FR2590757A1 - System for creating a pseudo-stereophonic effect in monophone sound reproduction - Google Patents

Abstract

The system comprises a single active audio filter (FILPT) of the all-pass type which sends the audio signal to one of the two channels 1, 2 of the stereophonic apparatus while inserting a phase shift which varies progressively with the frequency and which has a central frequency, preferably adjustable, within the frequency band corresponding to the middle tones. The frequency response of the all-pass filter can also be modified so as to insert an amplitude modification of the audio signal sent to the said channel, enabling the pseudo-stereophonic effect to be heightened.

Description

SYSTEM FOR THE CREATION OF A PSEUDO-STEREOPHONIC EFFECT
IN THE REPRODUCTION OF HIS MONOPHONIC
The present invention relates to a system for creating a pseudo-stereophonic effect (for short, pseudostereo) in the reproduction of monophonic sound by means of a stereophonic reproduction and broadcasting apparatus.

 When reproducing monophonic sound using a stereophonic device, it is often desirable for more pleasant listening to process the audio signals in such a way as to give the listener a sensation similar to that offered by the reproduction of a its stereophonic.

 It is obvious that the processing of a monophonic audio signal cannot give the sound itself a truly stereophonic content, but, by acting appropriately, it is possible to obtain from such a signal two different audio signals as they simulate those making up stereophonic sound, to be sent to the speakers of the two channels of the reproducing and broadcasting apparatus.

 This processing produces what is called "pseudo-stereophonic image" or the pseudo-stereophonic effect ".

 A technique commonly used to create the pseudo-stereophonic effect in the reproduction of a monophonic sound is that of dividing into bands the spectrum of the monophonic audio signal to be reproduced, and of obtaining therefrom the signals to be sent. to the speakers of the two audio channels of the reproduction and broadcasting device (for short, reproduction device).

 According to a solution based on this technique, part of the spectrum is sent in one of the two channels of the reproduction device, while the complementary part of this is sent in the other channel.

 The acoustic signals broadcast by the loudspeakers of the two channels will therefore be different from each other, and, if the audio frequency spectrum is divided into bands according to appropriate criteria, this will give the listener impression of being in the presence of the reproduction of a stereophonic sound.

A system based on this type of technique is described, for example, in the publication: - "DIGIT 2000 - VLSI Digital
TV System ", ITT Semiconductors, Publication order number: 6251-190-2E, p. 6.13, August 1982.

 The general diagram of a system of this type is shown in FIG. 1, in which the block indicated by the symbol FIL represents the filter whose object is to subdivide the spectrum of the audio signal as desired.

 The output of the filter is sent to channel 1. The complementary part of the spectrum is obtained by means of the adder block 1, which delivers as output the difference between the two input signals, which is then sent to channel 2.

 A drawback of this kind of technical solution is that, especially when the number of bands into which the audio frequency spectrum is subdivided is very small (up to the borderline case where a single band, for example that of the frequencies corresponding to the medium tones, is sent in one channel and where the remaining part of the spectrum is sent in the other channel, as is the case, for example, in said "DIGIT 2000" system), the acoustic powers sent to the two channels can be very different l relative to each other, especially when sounds with particular musical and / or acoustic contents are reproduced, which constitutes a fact which can make listening less pleasant.

 Naturally, to a greater number of bands into which the audio frequency spectrum is subdivided corresponds, if the selection of the bands is appropriate, a pseudo-stereophonic effect of better quality. On the other hand, this increases the complexity of the reproduction apparatus, and therefore its cost, and, in general, the difficulty of its adjustment also.

 According to another proposition, the bands into which the spectrum of the audio signal is subdivided are mixed together with an appropriate phase relation, then sent in a channel. Instead, the original monophonic signal is sent in the other channel.

A system of this type is described, for example, in the publication: The German 2-Carrier System for Terrestrial TV-Sound Transmission System and Integrated Circuits for "High Quality" TV-Receivers, by U. Buhse, in IEEE
Transactions on Consumer Electronics, Vol. CE-28, NO 4, p. 489, November 1982.

The general diagram of this system is shown in
Fig. 2. The two blocks FIL I and FIL 2 represent two filters with complementary types of amplitude frequency response. More precisely, these are, respectively, a band-pass filter and a band-repair filter having their center frequencies f0 and their bandwidths BW substantially identical.

 The central frequency and the bandwidth are understood here as being the central frequency and the width (at - 3 dB) of the passband for the bandpass filter and as being the central frequency and the width (at - 3 dB) of the band removed for the band rejection filter.

