Method, medium, and system encoding/decoding a multi-channel audio signal, and method medium, and system decoding a down-mixed signal to a 2-channel signal

Abstract

A method, medium, and system encoding and/or decoding a multi-channel audio signal, and a method, medium, and system decoding a signal down-mixed from multi-channels to a 2-channel signal. The method of encoding an audio signal may include generating spatial cues indicating directivity information of a virtual sound source generated by at least two channel sound sources among a plurality of channels, and down-mixing the plurality of channel signals. The method of decoding an audio signal may include receiving inputs of spatial cues indicating directivity information of a virtual sound source generated by at least two channel sound sources among sound sources of a plurality of channels, and a signal down-mixed from the plurality of channel signals, and restoring the down-mixed signal to a plurality of channel signals by using the spatial cues. According to such systems, media, and methods, a multi-channel audio signal can be accurately encoded and/or decoded regardless of frequency bands.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

These and/or other aspects and advantages of the invention will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 illustrates a system to encode a multi-channel signal into a down-mixed mono signal and the generation of decoded 2 channels from an up-mixing of the down-mixed mono signal, according to an embodiment of the present invention;

FIG. 2A illustrates a method of generating spatial cues indicating directivity information of virtual sound sources generated for a plurality of channels, according to an embodiment of the present invention;

FIG. 2B illustrates a one-to-two (OTT) encoder having inputs of 2 channels, and outputting channels directivity differences (CDDs) and the energy and direction information of a sound source, according to an embodiment of the present invention;

FIG. 3A illustrates a system encoding a multi-channel audio signal by using a 5-1-5 tree structure, according to an embodiment of the present invention;

FIG. 3B illustrating a channel layout explaining an encoding method for encoding a multi-channel audio signal, such as with the system illustrated in FIG. 3A, according to an embodiment of the present invention;

FIG. 4 illustrates a method of encoding 5.1 channels, according to an embodiment of the present invention;

FIG. 5 illustrates a system for decoding a multi-channel audio signal by using a 5-1-5 tree structure, according to an embodiment of the present invention;

FIG. 6 illustrates a method of decoding a mono signal down-mixed from 5.1 channels, according to an embodiment of the present invention;

FIG. 7 illustrates a decoding system outputting a 2-channels signal from a mono signal down-mixed from a plurality of channels, according to an embodiment of the present invention; and

FIG. 8 illustrates a decoding method of outputting a 2-channel signal from a mono signal down-mixed from a plurality of channels, according to an embodiment of the present invention.

DETAILED DESCRIPTION

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. Embodiments are described below to explain the present invention by referring to the figures.

FIG. 1 illustrates an end-to-end system showing an encoding of multi-channel signals into a down-mixed mono signal, and the generation of decoded 2 channels from an up-mixing of the down-mixed mono signal, according to an embodiment of the present invention.

The system may include a binaural decoder 120 including a decoding unit 130 and a 2-channel-synthesis unit 140, for example.

First, a plurality of channel signals may be input to the encoding unit 110, as the multi-channel signals. Referring to FIG. 1, an example of the plurality of channel signals, in a 5.1 channel system, may include a front center (C) channel, a front right (Rf) channel, a front left (Lf) channel, a rear right (Rs) channel, a rear left (Ls) channel, and a low frequency effect (LFE) channel, noting that embodiments of the present invention are not limited to the same, e.g., embodiments of the present invention may also be applied to a 7.1 channel system, only as an example.

Thus, the encoding unit 110 may generate spatial cues indicating frequency independent direction information of a virtual sound source generated by at least two channel sound sources among the sound sources of the plurality of channels, during the down-mixing of the plurality of channel signals to eventually generate the resultant down-mixed mono signal.

Below, for convenience of explanation, such spatial cues will also be referred to as channel directivity differences (CDDs), noting that alternative spatial cues with direction information may be available.

Thus, according to an embodiment of the present invention, the binaural decoder 120 may receive an input of such CDD spatial cues and the down-mixed mono signal, and by using the CDD spatial cues, up-mix the down-mixed mono signal to the multi-channel signals, and then further up-mix each multi-channel signal to synthesize a 2-channel signal.

Thus, here, the decoding unit 130 may receive the CDD spatial cues and the down-mixed mono signal, and by using the CDD spatial cues, restore a plurality of channel signals as the up-mixed multi-channel signals.

