Speaker array audio signal processing method

Abstract

The invention relates to a computer-implemented audio signal processing method for steering a speaker array, preferably a subwoofer array, the array comprising at least N fixed spaced speakers.

Description

[0001] SPEAKER ARRAY AUDIO SIGNAL PROCESSING METHOD

[0002] FIELD OF THE INVENTION

[0003] The invention relates to speaker arrays, preferably subwoofer arrays, and to computer- implemented audio signal processing methods for steering and balancing said speaker arrays.

[0004] BACKGROUND

[0005] A subwoofer, herein also referred to as a sub, is a loudspeaker designed to reproduce low-pitched audio frequencies known as bass and sub-bass. The typical frequency range for a subwoofer is about 20 to 200 Hz for consumer products, below 100 Hz for professional live sound, and below 80 Hz in THX-certified systems. Subwoofers are rarely used alone, as they are intended to augment the low-frequency range of speakers that cover the higher frequency bands.

[0006] A subwoofer array, herein also referred to as a sub array, is a set of two or more subwoofers (for example 3 or more) aligned each one close the other in orderto improve the sub bass response of the system, i.e., to adapt the directivity of the sound system to the acoustical requirements of the venue. A flown sub array is typically suspended at a certain height. A subwoofer stack, herein also referred to as a sub stack, is a set of two or more subwoofers placed on top of each other in order to improve the sub bass response of the system.

[0007] Ordinary bass speakers are very nearly omnidirectional over their working ranges, but when one aligns a few of them, the pattern becomes more directional and more complex. When multiple woofer arrays are present, stage left and stage right, for example, one obtains wave interference (also called "comb filtering"), causing peaks and nulls in different places in the room at different frequencies. Even beyond that, there is the problem of reverberation, which adds its own kinds of confusion and coloration in the time dimension.

[0008] Many, if not most, subwoofer installations use two separate arrays on opposite sides of the stage. Sometimes these arrays are stacked on the floor, sometimes they are flown. Either way, the multiple sources exhibit what physicists call "wave interference", and what audio people call "comb filtering" or "lobing". These lobes will produce uneven bass tonal balance and level in the venue. In indoor venues, the tonal balance problems are partly masked by reverberation, but the lack of clarity remains. Outdoors, there is no reverberation, and the problem is usually quite obvious.

[0009] In the face of all these phenomena, there is a need for audio professionals to design subwoofer arrays and drive schemes that provide required qualities of coverage and fidelity.

[0010] EASE Focus Version 3 (2019) from AFMG is a three-dimensional, acoustic simulation software for the configuration and modelling of line array systems and loudspeakers. "Focus your Sub arrays" from AMFG (2017) is a presentation illustrating the tools of EASE Focus 3. Slide 23 uses delay times automatically calculated by EASE Focus for outer subs, allowing to widen the opening angle. "Modeling point-source loudspeaker arrays in AFMG's EASE, Focus 3, and GLLViewer programs" by Keele and Sarvis describes a pair of free AFMG-supplied GLLs for creating point-source cluster and array-type configurations. The comparative examples of the present application provide the classic time delay as described in these documents.

[0011] Previous technologies in arcing a sub array introduce delay, such as time delay, either electronical or physical, into the sub array set-up. The amount of delay typically increases starting from the centre of the array moving outwards. There are a variety of calculation models to predict the desired effect. However, as shown in the example section, these methods typically only work well for the one frequency to which they were optimised, but the artefacts remain for other frequencies. The artefacts on the other frequencies can result in the opposite effect that was intended. Especially when looking at noise pollution.

[0012] Therefore, there is a need for methods to properly steer or balance a subwoofer array. More specifically, there is a need for methods that allow the subwoofer array to be steered for all frequencies or the entire frequency range, rather than just for a single optimised frequency or frequency sub-range.

[0013] When looking at modern live immersive setups, in many cases a frontal system is being used. A frontal system is a system that contains at least 3 sources on which a soundscape is represented. Since the sources of the frontal system are typically full range they also affect the frequency region identical to the subwoofer range. Therefore, there is also a need for methods to properly steer or balance a frontal system.

[0014] SUMMARY OF THE INVENTION

[0015] The inventors have surprisingly found that one or more of these problems can be solved by the present invention and embodiments thereof.

[0016] According to a first aspect, the invention relates to a computer-implemented audio signal processing method for steering a speaker array. The speaker array preferably comprises at least N fixed spaced speaker. The method preferably comprises one or more, preferably all, of the steps of: receiving an original signal to be transmitted by each speaker; whereby the original signal for all N speakers is identical; for one or more selected individual speakers or one or more groups of selected individual speakers, applying a relative phase shift to the full frequency range or a partial frequency range, wherein the phase shift is proportional to relative position, preferably proportional to relative distance, compared to a reference speaker or a group of reference speakers, thereby obtaining a modified signal for each selected individual speaker or group of selected individual speakers; and, emitting the modified signal by the speaker array.

[0017] In some preferred embodiments, the speaker array comprises a subwoofer array. In some preferred embodiments, the speaker array comprises a full-range speaker array. In some preferred embodiments, the speaker array comprises subwoofers and fullrange speakers.

[0018] In some preferred embodiments, the difference in phase shift applied for adjacent speakers is approximately constant, preferably does not exceed a variation between frequency difference of at most +- 30%, preferably of at most +- 20%, preferably of at most + / - 10%, preferably of at most +- 5%.

[0019] In some preferred embodiments, the method comprises one or more, preferably all, of the steps of: ■ obtaining a desired time delay At for at least one selected frequency f;

[0020] ■ calculating the phase angle shift as: phase angle (deg) p = time delay At xfreguencyfx 360; and,

[0021] ■ applying the phase angle shift to all other frequencies.

[0022] In some preferred embodiments, the step of obtaining a desired time delay At for at least one selected frequency f is performed by using computer simulations.

