EP2499843B1 - Method for mixing microphone signals of a recording using multiple microphones - Google Patents

Description

The invention relates to a method for mixing microphone signals of a sound exception with multiple microphones according to the preamble of claim 1. Such a method is known from WO 2004/084 185 A1 known.

In order to record an extensive acoustic scene in the production of sound recordings for music conserves, films, radio broadcasts, sound archives, computer games, multimedia presentations or Internet presences, it is from the " Handbuch der Tonstudiotechnik "by Michael Dickreiter et al., ISBN 978-3598117657 known to use multiple microphones instead of just a single microphone. For this purpose, the term "multi-microphone sound recording" is generally used. An extended acoustic scene may be, for example, a concert hall with an orchestra of a variety of musical instruments. To record the tonal details, each individual musical instrument is recorded with a single, closely positioned microphone and also positions further microphones at a greater distance to capture the overall acoustic image, including the reverberation in the concert hall and the audience noises (in particular applause). Another example of an extended acoustic scene is a drum kit consisting of several percussion instruments recorded in the recording studio. In the case of the "multi-microphone sound recording", in each case a microphone is positioned in close proximity in front of the individual percussion instruments and an additional microphone is mounted above the percussionist.

Such multimicrophone sound recordings make it possible to capture as many acoustic and sound properties of the details as possible as well as of the overall picture of the scenery in high quality and make them aesthetically satisfactorily shapable. Each microphone signal of the plurality of microphones is usually recorded as multi-track recording. In the subsequent mixing of the microphone signals further creative work is done. In special cases it is also possible to mix "live" immediately and record only the result of the mixdown.

The design goals of the mix are usually a balanced ratio of the volumes of all sound sources, a natural sound and a realistic spatial impression of the overall acoustic image.

100

ein erstes Mikrofonsignal

101

ein zweites Mikrofonsignal

110

eine auf Addition basierende Summierungsstufe

111

ein Summensignal

199

ein Ergebnissignal

200

ein n -tes Summensignal

201

ein n+2 -tes Mikrofonsignal

210

eine n+1 -te, auf Addition basierende Summierungsstufe

211

ein n+1 -tes Summensignal

In the conventional mixing technique in a sound mixing console or in the mixing function of digital sound editing systems, summation of the supplied microphone signals is carried out by a summer ("bus"), which is a technical realization of ordinary mathematical addition. In FIG. 1 By way of example, a single summation in the signal path of a conventional audio mixer or digital sound system is shown. A series connection of summations in the summer ("bus") in the signal path of a conventional sound mixer or digital sound system is in FIG. 2 exemplified. In the FIGS. 1 and 2 the reference numerals

100

a first microphone signal

101

a second microphone signal

110

an addition-based summation level

111

a sum signal

199

a result signal

200

an n-th sum signal

201

a n + 2 -th microphone signal

210

an n + 1th addition-based summation level

211

an n + 1-th sum signal

In multi-microphone sound recordings contain at least two microphone signals due to the unavoidable multipath propagation of sound portions of sound that come from the sound of one and the same sound source. Since these sound components arrive at the microphones as a result of the different sound paths with different transit times, comb filter effects, which are audible as sound changes and run counter to the intended naturalness of the sound, are produced in conventional mixer technology in the summer. In conventional mixing techniques, such sound variations due to comb filter effects can be reduced by adjustable gain and optionally adjustable delay of the recorded microphone signals. However, such a reduction is only possible to a limited extent if there is a multipath sound propagation of more than a single sound source. In any case, however, a considerable adjustment effort on the mixer or digital sound system for finding the best compromise is required.

In the older one DE 10 2008 056 704 is a downmixing (so-called "downmixing") described for the generation of a two-channel sound format from a multi-channel (eg, five-channel) sound format, are shown with the phantom sound sources. In each case, two input signals are summed, wherein a weighting of the spectral coefficients of one of the two input signals to be summed takes place with a correction factor; that input signal which is weighted by the correction factor is prioritized to the other input signal. The in the DE 10 2008 056 704 However, the determination of the correction factor described results in that in cases where the amplitude of the prioritized signal is low compared to that of the non-prioritized signal, disturbing background noises can be heard. The probability of occurrence of such disturbances is low, but not influenced.