In an embodiment, and as noted above, in addition to the up-mixing of the multi-channel signals, the 2-channel-synthesis unit 140 may localize the up-mixed multi-channel signals, according to the positions of the respective channels, by using the CDD spatial cues and corresponding head related transfer functions (HRTFs), and thus, generate the 2-channel signal.

According to only an example, FIG. 2A illustrates a method of generating CDD spatial cues indicating directivity information of virtual sound sources generated by at least 2 channel sound sources among a plurality of channels, according to an embodiment of the present invention. According to one embodiment, such generation of the CDD spatial cues is performed during the down-mixing of input multi-channel signals by the encoder, with such CDD spatial cues being forwarded to the decoder for use in the decoding of the down-mixed mono signal.

Referring to FIG. 2A, as only convenience for explanation, only channel i 11 and channel j 12 are illustrated, noting that other channels (not shown) may also be distributed about the illustrated listener 13.

As illustrated, when a multi-channel audio signal is encoded, different magnitudes of energy of respective channels (channel i 11, channel j 12, and other channels) are distributed at a given point in time. In this case, assuming that other channels, other than channels l 11 and j 12, are not considered and a virtual sound source x 14 is generated only by the sound source of channel i 11 and the sound source of channel j 12, the energy of the virtual sound source x 14 can be considered to be the sum of the energy of channel i 11 and the energy of channel j 12, as in the below Equation 1.

W _i ² +W _j ² =W _x ²   Equation 1

Here, Wi2 is the energy of channel i, Wj2 is the energy of channel j, and Wx2 is the energy of channel x.

If both sides of Equation 1 are divided by Wx2, the result is the below Equation 2.

CDD _xi ² +CDD _xj ²=1   Equation 2

Here, CDD_xi=W_i²/W_x², and CDD_xj=W_j²/W_x².

Meanwhile, relationships of CDD_xi, CDD_xj, and directivity information of channel i 21, channel j 22, and virtual sound source x 24 may be represented by the below Equation 3.

$\begin{array}{cc}\text{Equation}3\text{:}& \\ \frac{\tan \phi }{\tan \theta }=\frac{{\mathrm{CDD}}_{\mathrm{xi}}-{\mathrm{CDD}}_{\mathrm{xj}}}{{\mathrm{CDD}}_{\mathrm{xi}}+{\mathrm{CDD}}_{\mathrm{xj}}}& \end{array}$

Here, θ represents directivity information of a channel and the angle between each channel and a plane bisecting the channel and a neighboring channel. Since the channel layout may have already been determined when a multi-channel audio signal is encoded, the directivity information of the channel may also be a predetermined value. Further, φ represents directivity information of a virtual sound source, and the angle between the virtual sound source x 14 and the bisecting plane, for example. As can be observed from Equation 3, CDDxi and CDDxj indicate the directivity information of the virtual sound source x 14 formed by the two channels i 11 and j 12.

Thus, in a process of generating a CDD, according to an embodiment of the present invention, the energy Wx2 of the virtual sound source x 14, CDDxi, and CDDxj may be obtained through Equations 1 and 2, and the directivity information of the virtual sound source x 14 may be obtained through Equation 3.

Here, based on the illustrated technique shown in FIG. 2A, each or either of channel i 11 and channel j 12 could also be virtual sound sources. For example, assuming that a virtual sound source y (not shown) is generated from two channels, e.g., other than channels i 11 and j 12, then, another virtual sound source z (not shown) may be generated from the generated virtual sound source x 14 and the generated virtual sound source y. In this case, CDDzx and CDDzy may be obtained along with energy and directivity information φ of the virtual sound sources.

FIG. 2B illustrates a one-to-two (OTT) encoder, having inputs of two separate channels, outputting CDD spatial cues, the energy of a virtual sound source, and directivity information, according to an embodiment of the present invention. Such OTT encoder modules may be repeatedly used for performing sequenced down-mixing to eventually generate the down-mixed mono signal, for example, noting that, upon each down-mixing, respective CDD spatial cues, energy, and directivity information may also be generated.

Here, referring to FIG. 2B, the OTT encoder 17 may, thus, receive input signals of two channels i and j, and output CDDxi, CDDxj, the energy Wx of a virtual sound source, and directivity information φ, for example. In addition, such a generated virtual sound source may also be input to another such OTT encoder 17.

