OBTAINING A HEAD-RELATED TRANSFER FUNCTION

Abstract

 A method and system for determining a response of a microphone (8) to a sound emitted by a sound source (2) at different azimuth angles, wherein the microphone (8) is part of a dummy (1), wherein the dummy (1) is at least partially simulating a human head including at least an auditory canal having an entrance (12), wherein the method comprises at least the following steps: recording a first response at a first azimuth angle; pivoting the dummy (1) about a pivot axis (10) for changing the azimuth angle; recording a second response at a second azimuth angle; wherein a radial distance between the entrance (12) of the auditory canal and the pivot axis (10) is less than half of an interaural distance of the dummy (1), wherein the diaphragm (15) of the microphone (8) is arranged along the auditory canal.

Description

[0001] The invention concerns a method for determining a response of a microphone to a sound emitted by a sound source at different azimuth angles, wherein the microphone is part of a dummy, wherein the dummy is at least partially simulating a human head including at least an auditory canal having an entrance, wherein the method comprises at least the following steps: recording a first response at a first azimuth angle; pivoting the dummy about a pivot axis for changing the azimuth angle; and recording a second response at a second azimuth angle. Correspondingly, the invention also concerns a system comprising a sound source and a dummy, wherein the dummy is at least partially simulating a human head including at least an auditory canal having an entrance, wherein the dummy is equipped with a microphone, wherein the microphone is arranged for determining a response to a sound emitted by the sound source, wherein the system comprises a drive for pivoting the dummy about a pivot axis at least from a first azimuth angle to a second azimuth angle.

[0002] More specifically, the invention can be implemented as a robot-assisted measurement method for obtaining a head-related transfer function (HRTF). An HRTF describes the direction-dependent transmission between a sound source and the auditory inputs of a person. The brain constructs a corresponding directional impression from the transfer properties.

[0003] Different possibilities for measuring an HRTF are known.

[0004] EP 0156333 A2 discloses an artificial head measuring system with shoulder, head and ear simulations and with microphones on both sides in the replicated auditory canal.

[0005] US 2014/0198918 A1 concerns a 3D sound processing application which measures head related transfer functions (HRTFs) in communication with a simulator apparatus that simulates a human's upper body. The simulator apparatus is automatically rotated via the turntable for varying the azimuths and positions of the simulator apparatus for enabling the microphone to record a head related impulse response (HRIR). The 3D sound processing application truncates the computed HRIRs using a filter and applies a Fourier transform on the truncated HRIR to generate final head related transfer functions (HRTFs).

[0006] Lindau et al in their article "FABIAN-An instrument for software-based measurement of binaural room impulse responses in multiple degrees of freedom." (24. Tonmeistertagung; 2006) discuss a head and torso simulator whose orientation can be controlled in multiple degrees of freedom via a servo-motorized neck joint, while the whole torso can be rotated about a central axis on a motorized turntable device.

[0007] Bernschütz in his article "A spherical far field HRIR/HRTF compilation of the Neumann KU 100." (Proceedings of the 40th Italian (AIA) annual conference on acoustics and the 39th German annual conference on acoustics (DAGA) conference on acoustics. Berlin: German Acoustical Society (DEGA), 2013.) discusses measurements carried out with a Neumann KU 100 dummy head on a VariSphear motion system. Also here, the physical origin is in the center of the head. The distance between the employed speaker system and the center of the head in this case was approximately 3.25 meters, which is assumed to be far field relative to the speaker dimensions.

[0008] It is an object of the present invention to reduce the space requirements and provide a more compact setup without compromising the measurement accuracy.

[0009] The present invention provides that a radial distance between the entrance of the auditory canal and the pivot axis is less than half of an interaural distance of the dummy, wherein the diaphragm of the microphone is arranged along the auditory canal. In this disclosure, the radial distance is determined as the minimum distance between the pivot axis and the entrance of the auditory canal of the ear to be measured. The entrance is at the distal end of the auditory canal. The exact position of the entrance of the auditory canal can be defined as the bottom of the cavum conchae at the beginning of the ear canal entrance (see also Lindau et al). In order to realistically simulate and determine an HRTF, the dummy comprises a model of the auditory canal (i.e., the auditory canal with the entrance are modelled by the dummy) and the microphone is arranged to capture sound within the auditory canal similar to the human ear. The pivot axis does not coincide and is not parallel to an interaural axis. The interaural distance of the dummy is the interaural distance that a complete head would have that is most closely resembled by the dummy; i.e., the dummy does not need to have two ears, it may optionally be a half-head. More precisely, the interaural distance is determined as the minimum distance between the centroid of the outermost cross-section of the external auditory canal of the left ear and the centroid of the outermost cross-section of the external auditory canal of the right ear.

