JP2009520419A - Apparatus and method for synthesizing three output channels using two input channels - Google Patents

Abstract

In order to synthesize at least three output channels using two stereo input channels, the stereo input channels are analyzed (15) to detect signal components occurring in both input channels. The signal generator (16) is detected at least on the second channel (12b) in relation to the second speaker in the intended speaker scheme located between the first and third speakers in the speaker scheme. It works to introduce some of the signal components. However, if supplying a fully detected signal component results in a clipping situation, a portion of the detected signal component is supplied to the second channel as the true center channel, and the rest is , Located in the first and third channels as phantom center channels.
[Selection] Figure 1

Description

  The present invention relates to a multi-channel synthesizer, and more particularly to an apparatus that generates two or more output channels using two stereo input channels.

Multi-channel audio material is becoming increasingly popular in consumer home environments. This is mainly due to the fact that DVD movies provide 5.1 multi-channel audio, and therefore audio installations are often installed that allow multi-channel audio to be played even by home users. Such equipment is composed of, for example, three speakers L, C, and R at the front, two speakers Ls and Rs at the rear, and a low frequency enhancement channel (LFE). For example, there are the following well-known strengths.
-The center channel improves the stability of the front image, even outside the optimal center listening position. (Larger “sweet spot” = optimal listening position)
-Increased sense of “participating” by the listener, caused by the rear speakers.

  Nevertheless, for example, there are a large number of conventional audio contents that consist of only two (stereo) audio channels on a compact disc (CD), for example.

  5.1 There are basically two options for playing conventional 2-channel audio material on a multi-channel facility.

  1. Left and right channel stereo signals are reproduced in the L and R speakers, respectively. In other words, playback is performed by a conventional method. This solution does not take advantage of the characteristics of the extended loudspeaker equipment (center and rear loudspeakers).

  2. A method of converting two channels of content material into a multi-channel signal can be used (this can happen on-the-fly or by pre-processing), and the multi-channel signal uses all 5.1 speakers and this The method leverages the aforementioned strengths of multi-channel equipment.

  Solution 2 is clearly advantageous over Solution 1, but also, in particular, the front two channels (left and right = LR) are replaced by the front three channels (multichannel left, center, and right = L ′). C'R ') is problematic.

  A suitable LR to L'C'R 'conversion solution must meet the following requirements:

  1) L'C'R 'playback is similar to LR playback, but to reproduce a more stable front image, when the listener is at the “sweet spot”, the center channel is usually All audio events that are perceived as coming from the middle of the left and right loudspeakers must be reproduced. Furthermore, the signal located at the left front must be reproduced by L'C ', and the signal located at the right front must be reproduced by R'C' respectively (JM Jot and C. Avendano's "Audio" Spatial enhancement of recording ", see the 23rd AES meeting in Copenhagen, 2003).

  2) For L'C'R ', in order to obtain a sound effect equivalent to LR, the sum of the acoustic energy emitted by the channel L'C'R' is the sum of the acoustic energy of the source channel LR. Must be equal. Assuming equal characteristics in all playback channels, this translates to “the sum of the electrical energy of the L′ C′R ′ channels must be equal to the sum of the electrical energy of the LR source channels”. .

  Due to requirement 1, the left and right channel signals are mixed into one (single) central channel. This is particularly true when the left and right channel signals are substantially the same, that is, when the left and right channel signals represent a phantom sound source in the middle of the front sound stage. This phantom image is now replaced with a “real” image generated by the central speaker. For requirement 2, this central signal must convey the sum of the left and right energy. If the level of the left or right channel signal is close to the maximum amplitude transmitted by the channel (for 0 dBFS, dBFS is dB full scale), the sum of the levels of both channels exceeds the maximum level and / It can be expressed by a device. This usually results in an undesirable “clipping” effect.