[0023] According to a second aspect, the invention relates to a filter configured to perform the method according to the first aspect, and (preferred) embodiments thereof.

[0024] In some preferred embodiments, the filter is configured to adjust the phase in a predefined frequency region, preferably in the low frequencies, for example the region of from 20Hz to 100Hz.

[0025] In some preferred embodiments, the filter is configured to be adjusted in real-time.

[0026] In some preferred embodiments, the filter comprises one or more of the following variables:

[0027] ■ the frequency which defines the frequency range of the filter;

[0028] ■ the phase shift; and,

[0029] ■ the Q-factor or slope.

[0030] In some preferred embodiments, the filter comprises a stack of at least 2, for example at least 3, preferably at least 4, 1storder all-pass filters.

[0031] In some preferred embodiments, the filter comprises a combination of one or more 1storder maximum phase and one or more 1storder minimum phase all-pass filters. In some preferred embodiments, at least one 1storder maximum phase and one or more 1storder minimum phase filters are offset on the frequency scale.

[0032] In some preferred embodiments, the method according to the first aspect, and (preferred) embodiments thereof, is performed by using a filter according to the second aspect, and (preferred) embodiments thereof.

[0033] According to a third aspect, the invention relates to use of the method according to the first aspect, and (preferred) embodiments thereof, and / or of the filter according to the second aspect, and (preferred) embodiments thereof, in a live setting. In some preferred embodiments, the invention relates to use of the method according to the first aspect, and (preferred) embodiments thereof, and / or of the filter according to the second aspect, and (preferred) embodiments thereof, using warped FIR filters

[0034] BRIEF DESCRIPTION OF THE FIGURES

[0035] FIG. 1 illustrates a render of 6 sub speakers lm spaced apart, with no delay at 50Hz.

[0036] FIG. 2 illustrates a render of the 6 sub speakers lm spaced apart of FIG. 1, with no delay at 100Hz.

[0037] FIG. 3 illustrates a render of the sub array of FIG. 1, with 1ms step delay at 50Hz.

[0038] FIG. 4 illustrates a render of the sub array of FIG. 1, with 2ms steps at 50Hz.

[0039] FIG. 5 illustrates a render of the sub array of FIG. 1, with 1ms steps at 100Hz.

[0040] FIG. 6 illustrates a render of the sub array of FIG. 1, with 2ms steps at 100Hz.

[0041] FIG. 7 illustrates a render of the sub array of FIG. 1, with 4ms steps at 50Hz.

[0042] FIG. 8 illustrates a render of the sub array of FIG. 1, with 4ms steps at 100Hz.

[0043] FIG. 9 illustrates a render of the sub array of FIG. 1, with a 30° phase shift at 50Hz, according to a preferred embodiment of the invention.

[0044] FIG. 10 illustrates a render of the sub array of FIG. 1, with a 30° phase shift at 100Hz, according to a preferred embodiment of the invention.

[0045] FIG. 11 illustrates a minimum phase inverted 1kHz 1storder all-pass filter, which may be used according to a preferred embodiment of the invention.

[0046] FIG. 12 illustrates an all-pass filter 1 set to 100Hz.

[0047] FIG. 13 illustrates an all-pass filter 2 set to 60Hz.

[0048] FIG. 14 illustrates an all-pass filter 3 set to 35Hz.

[0049] FIG. 15 illustrates a stack of the all-pass filters of FIG. 12, 13, and 14.

[0050] FIG. 16 illustrates a render of an unprocessed flown sub array comprising 16 sub speakers at 50Hz.

[0051] FIG. 17 illustrates a render of the flown sub array of FIG. 16, with 2ms step delay in zone C and 4ms steps delay in zone D at 35Hz.

[0052] FIG. 18 illustrates a render of the flown sub array of FIG. 16, with 2ms step delay in zone C and 4ms steps delay in zone D at 100Hz.

[0053] FIG. 19 illustrates a render of the flown sub array of FIG. 16, with 0.5ms step delay in zone C and 1ms steps delay in zone D at 100Hz. FIG. 20 illustrates the offset between 2 1storder all-pass filters.

[0054] FIG. 21 illustrates a visual representation of the ideal filter.

[0055] FIG. 22 illustrates a speaker set-up comprising full-range speakers E and F, and subwoofer array A-D.

[0056] DETAILED DESCRIPTION OF THE INVENTION

[0057] The present invention will be described with respect to particular embodiments, but the invention is not limited thereto but only by the claims. Any reference signs in the claims shall not be construed as limiting the scope thereof.

[0058] As used herein, the singular forms "a", "an", and "the" include both singular and plural referents unless the context clearly dictates otherwise.

[0059] The terms "comprising", "comprises" and "comprised of" as used herein are synonymous with "including", "includes" or "containing", "contains", and are inclusive or open-ended and do not exclude additional, non-recited members, elements, or method steps. The terms "comprising", "comprises" and "comprised of" when referring to recited members, elements or method steps also include embodiments which "consist of" said recited members, elements, or method steps.

[0060] Furthermore, the terms first, second, third and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a sequential or chronological order, unless specified. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein.

[0061] The term "about" as used herein when referring to a measurable value such as a parameter, an amount, a temporal duration, and the like, is meant to encompass variations of + / -10% or less, preferably + / -5% or less, more preferably + / -1% or less, and still more preferably + / -0.1% or less of and from the specified value, insofar such variations are appropriate to perform in the disclosed invention. It is to be understood that the value to which the modifier "about" refers is itself also specifically, and preferably, disclosed.

[0062] The recitation of numerical ranges by endpoints includes all numbers and fractions subsumed within the respective ranges, as well as the recited endpoints. All documents cited in the present specification are hereby incorporated by reference in their entirety. Unless otherwise defined, all terms used in disclosing the invention, including technical and scientific terms have the meaning as commonly understood by one of the ordinary skill in the art to which this invention belongs.

