GB2418332A - Array loudspeaker with steered nulls - Google Patents

Abstract

A sound system uses steerable (phased) array techniques to generate a strong sound at a first location and a weak or null sound at a second location. The sound signal is split into a plurality of frequency bands and each frequency band is directed independently such that a trough in sound pressure level exists at a common location in space. Two such sound signals may be reproduced simultaneously, the first generating a strong sound at the first location and a null at the second location while the second generates a strong sound at the second location and a null at the first location. The two signals may be the left and right channels of a stereo signal, providing excellent cross-talk cancellation. Alternatively, the two signals may be different audio programmes, directed at two different listeners at the first and second locations.

Description

. 241 8332

ARRAY LOUDSPEAKER WITH STEERED NULLS

The present invention relates to a method and apparatus for creating an energy field, especially a sound field. In particular the invention is concerned with creating a sound field that has a location where a signal can be heard and a location where it can not.

Background to the Invention

Stereo and surround-sound audio systems which use multiple, widely separated loudspeakers to deliver a real spatially extended sound field are well known in the art. Also well known in the art are "pseudo-stereo" and "pseudo-surround-sound" audio systems which attempt to deliver to each of the ears of generally just one listener, with more or less success, the distinct sounds those ears would have heard had they been immersed in a real spatially extended sound field, whilst avoiding the need to create such an extended field. Perhaps the simplest delivery method is a pair of earphones which deliver the left ear signals just to the left ear, and the right ear signals just to the right ear. Where headphones are not an option for the audience, other more complex systems have been devised and are known in the art, which use the outputs of two or more loudspeakers to approximate the required two discrete sound signals at the listener's ears. In general such systems rely on more or less complex cross-talk cancellation schemes, usually involving complex filters and sometimes recursive schemes where the crosstalk caused by the first-order crosstalk cancellation, is regarded as second order crosstalk, and then a further crosstalk cancellation signal is introduced to minimise this, inturn causing third order crosstalk and so on. Also well known in the art are phased array antennae, their ability to produce steerable and focussable beams, and their ability to produce multiple simultaneous beams each carrying possibly independent content and each independently steerable and focussable, e.g. as described in WO 01/23104 and WO 02/078388. A method and system is also disclosed in PCT/GB05/000887.

The Present Invention In the present invention, a new approach to the problem of delivering two discrete signals to the ears of a remote listener (i.e. one some distance from the sound generation apparatus) with minimal crosstalk, is addressed, preferably using steerable (phased) array techniques, and the inherent advantages of this approach are explored.

. . . . .. . . . . a ce . . . . . . . a a a- ,, Firstly, in the known phased array antennae (PAA) art, it is conventional and often unstated, that the beam formed by the PAA has a specific direction (of transmission or reception) relative to the antennae, which is generally considered to be the direction of maximum radiation emission or sensitivity, and that to first order (with minor imperfections ignored) this beam direction is not a function of frequency, but only of the phases or delays distributed amongst the multiple array transducers. The beam width almost always varies with frequency and this feature is generally undesirable but practically unavoidable, although complex schemes do exist to minimise beam-width variations with frequency, and work to a greater or lesser extent (see e.g. WO 03/034780). However, in a first aspect of this invention, a PAA is produced which generates one or more beams whose beam directions do deliberately change with frequency, in a precisely determined manner.

This beam moving with frequency would normally be undesirable. But it is also a feature of PAA beam patterns, again often unstated but simply taken for granted, that they have peaks and troughs, or maxima or minima, and in general, directions in which the beam is particularly strong (the beam "direction"), and most importantly for our purpose, directions in which the beam is particularly weak, which hereafter we refer to as Troughs. For example, completely regular 2D and 3D PAAs when phased to point in a particular direction and focussed at infinity (far field), will above a certain cut-off frequency, also have in their beam pattern, directions in which the radiation strength is zero, a true null (only in theory of course - any real antenna will have non-zero Troughs and no true nulls).

