LU83854A1 - IMPROVEMENTS ON MOUNTINGS FOR MODIFYING A DYNAMIC RANGE - Google Patents

Description

The present invention relates generally to arrangements which modify the dynamic range of acoustic and other signals, namely compressors which compress the dynamic range and expanders which widen the dynamic range. More particularly, the invention relates to improvements in compressors and expanders, improvements which reduce their sensitivity to control by undesirable signals. These improvements are here called "modulation control" for 10 reasons which will be explained later.

Complementary compressors and expanders (compressors-expanders) are often used together to provide noise reduction when the signal is compressed before transmission or recording and expanded after reception or reproduction from the transmission channel. However, it is possible to use separate compressors (without expander) to reduce the dynamic range, for example in order to adapt it to the capacity of a transmission channel, without subsequent expansion when the compressed signal is suitable for the final purpose. Separate compressors are also used in some devices, in particular audio devices which are only intended to transmit or record radio broadcasts or pre-recorded signals already compressed. In some devices, in particular in recording and acoustic reproduction apparatuses, one and the same device is often arranged so as to allow a switchable operating mode, either as a compressor for recording signals, or as an expander to reproduce ies * 30 radio programs or pre-recorded, compressed signals.

The degree of compression or expansion can be expressed in dB, For example, the expression "10 dB of compression" means that a dynamic input range of! F £ Ξ

AT

2 is compressed into an output range of (N — 10) dB. In a noise reduction device, it is said that 10 dB of compression followed by 10 dB of additional expansion provides 10 dB of noise reduction.

The present invention relates in particular to improvements in circuits allowing the dynamic range of an input signal to be modified, circuits which have a bilinear characteristic (the term "linear" in this context means "constant gain") constituted per s 10 l) a low-level linear part rising to a threshold, 2) a non-linear (variable gain) part of intermediate level extending above the threshold and up to a termination point and ensuring a predetermined maximum compression or expansion ratio 15, and 3) a high level linear part having a gain different from that of the low level part.

The characteristic is called "bilinear characteristic" because it comprises two parts with substantially constant gain.

In practice, the threshold and the termination point are not always well-defined "points". The two transition regions in which the intermediate level part merges into the low level and high level linear parts can each have a variable shape ranging from a soft curve to a sharp curve, depending on the control characteristics of the compressor and expander.

It should also be noted that assemblies with 30 bilinear characteristics are distinguished from two other known assembly categories, namely t (a) a logarithmic or nonlinear assembly with fixed or variable slope and having no linear part * the gain varies in the whole dynamic range.

(b) assemblies having a characteristic comprising 3 two or more two parts, only one of which is linear ("unilinear). The invention is also applicable to unilinear circuits, as described below in more detail.

5 An assembly with bilinear characteristic offers particular advantages and is widely used. The threshold can be adjusted above the input noise level or the noise level of the transmission channel to exclude the possibility of noise control of the circuit. The high level portion of substantially constant gain avoids non-linear processing of high level signals which would otherwise introduce distortion.

Two well-known types of bilinear circuits are respectively called "sliding band circuits" 15 and "band, fixed circuits" (or fractional or multiband bands) ·

Sliding band circuits create the desired characteristic specified for high frequency acoustic or sound compression or expansion 20 by applying high frequency boost (for compression) or high frequency cutoff (for l 'expansion) using a high pass filter with variable lower angle frequency. As the signal level in the high frequency band rises, the angle frequency of the filter slides upward so as to narrow the accented or cut band and to exclude the wanted signal from emphasis or the cut. Examples of such circuits can be found in US Patent Re 28,426, US Patent 3,757,254, US Patent 4,072,914, US Patent 3,934,190 and Japanese Patent Application 55,529 / 11 · These circuits can also be configured to act at low frequencies, in which case boosting or cutting of low frequencies is ensured by means of a low-pass filter with frequency of variable angle supérieure k greater.

In fixed band circuits, the frequency spectrum is split into a plurality of bands by corresponding bandpass filters and compression or expansion is performed in each band by a gain control device (which can be, either a limiting device of the type with automatic response diodes, or a controlled limiting device) in the case of a compressor, with some form of reverse or complementary circuit 10 for an expander. Examples of such circuits can be found in U.S. Patents 3,846,719 and 3,903,485 and in the Journal of the Audio Engineering Society, Vol. 15 "No. 4 ^ October 1967" pages 383 to 388.

