EP1024679A2 - Method for decoding distorted radio signals in multichannel audio signals - Google Patents

Abstract

The present invention relates to a method for decoding multi-channel audio broadcasts, in particular two-channel stereo audio broadcasts, with a respective useful signal per channel, with a signal source for each frequency spectrum transmitted by the useful signals due to different occurrence in time and different levels in the different channels spatial impression or location information is generated for the corresponding signal source. Here, during the reception of the multichannel audio broadcast in periods in which reception is disturbed in such a way that direct multichannel playback is no longer possible, the location information is extracted from the useful signals and with this current location information an artificial one is obtained from a mono signal containing the useful signals of all channels Surround sound by distributing different frequency bands on the number of channels of the multi-channel audio broadcast corresponding channels with different time delays and / or different attenuation of the levels in the different channels. <IMAGE>

Description

Technical field

The invention relates to a method for decoding multi-channel audio broadcasts, especially of two-channel stereo audio broadcasts a respective useful signal per channel, for each of the useful signals transmitted frequency spectrum of a signal source by different temporal occurrence and different levels in the different Channels a spatial impression or location information is generated for the corresponding signal source, according to the preamble of claim 1.

State of the art

Radio receiver for multi-channel audio signals, e.g. Stereo broadcasts, should with different signal levels, with level fluctuations of the high-frequency carrier signal and, in the case of radio interference, the low-frequency Play the (audio) original signal as realistically as possible. This also includes the spatial arrangement of the different ones reproduced sound sources. This spatial arrangement will hereinafter also referred to as location information. With a multi-channel audio broadcast, especially two-channel stereo audio broadcasting is one of them a respective useful signal is provided per channel, for each of the frequency spectrum of a signal source transmitted to the useful signals due to different times and different levels in the different channels a spatial impression or location information is generated for the corresponding signal source. With some Reception situations deteriorate the audio quality of a Multi-channel signal, e.g. at low useful signal levels and when fading or multi-way reception. To be acceptable in these situations To be able to continue providing sound quality is a remedy in conventional Receivers reduced the channel separation, up to single-channel (mono) operation. This gives the listener the impression that all of them are before Signal sources distributed in space are moved together at one point. Such a change in the original signal is so far called least disturbing measure in difficult reception situations accepted.

The strategy described above is particularly promising if the single-channel signal obtained is due to the modulation method used a lower susceptibility to the mentioned interferences having. This is particularly the case with the widely used frequency modulated FM radio (multi-plex signal using the pilot tone method) the case. The mono information in the low-frequency, parts of the useful signal that are less affected by the system transfer.

Methods are also known for processing a mono signal in such a way that that the impression of a spatial distribution of the signal sources arises. For this purpose, the mono signal is broken down into several frequency ranges. This Areas become different in strength and / or with different delays distributed over different audio signal paths (+ reverb or others known measures for producing an artificial surround effect). The more complex this processing is, the more and closer Frequency segments can be differentiated. Always going on Splits ultimately lead to continuous functions for Attenuation-over-frequency and delay-over-frequency, each one separate function for each signal path. With discrete frequency ranges their width can be chosen differently, adapted to the Stereo sensitivity of the human ear at different frequencies (e.g. no channel separation for basses, narrow segments for approx. 1kHz, large segments at high frequencies).

Conventional stereo decoders are relatively simple and therefore inexpensive. So far with stereo receivers with such stereo decoders method used with mono stereo blend based, like previously explained, only on the control of the channel separation. At occurring faults are switched to single-channel operation very early, whereby any location information that may still be present is no longer reproduced becomes. The known simple stereo decoders have noisy Stereo signals only the choice, the location information entirely or partially discard.

Presentation of the invention, task, solution, advantages

It is an object of the present invention to provide an improved method and an improved receiver of the type mentioned above to provide, which significantly reduces the disadvantages mentioned above.

This task is accomplished by a procedure of the above. Kind with the in claim 1 characterized features and by a recipient of the above Kind with solved the features characterized in claim 12.