The invention provides a method of creating a sound field comprising a strong first sound signal at a first location and a weak or null first sound signal at a second location, said method comprising: determining a first direction in which to direct a sound beam of a first frequency such that the sound pressure level at said first frequency at said second location is weak or null; directing frequency components of said first sound signal at said first frequency in said first direction; determining a second direction in which to direct a sound beam of a second frequency such that the sound pressure level at said second frequency at said second location is weak or null; directing frequency components of said first sound signal at said second frequency in said second direction.

The invention thus allows different frequencies, or bands of frequencies, to be directed in . . ë.. ee c . . ..

. . . as- . e different principal directions. The directions are chosen such that the strongest "trough" exists at the second location for each frequency. This keeps crosstalk down to a minimum.

There is no limit on the number of frequencies, or frequency bands to which the invention can be applied. Preferably, directions are determined for a plurality of frequencies and frequency components of the first sound signal at or around these frequencies are directed in the so- deterrnined direction.

The invention might therefore be thought of as splitting the sound signal into frequency bands and directing components of the sound signal of any particular frequency band in a direction such that the sound pressure level at the second position is weak or null.

The linear nature of the system means that while a first sound signal can be directed so as to be heard at a first location and to be weak or null at a second location, a second sound signal can be directed so as to be heard at a second location while being weak or null and the first location. The same principles are enumerated above in respect of the first sound signal are carried out for the second sound signal, using the same apparatus.

Directing the frequency component in accordance with the null position, rather that in accordance with the maximal beam positions, can mean that frequency components arrive at the first location with relatively different amplitudes. This can be corrected by adjusting the amplitude of the component prior to transmission such that when they arrive at the first location they have the correct relative amplitudes. The same can be done in respect of a superimposed second or third sound signal if present.

The invention also provides apparatus for creating a sound field that comprises a strong first sound signal at a first location and a weak or null first sound signal at a second location, said apparatus comprising: a plurality of output transducers capable of directing more than one sound beam, each beam being different; a processor arranged to determine a first beaming direction for a first frequency of sound such that the sound pressure level at said first frequency at said second location is weak or null; said processor also being arranged to determine a second beaming direction for a second frequency of sound such that the sound pressure level at said second frequency at said second location is weak or null; wherein, in use, components of said first sound signal at said first frequency are beamed in said first . c . . . . direction and components of said first sound signal at said second frequency are beamed in said second direction.

The transducers are preferably arranged in a horizontally disposed array and the system may be embodied inside a laptop personal computer, for example.

Further aspects of the invention are detailed below.

First Aspect of the Invention For the purposes of this invention, what is important is that when the PAA beam direction is changed by adjusting the element phases (or equivalently for broadband use, their delays), the direction(s) in which the Trough(s) are located also change, but in general by a different amount. So it will be seen that a steerable beam direction also implies a steerable Trough direction too. So that in this first aspect of the invention, a PAA is produced which generates one or more beams whose beam directions deliberately change with frequency, in a precisely determined manner, and whose beam Trough directions also change with frequency in a controlled and precise manner.

In what follows we will describe only the PAA transmitting case. However it will be obvious to those skilled in the art that the processes and devices can be made to operate similarly in reception mode, by suitable adjustment of the equipment.

Second Aspect of the Invention The utility of this device is that if there are two targets A and B in the field of the PAA, and it is desired to irradiate strongly (or receive from with high sensitivity) target A, and to irradiate B (or receive from B) as little as possible, then the following procedure may be carried out.

The PAA beam direction is first set directly to target A and the required amplitude at A adjusted by varying the transmitted PAA amplitude. In general, because real PAAs do not have pure pencil beams, significant radiation will also reach B. especially if B is close to A. However, if the PAA beam is then steered away from A in the direction opposite to the direction to B. above a certain cutoff frequency, a first (potentially one of several) Troughs in the beam shape will be directed closer to and eventually directly at B and the irradiation of B will be very low. However, the position of this Trough is, as described above, a function of frequency. So it is arranged that the PAA beam direction varies with frequency (along the .e Be/ec. . ....

