SE528706C2 - Device and process method for surround sound - Google Patents

Abstract

A surround sound processing arrangement where directionally encoded audio signals in a B-format (W, X, Y, Z) or B-format-like signals are via a primary matrix, synthesizing virtual arbitrary microphones, encoded to intermediate channels used as virtual loudspeaker feeds. These virtual loudspeaker feeds are via a secondary matrix transcoded to multiple channels corresponding to coincident or non-coincident virtual arbitrary microphone signals suitable for subsequent mono, stereo or surround sound loudspeaker replay. The described transcoding process enables variation of the properties of the reproduced sound field regarding parameters such as localization, depth, image size, spaciousness, envelopment, and particularly can offer a stable frontal image and a large listening area size.

Description

20 25 30 35 5-channel recommendation [l 1], however, ambisonic decoding is found to have serious problems typically resulting in a small listening area (small "sweetspot"), unstable frontal sound image and color staining during main movement [12] [l 3 ] and is currently uncommon in practical surround finishing. I Gerzon m. Fl. [5], the previous limitation in ambisonic theory regarding the requirement for regular or rectangular speaker arrays was removed and a large number of numerical examples for irregular arrays were presented where no one had such a high degree of asymmetry as that chosen in the ITU recommendation: O °, = l = 30 °, i (100 ° -120 °) from the frontal center line. Although “ambisonicfïavkodning is possible for an ITU setup, as reported in [5], see e.g. [l2], so the strong front / back asymmetry accentuates the "sweetspot" problems. Partly p. G.a. these problems, it is more common in current recording practice to synthesize microphones from B-format without explicit consideration of the arnbisonic theory, but even in this case it leads to an unstable frontal sound image and a small listening area, see e.g. [l2], [l 3].

Separated microphone arrays, on the other hand, tend to offer a larger listening area, a stable frontal sound image [12] [l3], and to generally provide a sound that is perceived as more spatial.

However, each separated microphone array gives a different result when the recorded signals are played over a speaker array, typically the ITU recommendation [1l], see Fig. individual sound sources, general sound character, and front / rear / side balance. In addition, within each arrangement, fl your variants are possible such as variation of the main recording angle, see [l]. A good example of this is a conference paper on Williams MMA that describes over 220 ways to configure an array of five cardioid microphones [8]. This conference contribution, in itself, clearly shows the need for an approach that does not force a decision regarding the microphone setup already at the recording stage where an unfortunate choice can lead to unsatisfactory results and even to a costly second recording opportunity.

The third type of arrangement, with a physical object separating the microphones, is not particularly suitable for use with the invention. Listening tests [13] also point to sweetspot problems similar to those of coincidental lineups and are not discussed further.

Brief Description of the Invention The invention offers an opportunity to overcome real and perceived problems with incidental microphone arrays so that, after the rework described, the benefits of separated arrays are mainly achieved, especially the more stable frontal sound image and the larger listening area. also a more spacious sound, at the same time as the decision on exactly which separate arrangement to use can be postponed until the finishing.

