Method for representing a stereo source in a spatial audio mix

Abstract

A computer-implemented audio signal processing method for representing a stereo source in a spatial audio mix, comprising the steps of: - providing a first and a second input signal defining an original stereo image; - positioning an aggregate sound object, wherein the aggregate sound object comprises at least one centre object and at least one pair of ghost objects, the first ghost object and second ghost object positioned symmetrically with respect to a vertical plane through the centre object; - generating an audio signal for the at least one centre object, comprising summing the first and second input signal; - generating audio signals for the at least one pair of ghost objects, comprising differentiating the first and second input signal, and upmixing the differentiated signal; and, - using the generated audio signals in a spatial audio mix, thereby obtaining a stereo source represented by the aggregate sound object.

Description

[0001] METHOD FOR REPRESENTING A STEREO SOURCE IN A SPATIAL AUDIO MIX

[0002] FIELD OF THE INVENTION

[0003] The invention relates to a method and system for creating a three-dimensional spatial audio mix. In particular, the invention relates to a computer-implemented audio signal processing method for representing a stereo source in a spatial audio mix, wherein the original stereo-wide experience is maintained while avoiding unwanted artefacts.

[0004] BACKGROUND

[0005] In stereo sound, a pair of two different audio signals is used to recreate a multidirectional audible perspective. In some types of stereo sound, e.g., originating from a stereo piano with the high notes on the right and the low notes on the left, the left and right stereo signal almost have a unique content. In other types of stereo sound, e.g., an original voice with a stereo effect over it, the left and right signal have a largely overlapping content. Typically, stereo is based on the creation of a phantom centre, the latter referring to the psycho-acoustic phenomenon of a sound source appearing to emanate from a point between two speakers in a stereo configuration.

[0006] In many applications, it is desirable to generate an even more immersive and three-dimensional audio experience. For this purpose, a spatial audio mix is created in a 3D mixing environment. Typically, the mixing environment provides a virtual sound space, wherein virtual sound objects may be positioned and moved around. Upon creating a spatial audio mix, specific audio signals may be assigned to individual sound objects, and the information describing object positions is stored as metadata. Afterwards, a renderer is used to turn the spatial audio mix into a channel-based output, which fits the physical speaker layout used for playback.

[0007] In general, virtual sound objects in a spatial audio mix are treated as mono-objects, wherein every object is independent and represents a unique audio signal. In this, the use of available stereo sound, wherein the pair of stereo signals would need to be integrated in the 3D spatial mix, is often avoided, because of the risk of unwantedartefacts. When a left and right object with corresponding left and right stereo signal would be defined in the 3D virtual space, interference may occur between the two signals having partly overlapping content, wherein such interference may result in unwanted artefacts. Moreover, if one would assign the left and right stereo signal to two individual objects in a standard spatial mix, e.g., a Dolby Atmos mix, there is no guarantee that the phantom centre of the original stereo sound is still correctly represented. Because of the flexible output formats offered by spatial audio like Dolby Atmos, the phantom centre may shift depending on the selected output format, e.g., 5.1 Dolby Atmos versus 7.1.4. Moreover, defining two individual objects in a standard spatial mix may result in an unwanted variation of the sound level; for example, when the two objects are too close to each other, the level of their sum, i.e. the phantom centre, will be higher than originally intended.

[0008] Therefore, there is a need for a method that allows to use stereo sound in a 3D spatial audio mix, wherein the original stereo-wide experience is maintained while avoiding unwanted artefacts. Moreover, there is a need for a method that allows to use stereo as a true 3D source.

[0009] SUMMARY OF THE INVENTION

[0010] The inventors have surprisingly found that one or more of these problems can be solved by the present invention and (preferred) embodiments thereof.

[0011] According to a first aspect, the invention relates to a computer-implemented audio signal processing method for representing a stereo source in a spatial audio mix, the spatial audio mix comprising audio signals represented by sound objects positioned in a sound space. The method preferably comprises the steps of:

[0012] providing a first and a second input signal together defining an original stereo image;

[0013] positioning an aggregate sound object in the sound space, wherein the aggregate sound object comprises at least one centre object and at least one pair of ghost objects, said pair preferably comprising a first ghost object and second ghostobject positioned symmetrically with respect to a vertical plane through the centre object;

[0014] generating an audio signal for the at least one centre object, preferably comprising summing the first and second input signal;

[0015] generating audio signals for the at least one pair of ghost objects, preferably comprising differentiating the first and second input signal, and preferably upmixing the differentiated signal to obtain at least two upmixed signals; and, using the generated audio signals in a spatial audio mix, thereby obtaining a stereo source represented by the aggregate sound object.

[0016] In some preferred embodiments, the computer-implemented audio signal processing method comprises the steps of:

[0017] providing a first and a second input signal together defining an original stereo image;

[0018] defining an aggregate sound object, wherein

[0019] o the aggregate sound object comprises at least three sub-objects, the at least three sub-objects consisting of at least one centre object and at least one pair of ghost objects, said pair comprising a first ghost object and a second ghost object, and

[0020] o the first ghost object and the second ghost object are positioned symmetrically with respect to a vertical plane through the centre object; positioning the aggregate sound object in the sound space;

[0021] generating an audio signal for the at least one centre object, comprising summing the first and second input signal;

[0022] generating audio signals for the at least one pair of ghost objects (101, 102), comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and,

[0023] using the generated audio signals in a spatial audio mix, thereby obtaining a stereo source represented by the aggregate sound object.In some preferred embodiments, the first and second input signal originate from stereo sound; In some embodiments, the first and second input signal originate from two channels available in stereophonic sound, for example quadrophonic sound.

[0024] In some preferred embodiments, the method comprises performing Mid-Side processing, comprising:

[0025] summing the first and second input signal, thereby obtaining a mono signal; and, polarity reversing the first input signal, and summing the polarity reversed first input signal to the second input signal, thereby obtaining a difference signal; preferably wherein:

[0026] the mono signal is used forgenerating the audio signal for the at least one centre object; and,

[0027] the difference signal is used for generating the audio signals for the at least one pair of ghost objects, preferably wherein the difference signal is upmixed to obtain the at least two upmixed signals.