[0063] By means of further guidance, definitions for the terms used in the description are included to better appreciate the teaching of the present invention. The terms or definitions used herein are provided solely to aid in the understanding of the invention. Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention.

[0064] Thus, appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner, as would be apparent to a person skilled in the art from this disclosure, in one or more embodiments. Furthermore, while some embodiments described herein include some, but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would be understood by those in the art. For example, in the following claims and description, any of the claimed or described embodiments can be used in any combination.

[0065] According to a first aspect, the invention relates to a computer-implemented audio signal processing method for steering a speaker array. The speaker array preferably comprises at least N fixed spaced speaker. The method preferably comprises one or more, preferably all, of the steps of: receiving an original signal to be transmitted by each speaker; whereby the original signal for all N speakers is identical; for one or more selected individual speakers or one or more groups of selected individual speakers, applying a relative phase shift to the full frequency range or a partial frequency range, wherein the phase shift is proportional to relative position, preferably proportional to relative distance, compared to a reference speaker or a group of reference speakers, thereby obtaining a modified signal for each selected individual speaker or group of selected individual speakers; and, emitting the modified signal by the speaker array.

[0066] According to a second aspect, the invention relates to a filter configured to perform the method according to the first aspect, and (preferred) embodiments thereof.

[0067] In some preferred embodiments, the method according to the first aspect, and (preferred) embodiments thereof, is performed by using a filter according to the second aspect, and (preferred) embodiments thereof. (Preferred) embodiments of the first aspect are also (preferred) embodiments of the second aspect, and vice versa.

[0068] A time delay, as used in the prior art, applied to a frequency range will provide a different phase shift for each frequency within the range. Lower frequencies will have lesser phase shift where higher frequencies have more phase shift. The present invention, on the other hand, provides a fixed phase shift for all frequencies within a selected range. This is the main difference with the methods of the prior art that merely focus on a fixed time delay to obtain a similar goal. By using a fixed phase shift between different sources of a sub array, every frequency will be projected the same whereas when using time delay the projection of each frequency varies.

[0069] As used herein, the terms "speaker", "sound source", or "source" are used interchangeably.

[0070] In some preferred embodiments, the speaker array comprises a subwoofer array. In some preferred embodiments, the speaker array comprises a full-range speaker array. In some preferred embodiments, the speaker array comprises subwoofers and fullrange speakers.

[0071] In some embodiments, the method comprises the step of: receiving an original signal to be transmitted by each speaker; whereby the original signal for all N speakers is identical.

[0072] Preferably, the original signal is a signal configured for a subwoofer array, or a signal configured for a full-range speaker array. The original signal is the originally intended signal to be transmitted by each speaker, which, prior to applying the method of the invention, would have been identical for all N speakers. In some embodiments, the method comprises the step of: emitting the modified signal by the speaker array.

[0073] Preferably, the modified signal is emitted by the same a subwoofer array full-range speaker array as the original signal was intended for. The modified signal is dependent on the speaker, since the phase shift is dependent on the speaker. Therefore, the modified signal is a speaker-dependent modified signal. It is clear that, for those speakers where the original signal undergoes no phase shift, the modified signal to be emitted is identical to the original signal originally intended to be emitted.

[0074] A subwoofer array, herein also referred to as a sub array, is a set of two or more subwoofers (for example 3 or more) aligned each one close the other in orderto improve the sub bass response of the system (i.e., the low frequency response of a sound system). A subwoofer array is used to adapt the directivity of the sound system to the acoustical requirements of the venue. Preferablythe subwoofers are placed next to each other or above each other and are directed in the same direction, as opposed to a cardioid subwoofer arrangement. However, in some embodiments, the present technology can also be used in combination with cardioid subwoofer set ups. Instead of giving each subwoofer a different signal, the signal will vary in subwoofer groups of typically 4. 3 forward and one backward.

[0075] Subwoofers are typically designed to reproduce frequencies in the lower range of human hearing or even below the range of human hearing, and are often used in concert sound systems, movie theatres, and home theatre systems to add depth and impact to the audio experience. As used herein, the subwoofers will preferably at least operate in a frequency range of 20 Hz to 200 Hz, for example 20 Hz to 100 Hz.

[0076] There are two main types of subwoofer arrays: ground arrays and flown arrays. A ground array is a group of subwoofers that are placed on the ground and pointed upwards towards the listener. This type of subwoofer array is often used in outdoor concerts and other events where the subwoofers need to be portable and easy to set up.

[0077] On the other hand, a flown array is a group of subwoofers that are suspended in the air at a certain height, for example using rigging equipment. This type of subwoofer array is often used in indoor concerts and other events where the subwoofers need to be suspended above the audience for optimal sound coverage. Flown arrays are more complex to set up and require specialized rigging equipment, but they offer greater control over the directionality of the sound and can be adjusted to suit the specific needs of the event.

[0078] A subwoofer stack, herein also referred to as a sub stack, is a set of two or more subwoofers placed on top of each other in order to improve the sub bass response of the system.

[0079] For ground subs, N is typically at least 3. For flown subs, N is typically at least 2. The size of the sub array may also be determined by the room or area it needs to cover. A small sub array may consist of 8 to 12 subwoofers. Larger sub arrays may go up to 96 subwoofers.

[0080] Full-range speakers are speakers that are designed to reproduce the full range of human hearing, from the lowest frequencies to the highest frequencies. Full-range speakers are often used in home theatre systems and other high-fidelity audio systems to provide a complete and balanced sound experience. As used herein, the full-range speakers will preferably at least operate in a frequency range of 20 Hz to 20 kHz.