The two filters are of order two and are only formed by resistors and capacitors (which means that they are both passive RC filters). Blocks 1 and 2, indicated by the symbol INV, simply represent reversing blocks, and each consist of an operational amplifier. In addition to ensuring the attack on the lines of the two sound channels, they are also necessary, respectively, to obtain the phase inversion between the signals entering the two filters FIL 1 and
FIL 2 and to perform the addition function on channel 2. The value of fO is approximately 1 kHz.

Consequently, in addition to the phase inversion introduced in the two channels, while the original unmodified monophonic signal is sent to a channel, the same monophonic signal, after having been modified by the two filters and by the adder block, is sent to the other channel. More precisely, the sound components belonging to the intermediate frequency band (ie the frequency components around fO, which go from fO - BW / 2 to fO + BW / 2) are removed by the filter
FIL 2 of the monophonic signal sent to this second channel.

However, are also sent to this latter channel, in phase opposition with respect to the other components, the components belonging to the same intermediate band as that obtained by filtering by the block FIL 1.

 In this way, the signal components, whatever their frequency, have a substantially similar amplitude in the two channels. However, although the components of the signal having a frequency located inside the intermediate band have the same phase in the two channels (the inversion introduced by block 2 actually compensates for the inversion introduced by block 1) , the components of the signal with intermediate frequencies have, in the two channels, substantially different phases. The phase relationship between identical frequency components sent to the two channels in the intermediate band varies with frequency; in particular, at the frequency f0, the components sent to the two channels are in phase opposition. Thanks to that, the technical proposal shown in Fig. 2 produces a good impression of pseudostereophonic listening.

 In practice, the perception of the directional effect in listening does not exist for frequencies lower than a few hundred Hz. In addition, the presence at high frequencies also of a phase opposition between the acoustic signals in the two channels would decrease the effective quality of listening. To obtain a good pseudo-stereophonic effect, it is therefore desirable that there is no substantial phase difference between the components of the signal sent to the two channels except in the intermediate frequency band.

 To make a system based on this technical proposal, two audio filters are necessary.

 Furthermore, only a precise agreement of the characteristics of the two filters (in particular between the central frequencies and the bandwidth of the two filters as well as between their gains in the passband) makes it possible to optimize the pseudo-stereophonic effect. .

This may require careful selection of the components of the filters themselves in order to optimize the characteristics of the system.

In addition, it should be remembered that in a filter
Passive second order RC, because its blades are always real, what is called "the overvoltage factor"
Q is always rather low (Q <0.5), which means that it is not possible to make a filter of this type which has a very low value of ratio between the bandwidth BW and the central frequency fO ( we indeed know that, for these filters, in fact the relation fo / BW = Q) applies. It follows that if, as in the system quoted, two passive RC filters of order two are used to make the FIL 1 blocks and
FIL 2, it is not possible to make a system in which, once the value of fO is fixed, the frequency band affected by the processing (ie that within which a difference of substantial phase is introduced between the components of the signals sent to the two channels) either of reduced width or can be in one way or another capable of being fixed by the designer of the system within a sufficiently large interval , for example to allow a substantial adjustment of pseudo-stereophonic effect by one operator.

On the other hand, by using two active RC filters of order two which can have poles with imaginary parts different from zero, and, consequently, a "overvoltage factor" Q higher than that which can be obtained with filters Passive RC to make FIL 1 blocks and
IDF 2, we would make this technical proposal rather complicated and costly.

 The use of passive RC filters of higher order than order two, in addition to increasing the complexity and the cost of the system, would make the selection of components even more critical to obtain an optimal agreement between the characteristics of the blocks. WIRE 1 and WIRE 2.

 In addition, considerations of cost and size make it undesirable to use second order passive filters also formed with self induction coils (RLC passive filters), which can have Q values greater than those of passive filters. RC.

 A technical proposal of the latter type does not lend itself very well to total integration into a monolithic integrated circuit.

 The main object of the present invention is to provide a system, usable in a stereophonic device for acoustic reproduction and broadcasting, generating a pseudo-stereophonic listening effect when a monophonic sound is reproduced; the system maintaining substantially equal the acoustic powers diffused by the two channels whatever the frequency, being produced with a unique audio filter, allowing the determination of the magnitude of the effect itself in an easy way and being easily integrated in a monolithic circuit without requiring the use of components external to the circuit itself.

 According to the invention described here, such an objective is achieved by a system which sends in a channel the monophonic signal to be reproduced without any processing, and which sends in the other channel the same monophonic signal processed by means of an active filter. substantially unique of type for all frequencies.