These fixed band circuits exert an independent action in the various frequency bands.

It is known to build bilinear compressors and expanders, both of the sliding band type and of the fractionated band type, using only a single signal path. However, it is generally preferred to construct such devices by providing a main signal circuit linear with respect to the dynamic range, with a combination circuit in this main circuit, and an additional circuit which draws its input signal from the input or output of the main circuit 25 and the output of which is coupled to the combination circuit, The additional circuit includes a limiter (autonomous or controlled) and the limited signal of the additional circuit accentuates the signal of the main circuit in the circuit combination in the case of compression but attenuates the signal of the main circuit in the case of expansion. The limited signal of the additional circuit is weaker than the signal of the main circuit in the upper part of the dynamic input range. The main and additional circuits are preferably and

AT

(5 very advantageously constituted by separately identifiable signal channels. Several additional circuits are generally provided in the case of fixed band devices. A bilinear device comprising main and additional circuits 5 is often called "two-channel device".

These known two-way compressors and expanders are particularly advantageous, as they allow the desired type of transfer characteristic to be established in a precise manner without problems of high level distortion.

The portion of low level of substantially constant gain is determined by giving the additional circuit a threshold greater than the noise level) below this threshold, the additional circuit is linear. The intermediate level part is created by the region in which the limiting action of the additional circuit or channel becomes partially effective and the substantially constant gain level part increases after the limiter has become fully effective, from so that the signal of the additional channel stops growing and becomes negligible compared to the signal of the main circuit or main channel. In the highest part of the dynamic input range, the output signal of the assembly is in fact only constituted by the signal which the linear main channel passes, that is to say that it is linear with respect to the dynamic range.

Examples of these known circuits can be found in US Patent 3,846,719, US Patent 3,903,485 and US Patent Re 28,426. There are also known analogous circuits which provide similar results, but in which the additional channel offers characteristics opposite to the characteristics of the mixers, and in which the output of the additional channel C "“ '“·“ · “~“ i “6 compression and enhances the signal of the main channel for 1 * expansion (patents U, S, 3 828 280 and 3 875 537) · The invention can be applied to any one of these known bilinear circuits in order to obtain the 5 advantages which are inherent to it · The invention does not is not limited to bilinear circuits and it can also be used to improve the functioning of the previously mentioned unilinear circuits. As described in more detail below, the invention can also be applied to logarithmic circuits. , provided that a difference by relation to a logarithmic transfer function can be tolerated. However, the preferred embodiments relate to bilinear circuits and, unless otherwise indicated, reference will be made to bilinear circuits throughout the present description.

As previously mentioned, it is not essential to create the desired form of bilinear characteristic using such “two-way” techniques. XI exist variants operating with a single way, as described in US Patents 3,757,25k , 3,967,219, 4,072,91, i 3,909,733 and in Japanese patent application 55,529 / 71 "for example · Although these variant circuits are generally not capable of producing results as good as the circuits two-way, and although they may be less convenient and therefore less economical, they can produce essentially equivalent results, therefore the invention is also applicable to these known circuits.

The invention also applies to known compressors and expanders, in which bilinear * circuits connected in series (for example with multiple stages) are used. Such arrangements are described in Belgian patent 889 ^ 28.

In compressors and expanders and, in particular in frequency selective or multiband devices, ± 1 it is clear that it is desirable that strong signals located in a given frequency range do not unduly affect the behavior of signals located in another frequency range. Filtering and equalization in the various circuits have so far been the standard method for dealing with this problem, both in logarithmic devices and in specialized devices, such as unilinear and bilinear circuits which have have been described above. In these prior art circuits, the control DC signal, which controls the variable gain / loss device (e.g., a variable gain device such as a voltage controlled amplifier (VAC) or a variable loss device (variolosser (such as a TEC attenuator) or the variable filter is formed by additive linear combination of bandwidth signals and cut band signals reaching the control circuit. The invention effectively modifies this simple combination in a level dependent manner to optimize the performance of the compressor or expander with respect to the relationship between bandwidth signals to cut band signals. Non-linear operations are performed, including signal straightening in various parts of the spectrum and analyzes of the relative and / or absolute amplitudes. The final command can be obtained by selecting one of the signals, by combining two or more than two signals or by performing non-linear operations such as limiting on at least one of the signals.