FIG. 3A illustrates a system encoding a multi-channel audio signal by using a 5-1-5 tree structure, according to an embodiment of the present invention, briefly noting that alternative tree structures are equally available. FIG. 3B similarly illustrates a channel layout for explaining an encoding method for encoding a multi-channel audio signal, such as with the system illustrated in FIG. 3A, according to an embodiment of the present invention. FIG. 4 further illustrates a method of encoding 5.1 channels, according to an embodiment of the present invention. Such a method will now be explained with reference to FIGS. 3A and 3B, noting that such references should not be limited to the same. Such methods should also not be construed as being dependent on the referenced tree structure of FIG. 3A nor the illustrated directional channel layout of FIG. 3B.

In operation 310, a first OTT encoder 250 may receive inputs of the Lf channel and the Ls channel, e.g., corresponding to a plurality of available channel signals with determined direction information, generate CDD1Lf and CDD1Ls, and calculate the energy and directivity information of a first virtual sound source 210, as shown in FIG. 3B. In CDD1Lf, and CDD1Ls, the subscript 1 represents the virtual sound source, and Lf and Ls represent the front left channel (Lf) and rear left (Ls) channel, respectively. More specifically, by using the energies of the Lf channel and the Ls channel, the energy of the first virtual sound 210 and spatial cues CDD1Lf and CDD1Ls may be generated, and by using CDD1Lf, CDD1Ls, and directivity information of Lf and Ls channels, the directivity information of the first virtual sound source 210 may, thus, be calculated.

In operation 320, a second OTT encoder 255 may receive inputs of the Rf channel and the Rs channel, generate CDD2Rf and CDD2Rs, and calculate the energy and directivity information of a second virtual sound source 220.

In operation 330, a third OTT encoder 260 may receive inputs of the C channel and the LFE channel, generates CDD3C and CDD3LFE, and calculate the energy and directivity information of a third virtual sound source 230.

Further, in operation 340, a fourth OTT encoder 265 may receive inputs of the first virtual sound source 210 and the second virtual sound source 220, for example. Here, referring back to FIGS. 2A and 2B, operation 340 may be considered as corresponding to the case where the channel i 11 and the channel j 12 are replaced by the first virtual sound source 210 and the second virtual sound source 220, respectively. In operation 340, by using the energies of the first virtual sound source 210 and the second virtual sound source 220, the energy of a fourth virtual sound source 240 and CDD41 and CDD42 may be generated, and by using CDD41, CDD42, and the directivity information of the first virtual sound source 210 and the second sound source 220, the directivity information of the fourth virtual sound source 240 may be calculated.

In operation 350, a fifth OTT encoder 270 may receive inputs of the third virtual sound source 230 and the fourth virtual sound source 240, generate CDDm4 and CDDm3, and output a corresponding down-mixed mono signal, i.e., down-mixed from 5.1-channel signals. In such a method of encoding 5.1 channels, according to this embodiment of the present invention illustrated in FIG. 4, 5.1-channel signals can be down-mixed through operations 310 through 350, again noting that the reference to such a 5.1 channel system is only an example.

In operation 360, a multiplexing unit (not shown) generates and outputs a bitstream, including CDDs and the down-mixed mono signal.

FIG. 5 illustrates a system decoding a multi-channel audio signal by using a 5-1-5 tree structure, according to an embodiment of the present invention. Similarly, FIG. 6 illustrates a method of decoding a down-mixed mono signal, e.g., down-mixed from 5.1 channels, according to an embodiment of the present invention, and will now be explained with reference to FIG. 5, noting that such references should not be limited to the same. Such methods should also not be construed as being dependent on the referenced tree structure of FIG. 5.

In operation 505, a demultiplexing unit (not shown) may receive an input of an audio bitstream, including a down-mixed mono signal for multi-channel signals and CDDs, and may proceed to separate/parse the bitstream for the down-mixed mono signal and the CDDs.