  The clipping situation is shown in FIG. FIG. 6 illustrates the time waveform of the signal 60 processed by a processor having a maximum positive threshold 61a and a maximum negative threshold 61b. Depending on the capabilities of the digital processor processing the digital signal, the maximum positive threshold and the maximum negative threshold can be +1 and -1. Alternatively, if the digital processor is used with numbers shown as integers, the maximum positive threshold will be 32768 corresponding to 215 and the maximum negative threshold will be −32768 corresponding to −215.

  Since the signal of the time waveform is represented by an array of samples (each sample is a digital signal from -32768 to +32768), for a particular time example, the first channel has a fairly high value, It is readily apparent that if two channels also have fairly high values, and if these fairly high values are summed, a larger number is obtained. Theoretically, the maximum number obtained by summing the two channels can be 65536. However, digital signal processors cannot represent this high value. Instead, the digital signal processor represents only numbers that are equal to the maximum positive threshold or the maximum negative threshold. Thus, the digital signal processor performs clipping, such that numbers above the maximum positive threshold or maximum negative threshold are replaced by numbers equal to the maximum positive threshold and maximum negative threshold, so that the situation described with respect to FIG. 6 appears. Within the clipping time portion 62, the time waveform 60 does not have a natural (sinusoidal) waveform, but is smoothed or clipped. When this clipped waveform is evaluated from a spectral standpoint, it is clear that this time domain clipping results in a stronger harmonic component caused by the high gradient magnitude at the beginning and end of the clipping time portion 62.

  This “digital clipping” is not related to playback equipment, in other words, the amplifiers and loudspeakers used to represent the audio signal. However, each amplifier / loud speaker combination also has only a limited linear range. And when the processed signal exceeds this linear range, a kind of clipping also occurs, which can be avoided by using the concept of the present invention.

  In either case, the occurrence of clipping causes a large distortion in the audio signal and significantly reduces the perceived sound quality. Therefore, clipping must be avoided. This is also due to the fact that the improvement in sound quality by representing stereo signals with multi-channel equipment such as 5.1 speaker systems is small compared to very harsh clipping distortion. Therefore, if it cannot be guaranteed that clipping will not occur, it is preferred to use only the left and right speakers of the multi-channel equipment to represent the stereo signal.

  Prior art solutions exist to overcome this clipping problem.

A simple solution to overcome this problem is to reduce all channels equally to a level where no channel signal (especially the center signal) exceeds the 0 dBFS limit. This can be done statically with a predefined fixed value. In this case, the fixed value must be valid even in the worst case situation where the left and right channels have maximum values. For the average LR to L'C'R 'conversion, this leads to an L'C'R' type that is significantly quieter than the original stereo LR, which is particularly useful for users with stereo and multi-channel playback. This is undesirable when switching between. This behavior can be confirmed with commercially available matrix decoders (Dolby ProLogic II and Logic 7 decoders) that can be used as LR to L′ C′R ′ converters. See Dolby's announcement. "Dolby Surround ProLogic II Decoder-Principles of Operation"
http: // www. dolby. com / assets / pdf / tech_library / 209_Dolby_Surround_Pro_Logic_II_Decoder_Principles_of_Operation. pdf
Or “Multichannel Matrix Surround Decoder for Binaural Listeners” by D. Griesinger, 1996, Los Angeles, USA, 101st AES Conference, Sample 4402.

  Another simple solution is to use dynamic range compression to dynamically limit the peak signal (depending on the signal), sometimes referred to as a limiter. The disadvantage of this approach is that the pure dynamic range of the audio program is not reproduced but compressed (Udo Zölzer, edited by the digital audio effect DAFX “Limiter”, 2002 Wiley. & Sons, page 99).

  The reduction problem is undesirable because it reduces the level or amount of the audio signal relative to the original signal level. To completely prevent any theoretical occurrence of clipping, all channels would have to be reduced by a scale factor of 0.5. This leads to a significant reduction in the output level of the multichannel signal compared to the original signal. If only this reduced multi-channel signal is heard, this level reduction can be compensated for by increasing the amplifier amplification. However, when switching between several sources, playing back a multi-channel playback set using the same amplifier amplification settings, the (legacy) stereo signal sounds very loud to the listener.