For this purpose, it is provided according to the invention that during reception the multi-channel audio broadcast in periods when the reception is so disturbed that direct multi-channel playback is no longer possible is possible, the location information is extracted from the useful signals and with this current location information from one the useful signals of all channels containing mono signal an artificial spatial sound by distributing different frequency bands to the number of channels of the multi-channel audio broadcast corresponding channels with different times Delay and / or different levels of attenuation in the different channels is generated.

This has the advantage that, although a multi-channel decoder according to the state the technology no longer multi-channel decoding with acceptable quality can make the still existing in the disturbed useful signals Location information is extracted and playback with a corresponding "artificial" spatial sound takes place, this "artificial" Generated surround sound essentially corresponds to the original surround sound.

Further developments of the method are preferred in the claims 2 to 11 described.

In a preferred embodiment, spectral distributions of the Useful signals of the different channels and / or runtime differences or time differences of two or more different spectral components with each other in the respective useful signal of different channels compared. The comparison is for each channel for at least two or more different spectral components parameters for signal attenuation and / or a signal delay is determined and the corresponding Spectral components from one containing the useful signals of all channels Single-channel signal delayed according to the specific parameters and / or damped corresponding to the number of channels of the multi-channel audio broadcast Channels distributed in such a way that for a listener for the corresponding Spectral components produce a spatial sound impression, which essentially a spatial sound impression of the directly reproduced Audio signals of the channels. This has the advantage that not switched to single-channel operation during a reception disturbance must be and thus does not give the impression of the signal sources would collapse in the middle of the room. The adulteration by the The listener sees the artificial spatial sound that is shaped as a function of the signal less annoying than falsification by mono operation. Short disturbances are no longer perceived by the listener. You can with shorter time constants or delays for multi-channel operation return because of a pulsating effect (in the case of periodic disturbances the Signal source to jump quickly between different locations) or occurs to a much lesser extent.

An undesirable impression of a quick change of location in the audio signal reproduced signal source when the frequencies of this Signal source a range boundary between adjacent spectral components exceeding is avoided by the parameters for damping and / or delay as a continuous function of the level / transit time difference determined depending on the frequency become.

For digital signal processing, it is particularly suitable if the Spectral range is divided into several predetermined spectral components, where frequencies of a spectral component when determining the parameters weighted differently. Here the disturbing impression of a spatially jumping signal source reliably thereby avoided that the predetermined spectral components in the Frequency partially overlap and the frequencies of a spectral component lower in the overlap area to an adjacent spectral component be weighted. The division of the spectral components is expedient depending on the analysis of the useful signals dynamically changed.

To prevent short-term events in the useful signal from creating a surround sound certain parameters become inappropriately strong added over time using a weighting function. Here for example, a message about predetermined ones is given as a weighting function Periods or a summary of a predetermined Period with greater consideration of younger certain parameters.

To noise components in the useful signal from the design of an artificial Keeping room sounds away is important when determining the parameters only those spectral components are taken into account that have a predetermined level threshold or exceed a frequency-dependent threshold function. The within a spectral component or frequency range existing signal components thereby affect location determination by comparing the useful signals of different channels of the dominant signal no longer.

A good prediction of a future spatial location of a signal source is achieved in that for those spectral components in which a Determination of the parameters is not possible, these from neighboring Spectral components interpolated, certain parameters previously weighted if necessary further used, predetermined parameters or parameter functions used and / or random parameters are used.

Furthermore, a recipient of the above Type marked according to the invention by an analysis module, which spectral distributions of the useful signals the different channels and / or runtime differences or Transit time differences of two or more different spectral components in the respective useful signal of the different channels compares from the comparison for each channel for at least two or more different spectral components parameters for signal attenuation and / or a signal delay determined in this way, and a surround sound module , which the corresponding spectral components from a the useful signals of all channels containing single-channel signal according to the determined Parameters delayed and / or damped on the number of channels of the Multi-channel audio broadcast corresponding channels distributed such that a Listeners for the corresponding spectral components a spatial Sound impression, which is essentially a spatial Sound impression of the directly reproduced audio signals of the channels corresponds.

With this receiver the advantages of the achieved method according to the invention.

Further developments of the receiver are preferred in the claims 12 to 17.