[0010] By moving the pivot axis out of the center of the interaural axis and closer to the entrance of the auditory canal, the lateral movement (in other words, the gyration or turning radius) of the entrance of the auditory canal is reduced and the sound source can be moved closer to the pivot axis without loss of the far field character of the measurement. In the far field, there is a negligible change in the sound pressure level (SPL) due to changes in the distance between the entrance of the auditory canal and the sound source caused by the rotation of the head. Rather, the changes in sound pressure level arise from the shadow cast by the head. Reducing the distance between the sound source and the measurement position makes the measurement more sensitive to distance changes, resulting in a dependence of the sound pressure level on distance changes caused by head rotation. By reducing the lateral movement of the entrance of the auditory canal in dependence of head rotation, changes in SPL are mainly caused by shadowing effects. Due to the reduced distance between the sound source (for example and usually a loudspeaker) and the pivot axis, the entire setup becomes smaller. The angle enclosed between a line from the sound source to the entrance of the auditory canal and a line from the sound source to the pivot axis gets smaller the closer the pivot axis is to the entrance of the auditory canal. In this way, the incident angle of the sound arriving from the sound source at the position of the entrance of the auditory canal becomes less sensitive to the distance between the entrance of the auditory canal and the sound source. Additionally, the closer the sound source is to the entrance of the auditory canal, the larger the time shift between the arrival of the first wavefront from the sound source at the microphone and the arrival of the first expected reflections. This facilitates filtering reflections by time windowing the measurement signal, i.e., filtering out signal components arriving after a predefined time delay.

[0011] Optionally, the pivot axis may intersect the interaural axis of the dummy. In other words, the distance between the microphone and the pivot axis is on the interaural axis. This ensures symmetry of the distance between the sound source and the microphone relative to the pivot axis and consequently allows to capture a HRTF for both ears with only a single microphone. After the HRTF of one ear is captured, the HRTF of the other ear can be obtained by mirroring the captured HRTF at the sagittal plane of the head or by mirroring the second ear along the sagittal plane of the head, measuring the HRTFs of the mirrored ear and mirroring the captured HRTFs along the sagittal plane.

[0012] For example, the pivot axis may intersect the entrance of the auditory canal. In this case, the distance between the pivot axis and the entrance of the auditory canal is minimal. This setup achieves the least (essentially zero) lateral movement of the entrance of the auditory canal for different pivot angles and thus allows the smallest distance between the sound source and the pivot axis. Generally, it is assumed that the sound source is configured to emit sound toward the pivot axis; more specifically, the central axis of emission of the sound source intersects the pivot axis. Because the entrance of the auditory canal in this example stays on the central axis of emission of the sound source, far field characteristics are achieved even at smallest distances, minimizing the space requirements of the setup as dictated by the relative arrangement of the dummy and the sound source. Moreover, the modelling of the HRTF as a linear time invariant system (LTI system) is more accurate because the measurement position is constant for variable incident angles, resulting in a more accurate calibration. The sound arriving in the simulated auditory canal, that is, anywhere in the auditory canal and specifically in the entrance, will mostly enter through the canal entrance. Even if the microphone may be placed inside the auditory canal, the advantages discussed above regarding the incident angle and far field characteristics also apply in this case, because the microphone receives sounds mostly or only through the auditory canal and its entrance and the position of the microphone within the auditory canal can thus be varied without compromising these advantages. Moreover, the sound source can be arranged to emit sound directly at the auditory canal entrance irrespective of the pivot angle of the dummy.

[0013] In one embodiment, the diaphragm of the microphone may be arranged in the entrance of the auditory canal. In practice, as the HRTF can be used to reproduce sound close to the ear from a recording of the far field, it is practical to place the microphone in the same or a similar spot relative to the dummy head where head phones or ear phones used for reproduction will be placed. A suitable position can be in the entrance of the simulated auditory canal (i.e., of the dummy). There are also other reasons why it can be practical to choose the entrance as the placement position of the diaphragm of the microphone. For example, the auditory canal itself has specific resonances, which are not direction-dependent and therefore do not add any useful direction related information. These resonances are added on top of the HRTF, regardless of the direction of the sound. If one wants to compensate for these resonances, any HRTF can be distorted with the same filter. When using over-ear headphones, these resonances are naturally present and thus do not need to be added artificially. By effectively blocking the entrance with the microphone, these resonances can be suppressed. On the other hand, there are also cases where placing the microphone inside the canal is a more suitable choice. Placing the microphone inside the canal may bring advantages, e.g., for measuring in-ear headphones or hearing aids. One would need to place the microphone inside the canal because the device (headphone or hearing aid) sits in the entrance. To compare the effect of a hearing aid, one could measure the HRTFs without the hearing aid but with the microphone inside the canal. Next, one could repeat the measurements with the device in place and compare the results.

