die Möglichkeit besteht, einzelne Parameter in kürzester Zeit einfach zu ändern, kann beispielsweise der Einfluß von Hörgeräten oder die Auswirkungen von ein- und beidohriger Hörgeräteversorgung einfach simuliert werden;
• 2. im medizinisch-diagnostischen Bereich zur Überprüfung des Richtungshörvermögens oder der Sprachverständlichkeit in störschallerfüllter Umgebung. So sind Hörtests, insbesondere Richtungshörtests mit dem erfindungsgemäßen Gerät unter Freifeldbedingungen möglich ohne die Notwendigkeit, einen reflexionsarmen Raum vorzusehen, und ohne großen apparativen Aufwand;
• 3. im tontechnischen Bereich zur synthetischen Erzeugung einer kopfbezogenen Aufnahme, wobei es möglich ist, Signale für beliebige Schalleinfallsrichtungen etwa Kunstkopfaufnahmen oder dgl. zuzumischen.

Particularly suitable areas of application for the electronic artificial head according to the invention, which carries out the transformation of sound signals and ear signals instead of the natural outer ear, are, inter alia, in the following three areas:

• 1. in psychoacoustic research for the easy realization of special outer ear transmission properties, for example for the simulation of a hearing aid supply. Since it is possible to easily change individual parameters in the shortest possible time, the influence of hearing aids or the effects of single- and double-ear hearing aid supply can be easily simulated;
• 2. in the medical-diagnostic area for checking directional hearing or speech intelligibility in a noise-filled environment. Hearing tests, in particular directional hearing tests, are possible with the device according to the invention under free field conditions without the need to provide an anechoic chamber and without great expenditure on equipment;
• 3. in the field of sound technology for the synthetic generation of a head-related recording, it being possible to mix in signals for any sound incidence directions, for example artificial head recordings or the like.

Claims (12)

1. An electro-acoustic simulation method for simulating transmission characteristics corresponding to the human external ear in free-field acoustic irradiation, wherein, for example, by connecting high pass filters (24b), low pass filters (24a) all-pass filters, bass pass filters, resonance systems (24c), adders (15''), time delays (24d) in series and in parallel, electrical circuits are put into effect which simulate by approximation respective complex external ear transmission partial functions according to amount and phase, taking as the basis a model of each element comprising at least the elements of head and auricle, and wherein parameters of the electrical circuits are passed from a computer (18') to these in order to control the same, characterised in that the parameters are calculated by the computer from physically predetermined, geometric dimensions of the human external ear for any sound incidence directions.
2. A device for electro-acoustically simulating transmission characteristics corresponding to the human external ear in free-field acoustic irradiation, which device has electrical circuits such as, for example, high pass filters (24b), low pass filters (24a), all-pass filters, bass pass filters, resonance systems (24c), adders (15''), time delays (24d), connected in series and in parallel, which electrical circuits simulate by approximation respective complex external ear transmission partial functions according to amount and phase, taking as the basis a model of each element comprising at least the elements of head and auricle, which device further has a computer (18') which passes the parameters of the electrical circuits to these in order to control the same, characterised in that the computer has means which calculate the parameters from physically predetermined, geometric dimensions of the human external ear for any sound incidence directions.
3. A device according to claim 2, characterised in that the electrical circuits are divided into a directional part (12) and a non-directional part (13) connected thereto, wherein the time delays are extracted from the circuit elements of the directional part and are assigned to a basic time delay block (17) connected in front of them.
4. A device according to claim 2 or 3, characterised in that the electrical circuits for the head region (10a'), the region (10b') bordering the auricle, and the shoulder and upper body region (10c'), each of which contain a preset number of individual circuit elements (high pass filters 24b, low pass filters 24a, band pass filters 24c, coefficient elements 24d), depending on the model in a preset number according to the type and sequence in parallel and series connection, are interconnected by their outputs via adding elements (15a, 15b, 15c, 15d) and are connected to a final adding element (15') to which the ear canal and the circuit part region (16) comprising the cavum conchae is connected.
5. A device according to one of claims 2 to 4, characterised in that the parameters of the individual circuit elements (high pass filters 24b, low pass filters 24a, band pass filters 24c, coefficient elements 24d, time delays) are determined from physically predetermined, averaged geometric dimensions, and in that at least certain circuit elements include parameter-adjustment devices for the continuous generation of ear signals corresponding to each sound incidence direction in the free field.
6. A device according to one of claims 2 to 5, characterised in that digital delay lines are provided to represent the delay times (basic time delays).
7. A device according to claim 6, characterised in that in order to form the digital time delay only one analogue/digital converter (26) is provided from which the quantized scanning values of the analogue input signal are read into a shift register (27), such that the time difference arising between the reading in and reading out of the scanning values at different storage locations determines the respective delay times.
8. A device according to claim 6 or 7, characterised in that the shift register is a freely addressable delay memory (RAM 27).
9. A device according to one or more of claims 2 to 8, characterised in that in order to adjust the particular control parameters of the individual circuit elements of the respective circuit blocks, there is provided a central control and memory circuit (microprocessor system 18') to which the sound incidence directions desired at the time may be supplied from outside to determine the control parameters.
10. A device according to one of claims 2 to 9, characterised in that there is assigned to the microprocessor system (18') a central timing period control (22) which determines the digitization timing of the supplied analogue input signal, the determination of the time difference to produce the time delay at the delay memory and the signal recovery, and which triggers a signal multiplexer (29) connected in series such that to the particular channels (24) of the analogue circuit part connected in series which contain the circuit elements are supplied the time delays respectively assigned to them.
11. A device according to one of claims 2 to 10, characterised in that there is provided a voltage-controlled low pass-high pass filter with an adjustable cut-off frequency (f_g), to which a control voltage may be supplied in order to change its transmission characteristics.
12. A device according to one of claims 2 to 11, characterised in that there is provided a device controlled by the microprocessor (18') for generating the control voltages to be supplied to the particular circuit elements for the purpose of parameter determination, comprising an address- and a data register (43, 40), a digital-/analogue converter (41) connected in series and a multiplexer (44) controlled by the address register (43), which supplies the particular output control voltage to sample-and-hold circuits (45) assigned to each electrical circuit element.

Images