TW200921643A - A method of merging at least two input object-oriented audio parameter streams into an output object-oriented audio parameter stream - Google Patents

Abstract

A method of merging at least two input object-oriented audio parameter streams into an output object-oriented audio parameter stream is disclosed. Each object-oriented audio parameter stream comprises object-oriented audio parameter values. Said object-oriented audio parameter values represent statistical properties of audio objects as a function of time. Said method comprises the following step. First, calculating a synchronized stream for each said input object-oriented audio parameter stream takes place. Said synchronized stream has object-oriented parameter values at predetermined temporal positions. Said predetermined temporal positions are the same for all synchronized streams. For each synchronized stream said object-oriented parameter values at the predetermined temporal positions are calculated by means of interpolating of the object-oriented parameter values of the corresponding input object-oriented audio parameter stream. Second, creating of the output object-oriented audio parameter stream is performed. Said output object-oriented audio parameter stream has object-oriented parameter values at the predetermined temporal position obtained by combining the object-oriented audio parameter values of the synchronized streams at said same predetermined temporal position.

Description

200921643 IX. Description of invention: [Technical field to which the invention belongs]

[Prior Art]

Hand-based spatial audio coding based on objects. Within the framework of the technical attraction of the work of the first object, a working group has already taken place. The goal of the working group is to "explore new technologies and reuse of existing MPEG surround components and techniques for efficiently encoding multiple sound sources or objects to multiple downmix channels and corresponding spatial parameters at bit rates." In other words, The goal is to encode a plurality of audio objects in a limited set of downmix channels with corresponding parameters. On the decoder side, the user can interact with the content by, for example, reconfiguring the individual audio objects. Implemented in an object-oriented decoder. This is accomplished by including a translation step after the decoding process. The translation step is combined with the decoding process as a separate processing step to eliminate the need to determine a single object. For loudspeaker playback, The combination is described in Faller, C., "parametfie joint-coding of audio sources", 'The 120th AES Symposium Conference, Paris, France, May 2006. An effective combination of headphone playback, a decoding and head related transfer function processing is described in Breebaart, J., Herre, J. 'Villemoes, L. 'Jin, C. 'Kjorling, K. ' Plogsties 132374.doc 200921643 J · KoPPens, J. (2006) "Multi-channel goes mobile: MPEG Surround binaural rendering" ' Record of the 29th AES Symposium, Seoul, South Korea. When using the object-oriented approach, as described above, in teleconferencing, multiple object-oriented audio streams need to be combined, each audio stream containing its own downmix and object-oriented audio parameter streams originating from different remote ends. As a result, a single object-oriented day stream produces 'which can be further decoded by the object-oriented decoder. Generally, the down-mixed audio stream uses a frame-based structure. The object-oriented parameters provided by the object-oriented, A-code form a stream whose 乂 time function reflects the statistical properties of the audio objects. Therefore, the object-oriented audio parameters are valid for a particular frame or even a portion of a frame. In teleconferencing applications, the frames used by the plurality of object-oriented audio encoders at the far end are less likely to be ideally aligned in time. In addition, ^^ audio objects may result in different framed and different object-oriented audio parameter positions' because the position of the parameters is better determined by the content. - and the object-oriented audio streams (the audio streams consisting of a plurality of downmixes and corresponding object-oriented audio parameter streams) will require (4) the objects to be directed to at least one of the audio streams to facilitate alignment of the audio streams Frame boundary. The disadvantage of this integration approach is the introduction of an undesired additional delay in instant telecommunication/telecom conferencing systems. SUMMARY OF THE INVENTION One object of the present invention is to provide an enhanced method for combining at least two object-oriented audio parameter machines with a human-made, and the method does not need to delay the objects for aligning the frame boundaries. The stream of audio parameters is directed. 132374.doc 200921643 This object has been achieved by a method for combining at least two input object-oriented audio parameter streams into an output object-oriented audio parameter stream as defined in claim i of the present invention. Each object-oriented audio parameter stream contains object-oriented audio parameter values. The object-oriented audio parameter values represent the statistical properties of the audio objects in the form of a time function. The method includes the following steps. First, a synchronous stream of each of the input object-oriented audio parameter streams is calculated. The synchronized stream has an object orientation parameter value at a predetermined temporary location. These predetermined temporary locations are the same for all synchronized streams. Cough for each sync stream