In operation 510, a fifth OTT decoder 410 may restore the down-mixed mono signal to a down-mixed third virtual sound source and a down-mixed fourth virtual sound source, by using CDDm4 and CDDm3, for example

In operation 520, a fourth OTT decoder 420 may further restore the down-mixed fourth virtual sound source to a down-mixed first virtual sound source and a down-mixed second virtual sound source, by using CDD41 and CDD42, for example

In operation 530, a first OTT decoder 430 may restore the down-mixed first virtual sound source to an Lf channel and an Ls channel, by using CDDiLf and CDD1Ls, for example

In operation 540, a second OTT Decoder 440 may restore the down-mixed second virtual sound source to an Rf channel and an Rs channel, by using CDD2Rf and CDD2Rs, for example

In operation 550, a third OTT decoder 450 may restore the down-mixed third virtual sound source to a C channel and an LFE channel, by using CDD3C and CDD3LFE, again as examples.

Here, the Lf, Ls, Rf, Rs, C, and LFE channel signals, output by such a system for decoding a multi-channel audio signal illustrated in FIG. 5, may be represented by the below Equations 4 through 9.

Lf=CDD _m4 CDD ₄₁ CDD _1Lf m   Equation 4

Ls=CDD _m4 CDD ₄₁ CDD _1ILs m   Equation 5

Rf=CDD _m4 CDD ₄₂ CDD _2Rf m   Equation 6

Rs=CDD _m4 CDD ₄₂ CDD _2Rs m   Equation 7

C=CDD _m3 CDD _3c m.   Equation 8

LFE=CDD _m3 CDD _3LFE m   Equation 9

FIG. 7 illustrates a decoding system to generate a 2-channels signal from a down-mixed mono signal for multi-channel signals, according to an embodiment of the present invention.

Referring to FIG. 7, as an example of such multi-channel signals, e.g., in a 5.1 channel system, such channel signals may include C, Rf, Lf, Rs, Ls, and LFE channels. Here, it is again noted that embodiments of the present invention are not limited to such a system, e.g., embodiments of the present invention may be applicable to a 7.1 channel system.

Referring to FIG. 7, the decoding system may include of a time/frequency transform unit 710, a decoding unit 720, a 2-channel-synthesis unit 730, an HRTF generation unit 750, a reference HRTF DB 760, a first frequency/time transform unit 770, and a second frequency/time transform unit 780, for example.

Here, the 2-channel-synthesis unit 730 may further include sound localization units 731 through 740, a right channel mixing unit 742, and a left channel mixing unit 743, for example.

The time/frequency transform unit 710 may receive an input of the down-mixed mono signal for multi-channel signals, transform the mono signal into the frequency domain, and output the same as a respective frequency domain signal.

The decoding unit 720 may receive respective CDD spatial cues indicating directivity information of the respective virtual sound sources, e.g., generated by at least two channel sound sources among the sound sources of the multi-channels, and the frequency domain down-mixed mono signal, and restore the frequency domain down-mixed mono signal to Lf, Ls, Rf, Rs, C and LFE channel signals, by using the CDD spatial cues.

In FIG. 7, the HRTF DB 760 may store a set of HRTFs corresponding to any one channel, for example, of the Lf, Ls, Rf, Rs, and C channels, also as an example. Hereinafter, the HRTF stored in the HRTF DB 760 will be referred to as the reference HRTF. In FIG. 7, the HRTF DB 760, thus, may store a set of HRTFs corresponding to the Lf channel, and in an example case, a right HRTF (HRTFR,Lf) and a left HRTF (HRTFL,Lf).

The HRTF generation unit 750 may further receive the CDD spatial cues and HRTFs stored in the HRTF DB 760, and by using the CDD spatial cues and the HRTFs, generate HRTFs corresponding to other channels, i.e., Ls, Rf, Rs, and C channels, for example.

The HRTF generation unit 750 will now be explained in greater detail with reference to the aforementioned Equations 4 through 9. As can be observed from Equations 4 through 9, each channel signal output from the decoding unit 720 may be in a form in which the down-mixed mono signal m is multiplied by respective CDD spatial cues.

In an embodiment, the HRTF generation unit 750 may assign a weighting to a reference HRTF, with the weighting being a ratio of the product of CDD spatial cues corresponding to the channel of the reference HRTF, to the product of CDD spatial cues corresponding to the channel of an HRTF desired to be generated, among the products multiplied to the down-mixed mono signal in Equations 4 through 9. Thus, the HRTF generation unit 750 may generate the HRTF corresponding to the another channel other than the reference HRTF. That is, by convoluting the ratio of the products of the CDD spatial cues and the reference HRTF, a HRTF corresponding to the other channel, other than the reference HRTF, may be generated.

