CN101938686B - Measurement system and measurement method for head-related transfer function in common environment - Google Patents

Abstract

The invention relates to a measurement system and a measurement method for a head-related transfer function in a common environment. The measurement system comprises a loudspeaker array consisting of two or more horizontal loudspeakers; a microphone is a double-ear microphone array and is arranged in the sound wave emitting direction of the loudspeaker array; test signals are processed by a sound signal DC conversion device and a power amplifier and then fed to the loudspeaker array and played by each loudspeaker, and the test signals of each path are irrelevant; the double-ear microphone array picks the sound signals, and the sound signals are output to a data processing terminal through a sound signal AD conversion device; and the data processing terminal calculates a cross correlation function between the signals played by each loudspeaker and the signals picked by the microphone by using a correlation analysis method, and the cross correlation function is processed to obtain the head-related transfer function (HRTF) from each loudspeaker to the left and right ears of a human head. The head-related transfer function, head type parameters and posture parameters of a listener are quickly acquired by combined processing of the loudspeaker array and the double-ear microphone array.

Description

The measuring system of a related transfer function and method of measurement in a kind of conventional environment

Technical field

The present invention relates to the acoustical signal processing technology field, more specifically, the present invention relates to measuring system and the method for measurement of a related transfer function in a kind of conventional environment, can in the normal room environment, measure fast and obtain the personalized related transfer function of attentive listener and head dummy parameter and attitude parameter.

Background technology

In recent years, related transfer function (Head-Related Transfer Function is called for short HRTF) and virtual sense of hearing technology thereof relevant auditory psychology and physiological scientific experiment, multimedia with virtual reality, artificial intelligence, communicate by letter, room acoustics design, sound programme are made, family expenses are low voice speaking scientific research, engineering, consumer electronics field such as puts that important use value is arranged.Virtual sense of hearing technology is a kind of transmission of sound wave from the sound source to ears of simulating, thereby fictionalizes the technology of spatial sound picture in earphone or loud speaker repeating transmission.From the angle of physics, the sound wave that sound source is sent through remarkable head, trunk and auricle etc. partly reflect and scattering after arrive ears, its feature can be described to the transfer function of ears fully by sound source.This related transfer function HRTF has described in the free field sound source to the acoustic transmission relation of ears.Relating to parameters such as it and sound source position, frequency and have individualized feature, comprised level difference (the Interaural Level Difference that sound wave arrives two ears, abbreviation ILD), time difference (Interaural TimeDifference, be called for short ITD) and phase difference (Interaural Phase Difference, be called for short IPD) etc. important information, in spatial hearing and the virtual sense of hearing, play a part core.

Because HRTF and physiological structure and people's area of bed are closely related, thereby are the physical quantitys with obvious individualized feature.In the application that the virtual sense of hearing is reset, adopt the personalized HRTF of attentive listener can obtain better effect, the effect that adopts other HRTF then is the similarity degree that depends on these data and the personalized HRTF of attentive listener.At present, obtain personalized HRTF the most accurately method be laboratory measurement.Over past ten years, existing a plurality of seminars have carried out measurement and database work to HRTF both at home and abroad.Though the used method of measurement of different authors, device etc. in detail can be different, basic principle of measurement and mentality of designing are similar to.

The technology of common employing at first is near experimenter's duct entrance or places measuring microphone near the eardrum, then utilizes one to be placed in fixing horizontal angle θ, elevation angle φ and to play test signal s (n) apart from the loud speaker of number of people fixed range.The transmission of sound wave from the sound source to ears can be regarded the filtering of a linear time invariant as, and HRTF is exactly the transfer function of this linear time invariant system.Therefore, the various method of measurement that are applicable to the sound system transfer function, as: impulse method, Fourier analysis method, correlation method etc. all are applicable to the measurement of HRTF.HRTF is the continuous function of direction in space, and because spatial sampling, actual measurement can only be carried out in limited direction.

At present, more existing seminars measure HRTF, have set up the HRTF database of space different azimuth.

The Media Lab of the Massachusetts Institute of Technology (Gardner W G, Martin K D.HRTF measurement of aKEMAR.J.Acoust.Soc.Am., 1995,97 (6): 3907-3908) in the anechoic chamber of MIT, dummy KEMAR (Knowles Electronics model DB-4004) is tested.KEMAR by positioned vertical on a platform that can horizontally rotate.During each the measurement, loud speaker is fixed on a suitable elevation angle, then, changes the horizontal subtended angle of KEMAR successively.