[0028] In some preferred embodiments, the at least two upmixed signals are obtained by filtering the differentiated signal, preferably wherein one or more filtering banks are used to apply frequency division to the differentiated signal.

[0029] In some preferred embodiments, the one or more filtering banks are configured to have linear phase response in the frequency domain.

[0030] In some preferred embodiments, the method further comprises:

[0031] applying an attenuation to the first and second input signal, resulting in an attenuated first and second input signal.

[0032] Preferably, the step of generating audio signals for the at least one pair of ghost objects comprises:

[0033] adding the attenuated first and second input signal to the respective upmixed signals.

[0034] This allows to add a sense of directivity to the stereo source.In some preferred embodiments, the signal strength of the attenuated first and second input signal is at least 10 dB to at most 25 dB lower than the signal strength of the respective upmixed signal, preferably at least 15 dB to at most 20 dB lower, for example 18 dB lower.

[0035] In some preferred embodiments, the vertical plane through the centre object extends in a height direction and depth direction, and / or the first ghost object is positioned at the same height position and at the same depth position as the second ghost object.

[0036] In some preferred embodiments, the vertical plane through the centre object extends in a height direction and depth direction with respect to a listener position in the sound space, and / or the vertical plane crosses the listener position.

[0037] In some preferred embodiments, the aggregate sound object comprises multiple pairs of ghost objects, wherein every pair comprises two ghost objects positioned symmetrically with respect to the vertical plane. In some preferred embodiments, the two ghost objects of a respective pair both are at the same height position. In some preferred embodiments, the ghost objects of different pairs are at different height positions. Thereby a three-dimensional stereo source in the spatial audio mix is obtained.

[0038] In some preferred embodiments, the differentiated signal is upmixed to multiple upmixed signals, the number of upmixed signals corresponding to the number of ghost objects.

[0039] In some preferred embodiments, the aggregate sound object is movable through the sound space. Preferably, upon changing the position of the at least one centre object, the relative position of the ghost objects with respect to the at least one centre object is maintained.In some preferred embodiments, the aggregate sound object is adaptable, by adapting the position of the at least one pair of ghost objects with respect to the at least one centre object. Preferably, upon adapting the height position and / or depth position and / or width position of the first ghost object and second ghost object, the first ghost object and second ghost object remain positioned symmetrically with respect to the vertical plane.

[0040] In some preferred embodiments, the mutual distance between the first ghost object and second ghost object is configurable, for example in view of obtaining a desired width of the stereo source in the spatial audio mix.

[0041] According to a second aspect, the present invention relates to an audio signal processing system for representing a stereo image in a spatial audio mix. (Preferred) embodiments of the first aspect are also (preferred) embodiments of the second aspect, and vice versa. Preferably, the system of the second aspect is configured to perform the method of the first aspect, and (preferred) embodiments thereof. Preferably, the method of the first aspect is performed with the system of the second aspect, and (preferred) embodiments thereof.

[0042] The audio signal processing system preferably comprises:

[0043] an input channel for receiving a first and a second input signal together defining an original stereo image;

[0044] a positioning module configured to position an aggregate sound object in the sound space, the aggregate sound object comprising at least one centre object and at least one pairof ghost objects, said pair preferably comprising a first ghost object and second ghost object positioned symmetrically with respect to a vertical plane through the centre object;

[0045] a signal processing module configured to:

[0046] o generate an audio signal for the at least one centre object, preferably by summing the first and second input signal; and,o generate audio signals for the at least one pair of ghost objects, comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and, a mixing module, configured to produce a spatial audio mix comprising the generated audio signals.

[0047] In some preferred embodiments, the audio signal processing system comprises:

[0048] an input channel for receiving a first and a second input signal together defining an original stereo image;

[0049] a positioning module configured for

[0050] o defining an aggregate sound object, wherein the aggregate sound object comprises at least three sub-objects, the at least three sub-objects consisting of at least one centre object and at least one pair of ghost objects, said pair comprising a first and second ghost object, and wherein the first ghost object and the second ghost object are positioned symmetrically with respect to a vertical plane through the centre object, and

[0051] o positioning the aggregate sound object in the sound space

[0052] a signal processing module configured to:

[0053] o generate an audio signal for the at least one centre object, comprising summing the first and second input signal; and,

[0054] o generate audio signals for the at least one pair of ghost objects (101, 102), comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and, a mixing module, configured to produce a spatial audio mix comprising the generated audio signals.

[0055] According to a further aspect, the invention relates to use of the method according to the first aspect, and (preferred) embodiments thereof, and / or of the system according to the second aspect, and (preferred) embodiments thereof.According to a further aspect, the invention relates to a computer program or computer program product having instructions which when executed by a computing device or system cause the computing device or system to perform the method according to the first aspect, and (preferred) embodiments thereof.

[0056] According to a further aspect, the invention relates to a computer readable medium having stored thereon the computer program or computer program product.

[0057] According to a further aspect, the invention relates to a data stream which is representative of the computer program or computer program product.

[0058] BRIEF DESCRIPTION OF THE FIGURES FIG. 1, FIG. 3, FIG. 4, and FIG. 5 each illustrate an aggregated sound object, according to preferred embodiments of the invention.

[0059] FIG. 2 illustrates a block scheme, schematically representing a processing method according to a preferred embodiment of the invention.

[0060] FIG. 6 illustrates an MS matrix, for use in a preferred embodiment of the invention. FIG. 7 illustrates an upmixing process, for use in a preferred embodiment of the invention.

[0061] DETAILED DESCRIPTION OF THE INVENTION

[0062] The present invention will be described with respect to particular embodiments, but the invention is not limited thereto, it is only limited by the claims. Any reference signs in the claims shall not be construed as limiting the scope thereof.

[0063] As used herein, the singular forms "a", "an", and "the" include both singular and plural referents unless the context clearly dictates otherwise.

[0064] The terms "comprising", "comprises" and "comprised of" as used herein are synonymous with "including", "includes" or "containing", "contains", and are inclusiveor 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.

[0065] 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.