In its most general aspects, the invention relates to a circuit making it possible to modify the dynamic range of an input signal, circuit which comprises a frequency selective circuit determining a bandwidth of

U

at 8 frequency, in which the dynamic range modification takes place and a dynamic modification means for effecting a progressive dynamic modification of signal components in this bandwidth or for producing a progressive sliding of the frequency bandwidth, which modifies the dynamic range, the dynamic action of the dynamic modification means being sensitive to increasing levels of what corresponds substantially to the linear additive combination of the passband signal components 10 and the cut band signal components present in the editing, so that, for high level input signals, the dynamic action of the dynamic modification means becomes less sensitive to the cut band signal components. In other words, at low input signal levels, the mqntage behaves much like a conventional compressor or expander. On the other hand, at high input signal levels, the action of the compressor or of the expander is modified by the modulation control circuit according to the invention.

20, A secondary effect is to modify the characteristic of input level-output level transfer of the device at any frequency or combination of particular frequencies. The overall effect is insignificant and may even be negligible at the dominant frequency in 23 bilinear systems. In contrast, in logarithmic systems, the effect of the modulation control, which H.

is exercised mainly in the high-level part of the dynamic range, is to cause a deviation from a purely logarithmic characteristic. This may or may not be important depending on the particular applications envisaged. The invention is based on the observation that ideally, in compressors and expanders, compression or expansion are sensitive only to the levels of the signals located in bandwidths of

L

9 desired frequency and not at signal levels of other frequencies, the latter can be considered as being in the cut bands. For example, in an ideal circuit, compression or expansion should not be affected by level signals outside the bandwidth of the fixed band or the bandwidth of the sliding band (whether the circuit is or not in the rest position). In the case of a sliding band circuit according to the invention, the degree of frequency sliding of the variable band does not exceed what is "necessary to prevent a dominant control signal from being emphasized (in the case of compression) above a reference level.

In its application to bilinear circuits, in particular those having the two-channel configuration, the invention takes another advantage of a characteristic inherent in such circuits t at high input signal levels, the signal of the channel main is significantly larger than the signal (s) from the additional (or secondary) channel (s). Consequently, manipulations of high level signals on the additional channel are essentially inaudible and, except as regards the phase shifts, are practically not measurable (negligible level variations). We will better understand this property of bilinear circuits in the context of a two-lane circuit. The principle, however, also applies to single-path bilinear circuits in which there are two or more than two signal components on the same channel instead of these components being on separately identifiable channels.

The invention takes advantage of the above observations concerning the characteristics of bilinear circuits. Compared with the prior art bilinear compressors and expanders, the invention allows 35. further signal manipulation (modulation control) in the high level region of the input signal where the overall response of the compressor or expander is linear. The relatively low level noise reduction component of the signal can only give rise to these additional manipulations at high signal levels, which ensures that any significant effects on the signal channel will be overshadowed by the large main signal component.

* In two-way bilinear circuits, an effect of the invention is to modify the transfer characteristic of the secondary (or additional) channel so that the characteristic of the secondary channel itself becomes bilinear instead of s flatten or bend downward at high levels of the input signal. This is a consequence of the proportionality aspect of the modulation control. That is, at high input levels, the level of the secondary channel does not fall below a selected fraction the level of the main track (for example a quarter or a tenth).

20 This is acceptable because the signal of the secondary channel still remains significantly weaker than the signal of the main channel at high input signal levels and because the cut band is generally significantly out of phase with the signal channel of the main track “25 pale.

For these same reasons, the invention can be ** adopted in unilinear circuits having a linear response to high signal levels.

From another point of view, the action of the invention is to raise the levels of cut-band signal components at the output of the device to high signal levels, but not to the extent of causing problems. with the recording or transmission channel, i.e. these levels still remain relatively

U

The low π. Raising the levels of such cut-band signal components is not in itself particularly advantageous, but it is necessary to obtain improved dynamic action and noise reduction in the bandwidth. The levels of cut band signals at the device output are raised to high signal levels by reducing the levels of cut band signal components in the control signal channel at ~ high signal levels, or by adopting them. - 10 as an arrangement whereby the control signal is generated as if there was a reduced level of cut-band signal components in the signal used to produce the control signal at high signal levels (for example by filtering and limitation in the control circuits, or by means of attenuation provisions of the control signal as a function of the frequency).