The object orientation parameter values of the predetermined temporary position are calculated by interpolation of the object-oriented parameter values of the corresponding input object-oriented audio parameter stream. The generation of the first-output object-oriented audio parameter stream is performed. The output object-oriented audio parameter stream has an object-oriented parameter value at a temporary position, and the parameter value such as "Hai" is obtained by combining the object-oriented audio parameter values of the synchronous stream of the same predetermined temporary position. An advantage of the method of combining at least two input objects to direct audio parameter streams in accordance with the present invention is that the object-oriented audio parameter streams need not be delayed in order to incorporate the streams. Instead, a very simple process is performed to obtain the object-oriented audio parameter values of the synchronized stream. The amount of the proposed method (4) is overcome by the problem of the position of the different object-oriented audio parameters across the flow of the object-oriented audio parameters (4). In an embodiment, chopping is applied to the object-oriented audio spring of the synchronous stream. For example, the simple piecewise linear interpolation has the effect that the resulting object-oriented audio parameter value of the interpolated value is low-pass filtering. Therefore, in the case of online interpolation, for example, high-pass filtering can be applied to reduce the low-pass chopping effect caused by the interpolated value of the 132374.doc 200921643. The advantage of applying an object that is filtered to the synchronized stream to the value of the audio parameter is to ensure that the dynamic behavior of one of the input object-oriented audio parameter streams is similar. In other words, it improves the quality of the synchronization: the object-oriented audio parameter values that help synchronize the flow simulate the original behavior of the parameters. The synchronization process provides a synchronized stream for the corresponding input #guided audio parameter stream. The filter applied in the embodiment is adaptive to match the statistical characteristics of the corresponding input object-oriented audio parameter values. This adaptive filtering is used to bring further improvements to the synchronization process. Since the object-oriented audio parameters reflect the statistical properties of the audio objects in the form of a time function, the synchronization process is required to take into account such fluctuations over time. In one embodiment, a linear predictive coding analysis is used to determine an envelope of the object-oriented audio parameter values of the input object-oriented audio parameter stream and then the envelope is assigned to the corresponding sync stream during the wave period. Object-oriented audio parameter values. The method of processing/filtering the object-oriented audio parameter values after the synchronization stream ensures that the resulting synchronization parameters have similar characteristics as the original input parameters. In an embodiment, mutually different frequency resolutions are matched by upsampling the objects to direct audio parameter values to a higher frequency resolution. These input object-oriented audio parameter streams may have different frequency resolutions that may occur. In this case, the frequency resolutions must be matched. j Sampling these object-oriented audio parameter values to a higher frequency resolution is simple and does not require extensive computational work. Since the majority of the 132374.doc 200921643 prior-frequency resolution has a common edge/阙 frequency for the parameter band, upsampling can be done by simply copying the appropriate parameter values. The present invention further provides a device request, a boot-programmable device for executing a computer program product in accordance with the method of the present invention, and a teleconferencing system comprising a device according to the present invention. [Embodiment] The above and other aspects will be clarified and explained with reference to the embodiments shown in the drawings. 1 shows a flow diagram of a method for combining at least two input object-oriented audio parameter streams into an output object-oriented audio parameter stream in accordance with the present invention: Each object-oriented a-parameter stream includes object-oriented audio parameter values. The fact that the object-oriented audio parameter values represent the statistical properties of the audio object in the form of a time function comprises the following steps. Step m includes calculating a synchronous stream of each of the input object-oriented audio parameter streams. The synchronized stream has object tracking parameter values at the pre-temporary temporary position. The predetermined temporary positions are the same for all synchronized flow systems. For each of the synchronized streams, the object-oriented parameter values of the predetermined temporary positions are calculated by interpolating the values of the object-oriented parameter values of the respective input-oriented audio parameter streams. Step (10) includes generating an output of the object-oriented audio parameter stream. The output object-oriented audio parameter stream has object-oriented parameter values at predetermined temporary positions, the parameter values being passed by the group: the object-oriented audio parameter values of the synchronized stream of the same predetermined temporary position. The combination is implemented by serially connecting the object-oriented audio parameter values corresponding to the input object-oriented audio parameter streams. Included in the object-oriented audio, the object-oriented audio parameters in the stream are set to each temporary position according to the time/frequency block I32374.doc -10- 200921643 (time/frequency tile). Each object-directed audio parameter is associated with an audio object, and each audio object is sequentially referred to as one of the time/frequency blocks. Thus, in step 12:: the object-directed audio parameter values corresponding to the input object-oriented audio parameter streams are each performed on each of the time/frequency blocks. The order of the parameter values of the concatenation of a particular time/frequency block is arbitrary. Figure 2 shows an example architecture in which two input object-oriented audio parameter streams are combined into one output object-oriented audio parameter stream. When applying the object