  Therefore, the user will have to consider reducing the amplification setting of the amplifier before switching from multi-channel playback of stereo signals to stereo playback of stereo signals in order not to damage the ears and equipment.

  Other prior art using dynamic range compression effectively prevents clipping. However, the audio signal itself is changed. Thus, dynamic compression results in an audio signal that is not genuine, which is problematic from a reliability perspective even if the introduced artifacts are not very harsh.

Giotto and C. Avendano, "Spatial Enhancement of Audio Recording", AES 23rd Conference, Copenhagen, 2003 Edited by Udo Zolzer, “Digital Audio Effect DAFX: Limiter”, Wiley & Sons, 2002, p. 99

  An object of the present invention is to provide an improved concept for multi-channel combining using two input channels.

  This object is achieved by a synthesizing device according to claim 1, a synthesizing method according to claim 14, a computer program according to claim 15, or three channel reproduction according to claim 16.

  In order to overcome the clipping problem and nevertheless gain the benefits received by playing a stereo signal using more than two channels of a multi-channel installation, the center channel is generated as usual. In other words, the center channel is based on the discovery that it receives a sound event located in the middle of the left and right loudspeakers and is also called the “true center” representation. However, if the true center falls within the clipping range, only a fraction of the signal component energy indicative of an event in the middle of the audio equipment is supplied to the central channel. The rest of the energy of these sound events is supplied to the first and third (or left and right) channels or remains there from the beginning.

  Thus, with respect to the period during which clipping occurs, if the 2/3 upmix procedure is performed without modification, the center channel is reduced to a level below the maximum level possible without clipping. Nevertheless, the missing part / energy of the signal that cannot be represented by the central channel is reproduced in the left and right channels as a “virtual center” or “phantom center”.

  The true center and virtual center signals are then acoustically combined during playback to reproduce the intended center without clipping. This mixture of true and virtual centers results in an improved, more stable front view of the stereo audio signal. In other words, it results in an increased sweet spot. However, the sweet spot is not as big as without a phantom center. However, the process of the present invention has no clipping artifacts. This is because the rest of the energy that is not processable in the second channel due to clipping problems is not lost, but is represented by the original left and right channels.

  Since the center channel energy is derived from the left and right channels, the energy of the left and right channels in a multi-channel installation is lower than the energy of the original left and right channels in any situation. Please be careful. Thus, in accordance with the present invention, there will never be a clipping problem in these channels even if the rest of the energy is fed back to the left and right output channels.

  A further advantage of the present invention is that, in a preferred embodiment, the total electrical or acoustic of the three generated output channels (and optionally additional output channels such as Ls, Rs, Cs, LEF). It is the signal shaping of the present invention that is performed in such a way that the energy is preserved against the energy of the original stereo signal. Regardless of how the signal is represented (in other words, whether the signal is represented using a stereo facility with only two speakers, or the signal is transmitted using a multi-channel facility with three or more speakers). The same overall volume can be guaranteed regardless of whether it is expressed.

  In addition, the present invention signal formation and distribution of sound energy to the center channel and the left and right channels is applied dynamically only when clipping is unavoidable. That is, in a situation that is not affected by clipping, in other words, if the sampling value of the second channel remains below the maximum threshold, the second center channel does not change at all.

  Furthermore, as a result of the acoustic coupling of the “true center” and “phantom center”, the optimum three channels, in other words, three channels without clipping, or three channels that allow sampling values without minimum / maximum thresholds. A signal closer to the structure is generated. Therefore, the sound image of the present invention does not differ in level compared to the stereo input signal in the preferred embodiment, and is not a fake unlike when using a limiter or simple clipper.