In a preferred embodiment, the analysis module for Each channel has a filter module that contains the respective useful signal or parts of the useful signal in several, in particular four spectral components disassembled. The analysis module points for each spectral component to be evaluated a level detector and one of the number of these spectral components corresponding number of level comparators, each level comparator the levels of an assigned spectral component in several, if necessary compares all channels. Every level comparator is one Signal converter stage downstream, which results from the comparison in the level comparator for each channel of the artificial spatial sound the parameter for signal attenuation and / or the parameter for signal delay certainly. The surround sound assembly has a filter assembly on which a mono signal containing the useful signals of all channels divided into several, in particular five, spectral components, whereby for at least one spectral component is one of the number to be processed Channels corresponding number of attenuator assemblies and / or Delay stages are provided, attenuator assemblies and / or delay stages in accordance with those for this channel and this Spectral component of parameters provided by the analysis module is delayed and / or generate damped output signal. For every channel of the artificial spatial sound, an adder combines the so obtained spectral partial signals.

A time offset between the segment of the multi-channel useful signal being analyzed and the segment of the Mono signal has the advantage that one only in the course of the signal segment clearly prominent spatial information even on the design of the spatial sound at the beginning of this signal segment.

Brief description of the drawings

Fig. 1

eine graphische Darstellung der Kanaltrennung in Abhängigkeit von der Empfangsqualität eines empfangenen Signals für verschiedene Dekodierverfahren,

Fig. 2

eine graphische Darstellung der Zeitauflösung in Abhängigkeit von der Empfangsqualität eines empfangenen Signals für verschiedene Dekodierverfahren,

Fig. 3

eine graphische Darstellung der Frequenzauflösung in Abhängigkeit von der Empfangsqualität eines empfangenen Signals für verschiedene Dekodierverfahren,

Fig. 4

eine graphische Darstellung der Reproduktion der aktuellen Ortsinformation bzw. des Raumklangs in Abhängigkeit von der Empfangsqualität eines empfangenen Signals für verschiedene Dekodierverfahren,

Fig. 5

ein schematisches Blockschaltbild einer bevorzugten Ausführungsform eines erfindungsgemäßen Empfängers,

Fig. 6

ein schematisches Blockschaltbild einer bevorzugten Ausführungsform einer Analysebaugruppe des erfindungsgemäßen Empfängers von Fig. 1,

Fig. 7

ein schematisches Blockschaltbild einer bevorzugten Ausführungsform einer Raumklangbaugruppe des erfindungsgemäβen Empfängers von Fig. 1 und

Fig. 8

verschiedene spektrale Gewichtungsfunktonen.

The invention is explained in more detail below with reference to the accompanying drawings. These show in

Fig. 1

a graphical representation of the channel separation as a function of the reception quality of a received signal for different decoding methods,

Fig. 2

a graphical representation of the time resolution as a function of the reception quality of a received signal for various decoding methods,

Fig. 3

a graphic representation of the frequency resolution as a function of the reception quality of a received signal for various decoding methods,

Fig. 4

a graphic representation of the reproduction of the current location information or the spatial sound as a function of the reception quality of a received signal for various decoding methods,

Fig. 5

2 shows a schematic block diagram of a preferred embodiment of a receiver according to the invention,

Fig. 6

2 shows a schematic block diagram of a preferred embodiment of an analysis module of the receiver according to the invention from FIG. 1,

Fig. 7

2 shows a schematic block diagram of a preferred embodiment of a surround module of the receiver according to the invention from FIGS. 1 and

Fig. 8

different spectral weighting functions.

Best way to carry out the invention

According to the invention during the reception of multi-channel transmissions Useful signals of different channels of the multi-channel audio signal analyzed for their spectral distribution. By comparing the Analyzes of different channels determine which spectral components at which point in the room have their origin. This according to the invention determined data to describe the original signal are hereinafter referred to as "location information". While disturbed Reception times are switched to an artificial surround sound or faded out, which is based on the currently from the disturbed Merkanal signal determined location information is designed, that is Spectral components of the mono signal distributed over the different channels so that the impression arises that the spectral components still have theirs Origin at the location determined for this. After the end of the disturbance will be on Multi-channel operation switched back or blinded. The parameters like e.g. Corner frequencies of sub-frequency bands, and control signals for Design of the artificial spatial sound are subsequently called "Surround sound parameters" called. Depending on the technical implementation Surround sound parameters can be identical to the location information. The location information and / or surround parameters collect data Individual frequencies or the signal components in frequency ranges together. The location information is recorded at certain times or for certain time intervals.

