EP1021062B1

EP1021062B1 - Method and apparatus for the reproduction of multi-channel audio signals

Info

Publication number: EP1021062B1
Application number: EP20000100054
Authority: EP
Inventors: Johannes Böhm; Andreas Aust; Jürgen Schmidt
Original assignee: Thomson Licensing SAS
Current assignee: THOMSON LICENSING
Priority date: 1999-01-13
Filing date: 2000-01-05
Publication date: 2012-12-12
Anticipated expiration: 2020-01-05
Also published as: EP1021062A2; EP1021062A3; DE19900961A1; JP2000228800A

Description

The invention relates to a method and an apparatus for the reproduction of multi-channel audio signals, in particular for reproduction by virtual surround systems.

Prior art

A spatial representation of audio signals is possible by means of multi-channel sound reproduction, called surround sound hereinafter. The multi-channel sound usually comprises three channels L, C, R (Left, Centre, Right), which are arranged spatially in front of the listener, and two channels Ls, Rs (Left surround, Right surround), which are arranged spatially behind the listener. In addition, a sixth channel for special effects may possibly be provided. In this case, each channel is reproduced by means of a separate loudspeaker. If such a multi-channel sound loudspeaker system is not available or cannot be used owing to installation problems, it is possible also to obtain approximately identical acoustics just using the front or main loudspeakers, by means of a so-called virtual surround system. The signal of the centre channel is in this case split equally between the two loudspeakers, with the result that the centre signal C seems to come from a so-called phantom point directly in front of the listener. The surround channels Ls, Rs are fed to the two main channels L, R, the virtual surround effect being attained by these surround channels being processed beforehand. This is done by taking account of the propagation time differences and a shift in the frequency characteristic which the surround signals experience when they reach the listener from the direction of the surround sound sources. This can be done with the aid of the so-called head related transfer functions (HRTFs), which describe the route to the auditory canal inputs, it being necessary to use the individually applicable HRTF for each of the sound incidence directions. Equally, it is possible to use a model listening space or a combination of HRTFs and model listening space for this purpose. In particular in the case of static objects or objects which only move slowly in the surround channels, the quality of the acoustic reproduction by known virtual surround systems is, however, behind the quality in the case of reproduction by a high-grade multi-channel audio system. A prior art method is disclosed in US 5 822 438 .

Invention

The invention is based on the object of specifying a method for improving the acoustic impression during the reproduction of multi-channel audio signals. This object is achieved by means of the method specified in Claim 1.
The invention is based on the insight that the spatial perception of sounds situated behind the listener is improved if the listener moves his head slightly. According to the invention, this improvement in the acoustics can also be achieved if, instead of the head movement, the virtual surround sources are moved.
In principle, the method for the reproduction of multi-channel audio signals comprising main channel signals and one or more surround channel signals, the surround channel signals being processed for the purpose of mapping the surround channels onto the main channels before being fed to the main channel signals and being reproduced by means of the main loudspeakers, consists in the fact that the virtual position of at least one surround loudspeaker is varied in order to improve the acoustics.
It is advantageous, if the main channel signals comprise two front channel signals, the surround channel signals comprise a centre channel signal and also two rear channel signals, and the surround loudspeakers have two rear loudspeakers, if only the virtual position of the rear loudspeakers is varied by alteration of the rear channel signals.
It is particularly advantageous in this case if the virtual position of the rear loudspeakers is varied with constant periodicity and/or as a function of the amplitude of the front channel signals.
In this case, the virtual position of the rear loudspeakers can be described in particular by a filtering of the rear channel signals.
In this case, the virtual position of the rear loudspeakers is advantageously varied by alteration of the filter parameters.
Equally, it may be advantageous for the virtual position of the rear loudspeakers to be varied exclusively or additionally by means of a different propagation delay, amplification or matrixing of the rear channel signals.
In this case, the multi-channel audio signals may be, in particular, MPEG2, Dolby AC3 or Dolby ProLogic signals.
In principle, an apparatus for the reproduction of multi-channel audio signals , having a processor unit for mapping the surround channels onto the main channels, after mapping the processed surround channel signals being fed to adders and being added to the main channel signals by means of the said adders, consists in the fact that the processor unit varies the virtual position of at least one surround loudspeaker in order to improve the acoustics.
In this case, the processor unit preferably comprises one or more filters, the filter parameters describing the virtual position of the surround loudspeakers.
In this case, it is particularly advantageous if the processor unit is fed a filter control signal which has a constant periodicity and/or is varied as a function of the amplitude of the front channel signals.
Finally, it may be advantageous if the processor unit additionally has, individually or in combination, one or more delay stages, amplifier stages and also matrix stages.