[0066] 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 + / -!% 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.

[0067] 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.

[0068] 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.

[0069] 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.

[0070] In a first aspect, the invention relates to a computer-implemented audio signal processing method for representing a stereo source in a spatial audio mix, the spatial audio mix comprising audio signals represented by sound objects positioned in a sound space. The method preferably comprises the steps of:

[0071] providing a first and a second input signal together defining an original stereo image;

[0072] positioning an aggregate sound object in the sound space, wherein the aggregate sound object comprises at least one centre object and at least one pair of ghost objects, said pair preferably comprising a first ghost object and second ghost object positioned symmetrically with respect to a vertical plane through the centre object;

[0073] generating an audio signal for the at least one centre object, comprising summing the first and second input signal;

[0074] generating audio signals for the at least one pair of ghost objects, comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and,using the generated audio signals in a spatial audio mix, thereby obtaining a stereo source represented by the aggregate sound object.

[0075] As used herein, the term "spatial audio" refers to a sound format allowing to obtain a surround effect or immersive sound experience, by replicating the experience of being surrounded by several audio sources. It may equally be referred to as 3D sound, 3D audio, (virtual) surround sound, 360-degree sound, or Immersive Audio.

[0076] As used herein, the term "sound space" refers to a virtual space or virtual environment, wherein spatial audio is defined with respect to a virtual listener position. Typically, it concerns a three-dimensional virtual space, but a two-dimensional environment is also possible. It may equally be referred to as virtual sound space, virtual acoustic space, 3D space, 3D environment, immersive soundscape, immersive audio environment, or immersive layout.

[0077] As used herein, the term "spatial audio mix" refers to a set of audio signals and accompanying information for defining spatial audio. Typically, spatial audio is defined in a 3D mixing environment or spatial mix engine, thereby generating a spatial audio mix.

[0078] As used herein, the term "sound object" refers to a virtual object or element representing a sound source in the virtual sound space. In a mixing environment, a sound object may be positioned with respect to a listener position.

[0079] As used herein, the term "position of a sound object" refers to a single point in a virtual sound space, e.g., characterized by an X-, Y-, and Z coordinate. A sound object may have a fixed position overtime, or may be moved around, thus having a variable position over time. Sound objects can thus be positioned and moved to make them sound like they are coming from any direction around the (virtual) listener. Typically, the information describing the position of a sound object is stored as metadata. Moreover, upon creating a spatial audio mix, an audio signal is assigned to the respective sound objects, the audiosignal representing the sound played by the virtual sound source corresponding to the sound object. Accordingly, a spatial audio mix created in a mixing environment typically comprises audio signals corresponding to the various sound objects, together with metadata providing position information for the different sound objects. In particular embodiments, the spatial audio mix may be provided in a standard surround sound format, e.g., defined in Dolby Atmos mix with 5.1 or 7.1.4 output format.

[0080] The spatial audio mix is adapted for playback on a suitable system with any speaker layout. For example, the speaker setup may comprise multiple individual loudspeakers positioned in a 3D layout, e.g., in a home theatre, movie theatre or live setting. In other examples, the playback system may be provided as headphones, comprising a left and right speaker, or as a spatialized stereo system comprising two individual loudspeakers at separate spatial positions. Playback of the spatial audio mix may require a renderer to turn the spatial audio mix into a channel-based output which fits the physical speaker layout used for playback. In such case, the renderer processes the signal information and metadata of the spatial audio mix, for providing a reliable reproduction of the intended mix based on the physical speaker configuration. In certain examples, e.g., when headphones are used for playback, a binaural renderer or Head-Related transfer functions may be required, for creating a two-dimensional channel signal that, when listened to over headphones, makes the signal sound as an externalized sound, resulting in a 3D immersive sound experience.

[0081] The audio signal processing method comprises providing a first and a second input signal, wherein the input signals together define an original stereo image. This implies, that an input is provided, the input being an original stereo sound composed of at least one pair of stereo signals. Typically, the first and second stereo signal have a partly overlapping content, meaning that some content is the same in both the first and second input signal, and some content differs between the first and second input signal. In some types of stereo sound, e.g., originating from a stereo piano with the high notes on the right and the low notes on the left, the left and right stereo signal may have almost aunique content. In other types of stereo sound, e.g., an original voice with a stereo effect over it, the left and right signal may have a largely overlapping content.

[0082] The original stereo sound, which is used as an input in the present processing method, may originate from a recording, e.g., music being recorded in a two-channel stereo format, or may originate from a live output, e.g., from a stereo instrument, e.g., an electronic piano with stereo output. In some embodiments, the original stereo sound is adapted - and originally intended - for playback or reproduction on a two-channel stereo system, e.g., a configuration of two loudspeakers or stereo headphones. In such embodiments, the original sound may be composed by a single pair of two different signals. In other possible embodiments, an original sound may be available which is composed of multiple pairs of stereo signals, such as in quadrophonic sound.

[0083] The first and a second input signal are such that - together - they define an original stereo image. As used herein, the term "stereo image" refers to the spacious feeling or spatial audible experience created by stereo sound, due to the interaction between the two stereo signals and a listener's ears and brain. It may also be referred to as stereo effect or stereo-wide experience. Typically, stereo sound is created such that some phantom centre is experienced in the music, i.e. the experience of a central origin of the sound, within the stereo-wide image. As used herein, the term "phantom centre" refers to the psycho-acoustic phenomenon of a sound source appearing to emanate from a point between two speakers in a stereo configuration.

[0084] Moreover, typically, stereo sound is created in view of obtaining a stereo image with a desired stereo width, the latter referring to the perceived size of the acoustic image. For example, by manipulating the frequency and / or level balance of the stereo signals, the stereo image may become more wide or narrow, wherein a wide stereo image is used for songs intended to sound big, and a narrow stereo image is used for small intimate songs. In another example, reverb or spatial special effects may be used to create a wide stereo image.Finally, in some examples of stereo sound, the stereo image may include localization information. In such case, the stereo signals are defined such that a perception of location is created for one or more sound sources, i.e. some sense of directivity is present in the stereo sound. In some embodiments, the input stereo signals are adapted to create such directivity in the stereo image, e.g., by specific differences between the two signals or by manipulating specific sound levels. In other possible embodiments, no such directivity is present in the provided original stereo input.