CIPIC interface experiment chamber (the Algazi V R in California, USA university Dai Weide branch school, Duda R O, ThompsonD M et al.The CIPIC HRTF Database.Proceedings of 2001 IEEE Workshop onApplication of Signal Processing to Audio and Acoustics, 2001,99-102) on 50 different vertical orientations of 25 different level orientation, each ear of 45 testers (comprising two artificial head of 43 experimenters and big-ear and little ear) is measured.In the experiment, it is 1 meter ball center position that the tester is sitting in radius, and loud speaker then is placed on the diverse location of sphere.Hearer's duct is blocked, receives signal with the microphone that is placed on the duct mouth.The measurement of more accurate physiological structure has also been carried out to each experimenter in the CIPIC laboratory.

Researcher (Kimitoshi Fukudome is arranged, Toshimitsu Suetsugu, the fast measurement of headrelated impulse response for all azimuth directions using the continuous measurementmethod with a servo-swiveled chair.Applied Acoustics 68 (2007) 864-884) also begins to pay close attention to the measurement efficient that how to improve HRTF.They have designed a chair that can rotate continuously that horizontal rotating speed is 6 °/s, utilize computer and motor device that the rotation of chair is controlled.Like this, the HRIR that measures all horizontal angle orientation at a fixing elevation angle only needs one minute, has improved testing efficiency.

In general, the measurement of HRTF at present needs complicated test environment and equipment, comprise anechoic chamber, high-quality audio frequency apparatus, head pose tracker or turntable and health fixture etc., and owing to need continuous mobile sound source position, the measurement of HRTF is work very consuming time, and setting up bigger storehouse then needs bigger manpower and materials.In addition, the method for measurement of this fixed angle is had relatively high expectations to condition and the attentive listener of experimental facilities, and long-time measurement causes the discomfort of attentive listener easily.Therefore, no matter still being that manpower and materials are expensive from experiment condition goes up, and in actual applications, so accurate personal personalized HRTF measurement is unpractical.

Summary of the invention

The objective of the invention is to, solve and obtain attentive listener head related transfer function in conventional environment how fast, and the problem of head dummy parameter and kinematic parameter, the test environment that overcomes the common survey method requires height, equipment complexity, testing time long, shortcomings such as the manpower loss is big, thereby based on the mode of loudspeaker array and ears microphone array Combined Treatment, provide measuring system and the method for measurement of a related transfer function in a kind of conventional environment.

In order to realize one of foregoing invention purpose, the invention provides the measuring system of a related transfer function in a kind of conventional environment, comprise: loud speaker, microphone, acoustical signal AD/DA conversion equipment, power amplifier, data storage terminal, data processing terminal and display device, it is characterized in that described loud speaker is the loudspeaker array that comprises that two or more loud speakers are formed; Described microphone is the ears microphone array, is arranged on the sound wave transmit direction of loudspeaker array, is worn on people's ears position;

Each drive test trial signal is at first given loudspeaker array through acoustical signal DA conversion equipment and power amplifier rear feed, plays via each loud speaker, and each drive test trial signal is uncorrelated mutually;

Described ears microphone array is used for picking up acoustical signal, outputs to data processing terminal through acoustical signal AD conversion equipment;

Described data processing terminal utilizes relevant function method to calculate the cross-correlation function of each loud speaker play signal and microphone pickup signal, obtains each loud speaker after the processing to a related transfer function HRTF of number of people left and right sides ear; And calculating each loud speaker thus to the ears time difference ITD of number of people left and right sides ear, the geometry site between recycling time difference ITD and the number of people and the loud speaker calculates the attitude data of number of people radius parameter a and each measuring position;

Described data storage terminal or display device are for HRTF and the attitude data of storing or show each measuring position.

In the technique scheme, described data storage terminal is or/and data processing terminal adopts ordinary PC or digital signal processor.