The determination of the location information of the multi-channel signal and / or the Surround sound parameters for designing the spatial sound are, for example as continuous functions of the level depending on the Frequency realized. A continuous function of the level depending of the frequency prevents the impression of a fast Relocation of a signal source when the frequency of the source has a range limit exceeds.

The frequencies of a jointly processed area flow optionally with different weightings in the calculations (frequency-dependent Evaluation function). The weighting can be in the location information be chosen differently than for the surround parameters. It will Parts of the spectrum examined or edited, which are clearly delineated are or partially overlap. In this for digital signal processing more suitable division into frequency ranges will be neighboring Frequencies recorded with lower weighting, so too here the disturbing impression of a jumping signal source is reliable is avoided.

In addition to or instead of analyzing signal levels, the time difference the spectral components (transit time differences) in the different Channels according to the procedures described above in the location information accepted. With different useful signals (e.g. rhythmic Pop music) provides the time difference between the channels a more reliable Statement about the location of the signal generation and is therefore suitable better than starting information for the replica of the artificial Room sound.

Alternatively, location information that has been recorded is not replaced by the next Measurement replaced, but supplemented by a weighting function, e.g. Averaged over longer periods or with greater consideration the recent measurements summarized. Thus, the registration the location information is also done continuously. With a weighting function you can prevent short-term events in the Useful signal inappropriately influence the surround sound

When determining the location information, only such are optional Spectral components taken into account that have a certain level threshold or exceed a frequency-dependent threshold value function. Noise components in the useful signal are thus affected by the design of the artificial Room sounds kept away. The within a frequency range existing small signal components also do not affect more the location of the dominant signal.

For frequency ranges, the analysis of which is insufficient information provides the required location information about the location of the signal origin or surround parameters from neighboring frequency ranges interpolated, previously determined values continue to be used (weighted if necessary), predefined values or functions are used (e.g. these gaps treated as mono, i.e. equally distributed across all channels) and / or (if necessary partially) replaced by random parameters. The reuse of the last parameters determined from usable signal components and the interpolation from neighboring frequency ranges often allow one good forecast of the future location of the signal source.

The division into frequency ranges (e.g. range limits), the evaluation functions and / or threshold functions are variable, for example designed, in particular they are based on the analysis of the useful signal dynamically changed. This procedure enables individual spectral focus of the useful signal as a whole to edit it prevents the splitting and spatial separation of such areas. In addition, the useful signal analysis can be smaller for quiet passages Record signal amplitudes. The timing behavior can better match the characteristics the current fault situation, e.g. Frequency of fading drops be adjusted. In the event of interference in the single-channel useful signal, e.g. Noise, can be reacted individually, e.g. Reduction of signal levels and / or the channel separation in the high frequency range, so that Noise components not in the one currently used for high-frequency reproduction Channel become particularly audible.

Even with good reception signals, it is optionally temporary or permanent maintain the artificial spatial sound. During these periods temporally overlapping and / or alternately determining the location information and surround sound parameters as well as the generation of the surround sound carried out. There is no need to switch from multi-channel operation to surround sound and back, or their number is reduced. It arises a harmonious impression of space and sound.

Common to all variants is that the distortion components are one disturbed signal no longer affect the signal amplitude and thus reduce the signal-to-noise ratio, but only the location information suffers when the signal gets worse. A complete multi-channel audio signal contains at any time or smallest time interval and for each individual frequency provides information about the location of the signal generation or via a spatial arrangement of different signal sources regarding a recording microphone and thus regarding the listener. The human ear is unable to complete this wealth of information evaluate. Quickly switch from one channel to another do not usually occur and would otherwise be in full Speed not recorded by the ear. Now if the multi-channel audio signal, for example, through noise components when the Signal that loses information, remains enough for a long time Remaining information left to give the ear an acceptable spatial impression to provide. The invention uses this residual information to artificially to restore a spatial sound.