[0085] The present audio signal processing method comprises positioning an aggregate sound object in the sound space. As used herein, the term "aggregate sound object" refers to a composite object, namely a set or group of sound objects, wherein the latter are mutually related, e.g., via specific restrictions on their mutual positions in the sound space. The constituent sound objects, together forming the aggregate object, may also be referred to as sub-objects. In the present method, the aggregate sound object is defined such that it comprises at least three sub-objects, namely at least one centre object and at least one pair of ghost objects, the pair comprising a first ghost object and a second ghost object. The sub-objects are mutually related due to their relative position with respect to one another. In particular, in the aggregate sound object, the first ghost object and second ghost object are positioned symmetrically with respect to a vertical plane through the centre object. As used herein, the terms "centre object" and "ghost object" merely refer to their mutual positions, without any other specific characteristics being deferred therefrom. As used herein, symmetrically positioned objects, with respect to a plane, implies that the distance of each of the respective objects to the plane is equal, wherein a deviation of the actual distance may apply of maximum 10%, preferably maximum 5%.

[0086] The pair of ghost objects comprises a first ghost object and second ghost object positioned symmetrically with respect to a vertical plane through the centre object. As used herein, a "vertical plane" refers to a YZ-plane, namely a plane extending according to a Y-direction and Z-direction. As used herein, a "Y-direction" refers to a depth direction in the virtual sound space, a Z-direction refers to a vertical or height directionin the virtual sound space, and an X-direction refers to a width direction in the virtual sound space. The X-, Y-, and Z-directions are mutually perpendicular and are schematically indicated in FIG. 1. A vertical plane through the centre object implies that the plane crosses the position of the centre object, or- in other words -that the position of the centre object lies in the vertical plane.

[0087] The positions of the sub-objects in the aggregate object are thus not completely independent, but restricted such that when a vertical YZ-plane is defined through the position of the centre object, the positions of the first ghost object and second ghost objects are mirrored with respect to the vertical plane. This implies that, measured along the X-direction, the distance between the first ghost object and the vertical plane through the centre object is equal to the distance between the second ghost object and the vertical plane through the centre object. Moreover, the first ghost object is positioned at the same height position and at the same depth position as the second ghost object. In possible embodiments, however, the depth position of the two ghost objects, measured along the Y-axis may differ from the depth position of the centre object, and the height position of the two ghost objects, measured along the Z-axis may differ from the height position of the centre object. In some particular embodiments, the vertical plane may run through the listener position, while in other embodiments the vertical plane does not run through the listener position. A particular embodiment of an aggregate sound object is shown in FIG. 1, wherein a first ghost object 101 and a second ghost object 102 have a symmetrical position with respect to a vertical plane 103 through a centre object 100.

[0088] The present audio signal processing method further comprises generating an audio signal for the at least one centre object and generating audio signals for the at least one pair of ghost objects. As used herein, the term "generating the respective signals" refers to processing the first and second input signal comprised in the input stereo sound, for obtaining signals which may be assigned to the respective sub-objects of the aggregate object when defining the spatial audio mix. This is schematically represented in FIG. 2 for a particular embodiment of the invention, wherein input signals S_input_l andS_input_2 are processed to generate an audio signal S_center for a centre object, an audio signal S_ghost_l for a first ghost object and an audio signal S_ghost_2 fora second ghost object.

[0089] The step of generating an audio signal for the at least one centre object preferably comprises summing the first and second input signal. In other words, an audio signal for the centre object is obtained based on summing the two input stereo signals. Apart from this summing step, in possible embodiments, other additional processing steps may be applied to obtain a signal for the centre object.

[0090] The step of generating audio signals for the at least one pair of ghost objects preferably comprises differentiating the first and second input signal. In other words, an audio signal for the first ghost object and for the second ghost object is obtained based on differentiating the two input stereo signals, resulting in a differentiated signal. The latter is indicated as S_diff in the embodiment of FIG. 2. Moreover, generating audio signals for the at least one pair of ghost objects comprises upmixing the differentiated signal to obtain at least two upmixed signals. As used herein, the terms "upmixing", "upmix", or "upmixed", refer to the process of generating additional audio signals from an available sound signal with fewer channels. In particular, in the present method, one differentiated signal is upmixed to obtain two or more signals. In other words, a one-channel differentiated signal is converted into two or more upmixed signals. An audio signal for the first ghost object and for the second ghost object is thus obtained based on differentiating the two input stereo signals, and based on upmixing the differentiated signal. Apart from these steps, in possible embodiments, other additional processing steps may be applied to obtain signals for the ghost objects.

[0091] The present audio signal processing method further comprises using the generated audio signals in a spatial audio mix. This implies that when a spatial audio mix is defined, e.g., in a mixing environment, the signals generated for the centre object, for the first ghost object and for the second ghost object, are assigned to the respective sub-objects of the aggregate sound object. Accordingly, a spatial audio mix is obtained, wherein thismix comprises at least one centre signal, assigned to the centre object, and at least two ghost signals, assigned to the ghost objects. Apart from these signals, the spatial audio mix may comprise metadata representing position information for the sub-objects of the aggregate object. Moreover, the spatial audio mix may comprise additional audio signals, assigned to other sound objects in the audio space than the aggregate sound object.

[0092] In the created spatial audio mix, the aggregate sound object represents a stereo source. In particular, the aggregated sound object represents a sound source in the spatial mix that allows to recreate the stereo image being present in the original stereo input sound.

[0093] Indeed, in the original stereo sound, the phantom centre is represented by the left and right signal equally. Accordingly, by summing the left and right input stereo signal, thereby creating a so-called centre mono, the obtained mono maximally contains the audio information responsible for creating the centre perception. By assigning this mono signal to the centre of the aggregate object in the spatial audio mix, an object containing the true centre of the stereo track is obtained. The latter is much more reliable in representing the correct centre experience, than when merely relying on the phantom centre in the original sound; if one would just place the original stereo sound as two objects in the spatial mix, the phantom centre may shift according to a specific output format, resulting in an incorrect centre perception.