From the WO 2004/084 185 A1 For example, in a method for mixing microphone signals of a multi-microphone sound recording, it is known to form the spectral values of overlapping time windows of samples from a first microphone signal and a second microphone signal. The spectral values of the first microphone signal are distributed to the spectral values of the second microphone signal in a first summation stage to form spectral values of a first sum signal, thereby dynamically correcting the spectral values of one of the two microphone signals. From the spectral values of the first sum signal, spectral values of a result signal are formed, which are subjected to inverse Fourier transformation and block merging. For each block of samples, individual correction factors can be determined in this way. The dynamic correction by signal-dependent weighting of spectral coefficients, rather than ordinary addition, reduces undesirable comb filter effects in multimicrophone tone-mixing that arise in the summators of the sound mixer or tone-cutting system by ordinary additions. Meanwhile, even in this method disturbing noise is audible if the amplitude of the prioritized signal is low compared to that of the non-prioritized signal.

The diploma thesis written at the IRT " Automatic stereo downmix of 5.1 multichannel productions "by B. Runow (6.7.2008 ) shows a method for mixing microphone signals, in which one of the signals after spectral transformation is weighted by a factor and then added to another spectrally transformed microphone signal. The factor is calculated from the real and imaginary parts of the transformed signals. After the addition, the result signal is transformed back.

The object of the invention is to largely compensate for the sound changes resulting from the mixing of multi-microphone sound recordings as a result of multipath propagation of sound components.

The solution to this problem arises from the features of claim 1.

Advantageous embodiments and further developments of the method according to the invention are specified in the subclaims.

Figur 3

ein generelles Blockschaltbild einer Anordnung zur Durchführung des erfindungsgemäßen Verfahrens;

Figur 4

ein ähnliches Blockschaltbild wie in Fig. 3, jedoch mit dem Unterschied, dass die erste Summierungsstufe um eine Anzahl von weiteren Summierungsstufen erweitert ist;

Figur 5

ein Blockschaltbild einer in den Figuren 3 und 4 vorgesehenen ersten Summierungsstufe, und

Figur 6

ein Blockschaltbild einer in Figur 4 vorgesehenen weiteren Summierungsstufe.

The invention is based on the in the FIGS. 3 to 6 illustrated embodiments explained. It shows

FIG. 3

a general block diagram of an arrangement for carrying out the method according to the invention;

FIG. 4

a similar block diagram as in Fig. 3 , but with the difference that the first summation stage is extended by a number of further summation stages;

FIG. 5

a block diagram of one in the Figures 3 and 4 provided for the first summation stage, and

FIG. 6

a block diagram of an in FIG. 4 provided further summation stage.

100

ein erstes Mikrofonsignal

101

ein zweites Mikrofonsignal

199

ein Ergebnissignal

201

ein n+2 -tes Mikrofonsignal

300

Spektralwerte des ersten Mikrofonsignals

301

Spektralwerte des zweiten Mikrofonsignals

310

eine erste Summierungsstufe

311

Spektralwerte eines ersten Summensignals

320

eine Blockbildungs- und Spektraltransformationseinheit

330

eine inverse Spektraltransformations- und Blockzusammenführungseinheit

399

Spektralwerte eines Ergebnissignals

400

Spektralwerte eines n -ten Summensignals

401

Spektralwerte eines n+2 -ten Mikrofonsignals

410

eine n+1 -te Summierungsstufe

411

Spektralwerte eines n+1 -ten Summensignals

500

Zuordnungseinheit

501

Spektralwerte A(k) des zu priorisierenden Signals

502

Spektralwerte B(k) des nicht zu priorisierenden Signals

510

Berechnungseinheit für Korrekturfaktorwerte

511

Korrekturfaktorwerte m(k)

520

Multiplizierer-Addierer-Einheit

700

eine n -te Baugruppe bestehend aus der Einheit 320 und der n+1 -ten Summierungsstufe 410