For example, in Equation 4, the Lf channel signal, corresponding to the reference HRTF, may be in a form in which the down-mixed mono signal m is multiplied by CDDm4CDD41CDD1Lf. Meanwhile, in Equation 7, the Rs channel signal may be in a form in which the down-mixed mono signal m is multiplied by CDDm4CDD42CDD2Rs. In this case, the HRTF corresponding to the Rs channel may thus be generated by assigning a weight of

$\frac{{\mathrm{CDD}}_{m4}{\mathrm{CDD}}_{42}{\mathrm{CDD}}_{2\mathrm{Rs}}}{{\mathrm{CDD}}_{m4}{\mathrm{CDD}}_{41}{\mathrm{CDD}}_{1\mathrm{Lf}}},$

to the HRTF of the Lf channel, which is the reference HRTF.

The 2-channel-synthesis unit 730 may, thus, receive an input of an HRTF corresponding to each channel from the reference HRTF DB 760 and the HRTF generation unit 750, for example.

In an embodiment, the sound localization units 731 through 740, included in the 2-channel-synthesis unit 730, may further localize channel signals to the positions of the respective channels, by using a respective HRTF, and generate the localized channel signals. Since the reference HRTF is that of the Lf channel in FIG. 7, the Lf channel sound localization units 731 and 732 may receive the HRTF from the reference HRTF DB 760, and the sound localization units 733 through 740, for channels other than the Lf channel, may receive inputs of HRTFs from the HRTF generation unit 750.

As illustrated, the right channel mixing unit 742 may then mix signals output from the right channel sound localization units 731, 733, 735, 737, and 739, and the left channel mixing unit 743 may mix signals output from the left channel sound localization units 732, 734, 736, 738, and 740.

The first frequency/time transform unit 770 may further receive an input of the signal mixed in the right channel mixing unit 742, transform the signal to a time domain signal, and output the right channel signal, thereby achieving a synthesizing of the right channel signal.

Similarly, the second frequency/time transform unit 780 may receive an input of the signal mixed in the left channel mixing unit 743, transform the signal to a time domain signal, and output the left channel signal, again thereby achieving a synthesizing of the left channel signal.

FIG. 8 illustrates a decoding method for generating a 2-channel signal from a down-mixed mono signal for multi-channel, according to an embodiment of the present invention. In one embodiment, the decoding method may be performed in a time series in a decoding system, such as that illustrated in FIG. 7. Here, though the decoding system of FIG. 7 may be referenced below as an example of the operations of FIG. 8, embodiments of the present invention should not be limited to the same. In addition, embodiments of the present invention may further include features represented/performed by the elements shown in FIG. 7, even is not particularly referenced below.

In operation 810, as an example, the time/frequency transform unit 710 may receive a down-mixed mono signal for multi-channels, and transform the down-mixed mono signal to a respective frequency domain signal.

In operation 820, the decoding unit 720 and the HRTF generation unit 750, for example, may receive CDD spatial cues indicating directivity information of a virtual sound source generated by at least two channel sound sources, among sound sources for the multi-channels.

In operation 830, the decoding unit 720, for example, may restore the frequency domain down-mixed mono signal to respective multi-channel signals, by using the CDD spatial cues.

In operation 840, the HRTF generation unit 750 may receive an HRTF corresponding to a predetermined channel, among the multi-channels, e.g., from the reference HRTF DB 760, and by using the input HRTF and the CDD spatial cues, the HRTF generation unit 750 may generate an HRTF corresponding to a channel other than the predetermined channel.

In operation 850, the 2-channel-synthesis unit 730 may then localize the decoded multi-channel signals to respective positions, by using the HRTF corresponding to the predetermined channel and the generated HRTFs, thereby generating a 2-channel signal.

In operation 860, the first frequency/time transform unit 770 and the second frequency/time transform unit 780 may transform the 2-channel signal to time domain signals.

Thus, according to an embodiment of the present invention, information spatial cues indicating the directivity information of virtual sound sources may be generated for multi-channels and a corresponding down-mixed mono multi-channel audio signal may be encoded and/or decoded.