[0094] Moreover, the differentiated signal, obtained by differentiating the left and right signals of the original stereo sound, contains all the information that actually differs between the left and right signals, also referred to as side information. In music production, the width or spacious feeling of stereo sound is defined by the amount of side information. In other words, the differentiated signal usually contains all the sounds that are used for obtaining a wide stereo image like e.g., reverb or spatial sound effects. Accordingly, by assigning the differentiated signal to a pair of ghost objects in the spatial mix, positioned symmetrically with respect to the centre object, the spacious side information available in the original stereo sound is correctly represented in the spatial audio mix.Consequently, the stereo-wide experience defined in the original stereo sound is maintained in the spatial audio mix.

[0095] Finally, by using the aggregate object in the spatial audio mix, instead of using just a left and right object, the risk of interference between signals with partly overlapping content is avoided. Thus, the present method allows to maintain the stereo image present in the original stereo input sound, while avoiding unwanted artefacts.

[0096] In some preferred embodiments, the first and second input signal originate from stereo sound, or from two channels available in stereophonic sound, for example quadrophonic sound. As used herein, the term "stereo sound" refers to two-channel audio, composed of one pair of stereo signals, and the term quadrophonic sound refers to four-channel audio. As used herein, the term "stereophonic sound" refers to multi-channel sound, composed of multiple audio signals for creating a multi-directional audible perspective. Both stereo sound and quadrophonic sound may be considered as types of stereophonic sound.

[0097] In the present audio signal processing method, an input is provided comprising a first and a second input signal. In an embodiment wherein an original stereo sound is provided as input, the first and second input signal correspond to the pair of signals composing the original stereo sound. In an embodiment wherein an original quadrophonic sound is provided as input, the first and second input signal correspond to one pair of signals available in the original quadrophonic sound. In such a case, the present processing method may be applied multiple times, wherein a first aggregate object is created corresponding to a first pair of input signals, and a second aggregate object is created corresponding to a second pair of input signals. Similarly, the present processing method may be applied to any type of stereophonic sound wherein pairs of input signals may be identified, provided that a selected pair of input signals is adapted for - together - defining a stereo image.In some preferred embodiments, the vertical plane through the centre object extends in a height direction Z and depth direction Y with respect to a listener position in the sound space. As used herein, the term "listener position" refers to a single point in a virtual sound space, e.g., characterized by an X-, Y-, and Z-coordinate, representing the position of a virtual person hearing the various sound sources defined in the spatial audio. For example, the X-, Y-, and Z-directions may be defined with respect to a listener position, wherein a virtual listener is placed on the Y-axis, and depth, width, and height of an object are measured with respect to the listener position, according to the Y-, X-, and Z-axis respectively. For example, the listener position may correspond to the origin of a coordinate system, being at the intersection of the X-, Y-, and Z-axis.

[0098] In some preferred embodiments, the vertical plane through the centre object crosses the position of the centre object and crosses the listener position. In other words, in such embodiments, the position of both the centre object and the virtual listener are lying in the vertical plane. As such, they define the position of the vertical plane in the sound space, the latter functioning as symmetry plane for placing the two ghost objects. The vertical plane crossing the listener position in the virtual sound space is optional, being preferred but not required in all possible embodiments.

[0099] This is schematically represented in FIG.3 and FIG.4, fora particular embodiment of the invention. A virtual listener 200 is positioned in the origin of the coordinate system, and depth, height, and width are measured with respect to the listener position, according to the Y-, Z-, and X-axis, respectively. A centre object is positioned in a YZ-plane through the origin, at a depth position 201 and a height position 301 with respect to the listener position 200. Two ghost objects 101, 102 are at a depth position 202 with respect to the listener position 200, wherein the depth position 202 of the ghost objects differs from the depth position 201 of the centre object. The two ghost objects 101, 102 are at a height position 302 with respect to the listener position 200, wherein the height position 302 of the ghost objects differs from the height position 301 of the centre object. The ghost objects 101, 102 are positioned symmetrically with respect to a vertical YZ planethrough the centre object 100 and through the listener position 200, wherein the width position 203 of the first ghost object 101 is equal to the width position 204 of the second ghost object 102, and the height position 301 of the first ghost object 101 is equal to the height position 302 of the second ghost object 102. As used herein, an equal width position of two objects refers to the distance between the respective objects and the YZ-plane being equal, wherein a deviation of the actual distance may apply of maximum 10%, preferably maximum 5%. As used herein, an equal height position of two objects refers to the distance between the respective objects and the XY-plane being equal, wherein a deviation of the actual distance may apply of maximum 10%, preferably maximum 5%. As used herein, an equal depth position of two objects refers to the distance between the respective objects and the XZ-plane being equal, wherein a deviation of the actual distance may apply of maximum 10%, preferably maximum 5%.

[0100] In some preferred embodiments, the method comprises performing Mid-Side processing. Preferably, the method comprises:

[0101] summing the first and second input signal, thereby obtaining a mono signal; and, polarity reversing the first input signal, and summing the polarity reversed first input signal to the second input signal, thereby obtaining a difference signal; preferably wherein:

[0102] the mono signal is used forgenerating the audio signal for the at least one centre object; and,

[0103] the difference signal is used for generating the audio signals for the at least one pair of ghost objects, wherein the difference signal is upmixed to obtain the at least two upmixed signals.

[0104] This implies that for processing the input signals, to generate signals for the aggregate object, a technique called Mid-Side processing is used, also referred to as MS processing MS encoding, or use of an MS matrix. An example of an MS matrix is illustrated in FIG.

[0105] 6. MS processing allows to manipulate the mid and side information available in stereo sound separately. In particular, the MS matrix may be applied to the original stereo sound in order to create a mono (summed) of the original left and right signal and adifference (sides) signal. The mono sum may be created by simply adding the right and left signal together. The difference signal may be created by subtracting the right signal from the left signal, or vice versa. The subtraction may be done by reversing the phase of the right signal, creating a negative right signal, and adding the reversed to the left signal. Alternatively, the subtraction may be done by reversing the phase of the left signal, creating a negative left signal, and adding the reversed to the right signal. As a result, one obtains a difference signal which contains the signals that are not identical in the left and right signal. These signals typically contain the spatial information of a stereo track. Sounds such as reverb information or extreme panoramic sounds are therefore present in the difference signal.