[0081] Full-range speakers differ from subwoofer arrays in that they do not specialize in reproducing only the low frequency range of sound. While subwoofers are designed specifically to enhance the low frequency response of a sound system, full-range speakers are designed to reproduce all frequencies across the entire range of human hearing. This allows them to provide a more complete and balanced audio experience, as they are able to reproduce both low and high frequencies with equal clarity and precision.

[0082] However, full-range speakers may not provide the same level of impact and depth as subwoofer arrays, which are specifically designed to enhance the low frequency response of a sound system.

[0083] When looking at modern live immersive setups, in many cases a frontal system is being used. A frontal system is a system that contains at least 3 sources on which a soundscape is represented. Since the sources of the frontal system are typically full range they also affect the frequency region identical to the subwoofer range. Therefore, the present invention may also be applied to frontal systems, resulting in a more equal low frequency spread. In some embodiments, the speaker array is a frontal system. A frontal speaker system is a type of audio setup that consists of multiple speakers positioned in a horizontal row, typically with the intention of creating a more immersive listening experience. The speakers may be full-range or subwoofers.

[0084] In some embodiments, the speaker array is a soundbar. A soundbar is a type of speaker system that is designed to be compact and easy to install, and is typically used in home theatre setups. Soundbars typically consist of a single, elongated enclosure that contains multiple speakers, and are designed to be mounted on a wall or placed on a shelf below a television. They are intended to provide an improved audio experience compared to the built-in speakers on a television, and may also include additional features such as wireless connectivity and virtual surround sound.

[0085] In some embodiments, the method comprises the step of: for one or more selected individual speakers or one or more groups of selected individual speakers, applying a relative phase shift to the full frequency range or a partial frequency range, wherein the phase shift is proportional to relative position, preferably proportional to relative distance, compared to a reference speaker or a group of reference speakers, thereby obtaining a modified signal for each selected individual speaker or group of selected individual speakers.

[0086] As used herein, the term "phase shift" refers to a "phase angle shift". Phase is a measure of the position of a waveform in time, and it is typically expressed in degrees or radians. A phase angle shift refers to the process of altering the phase relationship between two or more audio signals.

[0087] Preferably, the relative phase shift is applied to the full frequency range of the subwoofer array, for example at least to the range from 20 Hz to 100 Hz, for example at least to the range from 20 Hz to 200 Hz. In some preferred embodiments, the filter is configured to adjust the phase in the full frequency range of the subwoofer array. In some preferred embodiments, the filter is configured to adjust the phase in a predefined frequency region, preferably in the low frequencies, for example the region of from 20Hz to 200Hz, for example the region of from 20Hz to 100Hz. Preferably, the relative phase shift is applied to the full frequency range of the full-range speaker array, for example at least to the range from 20 Hz to 20 kHz. In some preferred embodiments, the filter is configured to adjust the phase in the full frequency range of the full-range speaker array. In some preferred embodiments, the filter is configured to adjust the phase in a predefined frequency region.

[0088] In some preferred embodiments, the difference in phase shift applied for adjacent speakers is approximately constant. Preferably the difference in phase shift applied for adjacent speakers does not exceed a variation between frequency difference of at most +- 30%, preferably of at most +- 20%, preferably of at most + / - 10%, preferably of at most +- 5%. The phase shift between sources is preferably adapted to the desired spread of the low frequencies.

[0089] In the context of a sound speaker, a phase angle shift can be applied to a signal by delaying the timing of one of the signals relative to the other. This can be achieved using digital processing techniques, such as those used in audio software or hardware processors. Therefore, the phase shift may be calculated from a desired time delay. In some preferred embodiments, the method comprises one or more, preferably all, of the steps of:

[0090] ■ obtaining a desired time delay At for at least one selected frequency f;

[0091] ■ calculating the phase angle shift as: phase angle (deg) p = time delay At xfreguencyfx 360; and,

[0092] ■ applying the phase angle shift to all other frequencies.

[0093] In the present invention, the time delay is obtained for 1 frequency, but the resulting phase shift is applied to all other frequencies as well. This conceptually differs from the prior art, whereby a fixed time delay is applied to all frequencies, which results in a different phase shift for each frequency individually.

[0094] In some preferred embodiments, the step of obtaining a desired time delay At for at least one selected frequency f is performed by using computer simulations. In other embodiments, the desired time delay can be measured using a computer based software. The optimal phase shift cannot be predefined, since it varies depending on the desired dispersion which varies depending on the location / setup. Larger distances between the speakers result in large phase shifts. The phase shift is at most 180°. The phase shift is usually at most 90°, but may be more.

[0095] In some embodiments, the phase shift is at least 10° and at most 50°, preferably at least 20° and at most 40°, for example about 30°. In some other embodiments, the phase shift is at least 5° and at most 25°, preferably at least 10° and at most 20°, for example about 15°.

[0096] For a sub source it is preferred to only modify the frequency range of the subs and not the entire frequency spectrum. In some preferred embodiments, there is no phase shift outside the sub array frequency range. For example, in some embodiments, no phase shift is applied to frequencies below 10 Hz, preferably below 15 Hz, preferably below 18 Hz, preferably below 20 Hz. For example, in some embodiments, no phase shift is applied to frequencies above 200 Hz, preferably above 150 Hz, preferably above 120 Hz, preferably above 100 Hz.

[0097] For a full-range source it is also preferred to only modify the frequency range of the subs and not the entire frequency spectrum. For these reasons, preferably a filter is used, to apply a phase shift only in the sub range. In some preferred embodiments, there is no phase shift on the full-range source outside the sub array frequency range. For example, in some embodiments, no phase shift is applied to frequencies below 10 Hz, preferably below 15 Hz, preferably below 18 Hz, preferably below 20 Hz. For example, in some embodiments, no phase shift is applied to frequencies above 200 Hz, preferably above 150 Hz, preferably above 120 Hz, preferably above 100 Hz.