Another advantage of the invention resides in that during listening tests, the "pumping effects" of compressors and expanders with common termination are notably reduced if not eliminated. Consequently, in addition to its use in additional noise reduction devices, the invention is particularly effective when used in compressors acting alone and in expanders acting alone (ie * In compressors intended to be used to compress signals which are not subsequently enlarged and in expanders intended to be used to enlarge signals which have not been previously compressed).

As a basis for the invention, it may be mentioned that, although various practical embodiments of noise reduction circuits have proven to be effective, in service, such circuits deviate to a certain extent from the because of the signal problem

AT

12 cut strip unduly controlling compression and expansion. The effect of these shortcomings manifests itself in several interrelated ways: 1) a weakening of the noise reduction effect in part of the bandwidth of the noise reduction device; 2) noise modulation effects (eg modulation, by the level of a signal of a given frequency, of the noise level in a different part of the frequency spectrum); 3) signal modulation effects (eg modulation, by the level of a signal of a given frequency of the level of a signal of another frequency); 4) transmodulation effects (for example parasitic modulation products resulting from one and / or the other of the last two modulation effects listed).

The extent to which these deficiencies are observable depends on the type of circuits used in the noise reduction device, the recording and re-production equipment, the channel or the recording / reproducing medium and the nature of the signals. In many cases, the gaps are virtually unobservable except with the aid of test instruments. Nevertheless, it is desirable to fill these gaps. Owing to the fact that the previously mentioned shortcomings of known compressors, expanders and noise reduction devices relate to modulation effects, I fc. either signal or noise, the invention described here which aims to reduce these gaps is characterized by the ex- | - "modulation control" pressure.

3Q The gravity of these modulation effects depends to a large extent on the uniformity of the transmission channel between the compressor and the expander. For example, in magnetic tape recording and reproducing apparatus there is a frequency response phenomenon known as "head bounce". Even in professional devices, / - / 13 especially in those operating at 76 cm / s, the reproduction response below 100 Hz is not uniform due to the relationship between the wavelength of the signal on the tape and the size of the reproducing head, which are of the same order. If the compressor / expander device is sensitive to signals in the rebound region of the heads, such signals, when reproduced, may control the expander in a non-complementary manner, so that signals or noise are heard. higher frequencies (eg up to 3 kllz) can be modulated by signals in the region of 100 Hz or below.

In the fixed band circuits (single band and multiband) of the prior art, different filtering techniques have been used to minimize the control of compression and expansion by unwanted signals. In these techniques, sharp filters (for example having a characteristic with steep sides) are placed on the signal path or in the control circuit (of the limiting device).

20 However, the use of channel filters with signals sharper than 6 dB / octave (for example unipolar filters), in multiband compressors and expanders, cause amplitude and phase effects such that, when When the overall signal spectrum is recombined, amplitude and phase errors occur. This problem is made considerably more difficult if filters * more pointed than 12dB / octave are used. Conversely, a filter slope of only 6 or 12 dB / octave may not provide adequate discrimination against all unwanted signals. In the examples of multiband (fixed band) bilinear circuit of U.S. Patent 3,846,719 and in the Journal of the Audio Engineering Çociety, vol. 15 No. 4, October 1967, pages 383 to 388, filters having a slope of 12 dB / octave are used in the signal path of three of the four fixed ban-35s. A flat overall frequency response 14 is only obtained by using a complex filter characteristic in the frequency band adjacent to the sharp filters. It is obvious that such a solution is not universally applicable.

5 In the logarithmic multiband (fixed band) compressor / expander circuit described in Rundfunktechn. Mitteitungen, 22 Jan. 1973, volume 2, pages 63 to 74, the input signal is subdivided into four bands by unipolar filters. However, the control circuits of each band use sharp filters at 18 dB / octave. A sharp control circuit filter (12 dB / octave) is also used in a single fixed band compressor / expander circuit sold under the trade designation "dbx II". However, the use of sharp control circuit filters results in excessive amplification of high level signals outside the bandwidth of the control circuit filters in the case where the signals of large amplitude do not are not present in the bandwidth of these filters, which can result in even over-excitation of the transmission channel, unless sharp cut-off filters are also used in the signal channel.