When directed to teleconferencing applications, multiple object-oriented audio streams from different remotes must be combined into a single stream. Each stream contains its own downmix stream and the corresponding object-oriented audio parameter stream. In the present invention, the object-oriented audio parameters can be interchanged with the object parameters (4). In two: the object-oriented encoders 21 and 22() at the respective sides encode the respective audio object groups 2〇1 and 202 on the two sides. Each of the decoders produces a downmix stream 211 and 212, respectively, and corresponding object parameter streams 221 and 222. The two object-oriented audio streams are fed to unit 230, which performs the combining of the streams. An example of a unit 230 for the field of electronic conferences is a multi-party control unit (= MCU). The unit 230 produces a combined downmix 213 and a combined object parameter stream. The merged object-oriented audio stream can then be decoded based on the user profile 224 in the object-oriented decoder 240 at the receiving participant. The user profile 224 is related to the positioning of objects, for example, in a three-dimensional space. The result of the decoding is a translated output 214 that is presented to the user. The object parameters reflect the statistical properties of the audio objects in the form of a time function. Therefore, the parameters vary over time and are only valid during certain time intervals. These time intervals are generally referred to as frames. The object parameters are valid during a particular frame or even a portion of a frame. In teleconferencing applications, it is unlikely that the frames of multiple object-oriented audio encoders will be perfectly aligned in time. In addition, different audio objects may result in different frame and different object parameter positions, because the object parameter position is better - depending on the content. Without the proposed synchronization process of the present invention, the combination of the object-oriented audio streams would require delaying at least one of the downmixes η and corresponding object parameters to facilitate alignment of the frame boundaries of the pending merge stream. An advantage of the method of combining at least two object-oriented audio parameter streams in accordance with the present invention is that there is no need to delay the flow of the object-oriented audio parameters to merge the streams. Instead, a very simple process is performed to obtain the object-directed audio parameter values for the synchronized stream. An additional benefit of the proposed method is that the problem of directing the position of the audio parameters across different objects within the frame of the object-directed audio parameter stream is overcome. 3 shows object-oriented audio parameter values 331, 332, and 333 of the synchronous stream, which are guided by the corresponding input objects at predetermined temporary positions 35, 352, and 353 to the object-oriented audio parameter values 32 1 , 322 of the audio parameter stream. Interpolated values between 323, and 324 are obtained. Figure 3 depicts the object parameter values for three adjacent frames 〇, 1, and 2 of the input object parameter streams 31〇 and 32〇. The input object parameter stream 3 is maintained as it is, and the second input object parameter stream 320 is synchronized such that the new position of the object parameters is aligned with the input object parameter stream 3 1 . . The frame of the object parameters of the two input streams 3 1 0 and 320 and the cloth I32374.doc 12 200921643 are misaligned. The block and parameter locations of the object parameter stream 3 10 are copied to the sync stream 330. The object parameter value of the object parameter stream 32 is interpolated between the object parameter values of the input object parameter stream 32, as indicated by the dashed lines. The object parameter values of the interpolated values of the temporary positions 311, 312, and 313 are copied to the synchronized stream 3 3 0 .