[0098] In some preferred embodiments, a full-range speaker uses the same filter as its nearest subwoofer.

[0099] The phase shift is preferably proportional to relative position, preferably proportional to relative distance, compared to a reference speaker or a group of reference speakers, As used herein, the reference speaker (or group of speakers) refers to a fundamental speaker, typically in the middle of the set-up. For example, the reference speaker may be a centrally located speaker, typically in the case of ground sub arrays. For example, the reference speaker may be the top located speaker, typically in the case of flown sub arrays.

[0100] In smaller arrays, each speaker is preferably processed individually. In larger arrays one can use groups to reduce processing power. In some embodiments, the phase shift is applied to one or more groups of selected individual speakers, for example wherein each group comprises at least 2, for example at least 3, for example at least 4 individual speakers.

[0101] In some preferred embodiments, the method performs the adjustments in real-time. In some preferred embodiments, the filter is configured to be adjusted in real-time.

[0102] In the context of computer processing, a method that is performed "in real-time" refers to a process that occurs as fast as it is needed, with minimal or no delay. In order for an audio process to be considered real-time, it must be able to complete within a predetermined timeframe, known as the "latency." The latency of a real-time process is typically measured in milliseconds or microseconds, and it must be small enough to meet the requirements of the application. For example, in a real-time audio processing system, the latency must be small enough to ensure that the processed audio is played back in sync with the original audio, without any noticeable delay. Preferably the latency is at most 20 ms, for example at most 10 ms.

[0103] Real-time processing requires a combination of fast hardware and efficient software algorithms in order to meet the strict latency requirements. The present invention has the advantage that the method is efficient enough to be performed in real-time.

[0104] In some preferred embodiments, the filter comprises one or more of the following variables:

[0105] ■ the frequency which defines the frequency range of the filter;

[0106] ■ the phase shift; and,

[0107] ■ the Q-factor or slope.

[0108] The Q-factor, herein also referred to as the "quality factor," of an audio signal refers to the narrowness or width of a peak or dip in the frequency response of an audio filter. In the context of an audio filter, the Q-factor is a measure of the sharpness or slope of the filter's frequency response curve. A high Q-factor indicates a narrow, steep slope, while a low Q-factor indicates a broad, shallow slope. A steep filter is a filter which reaches its point of maximum phase shift relatively fast, a shallow filter takes a longer frequency range to achieve its maximum offset. A shallow filter is preferably used on full range sources since one only wishes to offset the low frequency range, however when using steep filters this may create unwanted artifacts to the horizontal spread of the system. In some preferred embodiments, the filter comprises a stack of at least 2, for example at least 3, preferably at least 4, 1storder all-pass filters.

[0109] A first order all-pass filter is a type of electronic filter that allows all frequencies to pass through with equal gain, but introduces a phase shift that varies with frequency. These filters are commonly used in audio processing applications to alter the phase response of a signal. First order all-pass filters are designed to have a single pole in the transfer function, which results in a linear phase shift across the entire frequency range. Preferably all the sources that need to be aligned with the sub array are provided with the same all-pass filter as their target sub source. The identical filter of the closest subwoofer to the source is applied to the source as well so the phase difference between the source and the subwoofer stays as minimal as possible. This is illustrated in example 4 and in FIG. 22. FIG. 22 illustrates a typical set-up of a ground sub arc in combination with a left right full range speaker system. As shown, the inner sub arc sources receive a different all-pass filter in relation to the outer sub arc sources. This way, a low frequency phase shift is created over the frequency range of the subwoofers. The full range speakers need to be phase aligned with the sources of the subwoofer arrays closest to the full range speakers. In this case, the full range speakers receive the same all-pass filter as the outer sub array sources.

[0110] In some preferred embodiments, the filter comprises a combination of one or more 1storder maximum phase and one or more 1storder minimum phase all-pass filters. A first order maximum phase filter is a type of electronic filter that introduces a phase shift that increases with frequency, while a first order minimum phase filter is a type of electronic filter that introduces a phase shift that decreases with frequency. These filters are characterized by their transfer function, which describes the relationship between the input and output signals of the filter. The transfer function of a first order maximum phase filter has a single pole at the origin, while the transfer function of a first order minimum phase filter has a single zero at the origin. The difference in the position of the pole or zero in the transfer function determines the slope of the phase shift introduced by the filter. In some preferred embodiments, at least one 1storder maximum phase and one or more 1storder minimum phase filters are offset on the frequency scale.

[0111] The combination of filters as described above is not obvious in application to low frequencies.

[0112] According to a third aspect, the invention relates to use of the method according to the first aspect, and (preferred) embodiments thereof, and / or of the filter according to the second aspect, and (preferred) embodiments thereof, in a live setting. The present methods and filters of the present invention are particularly suitable for real-time processing, and are therefore particularly suitable for use in a live setting. In some preferred embodiments, the invention relates to use of the method according to the first aspect, and (preferred) embodiments thereof, and / or of the filter according to the second aspect, and (preferred) embodiments thereof, using warped FIR filters.

[0113] A warped finite impulse response (FIR) filter is a type of digital filter that is designed to have a linear phase response over a certain frequency range, but with a non-linear phase response outside of that range. This allows for improved performance frequencysensitive applications such as audio processing. The filter coefficients may be determined using a warping function that maps the desired frequency response to a linear phase response within the specified frequency range. This approach can be useful for achieving a desired frequency response with a smaller number of filter coefficients, resulting in a more efficient implementation.

[0114] In some embodiments, the method is a computer-implemented method. In some embodiments, one or more steps of the method are performed by a computer. In some embodiments, all steps of the method are performed by a computer.