A known technique which can be called "spectral inflection" is described in Belgian patent 889,427; in * 25 Audio, M.ai 1980, pages 20 to 26; and in article J-6 and the reprint presented at the November 1980 meeting of * the Audio engineering Society, New York, New York State, E.U.A. Spectral inflection is also concerned with the suppression of modulation effects resulting from a compressor / expander non-complementarity due to errors on the transmission channel. According to the teachings of spectral inflection, sharp filtering is provided at least in the compressor at a frequency clearly situated in the normal bandwidth of the device and in the flat response region of the transmission channel. If the spectral inflection is

L

'/ 15 effective in reducing the effects of spurious signal modulation caused by irregularities in the transmission channel / on the other hand it does not solve the problem of excessive frequency slippage in devices with ban-5 sliding, nor that of '' excessive attenuation in fixed band devices.

In view of the above, the invention aims to minimize the control of expansion and compression by unwanted signals while avoiding the associated side effects and / or the complexity of the prior art.

“Although measurable modulation effects are not entirely suppressed by the invention, the effects thereof in applications to audio devices are supplemented by psycho-acoustic masking effects such that the perceived effects are inaudible for most of the listeners and the musical material. That is, only the modulation of a signal (or signals) sufficiently spaced (or spaced) in frequency of the modulating signal is perceived by the human ear. This modulation is minimized by the present invention. Although the modulation of a signal (or signals) by another signal closely related in frequency is less likely to be affected or improved by the invention, it is likely that such phenomena will not be perceived by the ear due to two related effects: a) a weak signal near frequency of a strong signal is masked by the latter, so that said weak signal is inaudible, or b) if the closely approximated signal is audible before compression, or if its level is raised by the compressor in such a way that it becomes audible, then there is a psycho-acoustic tolerance to the effects of modulation due to the low frequency spacing.

U

16

Consequently, the human ear is not able to discern modulation effects of signals at closely spaced frequencies, and therefore the invention need not be fully effective for such signals.

The functional environment of the invention is a fixed band or sliding band compressor or expander circuit in which there is provided a variable circuit element generally controlled by a continuous control current signal, which is especially active in the lower part of the total dynamic garnish. According to the invention, modulation control means are used in the upper part of the dynamic range to prevent the action of the variable circuit element from becoming greater than what is necessary to ensure the desired nominal attenuation of dominant signals, whether the frequencies of these signals are located in the pass band or in the cut band. In practice, controlling the action of the variable circuit element generally consists in acting on the command signal controlling the circuit element under management.

Modulation control can be in the form of active or passive control signal limiting means which come into play at high signal levels, or in the form of means using circuits which detect the presence of high level signals and generate signals which oppose the rise in the level of the control signal; ^ Such limitation of the control signal can be effected on one or more frequency selective control signal channels; in the second case, means are provided to ensure selection or a combination of the control signals so as to apply the optimal control signal to the variable circuit element. When a circuit is used for this high level signal detection or a modulation control generator, this circuit or this generator can operate in various ways, which can be L · 17

Ff to give a measurement of signal levels at least in the upper part of the dynamic range. For example, the modulation control signal can be taken from the input signal or the output signal of the compressor or expander.

5 The modulation control signal in fact constitutes a reference for the direct current control signal applied to the variable circuit element (VCA amplifier or voltage-controlled filter). The reference signal is combined in phase opposition with the direct current control signal (for example by being of polarity opposite to that of the latter or by being suitable for attenuating it), control signal which is generated mainly in response to cut band signal components to establish a limit to the amplitude that the control signal applied to the variable circuit element can achieve in response to signals located in the cut band, "that is i.e. outside the bandwidth of the fixed band or the sliding band. In practice, this limit can be made relatively "sharp" or relatively "crazy". In other words, the "continued" increases in the control signal may be abruptly stopped or they may be allowed to continue at a reduced rate.

It is also possible to obtain the modulation control signal from the variable circuit element (VCA amplifier; or variable filter) by measuring the voltage or current components of said element and, if necessary, performing an equalization in order to ensuring the generation of a signal usable to establish a limit on the amplitude that the control signal applied to the variable circuit element can reach in response to signals located in the cut band.