To accomplish this, the invention utilizes the inherent generality and orthogonality of the B-format and teaches how the B-format can be reworked into almost any other type of coincidence or separate microphone array so that, after the actual recording of the B-format signals, or after the B-format format the signals are created in an artistic way by e.g. computer simulation or convolution with B-shaped impulse response, offers a choice between a coincidence or a separated microphone array to achieve a given purpose, thus offering a choice between the main advantages of coincidental and separated arrays and also between different separated arrays. In contrast, reprocessing of signals recorded with a separate set-up to another separate set-up is not practically possible because already at the time of recording, mutually uncorrelated and non-orthogonal signals have been recorded in varying ways and to varying degrees. B-format, on the other hand, consists of strongly correlated signals which, via the described reworking, during post-processing, can be decorrelated in the selected way by using a virtual synthesized primary coincidence microphone array that creates feeds (here called L-format) to a virtual speaker array whose outputs are recorded virtually by a selected secondary virtual coincidence or separate microphone array see Fig. 2. In this way, the described process represents a virtual re-recording of the original sound field where the choice of microphone array can be postponed to the post-processing in a similar way to current coincident methods , described e.g. in [3] [7], but where many degrees of freedom are offered because even separate arrangements can be chosen. In particular, this allows for a substantially stabilized frontal sound image and a larger listening area. The indisputable technical advantages of B-format are maintained (including, for example, the possibility of monocompatibility, arbitrary rotation or other sound tent transformations) because the original B-format is retained or can be recreated from L- format. In cases where the secondary virtual arrangement is not coincidental, of course the reworked signals cannot be identical to those that would have been recorded directly with the corresponding physical arrangement. The process introduces certain artifacts consisting of additional lj extraction due to the round-receiving components of the primary virtual microphones, which follows directly from the description below, but these artifacts are generally perceived as more positive than negative compared to purely coincidental arrangements. Physically separated arrays always create comb filter effects, objectively an artifact but which is perceived by many to give a spatial sound. The described reworking process introduces additional chamber effects but not to such an extent that it negatively affects the listening experience. The final speaker playback, see Fig. 3, sounds more like the one that would have been created by a physically separated setup of the same type than the one created by a purely coincidental decoding such as e.g. ambisonic.

A further advantage of the present invention is that the characteristics of the final playback can be varied in real time by varying during listening both the primary virtual microphone array and individual microphone characteristics, direction and position, of the secondary virtual microphones. In this way, additional artistic possibilities are also offered in the finishing, possibilities which by their nature are both more extensive and of a different nature than those offered by current coincident techniques.

The array can also be used to transform between two arbitrary separated microphone arrays by first performing a virtual B-shaped recording of the original signals from the separated array and then applying the array to these B-shaped signals but due to as previously mentioned, the result will be of limited quality because the original signals from the separated array are too decorrelated.

Due to the above, the first embodiment of the present invention comprises a processing device intended for processing direction-coded audio signals into audio signals intended for playback via loudspeakers to a listening surface.

Thus, the device consists of: inputs for the direction-coded audio signals; primary matrix device; interconnecting device; secondary matrix device; outputs for subsequent amplification and playback of the processed signals. According to a preferred embodiment, the directionally coded audio signals consist of one or fl era of: recorded B-shaped signals; artificially created B-format signals; saturated B-shaped space impulse response; artificially created B-shaped space impulse responses; recorded B-shaped signals; artificially created B-format similar signals; saturated B-shaped similar spatial impulse responses; artificially created B-shaped similar spatial pulse responses. The direction-coded B-formats or B-formats similar to the audio signals are those from substantially perfectly coincident microphones consisting of at least one round-receiving microphone and three orthogonal figure-S microphones. Preferably, the directionally coded B-format audio signals may be limited to consist of busy and horizontal components. The directionally coded audio signals may also consist of B-format or B-format similar signals which have undergone a spatial transform. Preferably, the primary matrix may contain linear coefficients with characters aj, bf, c, - configured to create individual polar microphone directives from said B-shaped signals; l0 15 20 25 30 35 786 11101) = hfUt fl / š WW) J "b / 'COWWXÛU * f Cfsínß fifl YW) where m, bf, c, primary matrix coefficients for calculating directivity of primary virtual microphone j; h, primary gain factor virtual microphone j; go, - the horizontal angle of the front axle for synthesized primary microphone j; n arbitrary sound sample nurrimer.

Preferably, 0 _ Preferably, the number (N) of primary virtual microphones and outputs from the primary array 2 is 3.

According to a preferred design, the time delays, from the N virtual speakers to the center of the secondary microphone array, are selected to be substantially and practically the same.