According to the invention, information is still provided even during a fault obtained from the disturbed signal to create the "artificial" surround effect to design according to the currently transmitted location information. In critical reception situations, only the mono signal reaches the Speakers, which is subject to an artificial spatial sound. It there is a detection of those still present in the disturbed signal Location information even in times when the signal is so disturbed is that it is no longer suitable for multi-channel playback, but is not yet as badly disturbed as the location information would have been greatly distorted. Thus, the information on Design of the artificial spatial sound effect from the continue available, albeit poor quality, multi-channel audio signal won.

In periods of particularly poor reception, these are after this The residual space information that can be determined is not sufficient to create an artificial spatial sound. At these times, the capture the room information temporarily interrupted and with the last determined values continued or a fade to Mono operation carried out.

It is expedient if there is sufficient reception quality Averaging of measurements over longer times or over frequency ranges, in order to eliminate the noise components in the disturbed signal. This connection show graphically FIGS. 1 to 4.

1 to 4, a reception quality is in each case on the horizontal axis 10 a multi-channel audio broadcast plotted normalized to one. With "1" the reception quality is optimal, whereas towards the origin of the coordinate system to the reception quality always subsides more strongly until there is no reception at "0". On the respective vertical axis 12 is a channel separation (Fig. 1), a Time resolution (Fig. 2), frequency resolution (Fig. 3) and quality the reproduction of the current location information or the spatial sound plotted normalized to one. The dashed line 14 shows the respective Characteristics of a conventional decoder with cross-fading from stereo to mono if signal interference occurs (mono-stereo blend). The crossed, solid line 16 shows the respective characteristic when using the location information before the occurrence of a Disturbance and no further evaluation of the location information in the useful signal during the disturbance. The solid line 18 shows the respective Characteristic when extracting the location information also from the disturbed Signal, according to the invention.

The location information or the location information can be distorted in three parameters be, namely in channel separation, time resolution and frequency resolution. The channel separation (Fig. 1) corresponds to the spatial Separation of the signal sources. The time resolution (FIG. 2) is shown, for example in the maximum speed at which a signal source changes their location. The frequency resolution (Fig. 3) indicates to what extent frequency-similar signal sources localized at different locations can be. The quality of the reproduction of location information (Fig. 4) is expressed simply the product of the three parameters. In This representation is initially ignored that the human Heard the parameters evaluated very differently, and this in turn the frequency, level and other parameters of the signal varies.

Figures 1 to 4 show basic for different decoding methods Relationships for the three parameters (Fig. 1 to 3) and the overall quality the reproduction (Fig. 4) depending on the quality of the received signal, which on the horizontal axis 10 in the previously described is applied. This quality can be as S / N ratio can be plotted for other types of interference, such as Shrinkage or multipath reception could be a different scaling make sense. Conventional stereo decoders are relatively simple and thus inexpensive. The previously used with stereo receivers Mono-stereo blend method is based solely on control the channel separation, (line 14). In the event of malfunctions, it is right faded to single-channel operation early, the location information still available are no longer reproduced.

If previous location information is used before the fault (line 16) the time resolution is reduced to zero even with slightly disturbed signals (Fig. 2), so that a spatial impression is retained, this but in the case of longer-lasting disruptions strongly from the original information can deviate. The reception disturbance is with false information bridges with the assumption that this bogus information is the actual one Signal comes very close initially. In contrast, shows up in Application of the method according to the invention (line 18) also with strong disturbances a good spatial sound impression (Fig. 4), if possible largely maps the original signal itself.

The method according to the invention uses as far as possible all those that are still present Information from (line 18). A suitable mix of time and Frequency averaging is based on the resolving power of the human Hearing and the computing power of the signal processing modules. As a rule, the temporal resolution is initially due to the sluggishness cause hardly any audible loss of quality. The abstract in frequency bands, however, is noticeable in simple systems Falsifications of the location information (arrow 20, fine resolution requires great computing power). If the information content declines you have to further reduce one or both resolutions. The channel separation is only withdrawn when the remaining information available in the signal is no longer sufficient for acceptable audio quality. The Method is particularly suitable for decoding weak, but largely stable signals, i.e. below the "mono" level more conventional Decoder. This area of gain in location information shows Arrow 22.