Drawings

Exemplary embodiments of the invention are described with reference to the figures.
In the figures:

Figure 1: shows a schematic sketch of the method according to the invention, in which the surround signals are rotated or moved back and forth;
Figure 2: shows a simplified block diagram of an apparatus without the centre channel being taken into account;
Figure 3: shows a simplified block diagram of an apparatus with the centre channel being taken into account;
Figure 4: shows a first exemplary embodiment with alteration of the filter coefficients;
Figure 5: shows a second exemplary embodiment with static filter coefficients and alteration of the propagation time;
Figure 6: shows a third exemplary embodiment with static filter coefficients and variation of the loudness;
Figure 7: shows a fourth exemplary embodiment with static filter coefficients and alteration of the parameters of a matrixing device.

Exemplary embodiments

Figure 1 illustrates two possible embodiments of the method according to the invention. In this case, an arrangement comprising two real loudspeakers SL, SR and three virtual loudspeakers SC, SLs, SRs is shown in each case for different instants 1a, 1b, 1c and 2a, 2b, 2c, respectively. In this case, the loudspeakers SL and SR are arranged in front of the listener Z on the left and right, the virtual loudspeaker SC is arranged between the former loudspeakers and the virtual loudspeakers SLs, SRs are arranged behind the listener on the left and right.
The listener Z is situated in the so-called "sweet spot", which is defined by the fact that the listener Z is situated exactly in the middle between the two front loudspeakers SL and SR, the front loudspeakers L and R and the listener Z approximately forming an equilateral triangle. This condition can easily be fulfilled for multimedia PCs, since the user usually sits in front of the monitor and the loudspeakers are situated to the sides of the monitor. It is likewise the case with stereo televisions, since television viewers prefer to look straight at the screen in order thus to obtain both a good visual impression and stereo sound.
In this example, the actual position of the virtual loudspeakers Ls, Rs is, in an exemplary manner, 125° and -125°, respectively, referring to the centre loudspeaker SC. In accordance with the upper embodiment, this configuration of the virtual loudspeakers can now be rotated for example firstly in the clockwise direction (1a), then back to the starting position (1b), then in the anticlockwise direction (1c) etc. In this case, the centre loudspeaker SC is likewise virtual, in which case the virtual position, as in the present example, can be fixed or else likewise rotated (not illustrated). In accordance with the second embodiment, the virtual loudspeakers SLs, SRs can instead be moved forwards (2a) and backwards (2c) referring to the starting position (2b).
Figure 2 illustrates a simplified block diagram of an apparatus without the centre channel being taken into account, in which a processor unit Proc converts the multi-channel audio signals L, C, R, Ls and Rs, in the manner according to the invention, into virtual surround signals L', R' with improved acoustics. In this case, the signal of the centre channel C is split equally between the two input signals L and R, it being usual for there to be attenuation by 3 dB. The addition to the channels L, R is effected by adder units Add1, Add2. The two rear surround channels Ls, Rs are processed by the processor unit Proc in such a way that the virtual surround sources are moved in the manner described in Figure 1. The resultant output signals Ls' and Rs' are then likewise added to the front channel signals L, R. The processor unit Proc is in this case driven by a filter control signal FiltCntr. This filter control signal may have a constant periodicity or be determined by the amplitude of the front channels.
Figure 3 illustrates a simplified block diagram of an apparatus with the centre channel being taken into account. In this case, the signal of the centre channel C is not split between the input signals, but rather is processed by the processor unit Proc in the same way as the rear surround signals, in which case particular attention should be paid to ensuring that an unstable sound impression is not produced.
Figure 4 illustrates a first exemplary embodiment of the processor unit Proc. In this case, the surround signals Ls, Rs are respectively fed to a filter Fls, Frs. These filters then contain filter sets which reproduce the arrangement of the virtual surround loudspeakers. A control unit (not illustrated in the figure) then transmits the filter control signal FiltCntr to the two filters Fls, Frs and simulates a dynamic slight movement of the virtual surround sources by corresponding variation of the filter sets of the two filters.
Figure 5 illustrates a second exemplary embodiment, in which the filter coefficients of the filters Fls, Frs are fixed and a propagation delay is instead effected in the two channels. For this purpose, delay stages T1a, T1b are provided for the left surround channel and delay stages T2a, T2b are provided for the right surround channel. In this case, a delay can be effected by the front delay stages T1a, T2a, the rear delay stages T1b, T2b or by both delay stages.
Figure 6 shows a third exemplary embodiment, in which the loudness of the two channels is altered instead of the filter coefficients being varied. In a manner corresponding to the second exemplary embodiment from Figure 5, in this case the variation of the loudness may be effected by the front amplifier stages A1a, A2a, by the rear amplifier stages A1b, A2b or by both amplifier stages.
Finally, Figure 7 shows a fourth exemplary embodiment in which the movement of the virtual surround sources is changed by variation of parameters of a matrix stage. In this case, too, the variation may be effected either upstream of the filters in the matrix stage M1, or downstream of the filters in the matrix stage M2 or in both matrix stages.
The configuration of the processor unit Proc is not restricted to the abovementioned exemplary embodiments; thus, by way of example, a combination of the abovementioned exemplary embodiments is also conceivable. Furthermore, the processor unit, filters, delay stages, amplifier stages and matrix stages do not have to be present as separate units. Finally, the method according to the invention can be carried out entirely or partly by corresponding software.
As already explained, the centre channel can also be processed in the same way as the rear surround signals. Equally, the invention can also be used if multi-channel audio signals with just one rear channel are present (so-called single surround, such as e.g. ProLogic). In this case, the rear channel is decomposed into two rear channels, decorrelation of the channels being recommended. In the course of further processing, the altered virtual starting position of the signals is then taken into account by adapted filters Fls, Frs.
The invention can be used for the reproduction of multi-channel-sound audio signals by a wide variety of audio reproduction apparatuses such as, for example, DVD players, computers with sound cards, headphones, radio receivers or television receivers. In this case, the signals may be, for example, MPEG2, Dolby AC3 or Dolby ProLogic signals.