In order to realize another purpose of foregoing invention, the present invention also provides the method for measurement of a related transfer function in a kind of conventional environment, this method is based on the mode of loudspeaker array and ears microphone array Combined Treatment, the ears microphone array is arranged on the sound wave transmit direction of the loudspeaker array that two or more loud speakers form; Related transfer function in the realization normal room and the quick measurement of attentive listener head dummy and attitude parameter comprise the steps:

(1) attentive listener wears number of people head-wearing type ears microphone, be positioned on the loudspeaker array sound wave transmit direction, test signal is given loudspeaker array through acoustical signal DA conversion equipment and power amplifier rear feed, the test signal that each loud speaker is play is uncorrelated mutually, the ears microphone is picked up acoustical signal, outputs to data processing terminal through acoustical signal AD conversion equipment;

(2) data processing terminal utilizes relevant function method to calculate the cross-correlation function of each loud speaker play signal and microphone pickup signal, obtains each loud speaker after treatment to a related transfer function HRTF of number of people left and right sides ear;

(3) calculate each loud speaker thus to the ears time difference ITD of number of people left and right sides ear;

(4) geometry site between recycling ears time difference ITD and the number of people and the loud speaker calculates the attitude data of number of people radius parameter a and each measuring position thereof;

(5) HRTF and the attitude data of each measuring position are stored.

In the technique scheme, employed test signal can be sinusoidal signal, swept-frequency signal, pulse signal, random noise signal or pseudo-random noise signal in the described step (1).In the technique scheme, described pseudo-random noise signal is MLS sequence or Golay Codes sequence.The MLS sequence is that cycle period is N _PΔ t, auto-correlation function are similar to a kind of random sequence of δ function, and it is little that it has the clean disturbance of input, advantages such as amplitude, cycle, the easy control of timeticks, and its statistical property is similar to white noise.Golay Codes sequence is that another kind is usually used in the pseudo-random noise signal that HRTF measures, and its characteristic and MLS sequence of measuring HRTF is similar.But Golay Codes sequence measuring is more responsive to the small movement of experimenter.

With the signal of each the loud speaker actual play cross-correlation function as reference calculated signals and microphone pickup signal, the delay error that can avoid system equipment self to cause, reference signal can utilize software that play signal is controlled acquisition with receiving signal, and the play signal that also acoustical signal AD/DA conversion equipment can be provided also feeds back to this equipment simultaneously and obtains.

In the technique scheme, ITD can ears acoustic pressure phase delay be poor by calculating, binaural sound flattens methods such as equal time difference, correlation method, rising edge method and obtains in the described step (3).

The invention has the advantages that: utilize loudspeaker array and ears microphone array Combined Treatment with attentive listener head related transfer function and head dummy parameter and attitude parameter in the quick acquisition conventional environment, compared with prior art:

1, be different from conventional HRTF measuring technique with " pure " of environment thus serve as to pursue the measurement result that obtains under the high signal to noise ratio, the present invention is conceived to the HRTF measurement under the normal room test environment, and is more tolerant to the requirement of test environment, has more universality;

2, replace common single loud speaker with loudspeaker array and test, carry out multimetering by a plurality of loud speakers, can obtain the HRTF of all directions simultaneously, improve test speed greatly;

3, need not manual measurement, can directly utilize acoustic method to obtain crucial head sizes parameter;

4, the present invention is not limited in ear canal mouth or the duct and measures, and obtains the HRTF directly related with equipment;

5, the present invention does not need accurate number of people positioning and tracking device, and the dependence system itself utilizes the acoustic method location.

6, the personalized HRTF of test provides effective method in bigger HRTF database and the average family acoustic surrounding in order to set up under the laboratory environment in the present invention.

Description of drawings

Fig. 1 is the schematic diagram of an embodiment of measuring system of the present invention;

Fig. 2 is the flow chart of an embodiment of method of measurement of the present invention;

Fig. 3 is number of people steel ball model and incident wave schematic diagram;

The rough schematic of the number of people and two-loudspeaker position when Fig. 4 is measurement number of people radius size.

Embodiment

Below in conjunction with the drawings and specific embodiments the present invention is described in further detail.