path direction passing through A and B) such that at each frequency or band of frequencies the Trough remains approximately pointing at B as the beam direction changes with frequency. A side effect of this beam direction change with frequency is that the amplitude of irradiation at A is now no longer fixed, but also varies with frequency, as the beam direction moves closer to and further from the direct position of A. So in a second aspect of the invention, a frequency varying beam-direction PAA as previously described in the first aspect, is also constrained to vary its amplitude with frequency, so as to reduce or even cancel completely, the variation in amplitude at target A. Note that varying the amplitude of the PAA beam does not affect the position of the beam direction, or the Trough direction and so this is an independent variable. Also, most Troughs in real PAA beams are quite deep (a depth of many 10s of dB is common) so even if the main beam has to be boosted by 3dB or more (perhaps even 10dB or more) then the level received at target B. in the Trough, can still be very much less than at target A, and often be many 10s of dB lower.

It should also be noted that both of these aspects of the invention work equally well for acoustic PAAs (e.g. sonic, ultrasonic, or infrasonic) used in reception mode (as microphones) or transmission mode (as loudspeakers), and for electromagnetic PAAs (e.g. radio, microwave, TeraHertz, optical or whatever) again as receivers or transmitters, and all these and other variants are to be understood as aspects of the invention.

Third Aspect of the Invention In a third aspect of the invention, a PAA as described in either of the first two aspects of the invention, is designed to produce two simultaneous beams S I and S2, each steered independently of each other and optionally and most usefully carrying different information in the beams, and at least one but preferably both being beams whose directions vary with frequency. In a particularly useful aspect of the invention, one of the two beams, S1 say, is directed as above so that the S1 beam is in the vicinity of A and the S1 Trough held in the vicinity of B and the beam-direction-change-with-frequency is set so that B stays in the vicinity or centre of the Trough of S I, and the converse is done with the second beam S2; i.e. the second beam S2 is directed to the vicinity of B and its corresponding S2 Trough is directed towards the vicinity of or at target A. With this arrangement, A dominantly receives radiation from S1 and B from S2, with very low crosstalk (i.e. A receives little energy from S2 and B little energy from S1). s

* .

e: I: e: : : : : :e c c c c ce. ace .

Fourth Aspect of the Invention One application for such a dual beam system as described in the third aspect of the invention is as follows. The PAA is a sonic acoustic PAA and may be mounted in a consumer electronics equipment, such as a TV set, a computer (a desktop PC or especially a portable laptop computer with no external loudspeakers). The targets A and B may be the two ears of a human listener/user of the equipment. Assuming the human user is within the beam steering range of the PAA and the user's ear positions can be targeted, then the apparatus described as aspect three is able to deliver two different sound channels one to each of the user's ears, with low crosstalk. Thus stereo reproduction is easily achieved with good fidelity and good imaging (because of the low crosstalk), and if the two channels are pre-processed to carry Head-Related-Transfer-Function (HRTF) sound information, then pseudo surround sound can be delivered to the user again with high fidelity and good imaging because of the low crosstalk. Any of the conventional HRTF schemes known in the art are applicable to this delivery system, and in particular those designed for headphone delivery would work well, but with the advantage of no headphones or other physical user contact being necessary. So a fourth aspect of the invention is a two-channel sound delivery system with low crosstalk comprising a two-beam PAA as in the third aspect of the invention, with the two beams arranged to have their most suitable Troughs located at opposite ears of a human listener within the overall beam accessible area of the PAA, and with the beam directions in the vicinities of the opposed ears of the listener (i.e. beam 1 direction in vicinity of left ear while Trough of beam 1 is in vicinity of right ear, and beam 2 direction in vicinity of right ear with Trough of beam 2 in vicinity of left ear), and the beam direction caused to change with frequency so that the Troughs of the two beams remain in the vicinity of the ears across the frequency band of interest.