Figure 5 shows a section of a block diagram of an FM stereo broadcast receiver as a preferred embodiment for an inventive Receiver. An antenna 1010 receives radio frequency signals from radio stations and routes them to a selection and demodulation assembly 1020. In this assembly 1020 the signal from a radio station is recorded and the modulation content extracted. An output signal 1021 of this assembly 1020 is the sum signal both stereo channels L + R (left plus right). Another output signal 1022 contains the difference signal of the two channels L-R. A third output 1023 indicates to what extent the signal reception Is subject to interference, whether due to too little or too quickly fluctuating signal strength, multipath reception or other events. The sum signal 1021 arrives at an assembly 1100 for generation an artificial spatial sound. Both the sum signal 1021 and the difference signal 1022 is fed to a stereo decoder 1030, of the two output signals 1031 with the right channel R and 1032 with the left channel L. These two signals arrive an analysis module 1300 for determining the room parameters. The Module 1300 is also connected to signal 1023. The determined Room parameters for signal attenuation and signal delay are over multiple leads 1301 and 1302 to surround module 1100 passed on. There they serve as spatial sound parameters for realistic purposes Replica of the room sound.

Via two signals 1101 and 1102, two come artificially from the single-channel signal 1021 generated signals Rs (right synthetic) and Ls (left synthetic) to a crossfade unit 1040. Signals 1031 and 1032 are led to the crossfade unit 1040. The signal 1023 for Information about reception interference also becomes the fade unit 1040 forwarded. Depending on this signal 1023 at poorer reception from signals 1031 and 1032 to the faded synthetic signals 1101 and 1102. Two signals 1041 and 1042 route the outputs of the crossfade unit 1040 via two amplifiers 1051 and 1052 to two loudspeakers 1061 and 1062. The modules 1030, 1100, 1300, 1040 and 1050 are examples in one digital signal processor (DSP) 1500 can be summarized, whereby the functions described are designed in particular as software are.

FIG. 6 shows an exemplary embodiment for the surround module 1100. The mono signal (L + R) 1021 is divided into five in a filter assembly 1110 spectral partial signals 1111 to 1115 decomposed, the highest frequency components are emitted via signal 1111, the lowest frequency via signal 1115. Signal 1111 arrives at a first attenuator module 1121, then goes through a first delay stage 1131 and arrives at a first summing point or adder 1141 The second path carries the signal 1111 over a second attenuator module 1122 and a second delay stage 1132 to a second Summing point or adder 1142. Signals 1112, 1113 and 1114 are each through two paths through the attenuators 1123 to 1128 and delay stages 1133 to 1138 to the summing points or adders 1141 and 1142. Output lines 1143 and 1144 of the two summing points 1141, 1142 form the outputs of the block 1100 and thus lead to the signals 1101 and 1102. The signal attenuation in the attenuators 1121 to 1128 and the throughput time of the delay stages 1131 to 1138 are via signal buses 1129 and 1139 controlled, which consist of eight lines, one line per module to be controlled. Because man is unable to find the place of origin of low or low frequency tones can be seen in signal division is not required in this frequency range. The signal 1115 is therefore directly at the two summing points or adders 1141 and 1142 headed.

FIG. 7 shows an exemplary embodiment for the analysis module 1300. The signals 1031 and 1032 coming from the stereo decoder 1030 for the right and left audio channels are in two filter assemblies 1310 and 1320 into four spectral partial signals 1311 to 1314 and 1321 to Disassembled in 1324, with the lowest ones that cannot be localized by human hearing Frequencies are not taken into account. The amplitude of the Signal line 1311 with the highest frequency components of the right channel is determined with a level detector 1331. A resulting signal 1351 reaches a level comparator 1371. Here it is with the corresponding level of the left channel compared, which via signal 1321 and a detector 1341 and line 1361 also to stage 1371 has arrived. A signal conversion stage 1381 generates from the result the comparison four control signals 1401 to 1404 to control the attenuators 1123 to 1128 and delay stages 1133 to 1138 in the Surround module 1100. Corresponding signals and processing stages are available for the signals of the other three spectral components.