[0106] After obtaining the mono signal from Mid-Side processing the original stereo sound, it may be used for generating an audio signal for the centre object of the aggregate sound object. The mono signal may be used directly as the signal assigned to the centre object, or some additional processing step(s) may be applied to the mono signal before assigning it to the centre object. Moreover, after obtaining the difference signal from Mid-Side processing the original stereo sound, it may be upmixed, to obtain at least two upmixed signals. The upmixed signals may be used directly as audio signals for the two ghost objects, or some additional processing step(s) may be applied to the upmixed signals before assigning them to the respective ghost objects.

[0107] In some preferred embodiments, the at least two upmixed signals are obtained by filtering the differentiated signal, wherein one or more filtering banks are used to apply frequency division to the differentiated signal. This implies that the differentiated signal, or difference signal obtained from Mid-Side processing, is filtered, resulting in a first filtered signal, corresponding to the first upmixed signal, and a second filtered signal, corresponding to the second upmixed signal. The first filtered signal for example comprises a first range of frequencies, and the second filtered signal for example comprises a second range of frequencies. Upmixing the differentiated signal is thus performed by splitting the differentiated signal into two signals by dividing the frequencies. A particular example of a possible upmixing process is illustrated in FIG. 7.In some preferred embodiments, the one or more filtering banks are configured to have linear phase response in the frequency domain. Whereas traditional filters would cause phase shifting between the obtained filtered signals, thereby leading to overprocessed sound and sound artifacts, applying linear phase filtering ensures that phase shifts are spectrally consistent, thereby resulting in an increased sound quality. Thus, by using frequency division and linear phase processing, the obtained upmixed signals remain compatible and will couple acoustically.

[0108] Filtering may be done according to various possible embodiments. A particular embodiment of a possible filtering algorithm and filter implementation is disclosed in W02023006945A1. In this embodiment, according to a preferred example, a magnitude of -9db is used, a width of l / 3rd octave is used, and the following example frequencies may be used: 502Hz, 652,6Hz, 848,8Hz, 1102,9Hz, 1433,7Hz, 1863,8Hz, 2423Hz, 3149, 9H, 4094,9Hz, 5323,3Hz, 6920,3Hz, 8999,4Hz. As exemplified herein, the spacing of the frequencies is preferably l / 3rd octave. This directly relates to the filter width used. If the filter width becomes more narrow, the spacing of the frequencies are preferably adjusted to match the filter width used. However, filters being too small or too wide are preferably avoided. According to some preferred examples, each of the filtering banks (PF) is configured to have a width between l / 9thof an octave and an octave.

[0109] In some examples the filtering banks (PF) are configured to operate around filter subbands (PSB), which are configured to operate within an operating frequency range spanning from Fi.and Fu, wherein FL to Fu is preferably from 350 Hz to 20 kHz, preferably from 400 Hz to 10 kHz, and more preferably from 500 Hz to 9 kHz. Furthermore, said filtering sub-bands may be configured to extract at least 4 (PSB) sub-bands signals, preferably 8 (PSB) sub-bands signals, more preferably 16 (PSB) per filtering bank. Extracting a sufficient number of sub-band signals used to reconstruct the audio feed, results in an output feed expressing enhanced dynamic range and improved spatial resolution.The average person is unable to locate the area below 500Hz and above 9kHz in a space. Therefore, the processing used to divide the frequencies between Processing Slot A and Processing Slot B, is preferably only active in this region. When the filtering would rise above 9kHz the open sound associated with these frequencies would become unequally divided over the listening area resulting in an unsatisfying coverage of the "open sound". If the filtering would drop below 500Hz the summation of the low frequencies would be next to nothing, resulting in a poor sound across from the upmixed signal that has a warm feel because of this low frequency summation.

[0110] In some preferred examples, amplitude compensation is performed outside the operating band of the sub-band filtering, wherein such compensation preferably has an amplitude of from -9dB to -3dB, more preferably about -6dB. In some preferred examples, compensation filters are low and high shelf filters. In the case of the height layer, low shelf filters may become high pass filters.

[0111] Since there may be an overlap of the filters, there may be a general amplitude reduction in the processing frequency range. Therefore, compensation filters are preferably introduced above and below the processing range. These filters are preferably low and high shelf filters, preferably also linear phase type. When using -9dB magnitude with a l / 3rd width and l / 3rd spacing, a compensation filter of -6dB is preferred.

[0112] The frequency of the compensation filter is the border frequency of the processing range. As seen in the list above, exemplified border frequencies are 502Hz and 8999,4Hz. Since the compensation filters are set to this frequency, they will reduce the magnitude by 3dB at that frequency. Therefore, the magnitude of the border frequency filters used in the processing is preferably set to -6dB so that the result of the sum of both will be -9dB.

[0113] In some preferred embodiments, the method further comprises:

[0114] applying an attenuation to the first and second input signal, resulting in an attenuated first and second input signal.

[0115] In some preferred embodiments, the step of generating audio signals for the at least one pair of ghost objects comprises:adding the attenuated first and second input signal to respective upmixed signals.

[0116] This adds a sense of directivity to the stereo source.

[0117] This implies that the first and second upmixed signal, obtained from upmixing the differentiated signal are not directly used as the respective signals for the ghost objects, but some additional processing is applied first. In particular, an attenuated first respectively second input signal is added to the first respectively second upmixed signal for obtaining the audio signals for the pair of ghost objects. As used herein, the term "attenuating" refers to obtaining a reduction in signal strength. Attenuation may be applied by any suitable attenuation known in the art.