Fifth Aspect of the Invention The variation of beam direction with frequency may be achieved in many ways. In a fifth aspect of the invention, the signals to be transmitted by the PAA are passed through a Fourier Transform (FT) process (in practice almost certainly implemented as a Fast Fourier Transform FFT or one of the many variants known in the art for their computational efficiency), which divides the signal into frequency bins. Given that direction of targets A & B for that beam (channel) are known, it is easy to compute the required delay at each frequency to steer the Trough of the beam towards target B (these delays are fixed for each * se..'. '

a c e ea. .ee frequency for a given Trough direction and a given PAA element (transducer)) and so may be applied for each FFT frequency bin for each transducer. The suitably delayed sets of frequency bins for each transducer are then Inverse Fourier Transformed (IF l) again using something like an FFT process, and the re-transformed time domain signal for each transducer then applied to the transducer element for transmission.

In a two beam system as per aspect 4 of the invention, a similar process (FT of each input signal, per-transducer and frequency selective delays applied, inverse FT of each group of delayed signals from each channel for each transducer then summed to make the resultant signal for each transducer) is applied to both input channels and the resulting two FTed and inverse FTed signals for each transducer summed and applied to the transducers.

Sixth Aspect of the Invention In a sixth aspect of the invention, the PAA is built with analogue or digital filters instead of delay elements on a per transducer basis, one analogue or digital filter per transducer per input channel, and where there is more than one input signal, say N input signals, the outputs of the N filters per transducer are summed, before application to the transducers (possibly via a DAC and power amplifier). The filters are designed such that they have a varying phase delay with frequency such that at each frequency, the ensemble of so-delayed signals arriving at the respective transducers steer the Trough of the beam to the requisite target A or B. and preferably, in addition, the filters have a variation of output amplitude with frequency that exactly or closely compensates the received amplitude at target A so that it remains predominantly constant with frequency despite the beam being steered away from the vicinity of A so as to keep the Trough of the beam closely in the vicinity of target B. Seventh Aspect of the Invention In a practical system, e.g. a stereo or surround sound system built into a laptop computer, the two beams and their troughs could be in fixed locations (Troughs fixed more or less, beam directions varying with frequency as described, so as to maintain the more or less fixed Trough locations) suitably placed for the ears of a user in front of and probably above the apparatus (e.g laptop PC), so that it was only necessary for the user to suitably position her head to experience the low crosstalk two channel sound. As described previously the two channels could carry stereo sound or additionally HRTF processed signals from several other sources to give a strong impression of surround sound. )e

t C ^ t Eighth Aspect of the Invention In an alternative practical system that is more flexible, the frequency-positioned beams and nominally fixed Troughs, could be steered en masse in conventional PAA manner so as to track that actual position of the user's head, making the whole system easier to use and more comfortable for the user. Such tracking could be done in several ways, but e.g. as laptop PCs already often contain miniature cameras, a simple image processing system could detect the user's head position from the camera image (assumed pointing in general direction of user and of wide field) and use this information to direct the nominal Trough locations at the user's ears, with the frequency changing beam system as described above then keeping the Troughs fixed with frequency, suitable displaced about the user's possibly changing head position.

Ninth Aspect of the Invention A further refinement to any of the previously described systems is to provide a phase- flattening filter to remove any gross variations of phase with frequency in the vicinity of the user's ears. Such a phase-adjusting filter would be required for each input signal channel and could best be applied just once at the point where the signal was input to the system, and would be designed to minimise phase variations with frequency introduced by the system.

Note that as the user's ears are instantaneously a fixed distance from the PAA even though the beam direction may instantaneously be different for different frequencies, there is no acoustic-path-length change with frequency induced phase error to correct - only phase change with frequency because of the frequency dependent beam steering corrections.