Preferred embodiments of the present invention will now be described with reference to the accompanying drawings.
FIG. 1 shows an apparatus for combining upper channels according to a preferred embodiment of the present invention.
FIG. 2a is a preferred embodiment of the signal processor of FIG. 1 having a post processor.
FIG. 2b is a preferred implementation of the post processor of FIG. 2a.
FIG. 3 is a further embodiment of the signal generator of the present invention with iterative upmixer control.
FIG. 4 is a further embodiment of the signal generator of the present invention operating fully in the parameter domain.
FIG. 5 is an example of a 5.1 sound system optionally also having a surround center channel Cs.
FIG. 6 is an explanatory diagram of a clipped waveform.
FIG. 7 is a schematic diagram of the energy status of the original 2-channel input signal and 3-channel output signal before and after clipping.
FIG. 8 shows a preferred input channel analyzer.

  FIG. 1 shows a preferred embodiment of the apparatus of the present invention for synthesizing three output channels using two input channels, the second channel of the three output channels is intended for a speaker in an audio reproduction facility, It is located between two speakers intended to receive a first output channel and a third output channel. The input channel is indicated by 10a (this channel can be, for example, the left channel L) and for the second channel is indicated by 10b (which can be the right channel R). The output channel is designated 12a for the right channel, 12b for the center channel, and 12c for the left channel. Additional output channels can be generated, such as left surround output channel 14a, right surround output channel 14b, and low frequency enhancement channel 14c. The arrangement of the corresponding speakers for these channels is shown in FIG. There is a sweet spot 50 in the center of these speakers 12a, 12b, 12c, 14a, 14b. If the listener is in the sweet spot, you will get the optimal sound impression.

  Further, a central surround channel 51Cs located between the left surround channel 14a and the right surround channel 14b can be added. The signal for the central surround channel 51 can be calculated using the same process as the signal calculation for the central channel 12b. Therefore, the method of the present invention can also be applied to the calculation of the center surround channel in order to avoid clipping in the center surround channel.

  Note the following: The process of the present invention can be used for the placement of each audio channel, where two input channels are used at the playback facility, targeting two different spatial locations, and three output channels are the two input channels. Generated using a channel. The second channel of the three channels is located between two additional speakers in a reproduction facility with first and third input channel signals.

  The inventive synthesizer of FIG. 1 includes an input channel analyzer 15 for analyzing two input channels in order to determine the signal components occurring in both input channels. These signal components occurring on both input channels are used to form the true center channel. In other words, these signal components can be expressed through the central channel shown in FIG. In general, a stereo signal is located in front of an orchestra when a speaker or music signal is considered, and therefore contains a large amount of mono signal components such as a singer or solo instrument located in front of the audience. Yes.

  Furthermore, the synthesizer of the present invention uses the information about the detected signal components generated in both input channels as supplied via two input signals 10a, 10b and line 13 to three output channels 12a. , 12b, and 12c are generated, and a signal-dependent signal generator 16 is further included. In particular, the signal generator of the present invention operates to send the detected signal component at least partially to the second channel. Moreover, in the situation where the complete detection signal component supply exceeds the maximum threshold, the generator operates to supply only a portion of the detection signal component in the second channel.

  Thus, the second output channel has a time portion that includes only a portion of the detection signal component to avoid clipping, while in another portion of the second output channel, the complete detection signal component is the second output signal. Has been supplied to. The remainder of the detection signal component is included in the first and third output channels, so if these channels are represented, for example, through the speaker equipment shown in FIG. Forms “Phantom Center”.

  Depending on the implementation of the inventive concept, the “part” of the detection signal component located in the second channel and the remainder of the detection signal component located in the first and third channels may be divided into an energy part, a frequency part or other part. So that the second channel contains only a portion of the detected signal component and does not have a value that exceeds the maximum threshold, and therefore does not cause clipping distortion.

  FIG. 2a illustrates a preferred embodiment of the signal analyzer 16 of the present invention of FIG. In particular, in the embodiment of FIG. 2a, the signal analyzer includes a 2-3 upmixer 16 that performs an upmixing process controlled by the input channel analyzer 15 of FIG. 2-3 Upmixer outputs L, R, and C are upmix channels. However, since channel C is generated using an addition process that adds signal components from the left and right channels, channel C may be susceptible to clipping.