The secondary matrix may consist of linear coefficients with signs, time delays and alteration devices which realize arbitrarily placed and directed, coincidental or non-incidental secondary virtual microphones of polar or non-polar type where the directivity of a polar type is described by F (0) = p + q + q + q 6l) where Û is the angle of incidence of the frontal axis of a microphone and where 0 5 p 5 l, p + q = l. A design may include control devices for adjusting the coefficients of the primary matrix and thus changing the type of primary microphones. A control device may be included to adjust coefficients and time delays of the secondary matrix and thus change the type, direction, position, filtering and delay of the secondary virtual microphones. The primary and secondary matrices can be integrated.

The invention also relates to a method for processing an audio signal in which direction-coded audio signals, via a primary matrix which synthesizes virtual coincident microphones, which convert into temporary signals which are transmitted to virtual speakers. The method comprises transmitting the output of these virtual speakers to a secondary array which realizes multiple coincidences or non-coincident virtual polar or non-polar microphones which are thus processed into multiple output signals for mono, stereo or surround recording. The directionally coded audio signals consist of one or more of: recorded B-form; artificially created B-format; saturated B-shaped space impulse responses; artificially created B-shaped space impulse responses; recorded B-format similar signals; artificially created B-format similar signals; measured B-shaped similar spatial impulse responses; and artificially created B-formats similar to runn impulse responses.

The invention also relates to a computer program, preferably realized as a "plug-in", for processing digital sound, where direction-coded audio signals, via a primary matrix which synthesizes virtual coincident microphones, are converted into temporary signals which are transmitted to virtual speakers. The program consists of a procedure for transmitting the output of these virtual speakers to a secondary array which realizes multiple coincidences or non-coincident virtual polar or non-polar microphone signals adapted for mono, stereo or surround playback.

The invention also relates to a computer program product with an instruction set for processing into digital sound where direction-coded audio signals, via a primary matrix which synthesizes virtual coincident microphones, are processed into temporary signals which are transmitted to virtual speakers. The instruction set includes a procedure for transmitting the output of these virtual speakers to a secondary array which realizes multiple coincidences or non-incidental virtual polar or non-polar microphone signals adapted for subsequent mono, stereo or surround playback.

Brief Description of the Figures The present invention is described in more detail, with examples of embodiments, in the accompanying figures, in which: Figure 1 shows a general diagram of a reworking system; Figure 2 shows an overview diagram of a 3 x 4 x 5 rework for ITU surround, see Fig. 3; Figure 3 shows an overview diagram of surround playback for ITU surround; Figure 4 illustrates the time delay of audio samples for individual secondary virtual microphones at a large distance r from virtual speakers; 10 20 25 30 528 766 7 Figure 5 angular ions for 9 ,, - regarding the directivity of secondary virtual microphones and for 12), regarding directivity for virtual speakers, examples with i = 2 and j Figure 6 schematically shows the main steps in a reworking system according to the invention.

Detailed Description of Embodiments The following examples are intended to illustrate, but not in any way limit, the invention according to the claims. For example, the invention may just as well use the height dimension (via the Z component in B-format) but for practical reasons with respect to speaker installations, the examples are based only on a round component (W) and horizontal components (X, Y).

The example also uses the commonly used 5-channel surround system but could just as well use e.g. 6, 7 or fl are channels or stereo.

Figure 1 illustrates a surround processing system (transcoding system) consisting of: input device (1), primary matrix device (2), virtual speaker inputs (3), secondary matrix device (4), and virtual microphone signals and output device (S).

The processing for lTU surround, which is illustrated in Fig. 2, consists of: W, X, Y input signals (6), primary matrix device (7), synthesized primary microphone directive (8), secondary primary matrix device (9), virtual speakers (10) , secondary virtual microphone (1 1), surround output (12), and the origin of the coordinate system (13).

The final playback via a surround speaker array, after appropriate amplification, exemplified by an LCD array, is illustrated in Fig. 3: listening room (14), real physical playback speakers (15), inputs from (12) :( 16), central listening position ( 17), and listening area (18).