Embodiment 1

As shown in Figure 1, measuring system of the present invention, comprised a twin loudspeaker array 101, number of people head-wearing type microphone 102, four I/O channels 103,104 and 105 of acoustical signal AD/DA conversion equipment, what the AD/DA conversion equipment adopted here is the multichannel analyzer of the pulse3560c model of BK company, and what data processing, storage and display device 106 utilized is ordinary individual's computer.Twin loudspeaker array 101 is used for playing test signal, and in the present embodiment, the test signal of the left and right loud speaker of feeding is the two-way white noise signal, and one the tunnel is the MLS sequence, and another road is the opposite sequence of this sequence, and therefore, this two paths of signals is incoherent mutually.In this example, the test signals samples frequency is 65536Hz, the sequence length of one-period is 32767 points, adopt the repetition in five cycles, this binaural signal is fed by I/O channel 103 by PC 106, the output of I/O channel 103 simultaneously feeds back to I/O channel 104 by external circuit, and the signal that receives with I/O channel 104 has so just been avoided the signal distortion error that is caused by the AD/DA equipment self as the reference signal in the subsequent treatment.Microphone 102 is picked up the acoustical signal from two loud speakers, and is converted into the digital signal PC 106 of feeding by I/O channel 105, and the signal of being picked up by 106 pairs of reference signals of PC and microphone gets related transfer function HRTF to the end as relevant treatment.

As shown in Figure 2, the concrete steps of a quick method of measurement of related transfer function HRTF of present embodiment are as follows:

Step 201:PC machine 106 reads the signal that microphone that I/O channel 105 feeds is picked up, and left and right sound channels is respectively LSig and RSig.

Step 202: owing to just begin recording before the play signal, therefore, LSig and RSig had also comprised a bit of noise before useful signal, normally found out the starting point of signal with manual methods, in the present invention, designed a kind of method of utilizing correlation peak picked up signal starting point.Because reference signal is known, when receiving signal and reference signal starting point and length and fit like a glove, the peak value of its correlation function just in time is positioned at the correlated series middle.According to this characteristic, can obtain the initial point position of useful signal among LSig and the RSig.

Step 203: reference signal is the reference signal Ref through AD/DA equipment that is fed back by I/O channel 104, and PC 106 reads reference signal Ref, and length intercepting LSig and RSig according to reference signal Ref remove the data in two cycles end to end;

Because the left and right two paths of signals of Ref is uncorrelated mutually, therefore the signal that receives of the left and right ear of microphone worn of people and the correlation function of left road or right wing reference signal are the HRTF of this direction, so can utilize correlation method to calculate the cross-correlation function of reference signal Ref and LSig and RSig respectively:

$R_{\mathrm{xy},L}\left(l\right)=\frac{1}{L}\underset{k=1}{\overset{L-\left|l\right|}{\mathrm{\Σ}}}x\left(k\right)y(k+l).$

Step 204: can produce bigger amount of calculation owing to directly calculate the estimated value of this correlation function, therefore, present embodiment adopts fast Fourier transform FFT to calculate correlation function, reducing amount of calculation, on average can obtain HRTF with what the result who obtains did three cycles.The duplicate measurements of getting a plurality of periodic signals on average can improve signal to noise ratio.

Step 205: HRTF is made the part of getting real part after the inverse Fourier transform, and length equals the length of one-period signal.This length is a lot of greatly than an effective length of coherent pulse response, contains the influence of composition and the reflected sound of many redundancies, therefore, goes intercepting with 512 Hanning windows, obtains 512 HRIR at last.

Step 206: again HRIR is done fast Fourier transform and obtain the related transfer function HRTF that left and right loud speaker branch is clipped to left and right ear.

On this basis, obtain number of people parameter and attitude parameter more in real time, concrete steps are as follows:

As shown in Figure 3, consider the effect of the number of people, head is approximately the spheroid 301 that radius is a, ears are approximately 2: 302,303 on the sphere.For the plane sound wave 304 of horizontal plane θ direction incident (can regard as the point sound source of infinity produce).Consider sound wave after the transmission of the surface of head-bent, ears time difference formula is:

When

Wherein, c is the velocity of sound, and ITD＞0 expression auris dextra is leading, and left ear falls behind; ITD＜0 item is opposite.