  The center channel C is input to a clipping detector 16b, which supplies a post processor 16c, which also receives information about the detected signal component. In particular, the clipping detector 16b operates to examine the time waveform of the central channel 12c.

  Depending on the implementation, the clipping detector can be configured in various ways. Assuming that the signal generator of FIG. 2a can process numbers larger than a predetermined maximum threshold, the clipping detector 16b knows whether there are numbers higher than the maximum threshold of the next processing stage. Simply examine the time waveform. If such a situation is detected, the post processor 16c reduces the center channel energy so that the three output channels 12a, 12b, 12c are finally output by the post processor 16c, and left and right It operates via the activity line 16d so as to initiate a post-processing such that the energy of the channel increases. Therefore, according to the embodiment of FIG. 2a, the conversion process from LR to LCR is performed as usual. In order to check if clipping occurs when the internal first stage center channel signal 20b has to be output as an external signal as in the case of AES / EBU or as in the case of SPDIF format, This internal first stage central channel signal 20b is analyzed. When this happens, part of the signal 20b is removed by the post processor 16c, resulting in a modified center channel 12b, which is instead distributed to the middle left and right channels 20a, 20c as a “phantom center” contribution. After post-processing, the center channel signal 12b is again below 0 dBFS.

  A preferred embodiment of the post processor 16c is shown in FIG. The central channel 20b behind the upmixer 16a is input to the partial extractor 25. The partial extractor receives information 13 about the detected signal component and the control signal from the clipping detector via line 16d, and the clipping detector can also include an indication of the extracted amount. On the other hand, the amount of extraction at each iteration stage is fixed independently of the generated clipping, and iteratively extracts the increase in the detected signal component in a stepwise manner until the clipping detector 16b no longer detects clipping. The trial and error process can be applied. Then, the modified center channel 12b is output by the partial extractor, and the remainder of the detection signal component corresponding to the extracted part is multiplied by 0.5 and then re-input to the output left and right channels 20c and 20a by the upmixer. Must be distributed. For this purpose, the post processor includes two multipliers 26 in each branch or a single multiplier before the branch, and a left adder 27a and a right / left adder 27b.

  If the detection of signal components occurring in both input channels is complete, the left and right channels 20a, 20c do not contain any “phantom center”. However, by adding the extracted components (after multiplying by 0.5) to these channels, the phantom center is added to the left and right channels.

  A further embodiment of the present invention will now be described, particularly in conjunction with FIG. 3, for an embodiment of the signal generator 16 of FIG. The input channel is input to a controllable 2-3 upmixer that receives information regarding the signal components for generating the three output channels in a first iteration stage controlled by the iteration controller 30. This first stage is equivalent to the upmixer operation in FIG. 2a. In other words, the center channel 20b may have a clipping problem. Such a clipping situation is detected by the clipping detector 16b. In contrast to the embodiment of FIG. 2a, the clipping detector 16b controls the upmixer 16a in a feedback manner via the upmixer control line 31 so that no further clipping occurs so that clipping occurs. 20b (after one or more iteration stages regulated by the iteration controller 30) changes the upmixing rules in some way so that only the permitted part of the detected signal component is received.

  Thus, the embodiment of FIG. 3 illustrates an iterative process. In the first stage of the iterative process, the upmixer operation is performed as usual. At the output, detector 16b checks whether clipping occurs. If clipping is detected, this time frame is processed again using the remapping procedure as the phantom center contribution and the rerouting of some central signal energy to the left and right channels.

  The embodiment of FIG. 4 operates fully in the parameter domain. For this purpose, an upmixer parameter calculator 40 connected to the parameter charger 41 is provided. In addition, a clipping detector 42 is provided. This clipping detector 42 operates to examine the original left and right channels or the calculated upmixer parameters to see if clipping occurs after the direct upmix process. If the clipping detector 42 detects the risk of clipping, the clipping detector 42 controls the parameter charger 41 via the control line 44 to supply the modified upmix parameters. This modified upmix parameter is then supplied to the direct upmixer 16a, which generates the first, second, and third output channels so that no clipping occurs on the second channel. . And during the time that the clipping detector 42 originally detected the clipping problem, the left and right channels 12c, 12a have a phantom center contribution.