In one embodiment, the predetermined temporary positions correspond to a temporary position of the object steering parameter value in one of the input object oriented audio parameter streams. However, other schemes for determining a predetermined temporary location are also possible. For example, two streams can be synchronized simultaneously at a temporary location that is missing from another object parameter stream. Another option is to select such temporary locations depending on the intensity and/or computational complexity of the synchronized locations. Further, although only two input streams are combined in the example discussed, a large number of input streams can also be combined. This is largely determined by the application. For teleconferencing applications, the number of merged streams can be taken as the number of participants communicating with each individual party. In one embodiment, the interpolated values are piecewise linear.斟 and (n, b, p + l) represent the object parameter values, where n is the object number, b parameter with the number 'p is the parameter index, then the interpolation value is calculated as follows. ·, σ(η, b, p) = Μ>ρ_λσ{η, b,pl) + wp+:a(n, b,p + \) WP· 1, Wp+, are the two interpolated weights whose sum is 1:

And wp-i, wp + 丨 are inversely proportional to the distance from the position (p_l) position (p), respectively (in the example). Single ' does not require a lot of calculation work. 132374.doc -13· 200921643 or 'Interpolation of these object parameters can be performed in a different field, for example: σ {n,b,p) = wp_}a2(n,b,p-\) + wp+ Ia2(n,b,p + l) or g( ^n,^>P))~^p^log(a(n,b,p-\))+wp^l〇g(a(n, b,p + \}) Circle 4 shows an architecture where ...."like' team:"to,,,",,_, 7 "IL~milk-oriented

Audio parameter value. In an embodiment, the applied data system is adaptive to the statistical properties of the corresponding input component-oriented audio parameter values. Elements 401 and 402 represent input object parameter streams, wherein stream 4〇1 provides a reference stream of temporary locations for the second object parameter stream 4G2. These temporary positions for synchronization are included in the control parameters indicated by 411. The object parameters having the respective temporary positions 412 of the object parameter stream 402 are fed to the same step 41G and the data processing list to 2Q. This unit example uses interpolation to synchronize the parameters of the object. Applications such as - simple piecewise linear interpolation = the resulting object-oriented audio parameter values of the interpolated values are low pass / effect. The resulting sync stream 431 is fed to the data processing unit Γν: the wave filter 44°. The chopper 44° is preferably a low-order high-pass according to the object-oriented audio parameter value of the wave object: the following:: - If the corresponding input object is directed to the audio parameter flow, the similar dynamic line guide is improved. The quality of the synchronization 'simulates the original behavior of the parameters of the input stream because of the value of the parameter values that help to synchronize the flow. The base 420 determines the statistics of the 2 sets, and the filter coefficients 432 are from the unit 4 and are filtered. The data filter example produces a 5-step object parameter value 441 at its output, which is similar to the similar dynamic behavior of the original input object 132374.doc 14 200921643 piece-oriented audio parameter stream. Figure 5 shows the use of a linear predictive coding analysis to determine the spectral envelope of one of the object-oriented audio parameter values of the input object-oriented audio parameter stream and then assign the envelope to the object-oriented audio parameter value of the corresponding synchronous stream during filtering. . The temporary positions 5 are derived from the reference input object parameter stream and fed to the synchronization unit 51A. The object parameters and corresponding temporary positions 512 are derived from the input object parameter stream 5〇2 and fed back to the synchronization unit 5 10. The 5th synchronization unit 5 synchronizes the object parameters of the streams at the locations. In order for the synchronized object parameters to obtain similar characteristics as the original input object parameters, an LPC analysis method is simultaneously performed in the units 520 and MO, respectively, on the original input character parameters and the isochronous object parameters. The Lpc analysis performed on the synchronized object parameters 531 in unit = 导致 results in so-called spectral whitening. When the filter unit 55 is used to perform a spectral whitening wave to obtain the inverse (integrated) data filter of the applied spectrum white input, the unit 570 will ... package the second data device as early as 57G will flow 5〇2 The original object is added to the whitening/2. The two filter stages 550 and 570 can be individually driven by the gentleman. The LPC analysis is preferably performed on a locale, the autocorrelation estimate using a self-correlation, autocorrelation. Preferably, the sliding window is implemented in the implementation of the change. The same frequency resolution component-oriented audio parameter stream has different object-oriented stone cutters for editing: and the object-oriented audio stream is encoded by the same frequency. '"Special encoders may use mutual exclusion--in the embodiment, the method of averaging the frequency resolution of the object-oriented audio parameters 佶, ^ ^ v Matches to # m more frequency resolution. ^ Obtain an extremely simple method that does not require an over-computation work to deal with the problem, and the object-oriented audio parameters include hierarchical information that defines a separate time/frequency block. The level information is right about the parameter value. The old time is 1 成 成 ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ : : : : : : : : : : : : : : : : : : :: Second, can have different splits to multiple frequency bands. The bandwidth can vary with different resolutions. The eye-catching and the month-to-month are used to upsample the values of the object-oriented audio parameters to a higher value: the resolution of the knife matches the mutually different frequency resolutions. Since the frequency resolution of the parameter band has a common edge frequency in most systems, the upsampling can be achieved by directing the appropriate object to the audio parameter value. The test wheel entry object parameter stream has nine frequency bands, as depicted by 610 in FIG. ^ f input object parameter stream has a lower frequency resolution 630 and 6 frequencies. If the common edge frequency of the two object parameter streams is aligned, the parameter values can be copied (as indicated by the dashed arrow) The frequency resolution of the upsampled second stream is obtained by overlapping the frequency band, as indicated by γ20 in the figure. The frequency band 51 in 630 is the two frequency bands of the 61 〇 and the like. Therefore, the value of the left object parameter of bl of 630 is copied to the frequency of 62 ;; 1 the value of the right object parameter of b1 of 63 同时 is copied to the frequency band b2 of 620. The proposed method of combining at least two input object-oriented audio parameter streams into an object-oriented audio parameter stream can be implemented by a device, the package 132374.doc -16· 200921643 including the synchronization member and the composite member. The processing component calculates a sync stream of each of the input object-oriented audio parameter streams. The sync stream has an object orientation parameter value at a predetermined temporary position. These predetermined temporary locations are the same for all synchronized streams. The object-oriented parameter values for each of the predetermined temporary positions for each of the synchronized streams are calculated by the object-oriented parameter value = interpolated value of the corresponding input object parameter stream. The combined component produces an output object parameter stream. The input parameter has an object-oriented parameter value at a predetermined temporary position obtained by combining the values of the =-directed audio parameters of the synchronized streams of the same predetermined temporary position. The combination is achieved by concatenating the material corresponding to the object-oriented audio (4) to the audio signal value. The computer program product performs the method according to the present invention.实施 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 Streams also need to be merged. A method similar to the proposed method of combining the object-oriented parameter streams can be combined with the downmix. It is to be noted that the above-described embodiments are intended to be illustrative, and not restrictive, and that the invention may be practiced in various alternative embodiments without departing from the scope of the appended claims. In the accompanying request, the circle (4) (10) and the reference mark ^ should limit the request. (d) "contains " does not exclude the existence of elements or steps other than the components listed in the request item 132374.doc 200921643 or steps. A A 1 dry stone surface feeding" or "-" does not exclude plural The existence of these components. The invention can be implemented as a hardware implementation with several discrete units' and by computer. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart showing a method for combining at least two input object-oriented sound m parameter streams into an output object-oriented audio parameter stream according to the spleen 5 rTWm h month of the present invention.