The output channels from the primary array (7) are created by N primary coincident virtual microphones (8) synthesized from incoming W, X, Y (6), and are connected to N virtual speakers (10), typically horizontally placed equidistant at a typically large radius r, see Fig. 2.

If air absorption is acidified, and in addition to a common propagation constant and time delay, the input signal to virtual speaker j can be approximately described by: uj- (n) = h, - (a fl / š W (n) + bjcos (ç1j) X (n) + c , - sin (çq-) Y (n)) (Eq. 1) where 10 l5 20 25 30 528 706 8 n arbitrary sound sample numbers uJ- (n) sound sample n for virtual speakers j, (the uJ signals are called here L-format ) N number of primary virtual microphones or speakers, N 2 3. N = 4 provides the most practical use with high channel separation. j index for primary virtual microphone and virtual speaker, 1 S j SN h] variable gain for primary virtual microphone j, typically h, - = 1 W (n) B-format W sound sample n (traditionally a W signal is that of a round recording microphone divided by the factor Ö, otherwise this factor is based on Equ. 1) X (n) B-format X sound sample n (signal from a frontally directed figure-S microphone) Y (n) B-format Y sound sample n ( signal from a laterally to the left-facing figure-S microphone) aj, 11, -, cj primary matrix coefficients for calculation of feed to virtual speaker j: Typically: N = 4 and for high channel separation a, - = 0.25, b, - = cj = 0.75 corresponding to the synthesized primary hypercardioid microphones (p, the horizontal angle from the origin (13) to each individual virtual speaker (10), and also the angle between the frontal axis of the corresponding synthesized primary polar microphone (8), where shown as 21: ( j - 1) / N radians and where an alternative realization is 21: (j - 1/2) / N radians .

The resulting sampled output signals created in the secondary matrix (hereinafter referred to as M-channel), for secondary virtual microphone i (11), for large r, as a result of a summation of output signals from each horizontally placed virtual speaker, except a common constant factor, is approximately: N an) i gfZ ur ”- 01, - e» fi »1 = 1 (Eq. 2) where e, ~ (n) sound sample n fi ir secondary virtual microphone signal i (12) in index for secondary virtual microphone, 1 S in SM 15 20 25 30 fš Co 528 706 9 number of secondary virtual microphones, M 2 1.

Example: M = 5 for ITU surround, M = 2 for stereo, M = 1 for mono variable gain for secondary virtual microphone in positive or negative sound sample delay (time delay) for secondary virtual microphone i (1 1) from virtual speaker j (10 ) truncated to the nearest integer value (å) = 0 for secondary virtual microphones placed in the origin (13)), see Fig. 4 for example with i = 1 where (19) indicates virtual speaker and (20) secondary virtual microphone. Fractional delay techniques can be used as an alternative design. sampling frequency, Hz the speed of sound in air, m / s additional positive time delay for secondary virtual microphone in truncated to the nearest integer value. Fractional delay techniques can be used as an alternative design. ideal first-order polar directivity of secondary virtual microphone in evaluated in the direction of virtual speaker j: FU = pi + q, - cos (i9,)) (Equ. 3) Simulation of real microphone properties via ÛÜ - dependent filter can be used as an alternative design (a physical cardioid is not a cardioid at all frequencies).

Eq. 3 is then mainly changed to: Ûjlf) = Fiffi + qiff) COSWIJ) (Eq. 4) lckepolar arbitrary directives secondary synthesized virtual microphones can be used as an alternative design. frequency, Hz weighting factors for secondary virtual polar microphones, 0 S p; S 1, p, - + qi = At a horizontal angle between the frontal axis of the secondary virtual microphone i (23) and a vector from the secondary virtual microphone i to the virtual speaker j (21), see Fig. 5 for a 528 TfÜâ 10 examples with j = 1 and i = 2.