In order to make the invention more easily understandable, the description will continue with the illustration of some embodiments thereof which are currently preferred and with reference to the accompanying drawings, in which
Figures 1 and 2, as already mentioned in the description of the existing technique, show two systems known for creating a pseudostereophonic effect;
Figure 3 is a general diagram of the system of the present invention;
Figures 4a and 4b are, respectively, the amplitude response graph and the phase graph of the active audio filter of the two-pass all-pass type (FILPT) usable in the diagram of Figure 3;
Figure 5 is a diagram of the filter circuit (FILPT) of Figure 3, according to a preferred embodiment;
Figure 6 is another embodiment of the filter (FILPT) of Figure 3, particularly preferred in the case of integration of the system of the invention in a monolithic integrated circuit manufactured according to technology
MOS (metal-oxide-semiconductor);
Figure 7 is a graphical illustration of a characteristic clock signal.

 The general diagram of a preferred system producing the pseudo-stereophonic effect according to the present invention is shown in Figure 3, in which the block indicated by FILPT represents an active filter of the pass-all type.

The transfer function f (s) of a non-inverting all-order filter of order two is given by
f (s) = - (wo / Q) s + wo2 s2 + (wn / Q) s + wg2 where s is the complex variable s = 6 + jw. Q and fO = w0 / 2tr are commonly defined, respectively, as the overvoltage factor and the central frequency of the filter.

 The frequency response, amplitude (A) as well as phase (P), of this filter, is shown in Figure 4.

In particular, the phase characteristic as a function of frequency depends on the value Q. For the system proposed here, a characteristic value of f0 is, for example, between approximately 800 Hz and 1 kHz.

 We can deduce from the observation of Figures 3 and 4 that, thanks to the system described here, the low and high frequency components of the signals sent to the two channels 1 and 2 are substantially equal to each other (and with the original monophonic signal) in terms of amplitude as well as in terms of phase (we must remember that a phase shift of 3600 is equivalent to a zero phase shift).

 The components located in the medium frequency range of the signals sent to the two channels, on the other hand, although having equivalent amplitudes, have substantially different phases. The phase difference between these components gradually increases with the frequency from 0 to 3600, and, at the center frequency fO, the components of the signals on the two channels are in phase opposition, thus providing the maximum effective phase shift between the audio signals in the two channels, as is necessary for the system to produce a good impression of pseudo-stereophonic listening.

 The central frequency fo and the characteristic of the variation of the phase shift introduced by the filter as a function of the frequency depend on the way in which the all-pass filter is produced, which can have different characteristics depending on specific requirements.

 By varying the Q value of the filter, it is possible to vary the bandwidth of the signal components which are phase shifted by the filter, and therefore, one can obtain an adjustment of the magnitude of the pseudo- effect. stereophonic. In particular, filters with a high Q value can be produced, to which corresponds a characteristic of abrupt variation of the phase with the frequency inside the band of the medium frequencies. In this case only, the components of a narrow frequency band have a substantial phase shift between the two channels and contribute to the creation of the pseudo-stereophonic effect perceived by the listener.

 According to a particularly sophisticated embodiment of the system of the present invention, the central frequency of the filter as well as the law of variation of the phase shift introduced by the filter as a function of the frequency can both be programmable separately or jointly, that is to say say adjustable within a certain range of adjustment, either during focusing operations, or even using appropriate control means accessible to the operator of the sound reproduction apparatus, in the aim to offer the possibility of adapting the system to different reproduction requirements. The ways in which such possibilities for adjusting the characteristics of an active audio filter of the type for all frequencies can be realized are well known to those skilled in the art and will not be the subject of a detailed description here.

 A practical embodiment of an active all-pass filter of the order two suitable for use in the system of the present invention is shown in Figure 5, in which the active audio filter of the all-pass type is composed inside the rectangle drawn in dashes (corresponding to the FILPT block in Figure 3). The blocks indicated by the symbol INV and designated respectively by the numbers 3, 4 and 5, are inversion blocks with unity gain, and blocks 6 and 7 also indicated by the symbol AMP are operational inverting amplifiers.

The configuration of the active filter of the order two of the all-pass type shown in FIG. 5 lends itself to the realization in a monolithic integrated circuit, and in an even more efficient way if the resistors R1,
R2, R3 and R4 are produced by what is called the "switched capacitors" technique. Furthermore, in an embodiment of this type, the need for inversion blocks 3 and 4 is eliminated thanks to the fact that the signal inversion produced by these can be carried out simply by using a switched capacitor structure of the inverting type, respectively, for the two resistors R2 and R3 in Figure 5.

The circuit diagram of the all-pass filter for the creation of the pseudo-stereophonic effect produced with this technique is shown in Figure 6, in which at each resistance (R1, R2, R3 and R4) of the circuit of Figure 5 has been replaced by an equivalent switched capacitor structure, respectively CR1, CR2,
CR3 and CR4, according to a well known method, familiar to those skilled in the art.