It should be noted that the schemes described above all have a natural lower cut-off frequency dependent on the linear extent of the PAA array (related to the wavelength of sound), and below this frequency the beam of the PAA has no nulls at all and thus there are no well defined Troughs to steer to the secondary target B. Tenth Aspect of the Invention Also, as frequency rises the PAA beam will in general have a multiplicity of Troughs (around the theoretical nulls) and the widths of these troughs becomes progressively less as frequency increases. It then becomes difficult to locate a null/Trough accurately enough around the target B and above another upper cut-off frequency, the scheme described above for r.

v e e.

e v c ^ . producing low-crosstalk becomes impractical. However, this ceases to be important, because the other effect of rising frequency on a PAA beam from an array of fixed length is that the directivity of the array progressively increases with frequency, one consequence of which is that the amplitude of the side lobes decreases with frequency. It then becomes adequate merely to steer the direction of the main beam towards target A, and simply through beam directivity and low side-lobe level, the amplitude at target B will generally be adequately low for low crosstalk. So in a tenth aspect of the invention, above a certain upper cutoff frequency Fu the mode of operation described above changes, and instead of steering a null towards target B. one simply steers the main beam towards target A, and thereafter relies on low sidelobe levels to produce low crosstalk at target B. If the PAA elements are spaced apart such that they are separated by more than a wavelength of sound, fullpower grating sidelobes will be produced by the PAA array and these may start to increase the crosstalk above some additional higher cut-off frequency, Fh. However as the grating sidelobes also move around with frequency these can be much less disturbing to the listener than might be expected, and in any case can be eliminated by suitable design of the array element layout.

Eleventh Aspect of the Invention In a further refinement applicable to any and all of the previously described schemes, one can operate the PAA in near-field mode rather than far-field mode. In this case it is possible to meaningfully focus the main PAA beam rather than merely point it in a given direction. So in this eleventh aspect of the invention, one proceeds as previously described with a direction varying-with-frequency main beam, and steers the most suitable Trough at target B. but in addition and simultaneously, one focuses the main beam so that its energy density is maximal in the vicinity of target A. Above the upper cut-off frequency, as described in aspect ten, it is trivially clear that the main beam should be focussed in the direction of and at the distance of target A. However, above the lower cut-off frequency and below the upper cut-off frequency, while the direction of main beam is determined as before (i.e. it is over-steered beyond target A to place a Trough in the vicinity of target B) the distance at which the main beam is focussed should be adjusted to maximise the amplitude of signal at target A. This in turn may slightly adjust the position of the Trough and some iterative selection of beam direction and focussing parameters can be helpful, or alternatively an analytical solution that satisfies the two constraints (maximise amplitude at A while minimising amplitude at B. or often better, maximising the difference in amplitude between targets A & B) may be used to control the PAA beam(s).

. ëhec. .. . ... ë. . . The invention will now be further described, by way of non- limitative example, with reference to the accompanying drawings, in which: S Figure I is a schematic plan view of an apparatus comprising four transducers and outputting a sound beam at a particular frequency; Figure 2 is a graph showing relative sound pressure levels at various angular displacements from the centre line of the apparatus; and Figure 3 is a schematic plan view of a laptop computer comprising apparatus according to the present invention and outputting two sound beams to a user's ears.

The invention will now be described in further detail, and with reference to preferred embodiments. Before this, however, it is useful to discuss some technical considerations that are useful to one of ordinary skill in the art when carrying out the present invention.

Consider first a continuous line array type of PAA antenna (i.e. not discrete elements but a continuous radiator), of length 2a and driven at frequencyf with wavenumber k, and with sound velocity c. For simplicity we assume all targets are in the far field though similar but more complex results apply in the near field tooWe have wavenumber k = (omega/c), where angular frequency omega=(2 pi f). Let rho >0 be the angle subtended between the user's right ear and the principal axis of the PAA antenna. For simplicity of description only, we assume the user's left ear is at angle -thO and that this is the position we want to produce a Trough (near null) for right ear signals.

The (acoustic pressure) radiation patten for a beam steered at angle threes is: P(th) = sin(k a(sin(th) -sin( ths,eer))/(k a (sin(th)-sin( the) ) which will have unit amplitude at th = ths,eer.