[0014] The dummy may be mounted eccentrically. Specifically, the dummy may be placed eccentrically on a turn table or it may be mounted on an excentric, e.g., an outrigger. The center of the dummy is arranged at a distance from the center and pivot axis of the turn table. The distance can be used to attenuate sound reflected at the turn table. Additionally, the turn table may be covered by a dummy shoulder in order to further reduce undesired reflections.

[0015] Optionally, the drive of the present system may be coupled to the dummy eccentrically. The drive causes a pivoting of the dummy around an eccentric pivot axis, that is, a pivot axis displaced from a central vertical axis of the dummy. The central vertical axis may be arranged in a plane of mirror symmetry of the dummy and may intersect the centroid of the dummy. Under the influence of the drive, the centroid of the dummy pivots around the eccentric pivot axis.

[0016] The method may optionally comprise tilting the dummy about a tilt axis, wherein the tilt axis is orthogonal to the pivot axis. The tilt axis may intersect the pivot axis or the two may be skew axes (i.e., that do not intersect and are not parallel).

[0017] In one embodiment, the tilt axis intersects the position of the entrance of the auditory canal. As for the pivot axis, this arrangement has the advantage, that the measurement position and in particular the distance between the entrance of the auditory canal and the sound source does not change under different tilt angles, which improves the accuracy (and predictive power) of the LTI system.

[0018] Referring now to the drawings, wherein the figures are for purposes of illustrating the present invention and not for purposes of limiting the same,

Fig. 1 schematically a measurement setup with a dummy and a sound source according to the arrangement used in the prior art;

Fig. 2 schematically a measurement setup with a dummy and a sound source according to the arrangement proposed with the present disclosure;

Fig. 3a-c schematically a dummy eccentrically mounted on a drive for pivoting the dummy about a pivot axis;

Fig. 4 schematically a dummy according to Fig. 3 with a mount for a microphone arranged in the entrance of the auditory canal; and

Fig. 5a-b schematically a tilting mechanism for a dummy according to Fig. 3 at different tilt angles.

[0019] Figures 1 and 2 schematically show a top view of two different measurement setups for determining a response of a microphone arranged in a dummy head 1 to a sound emitted by a sound source 2. In the first setup shown in fig. 1, the sound source 2 is arranged on a central axis 3 of the dummy. The central axes of the sound source and of the dummy are identical. The ears 4 of the dummy 1 define the interaural axis 5 of the dummy 1, which is orthogonal to the central axis 3. When the dummy 1 is arranged to look directly at the sound source 2, the incident angle of sound at each of the ears 4 is less than 90° by the angle β between the central axis 3 and the line 6 connecting the sound source 2 and the respective ear 4. In contrast, in the second setup shown in fig. 2, the sound source 2 is arranged such that the line 6 connecting the sound source 2 and a first one of the ears 4 is orthogonal to the interaural axis 5 of the dummy 1. Hence, the incident angle of sound at this ear is 90°.

[0020] The measurement setup in fig. 2 is a simplified schematic representation of a system 7 comprising a sound source 2 and a dummy 1. The dummy 1 is simulating a human head. The dummy 1 is equipped with a microphone 8 as shown in more detail in fig. 4. The microphone 8 is arranged for determining a response to a sound emitted by the sound source 2.

[0021] The system 7 comprises a drive 9 for pivoting the dummy 1 about a pivot axis 10. Fig. 3a-c show the dummy 1 mounted on an excentric 11 coupled to the drive 9, such that the drive 9 is coupled to the dummy 1 eccentrically. The pivot axis 10 intersects an interaural axis 5 of the dummy 1 at the entrance 12 of the auditory canal, which is also the position of the diaphragm of the microphone 8. The drive 9 is configured to pivot the dummy 1 to three different azimuth angles according to each of figures 3a-c.