As shown in Figure 4, be the rough schematic of the number of people and two-loudspeaker position during test number of people radius size in the present embodiment.The a pair of loud speaker of horizontal plane (sound source) 401,402 be arranged in the same horizontal plane of attentive listener on, loudspeaker position is fixed, between the two-loudspeaker distance be d (actual measurement as can be known).With head part's spheroid 301 that to be approximately a radius be a, ears are approximated to 2: 302,303 of symmetry on the sphere.Here convenient for signal, make that left speaker 401 is respectively x to people's left ear 302, the distance of auris dextra 303 ₁₁And x ₂₁, right loud speaker 402 is respectively x to the distance of people's left and right ear ₁₂And x ₂₂, left and right loud speaker is respectively r to the distance at number of people center ₁And r ₂Because the distance of people's area of bed much smaller than the number of people to loud speaker can obtain following approximate:

$r_1\≈\frac{1}{2}(x_{11}+x_{21}).$

$r_2\≈\frac{1}{2}(x_{12}+x_{22})$

Wherein, each loud speaker can be obtained by the time difference Δ t of device plays signal and microphone reception signal to the distance of left and right sides ear, for example, and x ₁₁=c Δ t ₁₁

During measurement, make us women's head-ornaments to a certain loud speaker, for example among the figure, in the face of left speaker 401.When the perpendicular bisector of the number of people overlapped with the sound ray of left speaker 401, measurement should obtain ITD ₁=0.Feed back ITD in real time by system ₁The result, when non-vanishing, the prompting experimenter carry out suitable adjustment.At this moment, right loud speaker 402 is θ with the angle of the horizontal perpendicular bisector of the number of people.The triangle geometrical relationship that is made of number of people central point S and left and right sides loud speaker can get:

At this moment, ITD2 can get by measurement result, so obtain the calculation relational expression of number of people radius a:

Wherein, utilize correlation method and rising edge method result to be averaged to obtain ITD; The Normalized Cross Correlation Function of definition ears impulse response HRIR is:

${\mathrm{\φ}}_{\mathrm{LR}}\left(\mathrm{\τ}\right)=\frac{\underset{-\∞}{\overset{+\∞}{\∫}}{h}_L(t+\mathrm{\τ})h_R\left(t\right)\mathrm{dt}}{{\left\{\right[\underset{-\∞}{\overset{+\∞}{\∫}}{h_L}^2\left(t\right)\mathrm{dt}\left]\right[\underset{-\∞}{\overset{+\∞}{\∫}}{h_R}^2\left(t\right)\mathrm{dt}\left]\right\}}^{\frac{1}{2}}},$

Can calculate function phi according to following formula _LR(τ) exist | τ | the maximum the in≤1ms scope, Dui Ying τ=τ therewith _MaxBe the ears time difference ITD of correlation method definition _Corre, i.e. ITD _Corre(θ, φ)=τ _Max

Some when the absolute value that defines the HRIR sequence of left and right ear for the first time surpasses its maximum absolute value 10% respectively is starting point, is designated as t _LAnd t _R, then rising edge ITD can be expressed as:

ITD _lead(θ，φ)＝t _L-t _R。

In the present embodiment, ITD is two kinds of results that method is averaged:

$\mathrm{ITD}(\mathrm{\θ},\mathrm{\φ})=\frac{1}{2}({\mathrm{ITD}}_{\mathrm{lead}}(\mathrm{\θ},\mathrm{\φ})+{\mathrm{ITD}}_{\mathrm{corre}}(\mathrm{\θ},\mathrm{\φ})).$

(3) definition attentive listener head pose parameter alpha is the angle of number of people perpendicular bisector and loud speaker sound ray direction, and on the basis that obtains number of people radius size, the α angle measuring method is as follows: by

Can obtain measuring the expression formula of number of people angle [alpha] constantly:

$\mathrm{\α}=g^{-1}(\frac{c}{a}\×\mathrm{ITD}\left(\mathrm{\α}\right))$

Wherein, g ^-1It is the inverse function of g (α)=sin α+α.Utilize the Chebyshev sequence to obtain g ^-1A polynomial expansion approximate expression:

Wherein

Polynomial approximation can greatly speed computational speed, improves computational efficiency.With the ITD that calculates substitution following formula as a result, get final product this moment the number of people and the angle α of loudspeaker array middle vertical plane.

What simulate in the present embodiment is twin loudspeaker, and test signal is the situation of MLS sequence white noise.When simulation be the multi-loudspeaker array, during the situation of other test signal, the inventive method is suitable equally.As long as guarantee that each loud speaker play signal is uncorrelated mutually, still can obtain each loud speaker to HRTF and experimenter's head dummy parameter and the number of people attitude parameter of this moment of left and right ear according to the above-mentioned steps method.