  In contrast to the embodiment of FIGS. 2 and 3, the processing of the present invention is based on the processing parameters used to obtain the output signals 20a, 20b, 20c or 12a, 12b, 12c from the input stereo signal. Executed. Thus, in order to achieve execution with still low computational complexity, the clipping detection and signal level manipulations or manipulations thereof are based on processing parameters. This is in contrast to the embodiment of FIGS. In the embodiment of FIGS. 2 and 3, after the possible clipping is detected, the process of the present invention is performed on the true audio channel signal already created for the center channel.

  The clipping detection / control of the present invention can be performed by post-processing. Thus, the intended transformation parameters are analyzed and modified according to the inventive concept to provide clipping after synthesis of the true output audio signal. Another way to control the parameter charger 41 is by an iterative method. The intended conversion parameters are analyzed. If clipping occurs after synthesis of the true audio signal, the conversion parameters are modified. The process is then started again, and finally the output channel signal is synthesized with true center and phantom center contributions without clipping in the associated channel.

  Subsequently, a preferred implementation of the input channel analyzer will be described. For this purpose, reference is made to FIG. 8 showing such a suitable input channel analyzer 15. First, at the output of block 80, successive subsequent or overlapping frames are such that there is a series of left channel numbers on line 81a and a series of right channel numbers on line 81b. Generated using window block 80. Next, frequency analysis is performed for each block individually. For this purpose, a frequency analyzer 82 is provided for each channel.

  A frequency analyzer is a device for generating a frequency domain representation of a time domain signal. Such frequency analyzers can include short-time Fourier transforms, FFT algorithms, or MDCT transforms, or other transform devices. On the other hand, the frequency analyzer block 82 includes a subband filter bank that generates, for example, 32 subband channels, or more or fewer subband channels from a series of input signal values. May be. Depending on the implementation of the subband filter bank, the functions of the framing device 80 and the frequency analysis block 82 are implemented in a single digitally implemented subband filter bank.

  Next, cross-correlation for each band is performed as shown in device 84. Thus, the cross-correlator determines the magnitude of the cross-correlation between corresponding bands, in other words between bands having the same frequency index. The magnitude of the cross-correlation determined by block 84 can have a numerical value from 0 to 1, with 0 indicating no correlation and 1 indicating complete correlation. If the device 84 outputs a low cross-correlation magnitude, this means that the left and right signal components in each band are different from each other, and this band occurs in both bands to be inserted into the central channel. This means that the signal component is not included. However, if the magnitude of the cross-correlation is high and the signals in both bands are shown to be very similar to each other, this band has signal components that occur in the left and right channels, and this band is the central channel Must be inserted into.

  A further criterion for determining whether the signals in a band are similar to each other is signal energy. Accordingly, the preferred embodiment of the input channel analyzer of the present invention includes a band-by-band energy calculator 85, which calculates the energy in each band and whether the energy in the corresponding band is similar to each other, Outputs energy-similar magnitudes indicating whether they are different.

  Both the energy-like magnitude output by the device 85 and the cross-correlation magnitude output by the device 84 are input to a final decision stage 86, which is a band i in a frame. A conclusion is reached as to whether or not it occurs in both channels. If the decision stage 86 determines that the signal occurs on both channels, this portion of the signal is fed to the central channel to produce a “true center”.

  FIG. 8 shows an embodiment implementing an input channel analyzer. Further embodiments are well known in the art, for example “Spatial enhancement of audio recording” by Jot and Avendano at the 23rd International AES Conference held in Copenhagen, Denmark, May 23-25, 2003. Explained. In particular, other methods of analyzing the two channels to find signal components in these channels include, for example, principal component analysis, independent subspace analysis, or other analysis known in the art of audio analysis, etc. Includes statistical or analytical analysis methods. All these methods are common in that the signal components occurring in both channels must be detected and fed to the central channel to produce a true center.