[0115] The invention also relates to a system or device, such as a data processing apparatus, comprising means for carrying out one or more steps, for example all steps, of the method according to the first aspect of the invention, and (preferred) embodiments thereof. The system or device may comprise different hardware and / or software aspects to provide the functionalities as illustrated. For example, the system or device may comprise a computer, a computing system, or a processor, e.g., a general-purpose computing platform specifically programmed, e.g., by a suitable executable code, for implementing all or some elements described herein. In embodiments, the system or device may operate as a standalone device or may be connected, e.g., networked, to other machines in a networked deployment. For example, the computer, computing system, or processor may comprise a filter as described herein.

[0116] The invention also relates to a computer program product directly loadable into the internal memory of a computer, or a computer program product stored on a computer readable medium, or a combination of such computer programs or computer program products, configured for performing a computer-implemented method according to the first aspect of the invention, and (preferred) embodiments thereof.

[0117] The invention also relates to a computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out one or more steps, for example all steps, of the method according to the first aspect of the invention, and (preferred) embodiments thereof. The invention also relates to a computer-readable medium comprising instructions which, when executed by a computer, cause the computer to carry out one or more steps, for example all steps, of the method according to the first aspect of the invention, and (preferred) embodiments thereof.

[0118] EXAMPLES

[0119] Comparative example 1: no corrections

[0120] Previous technologies in arcing a sub array introduce delay, either electronical or physical, into the sub array set-up. The amount of delay typically increases starting from the centre of the array moving outwards. There are a variety of calculation models to predict the desired effect. For this example, we will use a fixed step model meaning the amount of delay added to the separate sub signals is equal between all subs.

[0121] For this example, the L Acoustics Soundvision prediction software was used. A simple set up comprised 6x SB15m sub speakers spaced lm apart. Focus will be on 50Hz and 100Hz to prove the concept, but one can apply this theory on all frequencies of the sub low frequency range.

[0122] FIG. 1 and FIG. 2 illustrate the initial set-up, without delay, for 50 Hz and 100 Hz. One can already notice the difference in the radiation pattern of both frequencies. Each change in gradient of grey represents a change of 3dB in SPL.

[0123] Comparative example 2 refers to the use of a time delay as applied in the prior art. Electronic delay between the subs may be used to spread the radiation pattern of the sub array. For the next examples, the centre 2 subs have no delay. Starting from the centre and moving outwards to the sides each subwoofer has 1ms more delay than the previous one. This is called a 1ms step array, and is shown in FIG. 3. In a 2ms step array, each subwoofer has 2ms more delay than the previous one, and is shown in FIG 4. Both FIG. 3 and FIG. 4 show the 50 Hz frequency dispersion. FIG. 5 shows the same 1ms step array at 100 Hz, while FIG. 6 shows the same 2ms step array at 100 Hz.

[0124] As seen in FIG. 3 and FIG. 4, good results may be obtained by adding electronic delay to the subs. At 50Hz there is a more even spread of the low end. However, this is only looking at 50Hz. FIG. 5 (1 ms) still seems to provide reasonable results, but FIG. 6 (2 ms) clearly shows an undesired effect at 100Hz: there is energy loss in the centre of the array and focusing towards the outside of the array. This effect is already visible even when using small amounts of delay, but at even higher steps, the artefacts would become much more noticeable. For example, FIG. 7 shows a very good spread at 50Hz with 4ms steps. However, FIG. 8 demonstrates significant problems at 100Hz.

[0125] Example 3: phase shift

[0126] Example 3, according to the invention, is a modified version of comparative example 2, according to the prior art. In the present example, a 30-degree step is applied to the same prediction model. FIG.9 and FIG. 10 demonstrate that the pattern of 50Hz can be opened enough to get an even spread. Simultaneously, the 100Hz pattern is also spread decently without causing artefacts in the centre nor focusing to the sides. The phase shift results in a time difference: a phase shift of 30 degrees at 50 Hz corresponds to a little less than 2ms. The same phase shift of 30 degrees at 100 Hz corresponds to a little less than 1ms.

[0127] It has been found that if the phase shift at various frequencies is close to identical, this results in the most desirable patterns. Ideally the pattern would be identical on all frequencies, however there may be other factors involved which influence the radiation pattern. Therefore, the variation in phase shift at various frequencies is preferably as little as possible. This means that one would get lesser artefacts and better results with sub arrays when working with phase shifts instead of time shifts.

[0128] Example 4: filters

[0129] The controllable phase shift across a frequency range is preferably obtained by a filter. Preferably, some full-range sources can be controllable in the same way as for time delay.

[0130] It is desirable to control full-range sources with time delay, but for other benefits. If control of the low frequency range of a full-range source is required, this can be provided with a filter that introduces a phase shift only to the low frequency region of the fullrange source. Since the full-range source is designed to work as one source, this cannot be achieved by introducing time delay only to the low frequency region.

[0131] While time delay may be used for distance compensation, it cannot be used for beam steering. We want to be able to steer the subs with the full-range source. If the signal to the subs is modified, the full-range sources are preferably modified as well, since it is desirable to have the subs cooperate with the full-range sources. However, for a fullrange source it is preferred to only modify the frequency range of the subs and not the entire frequency spectrum. For these reasons, preferably a filter is used, to apply a phase shift only in the sub range.

[0132] Most preferably, the filter is capable of adjusting the phase only in the low region and can be adjusted in real-time. In this example, FIR Designer is used to design a 1storder all pass filter which is inverted and uses minimum phase.

[0133] FIG. 11 illustrates a minimum phase inverted 1kHz 1storder all-pass filter. The all-pass filter adjusts the phase from high to low starting from 0 degrees up to 180 degrees with its 90 degree point at 1kHz. Such a filter is subsequently used to create the ideal sub arc, by achieving the 30- degree offset as simulated above in our prediction model. Using FIR Designer, 3 all pass filters were designed:

[0134] FIG. 12: an all-pass filter 1 set to 100Hz;

[0135] FIG. 13: an all-pass filter 2 set to 60Hz; and,

[0136] FIG. 14: an all-pass filter 3 set to 35Hz. The advantage of these 1storder minimum phase all-pass filters is that they can be employed with minimal latency, pure processing power, only calculation power not calculation time, which is particularly useful for live sound.