Since such directivity information of virtual sound sources is determined according to information of channel layouts and is not dependent on frequencies of the channel signals, a multi-channel audio signal can be accurately encoded and/or decoded irrespective frequency regions.

In addition to the above described embodiments, embodiments of the present invention can also be implemented through computer readable code/instructions in/on a medium, e.g., a computer readable medium, to control at least one processing element to implement any above described embodiment. The medium can correspond to any medium/media permitting the storing and/or transmission of the computer readable code.

The computer readable code can be recorded/transferred on a medium in a variety of ways, with examples of the medium including magnetic storage media (e.g., ROM, floppy disks, hard disks, etc.), optical recording media (e.g., CD-ROMs, or DVDs), and storage/transmission media such as carrier waves, as well as through the Internet, for example. Here, the medium may further be a signal, such as a resultant signal or bitstream, according to embodiments of the present invention. The media may also be a distributed network, so that the computer readable code is stored/transferred and executed in a distributed fashion. Still further, as only an example, the processing element could include a processor or a computer processor, and processing elements may be distributed and/or included in a single device.

Although a few embodiments of the present invention have been shown and described, it would be appreciated by those skilled in the art that changes may be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the claims and their equivalents.

Claims

1. A method of decoding multi-channel audio signals, comprising: obtaining spatial cues at least indicating frequency independent directivity information for a virtual sound source generated from at least two sound sources among sound sources for a plurality of channels, and a down-mixed signal representing an encoding of the multi-channel audio signals; andrestoring the down-mixed signal to the plurality of channel signals by using the spatial cues.

2. The method of claim 1, wherein spatial cues for the virtual sound source generated from the at least two sound sources are generated based on corresponding energies of each of the at least two sound sources and an energy of the virtual sound source.

3. The method of claim 1, wherein the directivity information for the virtual sound source is directivity information calculated by using corresponding spatial cues and respective directivity information for each of the at least two sound sources.

4. The method of claim 1, wherein the restoring of the down-mixed signal to the plurality of channel signals by using the spatial cues comprises: restoring the down-mixed signal to a first virtual sound source and a second virtual sound source by using corresponding spatial cues; andrestoring the first virtual sound source to a third virtual sound source and a fourth virtual sound source by using other corresponding spatial cues.

5. The method of claim 4, wherein the restoring of the down-mixed signal to the plurality of channel signals by using the spatial cues further comprises restoring at least one of the first virtual sound source, second virtual sound sources, third virtual sound sources, and fourth virtual sound sources selectively to two channel signals among the plurality of channel signals by using additional corresponding spatial cues.

6. The method of claim 1, wherein in the obtaining of the spatial cues and the down-mixed signal, the spatial cues and the down-mixed signal are obtained from a parsing of a received bitstream.

7. The method of claim 1, wherein, in the generation of the virtual sound source in an encoder generating the down-mixed signal, the at least two sound sources comprise two sound sources corresponding to respective channels of the plurality of channels or two virtual sound sources each with directivity information different from directions corresponding to the plurality of channels.

8. At least one medium comprising computer readable code to control at least one processing element to implement the method of claim 1.

9. A method of encoding a multi-channel audio signal, comprising: generating spatial cues at least indicating frequency independent directivity information for a virtual sound source generated from at least two sound sources among sound sources for a plurality of channels;down-mixing a plurality of channel signals to a down-mixed signal through at least one operation of the generating of the spatial cues for at least one generation of a respective virtual sound source; andoutputting the down-mixed signal and generated spatial cues.

10. The method of claim 9, wherein, in the generation of the virtual sound source, the at least two sound sources comprise two sound sources corresponding to respective channels of the plurality of channels or two virtual sound sources each with directivity information different from directions corresponding to the plurality of channels.

11. The method of claim 9, wherein, in the generating of the spatial cues for the virtual sound source generated from the at least two sound sources, the spatial cues are generated based on corresponding energies of each of the at least two sound sources and an energy of the virtual sound source.

12. The method of claim 9, wherein the directivity information for the virtual sound source is calculated by using generated spatial cues and respective directivity information for each of the at least two sound sources.