[0118] Accordingly, a (small) portion of the original stereo signals is added in the respective signals generated for the ghost objects. In this way, a sense of directivity is added in the sound generated by the aggregate object. Indeed, in some examples of stereo sound, the stereo image may include localization information, wherein the original stereo signals are defined such that a perception of location is created. For example, specific differences between the left and right signals may be introduced or sound levels may have been manipulated to obtain the sense of localization. However, when using the (upmixed) difference signal directly for the pair of ghost objects, specific left-right differences for localization are lost in the aggregate object. By adding a portion of the original left and right stereo signal again, including the left-right differences for localization, localization information is re-introduced in the sound produced by the aggregated object.

[0119] In this, it suffices that only a (small) portion of the original stereo signals is added, to obtain the sense of localization. Moreover, as only a small portion of the original stereo sound, typically having partly overlapping content, is introduced, the risk of signal interference leading to unwanted artefacts is reduced. Adding such sense of directivity in the aggregate sound object is in particular advantageous for those types of originalstereo sound wherein localization of sound sources is important, i.e. original stereo signals with specifically relevant left and right information.

[0120] In some preferred embodiments, the signal strength of the attenuated first and second input signal is at least 10 dB to at most 25 dB lower than the signal strength of the respective upmixed signal, preferably at least 15 dB to at most 20 dB lower, for example 18 dB lower.

[0121] In some preferred embodiments, the aggregate sound object may comprise multiple pairs of ghost objects, wherein every pair comprises two ghost objects positioned symmetrically with respect to the vertical plane. The two ghost objects of a respective pair may both be at the same height position. Moreover, the ghost objects of different pairs may be at different height positions. In this way, an additional height layer is created in the stereo source represented by the aggregate object, thereby obtaining a three-dimensional stereo source in the spatial audio mix.

[0122] In such embodiments wherein multiple pairs of ghost objects are defined, the differentiated signal is upmixed to multiple upmixed signals, wherein the number of upmixed signals corresponds to the number of ghost objects. For example, the differentiated signal may be upmixed to three or four channels.

[0123] FIG. 5 illustrates a particular example of an aggregated object having multiple height layers. The aggregate object comprises a centre object 100, a first pair of ghost objects 101, 102, and a second pair of ghost objects 401, 402. The ghost objects 101, 102 of the first layer are positioned symmetrically with respect to a vertical YZ-plane though the listener position, at a height position 301. The ghost objects 401, 402 of the second layer are also positioned symmetrically with respect to the vertical YZ-plane, at a height position 501, the latter being larger than the height position 301 of the first layer.

[0124] In some preferred embodiments, the aggregate sound object is movable through the sound space. Preferably upon changing the position of the at least one centre object,the relative position of the ghost objects with respect to the at least one centre object is maintained. Moving a sound object through the sound space implies that the corresponding position in the sound space changes overtime. For example, such varying position or followed trajectory may be defined in a mixing environment for creating spatial audio. In these embodiments, when the centre object is moved through the virtual sound space, the ghost objects may follow accordingly, such that their relative position with respect to the centre object is maintained. Thus, when moving the centre object, the ghost objects move as a group. In other words, the aggregated object is moved as a whole, while keeping its symmetrical configuration. In this way, a movable stereo source can be defined in the spatial audio mix, while maintaining the stereo image as originally defined by the input stereo sound.

[0125] In some preferred embodiments, the aggregate sound object is adaptable, by adapting the position of the at least one pair of ghost objects with respect to the at least one centre object. Preferably upon adapting the height position and / or depth position and / or width position of the first and second ghost object, the first and second ghost object remain positioned symmetrically with respect to the vertical plane. Adapting the aggregate object may occur upon defining or positioning the aggregate object, or it may concern an adaptation over time. As used herein, the term "adapting the aggregate object" refers to changing the relative positioning of the two ghost objects with respect to the centre object. For example, the height position of the two ghost objects relative to the height position of the centre object may be changed, and / or the depth position of the two ghost objects relative to the depth position of the centre object may be changed, and / or the width position of the two ghost objects relative to the width position of the centre object may be changed. In such embodiments, when adapting a position of the pair of ghost objects, the adaptation is preferably such that the symmetry of the ghost objects with respect to the vertical plane through the centre object is maintained. In otherwords, after adapting the height / depth / width position of the ghost objects, the ghost objects are still positioned symmetrically with respect to the vertical plane. By ensuring that the position of one ghost object remains symmetrical to the position of the other ghost object, the sense of stereo is always maintained.1

[0126] In some preferred embodiments, the mutual distance between the first and second ghost object is configurable, in view of obtaining a desired width of the stereo source in the spatial audio mix. This implies that based on some input, e.g., a configuration parameter, a desired width of the stereo image produced by the aggregate object may be defined. In other words, by changing the position of the ghost objects on the X-axis, the stereo sense may be widened. Such freedom to position the ghost objects, results in a creative freedom to choose how wide the stereo signal is represented.

[0127] In some embodiments, the aggregate sound object is movable, adaptable, or configurable. In some embodiments, the aggregated sound object is simultaneously movable, adaptable, and configurable.

[0128] In a second aspect, the present invention relates to an audio signal processing system for representing a stereo image in a spatial audio mix. (Preferred) embodiments of the first aspect are also (preferred) embodiments of the second aspect, and vice versa. Preferably, the system of the second aspect is configured to perform the method of the first aspect, and (preferred) embodiments thereof. Preferably, the method of the first aspect is performed with the system of the second aspect, and (preferred) embodiments thereof.

[0129] The audio signal processing system preferably comprises:

[0130] an input channel for receiving a first and a second input signal together defining an original stereo image;

[0131] a positioning module configured to position an aggregate sound object in the sound space, the aggregate sound object comprising at least one centre object and at least one pair of ghost objects, said pair preferably comprising a first and second ghost object positioned symmetrically with respect to a vertical plane through the centre object;

[0132] a signal processing module configured to:o generate an audio signal for the at least one centre object, preferably comprising summing the first and second input signal; and,

[0133] o generate audio signals for the at least one pair of ghost objects, preferably comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and, a mixing module, configured to produce a spatial audio mix comprising the generated audio signals.

[0134] In a further aspect, the present invention relates to use of the method according to the first aspect, and (preferred) embodiments thereof, and / or of the system according to the second aspect, and (preferred) embodiments thereof.