It should be noted last that above embodiment is only unrestricted in order to technical scheme of the present invention to be described.Although the present invention has been described in detail with reference to embodiment, those of ordinary skill in the art is to be understood that, technical scheme of the present invention is made amendment or is equal to replacement, do not break away from the spirit and scope of technical solution of the present invention, it all should be encompassed in the middle of the claim scope of the present invention.

Claims (5)

1. the measuring system of a related transfer function in the conventional environment, comprise: loud speaker, microphone, acoustical signal AD/DA conversion equipment, power amplifier, data storage terminal, data processing terminal and display device, it is characterized in that described loud speaker is the loudspeaker array that comprises that two or more loud speakers are formed; Described microphone is the ears microphone array, is arranged on the sound wave transmit direction of loudspeaker array and is worn on people's ears position;

Each drive test trial signal is at first given loudspeaker array through acoustical signal DA conversion equipment and power amplifier rear feed, plays simultaneously via each loud speaker, and each drive test trial signal is uncorrelated mutually;

Described ears microphone array is used for picking up acoustical signal, outputs to data processing terminal through acoustical signal AD conversion equipment;

Described data processing terminal utilizes relevant function method to calculate the cross-correlation function of each loud speaker play signal and microphone pickup signal, obtains each loud speaker after the processing to a related transfer function HRTF of number of people left and right sides ear; And calculating each loud speaker thus to the ears time difference ITD of number of people left and right sides ear, the geometry site between recycling time difference ITD and the number of people and the loud speaker calculates the attitude data of number of people radius parameter a and each measuring position;

(31) head of establishing the number of people is approximately the spheroid that radius is a, and after the transmission of the surface of head-bent, ears time difference formula is for the plane sound wave of horizontal plane θ direction incident:

Wherein, c is the velocity of sound, and ITD＞0 expression auris dextra is leading, and left ear falls behind; ITD＜0 item is opposite;

(32) distance is d between the two-loudspeaker, makes that left speaker is respectively x to people's left ear, the distance of auris dextra ₁₁And x ₂₁, right loud speaker is respectively x to the distance of people's left and right ear ₁₂And x ₂₂, left and right loud speaker is respectively r to the distance at number of people center ₁And r ₂, because the distance of people's area of bed much smaller than the number of people to loud speaker obtains following approximate:

$\begin{array}{c}{r}_1\≈\frac{1}{2}(x_{11}+x_{21})\\ r_2\≈\frac{1}{2}(x_{12}+x_{22})\end{array};$

Wherein, each loud speaker is obtained by the time difference Δ t that device plays signal and microphone receive signal to the distance of left and right sides ear;

(33) make us women's head-ornaments to a certain loud speaker, when the perpendicular bisector of the number of people overlapped with the sound ray of this loud speaker, measurement should obtain ITD ₁=0; At this moment, the angle of another loud speaker and the horizontal perpendicular bisector of the number of people is θ; The triangle geometrical relationship that is made of number of people central point S and left and right sides loud speaker can get:

ITD ₂Can get by measurement result, obtain the calculation relational expression of number of people radius a thus:

$a=\frac{c\×{\mathrm{ITD}}_2}{\sin \left(\mathrm{\θ}\right)+\mathrm{\θ}};$

(34) definition attentive listener head pose parameter alpha is the angle of number of people perpendicular bisector and loud speaker sound ray direction, on the basis of acquisition number of people radius size, the α angle by

Obtain measuring the expression formula of number of people angle [alpha] constantly:

$\mathrm{\α}=g^{-1}(\frac{c}{a}\×\mathrm{ITD}\left(\mathrm{\α}\right));$

Wherein, g ^-1It is the inverse function of g (α)=sin α+α;

(35) utilize the Chebyshev sequence to obtain g ^-1A polynomial expansion approximate expression: $\mathrm{\θ}={\tilde{g}}^{-1}\left(x\right)=\frac{x}{2}+\frac{x^3}{96}+\frac{x^5}{1280};$ Wherein, $x=\frac{c}{a}\×\mathrm{ITD}\left(\mathrm{\α}\right);$

(36) with ITD substitution following formula as a result, get final product this moment the number of people and the angle α of loudspeaker array middle vertical plane;

Described data storage terminal or display device are for HRTF and the attitude data of storing or show each measuring position.

2. the measuring system of a related transfer function in the conventional environment according to claim 1 is characterized in that, described data storage terminal is or/and data processing terminal adopts ordinary PC or digital signal processor.