To analyze the signals transmitted in different channels alternatively no filters 1310/1320 are used for individual frequency ranges, instead, a function of the level above the is determined for each channel Frequency, from this a list of a parameter set with characteristic frequencies and associated level values. The frequencies in between can be obtained, for example, by interpolation. Artificial for everyone The channel to be generated is used instead of the assembly 1100, for example an analog multi-stage filter, for example from operational amplifiers, controlled such that it matches the incoming mono signal of these functions / parameters.

When filtering and determining the parameters for signal attenuation and Signal delay, for example, the frequency spectrum in different Parts divided, which according to the different shown in Fig. 8 Weighting functions can be weighted. How immediate 8 shows, some areas overlap. further become in some areas, frequencies in the peripheral areas are less strong weighted as frequencies in the middle of such ranges.

Alternatively, the levels in the frequency subsegments are measured using a measured over a longer period of time and these courses, for example, in one Ring buffer stored and fed to a correlation stage. This stage determined by various time shifts and subsequent Comparison of the channels for which time shift a pronounced Agreement is demonstrable. This time difference is called information about the origin of the signal used. The one previously described simple level comparison can continue to take place, with a subsequent one Level decides which of the two localization strategies in in this case better information (more credible, more pronounced, constant or other criterion) and processed.

In a preferred development for averaging and weighting the won Location information is in the in assembly 1371 and corresponding Assemblies en low pass provided.

The frequency ranges generated in stages 1310 and 1320 (e.g. signal 1311) are alternatively split into much finer frequency sub-spectra. All sub-spectra with low signal levels are discarded, the remaining parts added to signals that match those after the first Frequency division (e.g. 1311).

If there is a gap in the location signal, i.e. for this spectral part in all Channels are temporarily insufficient, the in DSP digitally calculated low pass function stopped for this measurement. When implemented in digital hardware, for example different measurements written in a shift register. Every new one Measurement generates a clock signal and shifts it during transmission oldest from the register. A weighted addition of all register components gives the surround sound parameter. For the duration of a gap in the Location signal, the clock to the shift register is interrupted. Alternatively First, all frequency ranges with sufficient information calculated and then the gaps in other frequency ranges linear interpolated from neighboring areas.

The levels in the frequency subsegments are in an alternative Embodiment measured over a longer period and this Gradients stored in a ring buffer. At the beginning of a malfunction a signal rise or fall from this record of the level curve calculated that for the duration of the disturbance in the surround parameters is incorporated.

In a continuation of the invention, the highest peaks in the Frequency spectrum determined. These are called the center frequencies of the filters 1110 used for spectral division.

The signal 1023 at the input of module 1040 is alternatively only in the switching direction from stereo to surround sound is switched through immediately. At the flashback from surround sound to stereo on the other hand Low pass led, so that there is a delay and thus for the artificial duration is limited even at times of good reception Surround sound remains activated.

At the input of the surround sound module 1100 carrying the signal 1021 a delay element can be added, e.g. in form of digital FIFO memory. As a result, this is the one on the antenna 1010 incoming signal sequence processed in the analysis module 1300, however, the results of the analysis affect a signal segment that was received earlier, according to the offset by the Time Delay. One clearly worked out only in the course of the analysis Location information can thus be applied to the entire signal sequence from the beginning Act. The time offset when fading from the artificial spatial sound to compensate for the reproduction of the original signals are also two inputs of the fade unit 1040, which the signals 1031 and 1031 lead to equip with delay stages of the same timing.

Claims (17)