Claims

A method of processing multi-channel audio signals for the reproduction by a virtual surround system, wherein virtual surround loudspeakers are assigned to surround channel signals, comprising the steps of:
receiving main channel signals (L, R) and one or more surround channel signals (C, Ls, Rs);

varying the position of the virtual rear loudspeakers (SLs, SRs) with constant periodicity and/or as a function of the amplitude of the front channel signals (L, R);

generating virtual surround signals from the processed surround channel signals and the main channel signals; and

outputting the generated virtual surround signals for reproduction.
Method according to Claim 1, wherein the main channel signals comprise two front channel signals (L, R), the surround channel signals comprise a centre channel signal (C) and two rear channel signals (Ls, Rs), the position of virtual rear loudspeakers being varied by alteration of the rear channel signals (Ls, Rs).
Method according to Claim 2 , wherein a filtering of the rear channel signals (Ls, Rs) describes the position of the virtual rear loudspeakers (SLs, SRs).
Method according to Claim 3, wherein the position of the virtual rear loudspeakers (SLs, SRs) is varied by alteration of the filter parameters.
Method according to one of the preceding claims, wherein the position of the virtual rear loudspeakers (SLs, SRs) is varied exclusively or additionally by means of a different propagation delay, amplification or matrixing of the rear channel signals (Ls, Rs).
Method according to one of the preceding claims, wherein the multichannel audio signals are MPEG2, Dolby AC3 or Dolby ProLogic signals.