  Subsequently, FIG. 7 will be referred to for explaining the energy situation before and after the 2-3 upmix processing is performed by the 2-3 upmixer 16a in the drawing. The left input channel L described at 70 in FIG. 7 has a specific energy. In this example, the right input channel of the two stereo input channels has different (lower) energy as described at 71. Assume that the channel analyzer finds that there is a signal component occurring in both channels. These signal components generated in both channels have energy as illustrated at 72 in FIG. If the total energy 72 is supplied to the central channel as shown at 73, the energy of the central channel will exceed the energy limit, and the energy limit will be such that the signal with such high energy is amplitude maximum. Having at least an outline has an amplitude value that exceeds a threshold. Thus, only a portion of the energy 72 is input to the true center, while the excess is equally (re) distributed to the combined left and right channels L ′ and R ′ as shown by arrow 76. .

  In this context, there are various ways to redistribute energy from the center channel to the left and right channels, or to introduce the correct amount of energy from the original left channel and the original right channel to the center channel. Please be careful. For example, all detection signal components can be reduced by a specific reduction factor and the reduced signal can be introduced into the central channel. This gives the same result with respect to signal components in each band when frequency selective analysis is applied. On the other hand, energy control for each band can be performed. This means that if 10 bands are detected, for example when 10 bands are detected, only 5 bands are introduced in the central channel to reduce energy in the central channel, and the left and right channels Can leave the remaining five bands.

  Depending on certain implementation requirements of the inventive methods, the inventive methods can be implemented in hardware or in software. This implementation can be performed using a digital storage medium, in particular a disk or CD that stores electronically readable control signals. The disc or CD can cooperate with a programmable computer system so that the method of the present invention is performed. Thus, in general, the present invention is a computer program product having program code stored on a machine-readable carrier so that the method of the present invention is performed when the computer program product runs on a computer. The program code is configured. In other words, the present invention is a computer program having program code for executing the method of the present invention when the computer program runs on a computer.

Fig. 3 shows an apparatus for combining upper channels according to a preferred embodiment of the present invention. A preferred embodiment of the signal processor of FIG. 1 having a post processor. Preferred implementation of the post processor of FIG. Further embodiments of the signal generator of the present invention with iterative upmixer control Further embodiments of the signal generator of the present invention fully operating in the parameter domain Example of 5.1 sound system optionally also having a surround center channel Cs Illustration of clipped waveform Schematic of the energy situation of the original 2-channel input signal and 3-channel output signal before and after clipping A preferred input channel analyzer is shown.

Claims (17)