 As can be seen in Figure 6, the two structures CR1 and CR4 are of the "non-inverting" type, while the structures CR2 and CR3 are of the "inverting" type, and therefore also carry out the phase inversion which , in the diagram of Figure 5, was carried out, respectively, by the reversing blocks 3 and 4.

 The switches of the switched capacitor structures are all synchronous with each other, and are designed, for example, to be originally in their quiescent condition, and in the opposite condition (indicated by the dash lines in Figure 6) during the period of their activation. The switches are controlled by means of suitable clock signals. These signals typically have a square waveform with a utilization factor of 50%, as shown for reference in the graph in Figure 7. The network switches of Figure 6, for example, are assumed to be in their quiescent condition when the control clock signal is at a high level, and in their actuated condition when this signal is at a low level.

The embodiment of the system shown in FIG. 6 is particularly suitable when such a system is produced in the form of a monolithic integrated circuit according to MOS technology. In such technology and in its various derived forms (N-MOS, C-MOS, etc.), in fact, as is well known, the operational amplifiers (6 and 7 in Figures 5 and 6) as well as the switches (which are simply made by means of MOS transistors whose gate electrode is attacked by appropriate clock signals), as well as the capacitors C1, C ?,
C3 and C4 or that the capacitors of the structures CR1, CR2,
CR3 and CR4 are all easily achievable. As far as capacitors are concerned, the monolithic integration MOS technology also makes it possible to obtain a very high precision of the ratio between the values of the capacitors of the network, which is an essential requirement. to be able to optimize the characteristics of the system. In addition, the capacitors necessary to make a pass-all filter with a center frequency of the required value (of the order of 1 kHz), using a clock signal with a characteristic frequency (for example, 125 kHz), are not excessively large, which means that the system can be integrated without requiring an excessive surface on the pad.

 In addition, in an MOS type integrated circuit, it is also easy to generate the necessary clock signals, which are typically obtained from general synchronization signals existing inside the MOS integrated circuit.

 Although for the sake of simplicity a practical embodiment of the system of the invention has been described in an output configuration, it is clear that the system of the invention can also be implemented in systems which use a "totally differential" processing of the audio signal. In the latter case, the original signals, as well as their inverses, being always available in the network, the production of the pass-all filter according to the configurations described in Figures 5 and 6 becomes even more simplified and compact.

In this case, in fact, the need for the inversion block 5 can also be eliminated since the inverse of the signal B (or S) is itself already available inside the system.

 Various modifications can easily be made by those skilled in the art in the circuit diagrams described above for illustration purposes only, while remaining within the scope of the applicability of the present invention.

 For example, when, for reasons of ambient equalization, the best pseudo-stereophonic listening effect is obtained by sending two audio signals, the amplitudes of which have slightly different characteristics, into the two channels of the reproduction and broadcasting apparatus. between them as a function of frequency, the system of the invention can be made by substantially following the same principle but by slightly varying the amplitude response as a function of the frequency of the filter for all the frequencies in such a way that it introduces , in the signal sent to one of the two channels of the stereophonic device, an amplitude distortion (modification) so as to provide the desired amplitude difference as a function of the frequency between the signals present on the two channels the sound reproducing apparatus.

 Finally, this 2 "distortion of amplitude can even be programmable in order to give the system greater flexibility of adjustment.

Claims (10)

Claims 1. System for creating a pseudo-stereophonic effect in the reproduction of monophonic acoustic signals by means of a stereophonic device, characterized in that the monophonic signal to be reproduced is sent substantially unmodified to one of the two channels (1, 2 ) of the stereophonic device, and to the other of said channels is sent the same monophonic signal processed by means of a single active audio filter (FILPT) substantially of the all-pass type which introduces into the components of the signal belonging to an intermediate frequency band a phase shift which varies gradually with the frequency. 2. The system of claim 1 characterized in that said active all-pass filter has a frequency response such that it does not introduce any amplitude variation in the audio signal. 3. The system of claim 1, characterized in that the frequency response of said active filter of the all-pass type is such that it introduces a modification of amplitude of the audio signal according to the frequency making it possible to make the pseudo-stereophonic effect. 4. The system of claim 1, characterized in that the central frequency of said active filter of the passetout type is adjustable. 5. The system of claim 1, characterized in that the law of variation of the phase shift as a function of the frequency of said active filter of the all-pass type is adjustable.  6. The system of claim 3 wherein said modification of amplitude of the audio signal as a function of the frequency introduced by the filter is adjustable. 7. The system of claim 1 characterized in that said active all-pass type filter is a second order filter. 8. The system of claim 1 characterized in that the system uses simple processing of the audio signal. 9. The system of claim 1 characterized in that the system uses a totally differential processing of the audio signal. 10. The system of claim 1 characterized in that said all-pass filter is produced with a switched capacitor technique.