Figure 2 shows the example architecture, in which two input character-oriented audio parameter streams are combined into one output object-oriented audio parameter stream; Figure 3 shows the object-oriented audio parameter values of the synchronous stream, which are at predetermined temporary positions. The interpolated value between the object-oriented audio parameter values of the corresponding input object-oriented audio parameter stream is obtained; FIG. 4 shows an architecture 'where the filtering is applied to the object-oriented audio parameter value of the synchronous stream; FIG. 5 shows the use of a linear predictive coding analysis And determining the spectral envelope of one of the object-oriented audio parameter values of the input object-oriented audio parameter stream and then assigning the envelope to the object-oriented audio parameter value of the corresponding synchronous stream during filtering; FIG. 6 shows that by upsampling Object-oriented audio parameter values to - higher frequency resolutions match different frequency resolutions. Doors In the drawings, the same reference numerals indicate similar or corresponding parts. Portions of the components referred to in the figures are generally implemented in software, and thus represent software entities such as software modules or objects. [Major component symbol description] 132374.doc 200921643 203 Audio object group 204 Audio object group 215 Object-oriented encoder 216 Downmix stream 217 Downmix stream 218 Combined downmix 219 Translation output 220 Object-oriented encoder 221 Object Parameter stream 222 object parameter stream 224 user profile 230 ΧΙΌ - early 240 object oriented decoder 310, 320 input object parameter stream 311, 312, 313 temporary position 321, 322, 323, object oriented audio parameter value 330 sync stream 331, 332, 333 object-oriented audio parameter value 35 1, 352, 353 predetermined temporary position 401, 402 input object parameter stream 410 synchronization unit 411 control parameter 412 temporary position 420 data processing unit 132374.doc -19- 200921643 431 synchronous stream 432 filter Coefficient 440 Filter 441 Synchronized object parameter value 501 Input object parameter stream 510 510 Synchronization unit - 511, 512 Temporary position 520, 540 xjO —· Early 兀 fs 531 Synchronized object parameter 550, 570 Filter unit 562 White material parameter