[0137] FIG. 15 illustrates a stack of these 3 filters zoomed in on the 50Hz - 100Hz region, illustrating the fixed phase shift needed. The phase difference in this example is the only relevant feature. There is nearly no audible difference when shifting the phase in one source. It is the phase difference between the sources that provides the desired results. Focussing on 100Hz we get these phase shifts:

[0138] • Filter 1: 90 degrees

[0139] • Filter 2: 61 degrees

[0140] • Filter 3: 37 degrees

[0141] The relative variation of 29 and 24 degrees is close to a stable difference of about 30 degrees. The variation between 24 and 29 degrees corresponds to 17%.

[0142] Focussing on 50Hz we get these phase shifts:

[0143] • Filter 1: 127 degrees

[0144] • Filter 2: 100 degrees

[0145] • Filter 3: 70 degrees

[0146] The relative variation of 27 and 30 degrees is close to a stable difference of about 30 degrees. The variation between 27 and 30 degrees corresponds to 10%.

[0147] Looking at the phase difference between the 3 filters, it can be seen that it is more or less stable around 30 degrees difference (some variation due to the character of the 1storder all-pass filter). Preferably, a filter is used that sets the low frequency region fixed to 30 degrees and the next one to 60 degrees and so on. Nevertheless, these 1storder all-pass filters already demonstrate a significant improvement.

[0148] Preferably all the sources that need to be aligned with the sub array are provided with the same all-pass filter as their target sub source. The identical filter of the closest subwoofer to the source is applied to the source as well so the phase difference between the source and the subwoofer stays as minimal as possible, as illustrated in FIG. 22. In FIG. 22, speakers E and F are full-range speakers, while speakers A to D for the sub array. Speakers E and A use the same all-pass filter. Speakers F and D also use the same all- pass filter. Speakers B and C use a different all-pass filter to obtain the 30 degree phase shift.

[0149] The phase shifts in the low end may also be used to direct the low-end coupling between main and off sources as desired. In a typical situation there is a main system pointing straight forward and an off-system pointing outwards from the stage. These 2 sources can work together in the low frequency region. If a phase difference is introduced between these two sources, the low frequency region can also be outwardly steered resulting in a much more equal low frequency dispersion.

[0150] Comparative example 5: flown array, no corrections

[0151] When looking at the recent changes in design with a higher focus on noise pollution, sound designers are frequently using flown sub arrays to focus the low frequencies more into the desired direction. Flown sub arrays, which are suspended in the air, have the advantage of limiting noise disturbance. However, the classic method of adding a time delay to a flown sub array, may create more noise disturbance than it solves.

[0152] A flown sub array of 16 speakers without any processing generates a 50Hz wavefront as shown in FIG. 16. This is not the desired propagation for a flown sub array. The back side is also illustrated, to demonstrate the noise disturbance. The back side in FIG. 16 is the left-hand side. This is also illustrated in the next figures, where the low frequencies are steered only to the front of the sub array.

[0153] Comparative example 6: flown array, time delay

[0154] Comparative example 6 refers to the use of a time delay as applied in the prior art. To solve this issue, a time difference was introduced at the bottom speakers. A variable time shifting was applied between speakers 9-12 (zone C) and speakers 13-16 (zone D), while speakers 1-8 remained unprocessed (Zones A to D run from top to bottom in FIG. 17). Optimizing the array for 32Hz, the most satisfying result was obtained when using 2ms steps in zone C and 4ms steps in zone D, as illustrated in FIG. 17. However, if we look at the other end of the frequency range of the sub array (100 Hz), we have an undesired effect as shown in FIG. 18: the 100Hz beam is split into two beams because of the high time differences. If, on the other hand, we would optimize this array for 100Hz, the result would be similar to FIG. 19, with 0.5ms steps in zone C and 1ms steps in zone D.

[0155] Example 7: phase shift

[0156] Example 7, according to the invention, is a modified version of comparative example 6, according to the prior art. By using the constant phase over frequency instead, the sound propagation of the flown sub array would be the combination of FIG. 17 for 50 Hz and FIG. 19 for 100 Hz.

[0157] Example 8: filters

[0158] By changing the frequency of an all-pass filter between each part of a subwoofer array, a variable phase shift may be achieved between two signals. However, the consequence of this processing technique is that the lowest frequencies in the system will often be close to 180 degrees shifted. There is no problem as long as the same processing is applied to all sources that need to work together.

[0159] By using this previous offset technique, we cannot be as variable as we would like, since a 1storder all-pass filter has a fixed slope and always ends up at positive or negative 180 degrees. Since we need an offset of the phase between all the frequencies of the low end, we can only go up in frequency until 2 all-pass filters reach 180 degrees and the phase difference between them is 0; at the lowest audible frequency (around 20Hz).

[0160] In FIG. 20, it is shown that the higher the offset between the filters start, the more risk there is of decreased offset at the very low frequencies. Ideally, the filter would only apply the phase shift desired to a certain frequency range of a signal.

[0161] The 1storder all-pass filter raises the phase up to +180 degrees as it moves down the frequency scale. Preferably, the phase offset is set to a fixed value starting from a defined frequency going down the frequency scale. This can be performed by shifting the minimum phase filter and the maximum phase filter.

[0162] Such a filter might resemble a low shelf filter in an equalizer, but instead of changes in gain it would change the phase of the low end without any gain changes. This filter could have three variables: the frequency which defines the frequency range of the filter; the phase shift; and, the Q-factor or slope. One example of how to achieve this type of filter is the combination of 1storder all-pass filters as illustrated in example 4 above. The filters used in FIG. 20 are 1storder all-pass filters which are inverted and are designed to be minimum phase.