Method for decoding multichannel audio broadcasts, in particular two-channel stereo audio broadcasts, with a respective useful signal per channel, whereby for each frequency spectrum of a signal source transmitted by the useful signals occur due to different temporal and different levels in the different channels, a spatial impression or a Location information is generated for the corresponding signal source,
characterized in that
during the reception of the multi-channel audio broadcast, at least in periods in which reception is disturbed in such a way that direct multi-channel playback is no longer possible, the location information is extracted from the useful signals and an artificial one with this current location information from a mono signal containing the useful signals of all channels Surround sound is generated by distributing different frequency bands on the number of channels of the multi-channel audio broadcast corresponding channels, each with a different time delay and / or different attenuation of the levels in the different channels. Method according to claim 1,
characterized in that
spectral distributions of the useful signals of the different channels and / or transit time differences or time differences of two or more different spectral components in the respective useful signal of different channels are compared with one another from the comparison for each channel for at least two or more different spectral components parameters for signal attenuation and / or a signal delay is determined in such a way and the corresponding spectral components are delayed and / or attenuated from a single-channel signal containing all the useful signals of the channels according to the specific parameters and / or attenuated on the number of channels of the multi-channel audio broadcast such that a spatial sound impression is produced for a listener for the corresponding spectral components which essentially corresponds to a spatial sound impression of the directly reproduced audio signals of the channels. The method of claim 1 or 2,
characterized in that
the parameters for damping and / or deceleration are determined as a continuous function of the level and / or time difference as a function of the frequency. Method according to one of the preceding claims,
characterized in that
the spectral range is divided into a plurality of predetermined spectral components, different frequencies or frequency ranges of a spectral component being taken into account with different weights when determining the parameters. Method according to claim 4,
characterized in that
the distribution of the spectral components is changed dynamically depending on the analysis of the useful signals. Method according to claim 4 or 5,
characterized in that
the predetermined spectral components partially overlap in frequency and the frequencies of a spectral component in the overlap region are weighted less to an adjacent spectral component. Method according to one of the preceding claims,
characterized in that
Once certain parameters are added over time using a weighting function. Method according to claim 7,
characterized in that
an averaging over predetermined periods of time or a summary of a predetermined period of time is carried out as a weighting function, taking more recent particular parameters into account. Method according to one of the preceding claims,
characterized in that
When determining the parameters, only those spectral components are taken into account that exceed a predetermined level threshold value or a frequency-dependent threshold value function. Method according to one of the preceding claims,
characterized in that
for those spectral components in which a determination of the parameters is not possible, these are interpolated from neighboring spectral components, previously determined parameters may be further used, possibly weighted, predetermined parameters or parameter functions used and / or random parameters used. Method according to one of the preceding claims,
characterized in that
by delaying the signal reproduction, a time offset between the analyzed time segment of the multichannel useful signal and the segment of the mono signal manipulated with this data is achieved. Receiver (100) for multi-channel audio broadcasts with a multi-channel decoder, which outputs several useful signals of different channels separately,
marked by an analysis module (1300) which compares the spectral distributions of the useful signals of the different channels and / or transit time differences or transit time differences of two or more different spectral components in the respective useful signal of the different channels, from the comparison for each channel for at least two or more different spectral components Parameters for a signal attenuation and / or a signal delay determined in such a way, and a surround sound module (1100) which delays the corresponding spectral components from a single-channel signal containing the useful signals of all channels in accordance with the determined parameters and / or attenuates corresponding channels over the number of channels of the multi-channel audio broadcast, that a listener receives a spatial sound impression for the corresponding spectral components, which essentially corresponds to a spatial sound impression of the directly reproduced audio signals of the channels. Receiver (100) according to claim 12,
characterized in that
it has a filter module (1310, 1320) for each channel (1031, 1032), which divides the respective useful signal into several, in particular four, spectral components (1311 to 1314, 1321 to 1324). Receiver (100) according to claim 13,
characterized in that
the analysis module (1300) has a level detector (1331, 1341) for each spectral component. Receiver (100) according to claim 13 or 14,
characterized in that
the analysis module (1300) has a number of level comparators (1371) corresponding to the number of spectral components, a level comparator comparing the levels of an assigned spectral component in at least two channels. A receiver (100) according to claim 15,
characterized in that
Each level comparator (1371) is followed by a signal converter stage (1381) which determines the parameter for signal attenuation and / or the parameter for the signal delay for each channel from the result of the comparison in the level comparator (1371). Receiver (100) according to claim 15 or 16,
characterized in that
the surround module (1100) has a filter module (1110), which divides a mono signal (1021) containing the useful signals of all channels into several, in particular five, spectral components (1111 to 1115), one of the number of channels corresponding to at least one spectral component Number of attenuator assemblies (1121 to 1128) and / or delay stages (1131 to 1138) is provided, attenuator assemblies (1121 to 1128) and delay stages (1131 to 1138) according to the parameters for signal delay and / or for this channel and this spectral component Signal attenuation generates a delayed and / or damped output signal, an adder (1141, 1142) connected downstream for each channel adding together all output signals of different spectral components of a channel

Images