In the FIGS. 3 to 6 the reference signs have the following meanings:

100

a first microphone signal

101

a second microphone signal

199

a result signal

201

a n + 2 -th microphone signal

300

Spectral values of the first microphone signal

301

Spectral values of the second microphone signal

310

a first summation level

311

Spectral values of a first sum signal

320

a blocking and spectral transformation unit

330

an inverse spectral transformation and block merging unit

399

Spectral values of a result signal

400

Spectral values of an n-th sum signal

401

Spectral values of a n + 2-th microphone signal

410

an n + 1th summation level

411

Spectral values of an n + 1-th sum signal

500

allocation unit

501

Spectral values A (k) of the signal to be prioritized

502

Spectral values B (k) of the signal not to be prioritized

510

Calculation unit for correction factor values

511

Correction factor values m (k)

520

Multiplier-adder unit

700

an n-th module consisting of the unit 320 and the n + 1-th summation stage 410th

FIG. 3 shows a general block diagram of an arrangement for carrying out the method according to the invention. A first microphone signal 100 and a second microphone signal 101 are each supplied to an associated blocking and spectral transformation unit 320. In the units 320, the supplied microphone signals 100 and 101 are first divided into blocks of time-overlapping signal portions, whereupon the formed blocks undergo a Fourier transform. This results in the spectral values 300 of the first microphone signal 100 and the spectral values 301 of the second microphone signal 101 at the outputs of the blocks 320. The spectral values 300 and 301 are then fed to a first summation stage 310, which generates the spectral values 311 of a first sum signal from the spectral values 300 and 301. The spectral values 311 also form the spectral values 399 of a result signal, which are first subjected to an inverse Fourier transformation in a unit 330. The inverse spectral values thus formed are then combined to form blocks. The resulting blocks of time-overlapping signal portions are accumulated into the result signal 199.

This in FIG. 4 illustrated block diagram is similar in structure as the block diagram in FIG. 3 , but with the essential difference that the spectral values 399 do not simultaneously represent the spectral values 311. Rather, it is in Fig. 4 between the spectral values 311 and the spectral values 399, a series connection of one or more identical assemblies 700 each consisting of a blocking and spectral transformation unit 320 and an n + 1-th summation stage 410 is inserted. From the assembly 700 is in Fig. 4 for simplicity, only a single assembly 700 is shown in the block diagram, which will be described below, where the count index n is for sequential numbering. The mentioned series connection of assemblies 700 should be understood as meaning that at the beginning of the series connection the spectral values 400 simultaneously form the spectral values of the first sum signal 311 and at the end of the series connection the spectral values 411 also form the spectral values of the result signal 399. In all other sections of the series connection, the spectral values 411 of a summation level 410 also form the spectral values 400 of the subsequent summation level 410. Each block formation and spectral transformation unit 320 of an assembly 700 of FIG A n + 2 -th microphone signal 201 is fed in series, in which it is divided into blocks of time-overlapping signal sections. The formed blocks of time-overlapping signal sections are Fourier-transformed, resulting in the spectral values 401 of the n + 2-th microphone signal. The spectral values 400 of the n-th sum signal and the spectral values 401 of the n + 2-th microphone signal are then supplied to the n + 1-th summation stage 410, which generates from them the spectral values 411 of the n + 1-th sum signal.

• Entweder wird der Korrekturfaktor m(k) wie folgt berechnet: $$\mathrm{eA}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{x}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{w}\left(\mathrm{k}\right)=\mathrm{D}\cdot \mathrm{x}\left(\mathrm{k}\right)/\mathrm{eA}\left(\mathrm{k}\right)$$
  
  $$\mathrm{m}\left(\mathrm{k}\right)={\left(\mathrm{w}{\left(\mathrm{k}\right)}^{\mathrm{2}}+\mathrm{1}\right)}^{\left(\mathrm{1}/\mathrm{2}\right)}=\mathrm{w}\left(\mathrm{k}\right)$$
  