With regard to the results, the invention, when applied to fixed band or sliding band devices, provides appreciable immunity against signals situated outside the bandwidth of the fixed band 35 or the band slippery. In the case of sliding strip devices, the invention provides yet another related advantage, namely that the sliding strip slides, in response to a dominant signal, only to the extent necessary to restore the gain to the frequency of the signal substantially equal to u-5 nity, at least for levels equal to or greater than a reference level. This reference level is located in the upper region of the dynamic operating range of the device or in the vicinity of this region, for example from 6 to 20 dB approximately of the maximum admissible level. Prior art sliding band circuits are susceptible to excessive slippage and have the effect of shifting the variable filter angle frequency more than is necessary for high level signals which not only causes potential modulation effects, but also results in a weakening of the noise reduction effect in part of the spectrum.

In its application to two-way sliding-band circuits, the invention provides, in its simplest embodiment, rectification and filtering or smoothing of the input signal or the output signal and the combination of the resulting direct current signal with the control signal applied to the variable filter. The level of the reference signal can be adjusted for a desired proportion limit on an additional channel signal dominating by; ; 25 relative to the corresponding component of the main channel signal. For example, the modulation control circuit <3e can be operated in such a way that, for example, in the upper 20 dB of the dynamic range, the limiter provides only the attenuation necessary to maintain the damaging signal component present in the additional channel at a relatively constant propoi "tior of this component in the signal of the main channel (for example at 15 dB below the last component mentioned).

In its application to two-channel fixed-band circuits, the invention is, in its embodiment

AT

19 the simplest, as in the case of sliding band circuits, rectifying and filtering the input signal or the output signal so as to ensure the generation of a modulation control signal which responds mainly to the 5 high level components of the input signal. However, in the case of fixed band circuits, a sharp filter is used in the bandwidth control circuit to produce a bandwidth control signal. In addition, a cut tape control circuit is used to generate a cut tape control signal. The mo- control signal. v dulation constitutes a reference for the cut-band control signal (i.e. it opposes the latter at high signal levels). The cut band control signal thus referred to a reference is compared with the bandwidth control signal and these two signals are combined, generally favoring the larger of them, via a circuit. maximum signal selection, which then controls the VCA amplifier. The general effect of the circuit is to provide the necessary attenuation (overall compression law) in the passband, while avoiding control of the attenuation in the passband by large signal components located in the cut band, j and while avoiding the risk of excessive amplifications of high level signals in the cut band, as they appear at the output of the overall compressor.

In these embodiments and in other variations where a reference attenuation signal is generated, this signal can be taken from the input or output, due to the fact that at signal levels the invention 50 is involved, the input and output levels are approximately identical. In some embodiments, the modulation control signal may be subject to filtering or equalization before rectification. This equalization acts concurrently with filtering or. fixed or variable equalization 35 λ used in signal circuits and in control circuits, to produce a global modulation command which is extremely effective in suppressing any control by signal components located in the cut band, all by interfering at the same time, as little as possible, with the control by signal components located in the bandwidth.

Other embodiments of the invention are described below. For example, the amplified alternating current output signal of the variable sliding band filter - 10 can be distributed over two or more than two bandpass channels, these various channels having selected limitation thresholds, after which the signals of the various channels are rectified and combined to produce a control signal. If appropriate thresholds are chosen, the DC control circuit of the compressor or the sliding band expander then exhibits a maximum frequency-dependent output characteristic, which has the effect of minimizing the control of the compressor or expander by signals outside the slippery bandwidth.

In a variant using only a single control circuit, a low frequency boost circuit is placed in the control amplifier. This circuit is followed by an amplitude limiter then by a high frequency boost circuit. The resulting alternating current signal is then rectified and filtered to form the control signal.

The invention will be better understood on reading the ** V ”detailed description which follows and on examining the accompanying drawings, which represent, by way of non-limiting examples, embodiments.