3. the method for measurement of a related transfer function in the conventional environment, this method is based on the mode of loudspeaker array and ears microphone array Combined Treatment, the ears microphone array is arranged on the sound wave transmit direction of the loudspeaker array that two or more loud speakers form, related transfer function in the realization normal room and the quick measurement of attentive listener head dummy and attitude parameter comprise the steps:

(1) attentive listener wears number of people head-wearing type ears microphone, test signal plays for loudspeaker array through acoustical signal DA conversion equipment and power amplifier rear feed simultaneously, the test signal that each loud speaker is play is uncorrelated mutually, the ears microphone is picked up acoustical signal, outputs to data processing terminal through acoustical signal AD conversion equipment;

(2) data processing terminal utilizes relevant function method to calculate the cross-correlation function of each loud speaker play signal and microphone pickup signal, obtains each loud speaker after treatment to a related transfer function HRTF of number of people left and right sides ear;

(3) calculate each loud speaker thus to the ears time difference ITD of number of people left and right sides ear, the geometry site between recycling ears time difference ITD and the number of people and the loud speaker calculates the attitude data of number of people radius parameter a and each measuring position thereof;

(31) head of establishing the number of people is approximately the spheroid that radius is a, and after the transmission of the surface of head-bent, ears time difference formula is for the plane sound wave of horizontal plane θ direction incident:

Wherein, c is the velocity of sound, and ITD＞0 expression auris dextra is leading, and left ear falls behind; ITD＜0 item is opposite;

(32) distance is d between the two-loudspeaker, makes that left speaker is respectively x to people's left ear, the distance of auris dextra ₁₁And x ₂₁, right loud speaker is respectively x to the distance of people's left and right ear ₁₂And x ₂₂, left and right loud speaker is respectively r to the distance at number of people center ₁And r ₂, because the distance of people's area of bed much smaller than the number of people to loud speaker obtains following approximate:

$\begin{array}{c}{r}_1\≈\frac{1}{2}(x_{11}+x_{21})\\ r_2\≈\frac{1}{2}(x_{12}+x_{22})\end{array};$

Wherein, each loud speaker is obtained by the time difference Δ t that device plays signal and microphone receive signal to the distance of left and right sides ear;

(33) make us women's head-ornaments to a certain loud speaker, when the perpendicular bisector of the number of people overlapped with the sound ray of this loud speaker, measurement should obtain ITD ₁=0; At this moment, the angle of another loud speaker and the horizontal perpendicular bisector of the number of people is θ; The triangle geometrical relationship that is made of number of people central point S and left and right sides loud speaker can get:

ITD ₂Can get by measurement result, obtain the calculation relational expression of number of people radius a thus:

$a=\frac{c\×{\mathrm{ITD}}_2}{\sin \left(\mathrm{\θ}\right)+\mathrm{\θ}};$

(34) definition attentive listener head pose parameter alpha is the angle of number of people perpendicular bisector and loud speaker sound ray direction, on the basis of acquisition number of people radius size, the α angle by

Obtain measuring the expression formula of number of people angle [alpha] constantly:

$\mathrm{\α}=g^{-1}(\frac{c}{a}\×\mathrm{ITD}\left(\mathrm{\α}\right));$

Wherein, g ^-1It is the inverse function of g (α)=sin α+α;

(35) utilize the Chebyshev sequence to obtain g ^-1A polynomial expansion approximate expression: $\mathrm{\θ}={\tilde{g}}^{-1}\left(x\right)=\frac{x}{2}+\frac{x^3}{96}+\frac{x^5}{1280};$ Wherein, $x=\frac{c}{a}\×\mathrm{ITD}\left(\mathrm{\α}\right);$

(36) with ITD substitution following formula as a result, get final product this moment the number of people and the angle α of loudspeaker array middle vertical plane;

(4) HRTF and the attitude data of each measuring position are stored.

4. the method for measurement of a related transfer function in the conventional environment according to claim 3, it is characterized in that employed test signal can be sinusoidal signal, swept-frequency signal, pulse signal, random noise signal or pseudo-random noise signal in the described step (1).

5. the method for measurement of a related transfer function in the conventional environment according to claim 3 is characterized in that, the middle ITD employing of described step (3) calculating ears acoustic pressure phase delay is poor, binaural sound flattens equal time difference, correlation method or rising edge method and obtains.

Images