A feature of this design is that if the secondary virtual microphone array is an ideal B-shaped microphone (M = 3, g.- = 1 0011 8: = 0 for i = 1 to M) placed in the origin (å- J- = O), and for at least N = 4, çoj = 21t (i - U / N, Oßh h) = 1 f exactly from L-shaped, e.g. if: 01:02 = ü3 = G4 = G> fÜ and bj = Cj = C # Û via the following relationship: e1 (n) = (secondary virtual microphone is recording: p, = I, q; = 0) = 4 = Z a, Wm fl š = 4 aßvr / (n) j = l => Wm) = mi) (4 aJš yl (Ekv- 7) 6201) = (secondary virtual microphone is figure-S in positive x-direction: p; = 0, q; = l) = = i (ajWU / z) Ja + bJ-cosüpj) X (n) + c, sin (ço, -) Y (n)) COS (9zj) = 20 XW) => Xm ) = 6201) (2 cyl (Ekv- 8) e 3 (n) = (secondary virtual microphone is fi gur ~ 8 in positive y-direction: pi = 0, q; = 1) = = i (a_, W (n ) J; + bj cos (çq) X (n) + cj sin (çßj) Y (n)) cos (63, -) = 2c Y (n) 1 = 1 => m) = 2301) (2 fy * (Eq- 9) Numerical example with hypercardioid L-format (a = 0.25, c = (175): Wm) = ernyiš (Eq. 10) X (n) == e2 (n) /1.5 (Eq. 11) Y (n) = e3 (n) /1.5 (Equ. 12) ie, the original B-format can simply be recreated via scaling of the secondary virtual microphone signals. 'S23 CK 11 An application of the above property is to use the N-channel L-format (uj) as a distribution form food from which many M-channel surround formats can be created, but which also allows reproduction of the original B-format in cases where e.g. an arnbisonic decoding [5] is desired or when the direction-coded sound field is to be pretreated e.g. via B-format transphones [5], before final decoding for loudspeaker playback. The following transformations are therefore possible: B-format <-> L-format -> a variety of M-channel setups (for mono, stereo and surround) where: e »indicates exact forward and backward transform -> indicates forward-transform only Examples of useful rework to stereo (M = 2) are with the virtual secondary microphones configured as ORTF, NOS, RAI, DIN or spaced omni [1].

The requirement according to the requirements does not address the problem of playback via a speaker array that differs from the ITU recommendation [ll], but offers an option to select a secondary virtual microphone array that gives a result that is less dependent on the exact speaker array or that is created for another lineup.

The primary matrix equations (Eq. 1) are based on prior art and are intimately associated with ambisonics [5] but the use of a primary (decoding) matrix corresponding to hypercardioid L-format, analyzed according to ambisonic theory, results in a velocity vector vector) V, exaggerated in relation to the energy vector E, and is therefore not useful for ambisonic decoding because the ambisonic theory stipulates ry = l for all coded angles of incidence and rg = constant for all angles and maximized against l [5 ]. For a decoding matrix corresponding to hypercardioid signals (a, - = 0.25, b, = 0.75) for N = 4, ry = 1.5 while rg = 0.5. In contrast, an ambisonic decoding for a square speaker array gives ry = 1.0 (which is stipulated by the ambisonic theory) while rg becomes = 0.75 using a, - = 0.333 and b) = 0.667. It must be emphasized that, unlike ambisonic, L-format is not a psychoacoustic decoding, which seeks to take into account aspects of human hearing and whose goal is direct speaker playback, but instead represents temporary signals adapted for virtual re-recording of the original sound field and the secondary virtual microphones not only record energy but also sound pressure including phase. The use of separated virtual microphones makes it impossible to characterize the final result in ambisonic terms because that theory is based on coincidental microphones. lO 15 20 25 30 \ 5, - 7.9 t C 12 How psychoacoustically relevant and aesthetic is the final playback of e; the signals (12) via a selected speaker array (typically lTU, see Fig. 3) are, depends largely on the characteristics of the secondary virtual microphone array and the speaker array used, and to a lesser extent on L-format, in a manner similar to an actual separated microphone array of the same type had been used to record the original sound field. It is worth noting that in cases where arnbisonic decoding is preferred, the use of B-format means that this possibility is still left open.