  oder der Korrekturfaktor m(k) wird wie folgt berechnet: $$\mathrm{eA}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{eB}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{x}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{w}\left(\mathrm{k}\right)=\mathrm{D}\cdot \mathrm{x}\left(\mathrm{k}\right)/\left(\mathrm{eA}\left(\mathrm{k}\right)+\mathrm{L}\cdot \mathrm{eB}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{m}\left(\mathrm{k}\right)={\left(\mathrm{w}{\left(\mathrm{k}\right)}^{\mathrm{2}}+\mathrm{1}\right)}^{\mathrm{1}/\mathrm{2}}-\mathrm{w}\left(\mathrm{k}\right)$$
  
  wobei
  • m(k) der k -te Korrekturfaktor
  • A(k) der k -te Spektralwert des zu priorisierenden Signals
  • B(k) der k -te Spektralwert des nicht zu priorisierenden Signals
  • D der Grad der Kompensation
  • L der Grad der Begrenzung der Kompensation
  bedeuten.

FIG. 5 represents the details of the first summation stage 310. In the summation stage 310, the spectral values 300 of the first microphone signal 100 and the spectral values 301 of the second microphone signal 101 are fed to an allocation unit 500, in which prioritization of the output signals 501, depending on the manufacturer's or a user's choice , 502 of the unit 500 takes place. Two alternative assignments are possible: Prioritizing the output signal 501, the spectral values A (k) of the signal 501 to be prioritized are assigned to the spectral values 301 and the spectral values B (k) of the signal 502 to be prioritized to the spectral values 300. Alternatively, the spectral values A (k) of the signal 501 to be prioritized are assigned to the spectral values 300 and the spectral values B (k) of the signal 502 that is not to be prioritized are assigned to the spectral values 301. The choice of Priorisierungszuordnung determines the spatial impression of the overall acoustic image and is made according to the design requirements. A typical possibility is to assign the signals of those microphones that are intended for recording the overall acoustic image (so-called main microphones) or the summation signals formed according to the invention to the prioritized signal path, and the Assign signals from those microphones positioned near the sound sources (called support microphones) to the non-prioritized signal path. The associated spectral values A (k) of the signal 501 to be prioritized and spectral values B (k) of the signal 502 that is not to be prioritized are then fed to a correction factor value m (k) calculating unit 510 which uses the spectral values A (k) and B (k) the correction factor values m (k) are calculated as an output signal 511 as follows:

• Either the correction factor m (k) is calculated as follows: $$\mathrm{eA}\left(\mathrm{k}\right)=\mathrm{real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)+\mathrm{imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{x}\left(\mathrm{k}\right)=\mathrm{real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{w}\left(\mathrm{k}\right)=\mathrm{D}\cdot \mathrm{x}\left(\mathrm{k}\right)/\mathrm{eA}\left(\mathrm{k}\right)$$
  
  $$\mathrm{m}\left(\mathrm{k}\right)={\left(\mathrm{w}{\left(\mathrm{k}\right)}^{\mathrm{2}}+\mathrm{1}\right)}^{\left(\mathrm{1}/\mathrm{2}\right)}=\mathrm{w}\left(\mathrm{k}\right)$$
  
  or the correction factor m (k) is calculated as follows: $$\mathrm{eA}\left(\mathrm{k}\right)=\mathrm{real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)+\mathrm{imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{eB}\left(\mathrm{k}\right)=\mathrm{real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)\cdot \mathrm{real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)\cdot \mathrm{imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{x}\left(\mathrm{k}\right)=\mathrm{real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{w}\left(\mathrm{k}\right)=\mathrm{D}\cdot \mathrm{x}\left(\mathrm{k}\right)/\left(\mathrm{eA}\left(\mathrm{k}\right)+\mathrm{L}\cdot \mathrm{eB}\left(\mathrm{k}\right)\right)$$
  
  $$\mathrm{m}\left(\mathrm{k}\right)={\left(\mathrm{w}{\left(\mathrm{k}\right)}^{\mathrm{2}}+\mathrm{1}\right)}^{\mathrm{1}/\mathrm{2}}-\mathrm{w}\left(\mathrm{k}\right)$$
  
  in which
  • m (k) is the k-th correction factor
  • A (k) is the k-th spectral value of the signal to be prioritized
  • B (k) is the k-th spectral value of the signal which is not to be prioritized
  • D the degree of compensation
  • L is the degree of limitation of the compensation
  mean.