13. The method of claim 9, wherein the generating of the spatial cues comprises: generating a first spatial cue indicating directivity information of a first virtual sound source generated from predetermined two sound sources, and calculating the directivity information of the first virtual sound source by using the first spatial cue and respective directivity information of each of the predetermined two sound sources; andgenerating a second spatial cue indicating directivity information of a second virtual sound source generated from other predetermined two sound sources, other than the predetermined two channels and calculating the directivity information of the second virtual sound source by using the second spatial cue and respective directivity information of each of the other predetermined two sound sources.

14. The method of claim 13, wherein the generating of the spatial cues further comprises generating a third spatial cue indicating directivity information of a third virtual sound source generated from the first and second virtual sound sources, and generating the directivity information of the third virtual sound source by using the third spatial cue and the directivity information of the first virtual sound source and the directivity information of the second virtual sound source.

15. The method of claim 9, wherein in the outputting of the down-mixed signal and the generated spatial cues, the down-mixed signal and the generated spatial cues are encoded into a bitstream.

16. At least one medium comprising computer readable code to control at least one processing element to implement the method of claim 9.

17. A method of decoding a down-mixed signal to a 2-channel signal, the method comprising: restoring the down-mixed signal to a plurality of channel signals by using spatial cues at least indicating frequency independent directivity information of at least one virtual sound source generated from at least two sound sources among sound sources for a plurality of channels; andlocalizing each of the plurality of channel signals to corresponding positions of respective channels based on a select 2-channel signal, and mixing the localized plurality of channel signals to generate the select 2-channel signal.

18. The method of claim 17, wherein, in the localizing of each of the plurality of channel signals, localizing is performed by using respective head related transfer functions (HRTFs).

19. The method of claim 18, further comprising generating select respective HRTFs corresponding to a channel other than a predetermined channel among the plurality of channels, by using a predetermined channel HRTF corresponding to the predetermined channel and respective spatial cues, wherein, when localizing a restored channel signal corresponding to the predetermined channel, the localizing is performed by using the predetermined HRTF corresponding to the predetermined channel.

20. The method of claim 19, wherein, in the generating of the respective HRTFs, spatial cues and the predetermined channel HRTF are convoluted to generate the respective HRTFs corresponding to the channel other than the predetermined channel.

21. The method of claim 19, wherein the predetermined channel is one of the select 2-channel signal.

22. The method of claim 17, further comprising: transforming the down-mixed signal into a frequency domain signal; andtransforming the select 2-channel signal into a time domain signal.

23. At least one medium comprising computer readable code to control at least one processing element to implement the method of claim 17.

24. A system decoding a multi-channel audio signal, comprising: a first decoder to decode a first virtual sound source into a first two sound sources among sound sources for a plurality of channels by using a first spatial cue; anda second decoder to decode a second virtual sound source into a second two sound sources, other than the first two sound sources, among the sound sources for the plurality of channels by using a second spatial cue,wherein the first spatial cue indicates frequency independent directivity information for the first virtual sound source, and the second spatial cue indicates frequency independent directivity information for the second virtual sound source.

25. A system encoding a multi-channel audio signal comprising: a first encoder to generate a first spatial cue indicating frequency independent directivity information of a first virtual sound source generated from a first two sound sources among sound sources for a plurality of channels, and to calculate the directivity information of the first virtual sound source by using the first spatial cue and respective directivity information of the first two sound sources; anda second encoder to generate a second spatial cue indicating frequency independent directivity information of a second virtual sound source generated from a second two sound sources, other than the first two sound sources, among the sound sources for the plurality of channels, and to calculates the directivity information of the second virtual sound source by using the second spatial cue and respective directivity information of the second two sound sources.

26. A system decoding a down-mixed signal, down-mixed from a plurality of channel signals to a 2-channel signal, the system comprising: a decoding unit to restore the down-mixed signal to the plurality of channel signals by using spatial cues at least indicating frequency independent directivity information of at least one virtual sound source generated from at least two sound sources among sound sources for a plurality of channels;an HRTF generation unit to generate HRTFs corresponding to a channel other than a predetermined channel among the plurality of channels based on a predetermined HRTF corresponding to the predetermined channel and the spatial cues; anda 2-channel-synthesis unit to localize the plurality of channel signals to corresponding positions of respective channels based on a select 2-channel signal by using the predetermined HRTF corresponding to the predetermined channel and the generated HRTFs, and mixing the localized plurality of channel signals to generate the select 2-channel signal.