Thus, the main steps of a method according to a preferred embodiment are summarized in a block diagram according to Fig. 6. In a first step, via the primary matrix (25), virtual coincident polar microphones are further synthesized by the direction-coded sound signals (24) coded as temporary signals intended to be used as input signals to virtual speakers (26). In a second step, the output of the virtual speakers is fed via a secondary array (27) which realizes multiple coincidental or non-incidental secondary virtual polar or non-polar microphones. The multiple output channels (28) are suitable for speaker playback of mono, stereo or surround sound.

In addition to the designs described above, further examples, applicable in arbitrary combinations, may include: Example 1: A distance dependence from the virtual speakers may be included so that the time delays above are instead calculated as: 5.1. = dUfs / CO (Equ. 13) d, = (xj. _ x,) 2 + y, _ y, 2 (Equ. 14) where x, x mode for virtual speakers jy, - y mode for virtual speakers jx, X mode for secondary virtual microphone iy, - y mode for secondary virtual microphone in Example 2: The geometric spread from each individual virtual speaker can be taken into account assuming that these act as point sources. Eq. 2 then reads instead: 10 15 20 25 30 35 528 705 13 N 2:01) = 8: 2 14101 - å; - Sf) »Fry / dy (Eq. 15) J = l Example 3: The primary virtual microphones can be of different type, a useful variant is e.g. to let the rearward-facing microphone (i = 3 in the example above and in Fig. 2) instead be a cardioid (a, = 0.5, bj = cj- = 0.5) because sound incident directly from the front will then not be present in u3.

Example 4: The virtual speakers can use directivities Gj, so that the level of secondary virtual microphones placed far from the axis of a virtual speaker will be lowered, see Fig. 5, where (21) is a virtual speaker, (22) a secondary virtual microphone , and (23) the axis of a secondary virtual microphone. Eq. 2 then reads instead: N 9:07) = gr Z "Ä" '511' Ei) Gjtfßjr) F11 / di] f = l (Eq. 16) Example 5: For some applications it may be advantageous to use different methods for reprocessing of low and high frequencies because human hearing has different properties below and above cza 500 Hz (main size in relation to the wavelength of sound). To achieve this with the presented invention, two of the converters described here can be used in parallel. The inputs (1) of the first converter are fed with direction-coded low-pass signals and the inputs (1) of the second converter are fed with high-pass signals. The first converter uses primary and secondary matrices whose settings are optimized or adapted for placement of low-frequency sound, while the second converter uses matrices optimized or adapted for placement of high-frequency sound. The crossover frequency may depend on the application but is typically between 300 and 700 Hz. The outputs (4) from both inverters are summed and form new signals for speaker playback and listening.

Example 6: A useful alternative to direct B-format recording of music or other performances is to measure, or create artificial, B-format room impulse responses via one of many available impulse response measurement methods ("Maximum Length Sequence", frequency sweep, etc.).

When these B-format impulse responses are folded with dry recorded music, the generated audio signals will be substantially identical to the signals that would have been recorded if the dry recorded music had been played back from the same speaker position used in the impulse response measurement. B-format signals created in this way can also be used with the presented invention, but it is also possible to apply the invention directly to the B-format impulse responses by seeing these as input signals (1). The outputs from the converter then become instead a new set of impulse responses corresponding to the secondary virtual microphones. This technique is useful for creating multiple types of fl-channel or stereo impulse responses for use in reverberation, from the same measured or simulated B-shaped room impulse response.