In these drawings: FIG. 1 is an example of a family of curves representing complementary bilinear compression and expansion characteristics; FIG. 2 is a simplified wiring diagram of a sliding band compressor of the prior art;

AT

Figure 3 is a simplified wiring diagram of a prior art sliding band expander; - Figure 4 is a simplified wiring diagram of a modification of Figures 2 and 3; Figure 5 is a block diagram of a prior art sliding belt compressor; - Figure 6 shows a family of probationary tone curves highlighting the sliding band effect of the circuit of Figures 2 and 4; ^ 0 “Figures 7 to 10 show a series of probative tone curves showing the effects of the modulation control according to the invention applied to a sliding band compressor; - Figure 11 is a block diagram of a preferred embodiment of the invention used in a sliding belt compressor; t - Figures 12 to 15 are block diagrams of other embodiments of the invention used in compres-

Slip band isers; “Figures 16 and 17 are block diagrams of a prior art compressor and fixed band expander; B - Figures 18 to 20 are response curves showing the effects of the modulation control according to the invention applied to a sliding-band compressor; - Figure 21 is a block diagram of a preferred embodiment of the invention used in a fixed-band compressor / and - Figure 22 is a block diagram of an alternative embodiment of the invention used in a fixed belt compressor.

Examples of complementary bilinear compression transfer and expansion characteristics (at a particular frequency) are shown in FIG. 1, which indicates (for the compression characteristic) the low-level part of the signal at substantially constant gain, the threshold, the part in which the dynamic action occurs, the termination point and the high level part with substantially constant gain.

Details of a particular two-way sliding strip bilinear circuit are shown in Figures 2, 3 and 4. The sliding strip embodiments of the present invention are described with reference to this circuit, although the invention is not limited to use in - * tr 10 of such circuits. Figures 2, 3 and 4 are respectively identical to Figures 4, 5 and 10 of U.S. Patent RE 28,426 and further details of said circuits, their operation and their theory can be found in this ü.S. Figure 5 is a block diagram of the circuit of Figure 2 (with or without the modification of Figure 4). The following description of Figures <2, 3 and 4 is largely taken from the Ü.S. RE 28 426.

The circuit of Figure 2 is specifically intended to be incorporated into the recording channel of a consumer tape recorder, two such circuits being necessary for a stereo tape recorder. The input signal is applied to terminal 10 at a follower emitter stage 12, which produces a low impedance signal. This signal is applied, firstly, via a main direct channel, constituted by a resistor 14, to an output terminal 16, and secondly, via a additional channel, - t the last element of which is a resistor 18 also at terminal 16. The resistors 14 and 18 add the output signals of the main and additional channels to provide 130 la. compression law required.

The additional channel is constituted by a fixed filter 20, by a variable cut-off filter 22 comprising a field effect transistor (TEC) 24 (these elements constituting the filter / limiter), and by an amplifier 26, of which I 35, the output is coupled with a limiter or clipper 28 with two diodes and with the resistor 18. The nonlinear limiter suppresses over-modulation of the output signal due to suddenly increasing input signals. The amplifier 26 raises the signal present on the additional channel to a level 5 such as the bend of the characteristic of the limiter or overmodulation suppressor 28, constituted by silicon diodes, is effective at the appropriate signal level under conditions transient. The effective threshold of the overmodulation suppressor is slightly higher than that ν'1 "10 at the syllabic filter / limiter. The resistors 14 and 18 are proportioned in such a way that the necessary degree of compensation for the attenuation is then ensured for the signal present on the additional track.

The output of the amplifier 26 is also coupled 15 with an amplifier 30, the output signal of which is rectified by a germanium diode 31 and integrated by a smoothing filter 32 to produce the control voltage of the T.E.C. 24.

We use two simple RC (resistance-capacitance) filters but we could also use IC (inductance-capacitance) or LCR filters

20 dndUctance-capacity-resistance (equivalent). The fixed filter 20 has a cutoff frequency of 1700 Hz (now 1500 Hz), below which a decreasing compression occurs. The filter 22 includes a series capacitor 34 and a shunt resistor 36 followed by a series resistor 38 and clu T.E.C. 24, the source-drain mesh of which is connected as a shunt resistor. In resting conditions, with a zero signal on the grid of the T.E.C. 24, it is blocked and has an almost infinite impedance; the presence of the resistor 38 can then be ignored; the cutoff frequency of the filter 22 is thus 800 Hz (now 750 Hz) which, as will be noted, is notably lower than the cutoff frequency of the fixed filter 20.