A null will occur whenever the numerator is zero (except for th = thS,ee,) ' i.e. when sin(k a(sin(th) -sin( ths,eer)) = 0 . ee .. . . c e. . Or when k a(sin(th) -sin( thS,eer)) = pi N. N= +/- 1, 2, 3, To produce a null at -thO we set th = thO and so we steer to: s thS,ee, = arcsin((-pi N/k a) - sin(thO)) The low frequency cut-off F' occurs when thS,eer = pi/2 which occurs when: 1 = (- pi N)/(ka) - sin(thO) or k = (pi N)/(a(l +sin(thO)) so that F. = (-N c)/(2 a (I +Sin(thO)) Also - there is an HE cut-off frequency Fh. First note the first null third' occurs when k a (sin (thrum) - s in (th5eer) = p i/2 so for a null to occur at thO then k a 2 sin(thO) = +/- N pi glvlng Fh = +/- N c/(4 a sin(thO)) Above the first upper cut-off frequency Fhl (N=1) in the above formulae, the main steered beam lobe directed towards target A begins to encroach on the target B. However, by using N =2 in the above formulae, a different beam null may usefully be steered to the target B maintaining good low crosstalk well above Fhl all the way to a higher cutoff frequency Fh2 (N=2 in the above equation). After this point N=3 may be selected so that yet another null in the beam is directed to target B. so that for practical purposes there is no upper cutoff frequency except that imposed by inter-transducer spacing (for regular arrays) as previously Ace . .e.. .. ...

. . . . . . . . . . . * described. This use of N = 1, 2, 3... to determine which (of potentially many) beam nulls is directed at target B. and which may be used with any of the previously described aspects of the invention, constitutes an aspect of the invention.

When a non-continuous, discrete array structure is used for the PAA, the results above are only approximations to the actual performance. However, they still give a flavour of how the various parameters interact, and in fact when the number of discrete transducers in the array is more than three, and the inter-transducer spacing is smaller than a wavelength of sound in air of the all of the frequencies of interest, the above formulae may be used as a good approximation to the behaviour of the discrete array.

Embodiments of the invention will now be described.

Fig.1 shows a schematic of a simple four element PAA comprising four transducers 4 driven from a common input signal l l via frequency sensitive delay devices to (which might e.g. be analogue or digital filters, or some kind of Fourier process), and with a nominal centre line l shown for descriptive purposes only. The listener's head 3 in this example is near to the centre line 1 (although this is not a necessary condition), and the listener's ears are shown as A (left ear) and B (right ear). A circle centred on the array and array centre line l of radius equal to the listener-head 3 distance is shown as 2, again for explanatory purposes only. The PAA is shown directing and focussing a beam of sound atone frequency F at point 5, and schematic sound "rays" 6 are merely shown for illustration. Note that the beam so formed converges at 5 and has maximum intensity in the region around 5, and diverges thereafter at greater distances from the array. The solid curve 7 is a very schematic representation of the polar diagram of the beamshape at frequency F and it will be seen that the beam has a maximum in the direction of 5 and a null or Trough 9 in the direction of the right ear B. Thus, at the frequency F (the "first frequency"), it is determined that directing the sound beam 7 in the direction of location 5 (the "first direction") causes the sound pressure level at the frequency F at the location of the listener's ear B to be weak or null. For other frequencies, the direction in which to direct the sound beam so as to achieve the weak or null signal at B will likely be different. Thus, the first sound signal is directed in a direction in accordance with its frequency components such that different frequency components are directed in different directions. In the direction of the left ear A marked schematically with dotted line 8, . . : : ::: :: :.

: : : : : : :. : at. . the polar diagram 7 shows considerable radiation (where line 8 cuts curve 7) and in this example it is at least as strong as the maximum at 5. So in this idealised situation at this one frequency F there is infinite signal to noise ratio (SNR) in the sense of signal at left ear A and noise at right ear B. or equivalently, zero crosstalk. Note the crucial step for this condition is not focussing and directing the beam at ear A, but rather, directing the null 9 in the beam at ear B. Not shown, for clarity, is how the beam direction 5 is changed with frequency by the frequency sensitive delay devices 10. Also not shown for clarity, is how a second input channel could be used to direct a second beam roughly towards ear B with its null directed to the vicinity of ear A, for a two channel zero-crosstalk system.