An apparatus for synthesizing three output channels (12a, 12b, 12c) using two input channels (10a, 10b),
The second channel of the three output channels can be fed to a speaker in the intended audio representation located between two speakers that can be fed by the first output channel and the third output channel, The device is
An analyzer (15) for analyzing said two input channels to detect signal components occurring in both input channels;
A signal generator (16) for generating the three output channels (12a, 12b, 12c) using the two input channels;
The signal generator introduces at least part of the detected signal component into the second channel (12b); and
If the complete detected signal component supply results in exceeding a maximum threshold for the second channel, it operates to supply only a portion of the detected signal component to the second channel. A device characterized by that. The signal generator (16)
The second channel comprises a 2-3 upmixer (16a) for generating three intermediate channels containing the detected signal component;
A clipping detector (16b) for detecting a portion of the second channel having an amplitude exceeding the maximum threshold;
Removing a portion of the detected signal component from the second channel in a portion detected by the clipping detector and adding a signal corresponding to the removed portion to the first channel and the second channel Device according to claim 1, characterized in that it comprises a post processor (16c). The signal generator (16)
A 2-3 upmixer (16a) for generating at least a second intermediate channel (12b) containing at least a portion of the detected signal component;
A clipping detector (16b) for detecting a portion of the second channel having an amplitude exceeding the maximum threshold;
Only the portion of the detected signal component is supplied to the second channel and the remaining portion of the signal component is located in the first and third output channels. Device according to claim 1 or 2, characterized in that it comprises a 2-3 upmixer controller (30, 31) for controlling the production. The signal generator (16)
A clipping detector (42) for determining a portion of the input channel in the case of potential clipping;
A second intermediate channel, a 2-3 upmixer (16a) for generating three intermediate channels including at least a portion of the detected signal component;
Control the 2-3 upmixer such that the second channel can always control the generation parameters for the portion of the upmixer determined by the clipping detector to have an amplitude below or equal to the maximum threshold. 4. A device according to any one of claims 1 to 3, characterized in that it comprises a controller (41) for the purpose. The signal generator (16)
Generating the three output channels such that for a period of time, the three output channels and potentially generated additional output channels are equal to the electrical or acoustic energy of the two input channels. Device according to any of claims 1 to 4, characterized in that it operates. The signal generator (16)
The second output channel always includes only the portion of the detected signal component that has a maximum amplitude below or equal to the maximum threshold (61a, 61b), the detected signal component being the second output channel's 6. An operation as claimed in any one of claims 1 to 5, characterized in that it operates to produce the second output channel so as to supply as much energy as possible to the second channel. apparatus. The signal generator (16)
7. The remainder of the detected signal component (73) that does not exist in the second channel and is contained in the first and third channels is applied. 7. The device described. The maximum threshold (61a, 61b) is
8. A device according to any one of the preceding claims, characterized in that it is a full-scale amplitude determined by a synthesizing, digital or analog processing device connected to the synthesizing device. The maximum threshold (61a, 61b) is
Device according to claim 8, characterized in that it is equal to the maximum allowable positive or negative sampling value in the time domain of the signal. The analyzer (15)
Determining a cross-correlation magnitude between at least a portion of the first input channel and the second input channel and detecting a portion of the cross-correlation magnitude that exceeds a threshold indicating similarity (86); The device according to claim 1, wherein the device operates in the following manner. The analyzer (15)
Detecting (85) the portion of the first channel and the portion of the second channel and operating to detect portions of the channel having different energies by being less than equal or equal threshold (86). 11. A device according to any of claims 1 to 10, characterized in that it is characterized in that The analyzer (15) and the signal generator (16) are:
12. Apparatus according to any one of the preceding claims, characterized in that it operates to perform frequency selective or time selective analysis and synthesis. The first and second channels are:
A left channel (L) and a right channel (R) of the representation of the audio signal,
The three output channels are a left front channel (L ′), a center channel (C ′) and a right front channel (R ′), or a rear left channel (Ls), a rear center channel (Cs), and a rear right channel (Rs). 13. Apparatus according to any of claims 1 to 12, characterized in that it is. A method for combining three output channels (12a, 12b, 12c) using two input channels (10a, 10b),
The second channel of the three output channels can be fed to a speaker in the intended audio representation located between two speakers that can be fed by the first output channel and the third output channel, The method
Analyzing the two input channels to detect signal components occurring in both input channels;
Generating the three output channels (12a, 12b, 12c) using the two input channels;
Said generating step introduces at least a part of the detected signal component (72) into said second channel; and
If the complete detected signal component supply results in exceeding a maximum threshold for the second channel, it operates to supply only a portion of the detected signal component to the second channel. A method characterized by that.   A computer program for executing the synthesizing method according to claim 14 when executed on a computer. Three channel representations (12a, 12b, 12c) of two channel input signals (10a, 10b),
The second channel (12b) of the three channel representations (12a, 12b, 12c) is a pair of two speakers that can be supplied with the first channel (12a) and the third channel (12c) of the three channel representations. Can be supplied to a speaker in the intended audio representation scheme located between, the three channel representations:
Said second channel having a time portion (62);
An expression comprising the first channel and the third channel having the remainder (73) of the signal component.   The medium of claim 16, wherein the three channel representations are stored on a computer readable medium.

Images