The invention is not limited to the designs shown but can be varied in a number of ways without departing from the scope of the appended claims and the device and method may be realized in various ways depending on application, functional units, needs and requirements, etc. [1] 10 [2 ] [3] 15 [4] 20 [5] [6] [7] 25 [3] 30 5 References “Multichannel natural music recording principles based on psychoacoustic principles” Günther Thiele Extended version of contribution presented at AES 19th International Conference, May 2001, httpz // wvvwirtde / IRT / F uE / as / mu1ti-rnr-ext.pdf US 4,042,779: COTNCIDENT MICROPHONE SIMULATION COVERING THREE DIMENSIONAL SPACE AND YIELDING VARIOUS DIRECTIONAL OUTPUTS Soundfield microphone / company: Sound fi eld Ltd.

Charlotte Street Business Center Wake fi eld, West Yorkshire, WFl lUH ENGLAND httpz // wvvwsound fi eldcouk US 4,086,433: SOUND REPRODUCTION SYSTEM WITH NON ~ SQUARE LOUDSPEAKER LAY-OUT US 5,757,927: SURROUND SOUND APPARATER ~ 4,4, software: David McGriffy 5208 Evans Ave.

Austin, TX 78751 USA http://mcgriffy.com/audio/ambisonic/vvmic/ “The Quick Reference Guide to Multichannel Microphone Arrays. Part 1: Using Cardioid Microphones ”25

[10] 1111

[13] 16 M. Williams, G. le Dü. 110m AES preprint 5336, Amsterdam, The NETHERLANDS, May 2001.

Sound technology Dn-lng.

Schoeps GmbH Spitalstrasse 20 D-76227 Karlsruhe (Durlach) GERMANY vvwwschoepsde / E / kfrnlš60.html Holophone® A Division of Rising Sun Productions Ltd. 258 Adelaide Street East, Suite 200 Toronto, ON, MSA lNl CANADA wwwholophonecom “Multichannel surround sound systems and operations” AES Technical Document AESTD1001 .0.01-05, based on ITU-R BS.775-1 “Empirical Comparison of Microphone Surveillance Procedures for 5.0 Surround ”J. Roland, M. Fleischer, S. Fuhrmarin, B. Steiglitch, U. Reiter, H. Kutsbach See also http://wWw.stucLtu-ilmenau.de/~proverdi/index.html ORF / Hauptmilcrofon (listening test ): “Classical Music in Radio and TV - a multichannel challenge. The ORF Surround Listening Test 2001 ”Florian Cramer and Christian Sodl, Austrian Broadcasting Cooperation.

See also httpz // wwwhauptmikrofon.de / or fi htm

Claims (1)