Degree of compensation is a numerical value that determines the extent to which the sound effects caused by comb filter effects are compensated. It is chosen according to the design requirements and the desired tonal effect and is advantageously in the range of 0 to 1. If D = 0, the sound is exactly the same as the conventional mix. If D = 1, this results in a complete removal of the comb filter effect. Values for D between 0 and 1 accordingly give a sound effect between that at D = 0 and that at D = 1.

The degree L of the limitation of the compensation is a numerical value which determines to what extent the probability of the occurrence of disturbing perceptible background noises is reduced. This probability is given if the amplitude of the microphone signal to be prioritized is small compared to that of the microphone signal which is not to be prioritized. It is L> = 0. If L = 0, there is no reduction in the probability of disturbing background noises. The degree L is chosen so that experience has shown that no background noises are perceived. Typically, the degree L is on the order of 0.5. The greater the degree L, the lower the probability of the disturbances, but this also partially reduces the compensation of sound changes determined by the setting of D.

The spectral values A (k) of the signal 501 to be prioritized are additionally supplied to a multiplier 520, while the spectral values B (k) of the signal 502 which is not to be prioritized are additionally supplied to an adder 530. In addition, the multiplier 520 receives the correction factor values m (k) of the output signal 511 from the calculation unit 510, where it is complex with the spectral values A (k) 501 (after real part and imaginary part). be multiplied. The result values of the multiplier 520 are supplied to the adder 530 where they are added complexly (after real part and imaginary part) with the spectral values B (k) of the non-prioritizing signal 502. This results in the spectral values 311 of the first summation signal of the first summation stage 310.

The decisive factor for the prioritization is thus the multiplication of the correction factor m (k) with exactly one of the two summands of the addition performed in the adder 530. Thus, the entire signal path of this summand is "prioritized" from the microphone signal input to the adder 530.

FIG. 6 represents the details of the n + 1-th summation stage 410. The n + 1-th summation stage 410 is similar in construction to the first summation stage 310, but with the difference that here the allocation unit 500 the spectral values 400 of the n-th sum signal and the Spectral values 401 of the n + 2 nd microphone signal, and that the result values of the adder 530 form the spectral values 411 of the n + 1 th sum signal.

Claims (3)

1. Method for mixing microphone signals of an audio recording with a plurality of microphones (multi-microphone audio recording), wherein a multipath propagation of sound portions is given and

   - a first microphone signal (100) and a second microphone signal (101) are subject to the building of blocks of samples and a Fourier-transformation, wherein the spectral values (300, 301) of the respective microphone signal (100, 101) are generated,

   - the spectral values (300) of the first microphone signal (100) are distributed onto the spectral values (301) of the second microphone signal (101) in a first summing level (310) while formation of spectral values (311) of a first sum signal, wherein a dynamic correction of the spectral values (300, 301) of one of the two microphone signals (100, 101) occurs,

   - in order to generate the spectral values (311) of the first sum signal of the spectral values (300) of the first microphone signal (100) and the spectral values (301) of the second microphone signal (101) the spectral values (300, 301) of one of the two signals is chosen, which is to be prioritized over the other signal,

   - the spectral values (311) of the first sum signal form spectral values (399) of a result value, and

   - the spectral values (399) of the result value undergo an inverse Fouriertransformation,

   - the so-formed inverse spectral values are merged into blocks of temporally overlapping signal segments and the hence resulting blocks are accumulated to the result signal (199),

   wherein the spectral values (A(k)) of the signal to be prioritized are multiplied with the respective corresponding corrective factors m(k), and that the spectral values (B(k)) of the signal not to be prioritized and the corrected spectral values m(k) · A(k) of the signal to be prioritized are added while formation of spectral values (399) of a result signal (199),
   characterized in that,
   the calculation of the corrective factors m(k) is as follows: $$\mathrm{eA}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)$$