Fig. 2 shows a graph of calculated performance for simple PAA designed according to one aspect of the invention. The horizontal axis is angular deviation in degrees from the normal to the PAA, so that Odeg, in the centre, represents the direction of a straight-ahead beam. The vertical axis represents beam amplitude in dB (as a function of angular position) at one frequency. The single component curve marked 101 represents the beam amplitude (in dB) for a beam at one frequency F steered at +lOdeg. In this example it is assumed that the PAA is being used to provide low crosstalk stereo listening to a listener a few hundred millimetres in front of a laptop PC in which the PAA is embedded. The listener's ears subtend an angle of +lOdeg (shown as 104) and -lOdeg (shown as 103) respectively with respect to the straight ahead position. So the first curve 101 shows the effect of directing a beam at the user's right ear 104 at +lOdeg. The right ear receives a unit amplitude (OdB) signal, but it will be seen that with this beam position the left ear 103 receives a similar signal only -2dB down on the right ear, very poor crosstalk performance. However, the second two-component curve 102, illustrates the effect of steering the beam at the same frequency F further away from 104 to a position close to +60deg marked as 105, with a 7db beam amplitude boost also provided (at this frequency F). The right ear at 104 is now still receiving a unit amplitude (OdB) signal as before, but the left ear at 103 is now sitting in a deep null in the beam shape 102 and the signal there is at least - 20dB down (on the right ear signal). By providing a similar beam angular position shift and amplitude correction at all frequencies between the lower cutoff FL and the upper cut-off frequency FH, the null or Trough is kept closely in the vicinity of left ear at 103 thus delivering minimal sound level, whilst the right ear at 104 receives nominally constant amplitude at all frequencies. The provision of a second frequency- positioned beam with a null centred on the right ear at 104 and the beam maximum in the vicinity of left ear . . . : : : : : : : :e . . . ... .. . . 103 then provides complete 2-channel low crosstalk stereo to the listener.

Fig. 3 shows a schematic drawing of a laptop PC 201 (drawn from above) with an embedded PAA 202. Sitting in front of the laptop is listener 203 with right ear 204 and left ear 205. It S will be seen that the ears 204, 205 subtend a considerable angle at the laptop PAA 202. The PAA directs two beams 206 and 207 one either side of the head of listener 203, and moves the beams left to right as a function of frequency, so as to keep the main (first) nulls of these beams in the direction of the opposing ears of the listener, providing two channel low crosstalk sound to the listener's ears.

Claims (28)

e* ë e. ee C e c c e . CLAIMS

1. A method of creating a sound field comprising a strong first sound signal at a first location and a weak or null first sound signal at a second location, said method comprising: determining a first direction in which to direct a sound beam of a first frequency such that the sound pressure level at said first frequency at said second location is weak or null; directing frequency components of said first sound signal at said first frequency in said first direction; determining a second direction in which to direct a sound beam of a second frequency such that the sound pressure level at said second frequency at said second location is weak or null; directing frequency components of said first sound signal at said second frequency in said second direction.

2. A method according to claim 1, wherein frequency components of said sound signal in a range around said first frequency are directed in said first direction.

3. A method according to claim 1 or 2, wherein frequency components of said sound signal in a range around said second frequency are directed in said second direction.

4. A method according to any one of the preceding claims, wherein said sound signal is split into frequency bands, and components of said sound signal in a particular frequency band are directed in a direction such that the sound pressure level at said second position is weak or null.

5. A method according to any one of the preceding claims, further comprising: determining a third direction in which to direct a sound beam of a third frequency such that the sound pressure level at said third frequency at said first location is weak or null; directing frequency components of a second sound signal at said third frequency in said third direction; determining a fourth direction in which to direct a sound beam of a fourth frequency such that the sound pressure level at said fourth frequency at said first location is weak or null; ::- ::e c:e directing frequency components of a second sound signal at said fourth frequency in said fourth direction.

6. A method according to any one of the preceding claims, further comprising: adjusting the amplitude of said components of said first sound signal at or around said first frequency so as to achieve a predetermined strong sound pressure level at said first frequency at said first location.