[0022] Fig. 4 shows that the dummy 1 is essentially hollow and encloses an inner volume 13. Inside the dummy 1, a microphone mount 14 is arranged holding the microphone 8 such that the diaphragm 15 of the microphone 8 is positioned in the entrance 12 of the auditory canal of the simulated ear 4 of the dummy. The dummy 1 may be a 3D printed simulation of a human head. The ear 4 may be a separately 3D printed object, which is replaceable such that the same dummy 1 can be used with different ear shapes for corresponding measurements.

[0023] In addition to the pivoting of the dummy 1 as illustrated by figures 3a-c, the system 7 may further comprise a tilting mechanism 16 as disclosed in figures 5a-b. The tilting mechanism 16 allows to tilt the dummy 1 together with the drive 9 and the eccentric 11 about a tilt axis 17. The tilt axis 17 is orthogonal to the pivot axis 10 and also intersects the position of the diaphragm 15 of the microphone 8 at the entrance 12 of the auditory canal. Consequently, the system 7 shown in fig. 5 can be used to measure the response of the microphone 8 to a sound emitted from a nearby sound source, such as a loudspeaker, at different azimuth angles relative to the pivot axis 10 and at different altitude angles α relative to the tilt axis 17. The tilting mechanism 16 can be implemented as a parallelogram guide as shown in figures 5a and 5b. The difference between Fig. 5a and Fig. 5b is the different tilt angle α1 and α2, respectively, to schematically illustrate the functioning of the tilting mechanism 16.

[0024] Using a system 7 as disclosed in Fig. 2-5, a head related transfer function can be sampled over a wide range of solid angles at a fixed measurement position defined by the position of the entrance of the auditory canal. To collect the corresponding measurements, for each sampling position including the respective azimuth and elevation, the sound source is controlled to emit a reference sound signal and the response of the microphone is recorded. Then, the dummy is pivoted and tilted to the next sampling position and the measurement is repeated.

Claims

1. A method for determining a response of a microphone (8) to a sound emitted by a sound source (2) at different azimuth angles, wherein the microphone (8) is part of a dummy (1), wherein the dummy (1) is at least partially simulating a human head including at least an auditory canal having an entrance (12), wherein the method comprises at least the following steps:

recording a first response at a first azimuth angle;

pivoting the dummy (1) about a pivot axis (10) for changing the azimuth angle;

recording a second response at a second azimuth angle;

characterized in that a radial distance between the entrance (12) of the auditory canal and the pivot axis (10) is less than half of an interaural distance of the dummy (1), wherein the diaphragm (15) of the microphone (8) is arranged along the auditory canal.

2. The method according to claim 1, characterized in that the pivot axis (10) intersects an interaural axis (5) of the dummy (1) .

3. The method according to claim 1 or 2, characterized in that the pivot axis (10) intersects the entrance (12) of the auditory canal.

4. The method according to any one of claims 1 to 3, characterized in that the diaphragm (15) of the microphone (8) is arranged in the entrance (12) of the auditory canal.

5. The method according to any one of claims 1 to 4, characterized in that the dummy (1) is mounted eccentrically.

6. The method according to any one of claims 1 to 5, characterized in that the method comprises tilting the dummy (1) about a tilt axis (17), wherein the tilt axis (17) is orthogonal to the pivot axis (10).

7. The method according to claim 6, characterized in that the tilt axis (17) intersects the entrance (12) of the auditory canal.

8. A system (7) comprising a sound source (2) and a dummy (1), wherein the dummy (1) is at least partially simulating a human head including at least an auditory canal having an entrance (12), wherein the dummy (1) is equipped with a microphone (8), wherein the microphone (8) is arranged for determining a response to a sound emitted by the sound source (2), wherein the system (7) comprises a drive (9) for pivoting the dummy (1) about a pivot axis (10) at least from a first azimuth angle to a second azimuth angle, characterized in that a radial distance between the entrance (12) of the auditory canal and the pivot axis (10) is less than half of an interaural distance of the dummy (1), wherein the diaphragm (15) of the microphone (8) is arranged along the auditory canal.

9. The system (7) according to claim 8, characterized in that the pivot axis (10) intersects an interaural axis (5) of the dummy (1).

10. The system (7) according to claim 8 or 9, characterized in that the pivot axis (10) intersects the beginning of the entrance (12) of the auditory canal.

11. The system (7) according to any one of claims 8 to 10, characterized in that the diaphragm (15) of the microphone (8) is arranged in the entrance (12) of the auditory canal.

12. The system (7) according to any one of claims 8 to 11, characterized in that the drive (9) is coupled to the dummy (1) eccentrically.

Drawing