1. 0 15 20 25 fv- Patent claim. An arrangement for processing, preferably for processing from direction-coded audio signals in the form of B-format (W, X, Y, Z) or B-format-like signals to audio signals reproducible via loudspeakers (10) over a listening surface (18), characterized by said arrangement comprising: - inputs for receiving direction-coded lj output signals (1); primary matrix device (2) arranged to create virtual individual microphone directives, - interconnection device (3); - secondary matrix device (4); and - output device for subsequent amplification and playback of processed audio signals (5). . Arrangement according to claim 1, wherein said direction-coded audio signals comprise one or more of: - recorded B-shaped (W, X, Y, Z); - artificially created B-format; - saturated B-shaped space impulse response; azti fi specially created B-shaped space impulse response; - recorded B-format similar signals (W, X, Y, Z); - artificially created B-shaped similar signals; - saturated B-shaped similar spatial impulse responses; - artificially created B-format similar to room impulse responses. . Arrangement according to claim 2, wherein said direction-coded B-format or B-format-like sound signals originate from non-perfectly coincident microphones consisting of at least one round-receiving microphone (W) and three orthogonal figure-X l0 15 20 25 (fl va? GO '* ~ 3 Arrangements according to either claim 2 or 3, wherein the directionally coded audio signals comprise B-shaped signals limited to round (W) and horizontal (X, Y) components. either claim 2, 3 or 4, wherein the directionally coded audio signals comprise B-format or B-format-like signals which have undergone a spatial transformation. linear coefficients with signs aj, bj, c, - configured to create virtual individual microphone directives (8) obtained from said B-format signals; 11,01) = 11,4 61.45 Vr / (n) + b, 003% X0 »+ a, sinwj) 1101)) where aj, bj, cjprimary matrix coefficients for calculating the directivity of the virtual primary microphone j (8); h] variable gain for primary virtual microphone j; (o, the horizontal angle of the frontal axis for each synthesized primary polar microphone (8); an arbitrary sound sample number. An arrangement according to claim 6, wherein: OSajSKandbj = cj = K-aj; where K is an arbitrary positive constant. Arrangement according to claim 6 or 7, wherein the number (N) of primary virtual microphones (8) and primary matrix outputs (3) is 2. Arrangement according to any one of the claims. 6 - 8, wherein the time delay of signals from the N virtual speakers to the origin (13) is substantially and practically the same Arrangement according to any one of claims 6 - 9, wherein the secondary matrix device (4) comprises linear coefficients with characters, delays and alteration devices realizing arbitrarily placed and directed coincidental and non-incidental secondary virtual microphones of any polar type with directivity F09) = p + q cos (0) where 9 is the angle of incidence towards the frontal axis of a microphone and where 0 Sp S1, p + q = 1 An arrangement according to any one of claims 6 to 9, wherein the secondary matrix device (4) comprises linear coefficients with signs, delays, and filtering devices realizing coincidental or non-coincidental secondary virtual microphones of any non-polar type. Arrangement according to any one of claims 6 - 1 1, wherein control devices are arranged for adjusting the coefficients (2) of the primary matrix device so as to change the primary virtual microphone types. Arrangement according to any one of claims 6 - 12, wherein control devices are arranged for adjusting the coefficients and delay of the secondary matrix device (4) and thus changing the type, direction, position, filtering device and delay of the secondary virtual microphones. Arrangement according to any one of claims 6 - 13, wherein said primary and 15. 16. 17. 18. secondary matrices are integrated. A method for processing sound in which direction-coded audio signals (24) in the form of B-format (W, X, Y, Z) or B-shaped signals via a primary matrix (25),. which synthesizes virtual coincidental polar microphones, is converted into temporary signals which are used as feed to virtual speakers (26), the method comprises feeding the signals from said virtual speaker to a secondary matrix (27) which realizes multiple coincidental or non-coincident secondary virtual microphones and thus reworked into multiple outputs (28) suitable for subsequent speaker playback of mono, stereo or surround sound. A method according to claim 15, wherein said direction-coded audio signals comprise one or more of: - recorded B-format (W, X, Y, Z); - artificially created B-format; - saturated B-shaped space impulse response; - artificially created B-shaped space impulse responses; - recorded B-shaped similar signals (W, X, Y, Z); - artificially created B-format similar signals; - saturated B-shaped similar spatial impulse responses; - artificially created B-shaped similar spatial impulse responses. The method according to any one of claims 15 or 16, wherein said secondary virtual microphones are one or more polar or non-polar types. A computer program for processing direction-coded audio signals in the form of B-shaped (W, X, Y, Z) or B-format-like signals into a digital surround sound in which direction-coded audio signals, via a primary matrix which synthesizes virtual see v, cr: ÖW polar microphones, converted into temporary signals which are used as feed to virtual speakers, the program comprising a procedure for feeding the signals from virtual speakers via a secondary matrix which converts to multiple channels corresponding to coincidental or non-incidental secondary polar or non-polar virtual microphones suitable for subsequent speaker playback of mono, stereo or surround sound. A computer-readable product with an instruction set for processing from direction-coded audio signals in the form of B-format (W, X, Y, Z) or B-format-like signals to a digital surround sound where direction-coded audio signals, via a primary matrix which synthesizes virtual polar microphones , converts to temporary signals which are used as input to virtual speakers, the instruction set comprises a procedure for feeding the signals from virtual speakers via a secondary matrix which converts to multiple channels corresponding to coincidental or non-coincidental secondary polar or non-polar virtual microphones suitable for subsequent speaker playback of mono, stereo or surround sound.