7. A method according to claim 6, wherein said adjusting the amplitude comprises amplifying.

8. A method according to any one of the preceding claims, wherein said first location is one ear of a listener and said second location is the other ear of said listener.

9. A method according to any one of the preceding claims, wherein said first sound signal is decomposed into frequency bins using a Fourier transform process in order that different frequency components of said first sound signal may be directed in different directions.

to. A method according to any one of claims l to 8, wherein filters are used to provide a varying phase delay to different frequencies of said first sound signal.

11. A method according to claim to, wherein said filters also vary the amplitude of said first sound signal with frequency such that frequencies received at said first location have the correct relative amplitudes.

12. A method according to any one of the preceding claims, further comprising tracking the location of a listener's ears and making said first location one ear of a listener and making said second location the other ear of said listener.

13. A method according to any one of the preceding claims, further comprising: directing components of said first sound signal above an upper cut-off frequency towards said first location.

a , cee eca a * a a a. a a sea A. tea

14. A method according to any one of the preceding claims, wherein the directed frequency components are focussed such that the sound pressure level is maximal at said first location.

15. A method of creating a sound field, said method comprising: splitting a sound signal into frequency bands; directing each frequency band of said sound signal independently such that a trough of sound pressure level exists at a common location in space.

16. A method according to any one of the preceding claims, wherein said sound field is created by a sound projector comprising a plurality of sonic output transducers, said transducers outputting suitably delayed sound signals to direct said frequency components.

17. A method according to claim 15, wherein said transducers are arranged in a horizontally disposed array.

18. A method according to any one of the previous claims, wherein an ultrasonic

or infrasonic sound field is created.

19. A method according to any one of claims 1 to 17, wherein an electromagnetic field is created instead of a sound field, said first signal being an electromagnetic signal.

20. A method according to any one of the previous claims, wherein a reception

field is created instead of a transmission field.

21. Apparatus for creating a sound field that comprises a strong first sound signal at a first location and a weak or null first sound signal at a second location, said apparatus comprising: a plurality of output transducers capable of directing more than one sound beam, each beam being different; a processor arranged to determine a first beaming direction for a first frequency of sound such that the sound pressure level at said first frequency at said second location is weak or null; c:.: - ; a.e e a a a, a, ,, ,,, it, said processor also being arranged to determine a second beaming direction for a second frequency of sound such that the sound pressure level at said second frequency at said second location is weak or null; wherein, in use, components of said first sound signal at said first frequency are beamed in said first direction and components of said first sound signal at said second frequency are beamed in said second direction.

22. Apparatus according to claim 21, further comprising an amplitude adjuster for adjusting the amplitude of frequency components such that frequency components received at said first location have the correct relative amplitude.

23. Apparatus according to claim 21 or 22, wherein said processor is further arranged to determine a third beaming direction for a third frequency of sound such that the sound pressure level at said third frequency at said first location is weak or null; said processor also being arranged to determine a fourth beaming direction for a fourth frequency of sound such that the sound pressure level at said fourth frequency at said first location is weak or null; whereby components of a second sound signal at said third frequency are beamed in said third direction and components of said second sound signal at said fourth frequency are beamed in said fourth direction.

24. Apparatus for creating a sound field, said apparatus comprising: means to split a sound signal into a plurality of frequency bands; a plurality of output transducers for directing each frequency band independently such that a trough of sound pressure level is created for said plurality of frequency bands at a predetermined location.

25. Apparatus according to any one of claims 21 to 24, wherein said apparatus is arranged to carry out the method of any one of claims 1 to 20.

26. A laptop personal computer comprising the apparatus of any one of claims 21 to25. .

- -

- ^

e e r e he eve c.* e

27. A method of creating a sound field, substantially as hereinbefore described or with reference to any one of Figures 1 to 3 of the accompanying drawings.

28. Apparatus for creating a sound field, constructed and arranged substantially as hereinbefore described or with reference to any one of Figures 1 to 3 of the accompanying drawings.