132374.doc -20-

Claims (1)

200921643 X. The scope of application for patents·· 1. Kind of at least two input object-oriented objects are directed to the direction of the audio parameter stream, and the round-robin-to-object-oriented audio parameter values are included, and the parent object is directed to the audio parameter stream for a time. The form of the function is expressed in the form of 44 object-oriented audio parameter values to include the following steps: Bamboo "玍 This method calculates the flow of each of the input character-oriented audio parameter streams - Synchronous production 1 == There is an object-oriented parameter value at a predetermined temporary position 1st: All synchronous flow systems are the same; for each synchronous flow, the object-oriented parameter values of the time-of-phase position are calculated by interpolating the object-oriented parameter values of the audio parameter stream by the corresponding input 7 to generate an output. An object-oriented audio parameter stream; the output object-oriented audio parameter stream having an object-oriented parameter value at the predetermined temporary position, the parameter values being the object-oriented audio by combining the synchronous streams of the same predetermined temporary position Obtained from the parameter value. The method of claim 1, wherein the predetermined temporary positions correspond to temporary positions of the object-oriented audio parameter values in one of the input object-oriented audio parameter streams. 3. The method of claim 1, wherein the interpolation is segmentally linear. 4. The method of claim 1, wherein the filtering is applied to the object-oriented audio parameter values of the synchronized stream. 5. The method of claim 4, wherein the applied filter is adaptive to facilitate matching the statistics of the corresponding input object-oriented audio parameter values to 132374.doc 200921643. 6 · If requesting the method 'which uses 'linear predictive coding analysis to determine the value of the object-oriented audio parameter stream, the object-oriented audio parameter, the cheek spectrum envelope, and then assign the envelope to the 7 during the filtering : The value of the object-oriented audio parameters of the two streams should be synchronized. The method of claim 1, wherein the input objects are directed to the audio parameter stream / and have different frequency resolutions. 8. If the request item 7 >, borrowing from ..., the different frequency resolutions of the object-oriented audio parameter values are averaged, a higher frequency resolution is obtained. 9:::: Method 7 of the '7' wherein the mutually different frequency resolutions 2 are matched by the method of sampling the values of the object-oriented audio parameters to the higher frequency resolution. The method of 立贝9' wherein the upsampling is achieved by copying the appropriate object to the value of the θ parameter. The m is configured to combine at least two input component-oriented audio parameter streams into an output object-oriented audio parameter stream, wherein each object-oriented audio parameter stream includes object-oriented audio parameter values, and the objects are oriented to the audio value. The form of the function represents the characteristics of the audio objects, and the apparatus comprises: a synchronization component for calculating a synchronization stream of each of the input objects to the audio stream; the synchronization stream has an object guide at a predetermined temporary position, the parameter value 'The predetermined temporary positions are the same for all synchronized flow systems; for the parent synchronous stream, the object-oriented parameter values of the predetermined temporary positions I32374.doc 200921643 are object-oriented parameter values directed to the audio parameter stream by the respective input objects Interpolating for calculation; 12. 13. 14. 15. 16. A composite member for generating an output object-oriented audio parameter stream; the j: object-oriented audio parameter stream having an object D parameter value at the predetermined temporary position (d) the parameter values are obtained by combining the values of the object-oriented audio parameters of the synchronized streams of the same predetermined temporary bits. . The device of item U, wherein the device further comprises a filtering component, wherein the objects of the synchronous streams direct the chopping of the audio parameter values. The apparatus, wherein the (four) wave member is configured to adapt to the statistical characteristics of the value of the audio parameter of the specific object. The device of claim 12 wherein the device further comprises matching components, and VIII is adapted to match the resolutions when the input objects are not the same. The frequency resolution of the parameter stream is used to execute any request item (1). A teleconferencing system' which includes - such as the devices of claims 11 through !4. 132374.doc