[0135] The present method may be used in (3D) mixing environment or a spatial mix engine, in particular in a 3D audio studio. It may also be used in a live setting. In some embodiments, the present method is used on a digital sound processor unit (DSP) or on a field-programmable gate array (FPGA).

[0136] 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.

[0137] The invention may be computer-implemented for studio mixing purposes. In this case, it could, for example, be provided as a plugin (VST) for a DAW system, or could form a part of a DAW system. In the live or broadcast sector, it could be implemented as a part of the processing in a live immersive mixing desk.

[0138] 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 maycomprise 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.

[0139] 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.

[0140] 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.

Claims

CLAIMS1. A computer-implemented audio signal processing method for representing a stereo source in a spatial audio mix, the spatial audio mix comprising audio signals represented by sound objects positioned in a sound space, the method comprising the steps of:providing a first and a second input signal together defining an original stereo image;defining an aggregate sound object, whereino the aggregate sound object comprises at least three sub-objects, the at least three sub-objects consisting of at least one centre object (100) and at least one pair of ghost objects, said pair comprising a first ghost object (101) and a second ghost object (102), ando the first ghost object (101) and the second ghost object (102) are positioned symmetrically with respect to a vertical plane (103) through the centre object (100);positioning the aggregate sound object in the sound space;generating an audio signal for the at least one centre object, comprising summing the first and second input signal;generating audio signals for the at least one pair of ghost objects (101, 102), comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and,using the generated audio signals in a spatial audio mix, thereby obtaining a stereo source represented by the aggregate sound object.

2. The method according to claim 1, wherein the first and second input signal originate from stereo sound, or originate from two channels available in stereophonic sound, for example quadrophonic sound.

3. The method according to any one of the preceding claims, wherein the method comprises performing Mid-Side processing, comprising:summing the first and second input signal, thereby obtaining a mono signal; polarity reversing the first input signal, and summing the polarity reversed first input signal to the second input signal, thereby obtaining a difference signal; and whereinthe mono signal is used forgenerating the audio signal for the at least one centre object (100); and,the difference signal is used for generating the audio signals for the at least one pair of ghost objects (101, 102), wherein the difference signal is upmixed to obtain the at least two upmixed signals.

4. The method according to any one of the preceding claims, wherein the at least two upmixed signals are obtained by filtering the differentiated signal, and wherein one or more filtering banks are used to apply frequency division to the differentiated signal.

5. The method according to claim 4, wherein the one or more filtering banks are configured to have linear phase response in the frequency domain.

6. The method according to any one of the preceding claims, wherein the method further comprises:applying an attenuation to the first and second input signal, resulting in an attenuated first and second input signal;and wherein the step of generating audio signals for the at least one pair of ghost objects (101, 102) comprises:adding the attenuated first and second input signal to the respective upmixed signals.

7. The method according to claim 6, wherein the signal strength of the attenuated first and second input signal is at least 10 dB to at most 25 dB lower than the signal strength of the respective upmixed signal, preferably at least 15 dB to at most 20 dB lower, for example 18 dB lower.

8. The method according to any one of the preceding claims, wherein the vertical plane (103) through the centre object (100) extends in a height direction and depth direction, and the first ghost object (101) is positioned at the same height position (301,302) and at the same depth position (201,202) as the second ghost object (102).

9. The method according to any one of the preceding claims, wherein the vertical plane (103) through the centre object (100) extends in a height direction and depth direction with respect to a listener position (200) in the sound space, and wherein the vertical plane crosses the listener position.

10. The method according to any one of the preceding claims, wherein the aggregate sound object comprises multiple pairs of ghost objects (101, 102; 401, 402), wherein:every pair comprises two ghost objects (101, 102; 401, 402) positioned symmetrically with respect to the vertical plane;the two ghost objects (101, 102; 401, 402) of a respective pair both are at the same height position (301; 501); andthe ghost objects (101, 102; 401, 402) of different pairs are at different height positions (301; 501),and wherein the differentiated signal is upmixed to multiple upmixed signals, the number of upmixed signals corresponding to the number of ghost objects (101, 102; 401, 402);thereby obtaining a three-dimensional stereo source in the spatial audio mix.

11. The method according to any one of the preceding claims, wherein the aggregate sound object is movable through the sound space, wherein upon changing the position of the at least one centre object (100), the relative position of the ghost objects (101, 102) with respect to the at least one centre object (100) is maintained.

12. The method according to any one of the preceding claims, wherein the aggregate sound object is adaptable, by adapting the position of the at least one pair of ghostobjects (101, 102) with respect to the at least one centre object (100), wherein upon adapting the height position and / or depth position and / or width position of the first and second ghost object (101, 102), the first and second ghost object (101, 102) remain positioned symmetrically with respect to the vertical plane.

13. The method according to any one of the preceding claims, wherein the mutual distance between the first and second ghost object (101, 102) is configurable, in view of obtaining a desired width of the stereo source in the spatial audio mix.

14. An audio signal processing system for representing a stereo image in a spatial audio mix, the spatial audio mix comprising audio signals represented by sound objects positioned in a sound space, the system comprising:an input channel for receiving a first and a second input signal together defining an original stereo image;a positioning module configured foro defining an aggregate sound object, wherein the aggregate sound object comprises at least three sub-objects, the at least three sub-objects consisting of at least one centre object (100) and at least one pair of ghost objects (101, 102), said pair comprising a first and second ghost object (101, 102), and wherein the first ghost object (101) and the second ghost object (102) are positioned symmetrically with respect to a vertical plane through the centre object (100), ando positioning the aggregate sound object in the sound spacea signal processing module configured to:o generate an audio signal for the at least one centre object (100), comprising summing the first and second input signal; and,o generate audio signals for the at least one pair of ghost objects (101, 102), comprising differentiating the first and second input signal, and upmixing the differentiated signal to obtain at least two upmixed signals; and, a mixing module, configured to produce a spatial audio mix comprising the generated audio signals.

15. The system according to claim 14, configured for carrying out the method according to any one of claims 1 to 13.

Images