[0163] If they were to be designed to be maximum phase the trace would start at 0° and decrease towards the - 180°. The minimum phase filter would increase towards the +180°. When we would combine a minimum phase 1storder all-pass filter and a maximum phase 1storder all-pass filter both having the same frequency, the sum would be 0 phase. However if we lower the frequency of the maximum 1storder all pass filter there would be an offset between the filters and there would be a momentary phase shift across a frequency region. However the sum of both will at one point return to zero as the target is to create constant phase offset, or at least down to 20Hz, the trick would be to add a second combination of minimum and maximum phase filters further down the frequency range. By adjusting the frequencies of all 4 filters there can be a wide variation of phase responses.

[0164] The frequency parameter defines the frequency range in which the filter would be active. For a 1storder all-pass filter this would be the 90 degrees point in the filter. However, since we have a variable phase shift the definition of the frequency cannot be fixed at 90 degrees. Preferably, the frequency would be set at half the phase shift. For example, on a 30 degrees phase shift the frequency would be defined at 15 degrees. This way we can vary the phase shift and always have a nice shelf filter-like interaction with the upper frequency range.

[0165] The phase shift defines the offset of the filter. The value is defined in degrees and the value can be both positive and negative. The negative values would be necessary in case of the need for more offset than 180 degrees. However, in real life situations the needed range for the uses as mentioned previous would vary between 10 and 45 degrees.

[0166] A potential third parameter could be a Q-factor which defines the slope of the "shelf". Shown in FIG. 21 is a visual representation of the most preferred filter. This filter would be a 95Hz 90° filter. The frequency is defined by half the max phase shift. In this case that is 90° and the filter is at 45° at 95 Hz.

[0167] When using these filters in sub arrays there is preferably an easy defined way to set the filters all on the same frequency and only vary the phase shift. One way to design this filter is the combination of both 1storder maximum phase and minimum phase filters. By offsetting these 2 filters on the frequency scale, the higher one being the minimum phase, a very wide bell-shaped phase offset is being created. By repeating the same combination at a lower frequency range the bell-shaped filter can be continued down to 2Ohz or lower resulting in a constant phase offset at a certain frequency range.

[0168] Example 9: full-range sources

[0169] In this example, a frontal system (such as a soundbar) consists of at least three full-range sources in a row. If these three full-range sources are provided with an identical signal, they will couple in the low-frequency region towards the middle. The frequency at which this occurs depends on the relative distance between the speakers. To spread the low frequencies evenly across the space, beam steering can be used. Normally, a time delay is applied to the outer arrays, but this is detrimental to the sound experience in the middle, since for high frequencies they should reach the listener at the same time. Therefore, using existing techniques a trade-off needs to be applied: evenly spread low frequencies OR maintain a good sound experience. By only applying a phase shift in the low range, there is no time delay between the arrays, but we can still evenly spread the low frequencies over the field to be played.

Claims

CLAIMS1. A computer-implemented audio signal processing method for steering a speaker array, the array comprising at least N fixed spaced speakers, said method comprising the steps of: receiving an original signal originally intended to be transmitted by each speaker; whereby the original signal for all N speakers is identical; for one or more selected individual speakers or one or more groups of selected individual speakers, applying a relative phase shift to the full frequency range or a partial frequency range, wherein the phase shift is proportional to relative position, preferably proportional to relative distance, compared to a reference speaker or a group of reference speakers, thereby obtaining a speaker-dependent modified signal for each selected individual speaker or group of selected individual speakers; and, emitting the speaker-dependent modified signal by the speaker array.

2. The method according to claim 1, wherein the speaker array comprises a subwoofer array.

3. The method according to any one of claim 1 or 2, wherein the speaker array comprises a full-range speaker array.

4. The method according to any one of the preceding claims, wherein the difference in phase shift applied for adjacent speakers is approximately constant, preferably does not exceed a variation between frequencydifference of at most +- 30%, preferably of at most +- 20%, preferably of at most + / - 10%, preferably of at most +- 5%.

5. The method according to any one of the preceding claims, comprising the steps of:■ obtaining a desired time delay At for at least one selected frequency f;■ calculating the phase angle shift as: phase angle (deg) p = time delay At xfreguencyfx 360; and,■ applying the phase angle shift to all other frequencies.

6. The method according to claim 5, wherein the step of obtaining a desired time delay At for at least one selected frequency f is performed by using computer simulations.

7. A processor comprising a filter configured to perform the method according to any one of the preceding claims.

8. The processor according to claim 7, configured to adjust the phase in a predefined frequency region, preferably in the low frequencies, for example the region of from 20Hz to 100Hz.

9. The processor according to any one of claims 7 or 8, configured to be adjusted in real-time, wherein the latency is at most 20 ms.

10. The processor according to any one of claims 7 to 9, comprising one or more of the following variables: the frequency which defines the frequency range of the filter;1 the phase shift; and, the Q-factor or slope.

11. The processor according to any one of claims 7 to 10, whereby the filter comprises a stack of at least 2, for example at least 3, preferably at least 4, 1storder all-pass filters.

12. The processor according to any one of claims 7 to 11, whereby the filter comprises a combination of one or more 1storder maximum phase and one or more 1storder minimum phase all-pass filters.

13. The processor according to any one of claims 7 to 12, whereby at least one 1storder maximum phase and one or more 1storder minimum phase filters are offset on the frequency scale.

14. The method according to any one of claims 1 to 6, wherein the method is performed by using a processor according to any one of claims 7 to 13.

15. Use of the processor according to any one of claims 7 to 14, in a live setting, preferably using warped FIR filters.

Images