   

   $$\mathrm{x}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$

   

   $$\mathrm{w}\left(\mathrm{k}\right)=\mathrm{D}\cdot \mathrm{x}\left(\mathrm{k}\right)/\mathrm{eA}\left(\mathrm{k}\right)$$

   

   $$\mathrm{m}\left(\mathrm{k}\right)={\left(\mathrm{w}{\left(\mathrm{k}\right)}^{\mathrm{2}}+\mathrm{1}\right)}^{\left(\mathrm{1}/\mathrm{2}\right)}-\mathrm{w}\left(\mathrm{k}\right)$$

   

   or calculated as follows: $$\mathrm{eA}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)$$

   

   $$\mathrm{eB}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$

   

   $$\mathrm{x}\left(\mathrm{k}\right)=\mathrm{Real}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Real}\left(\mathrm{B}\left(\mathrm{k}\right)\right)+\mathrm{Imag}\left(\mathrm{A}\left(\mathrm{k}\right)\right)\cdot \mathrm{Imag}\left(\mathrm{B}\left(\mathrm{k}\right)\right)$$

   

   $$\mathrm{w}\left(\mathrm{k}\right)=\mathrm{D}\cdot \mathrm{x}\left(\mathrm{k}\right)/\left(\mathrm{eA}\left(\mathrm{k}\right)+\mathrm{L}\cdot \mathrm{eB}\left(\mathrm{k}\right)\right)$$

   

   $$\mathrm{m}\left(\mathrm{k}\right)={\left(\mathrm{w}{\left(\mathrm{k}\right)}^{\mathrm{2}}+\mathrm{1}\right)}^{\left(\mathrm{1}/\mathrm{2}\right)}-\mathrm{w}\left(\mathrm{k}\right)$$

   

   and

   m(k) is the k^th corrective factor

   and

   A(k) is the k^th spectral value of the signal to be prioritized

   and

   B(k) is the k^th spectral value of the signal not to be prioritized

   and

   D is the grade of compensation

   and

   L is the grade of limitation of the compensation,

   that the grade L of the limitation of the compensation is a numeric value which determines in how far the probability of the occurrence of disturbing ambient noises is reduced, wherein this probability is given when the amplitude of the microphone signal to be prioritized is low in contrast to the microphone signal not to be prioritized,
   that the value of the grade L of the limitation of the compensation is bigger than or equal to zero, wherein for L=0 no reduction of the probability of disturbing ambient noises is given and the grade L is chosen according to experience so that just no more ambient noises can be heard,
   that the grade D of the compensation is a numeric value which determines in how far the sound changes due to comb-filter effects are balanced, wherein the value of D is chosen according to the creative demand and the intended tonal effect which lies in the range of 0 to 1, wherein for D=0 the sound is exactly the sound of mixing without the corrective factor m(k) and for D=1 the comb-filter effect is completely removed.
2. Method according to claim 1, characterized in that the grade L of the limitation of the compensation has a value of about 0.5.
3. Method according to claim 1, characterized in that the first summing level (310) is expanded by a number N of additional summing levels (410),
   in that respectively during the n+1^th summing level (410) an n+2^th microphone signal (201) undergoes a formation of blocks of samples and a Fourier-transformation, whereat the spectral values (401) of the n+2^th microphone signal (201) are generated,
   in that during the n+1^th summing level (410) the spectral values (400) of the n^th sum signals are distributed to the spectral values (401) of the n+2^th microphone signal (201) with generation of the spectral values (411) of an n+1^th sum signal, wherein a dynamic correction of either the spectral values (400) of the n^th summing level or the spectral values (401) of the n+2^th microphone signal (201) occurs,
   in that respectively during the n+1^th summing level (410) of spectral values (400) of the n^th sum signal and the spectral values (401) of the n+2^th microphone signal (201) the spectral values (400, 401) of one of the two signals is chosen, which is to be prioritized over the other signals,
   wherein
   n = [1 ... N] is the serial number of the summing level
   and
   N is the amount of expanded summing levels.

Images