When the signal on the grid increases enough for the resistance of the T.E.C. drops to less than, for example, 1 K (kilohm), resistor 38 actually shunts:

AT

24 the resistor 36 and the cutoff frequency rises, significantly narrowing the bandwidth of the filter. Raising the cutoff frequency is, of course, a gradual effect.

5 The use of a T.E.C. is advantageous because, in a suitable limited range of signal amplitude, such a device behaves essentially as a linear resistance (whatever the signal polarity) whose ohmic value is determined by the control voltage applied to Grid.

Resistance 36 and the T.E.C. are connected in return to an adjustable socket 46 of a voltage divider, which comprises a germanium diode 48 with thermal compensation. The socket 46 makes it possible to adjust the compression threshold of the filter 22.

The amplifier 26 includes complementary transistors, which provide a high input impedance and a low output impedance. Since the amplifier excites the diode limiter 28, a finite output impedance is necessary and is ensured by a coupling resistor 20 50. The dioses 28 are, as described above, silicon diodes and have a sharp bend at the surroundings! half a volt.

The signal applied to the limiter and, consequently, to the resistor 18, can be short-circuited to ground by a switch 53 when it is necessary to put the compressor out of action.

* The amplifier 30 consists of an NPN transistor

with a network at. transmitter time constant 52, which provides increased gain at high frequencies. High frequencies with strong signal strength (eg, a shot in the throat) consequently cause the band in which compression takes place to shrink rapidly, in order to avoid signal distortion.

\ The amplifier is coupled to the smoothing filter 32 i 35 via the rectifier diode 31. The filter! / n 25 includes a series resistor 54 and a shunt capacitor 56. The resistor 54 is shunted by a silicon diode 58 which allows rapid charging of the capacitor 56 for an abrupt attack, concurrently with good smoothing or filtering 5 under conditions of steady state. The voltage across the capacitor 56 is applied directly to the grid • of the T.E.C. 24.

A complete wiring diagram of the additional expander is given in Figure 3, but a full description of this diagram is not necessary since the circuit is substantially identical to Figure 2; as a result, the component values are mostly omitted from Figure 3.

The differences between Figures 2 and 3 are as follows:

In FIG. 3, the additional channel draws its input signal from the output terminal 16a, the amplifier 26a is an inverter, and the signals combined by the resistors 14 and 18 are applied to the input (base) of the emitter-follower stage 12, whose output (emitter) is coupled with terminal 16a. To ensure a low excitation impedance, the input terminal 10a is coupled to the resistor 14 via a follower emitter stage 60. Appropriate measures must be taken to prevent the appearance of 25 a bias in the expander.

The amplifier 26a has been made inverter by adopting as its output the emitter of the second transistor (PNP) instead of its collector. This modification consists in passing the 10 K resistance, 62 (FIG. 3) from the collector to the emitter (FIG. 4), which automatically ensures a suitable output impedance for the excitation of the limiter. Consequently, the resistor 50 is deleted in FIG. 3.

Note that it is important, when aligning a complete noise reduction device, to have

AT

/ "F 26 equal signal levels on the transmitters of the transistors 12, both in the compressor and in the expander. Measurement terminals M are shown connected to. these transmitters.

Figure 4 shows a preferred circuit for replacing the circuit between points A, B and C in Figures 2 and 3. When the T.E.C. 24 is blocked, the second RC network, 22, is inoperative and the first RC, 20, then determines the response of the additional channel. The improved circuit combines the phase advantages resulting from the fact that a single RC section is in quiescent conditions with the attenuation characteristics of 12 dB per octave of a two-section RC filter under cooling conditions. signal.

15 In the practical circuit using T.E.C. of the MPF 104 type, the resistance of 39 K, 36a, is necessary to produce a finite source impedance intended to act in the T.E.C. In this way, the compression ratio at all frequencies and at all levels is maintained at a maximum value of approximately 2. The resistance of 39 K, 36a, performs the same function of limiting the compression ratio in the circuit. perfected as the resistor 36 in the circuit of FIGS. 2 or 3. In addition, this resistor offers a low-frequency path to the signal.

Certain details of the circuits of Figures 2, 3 and 4 have evolved over the years and more modern forms of this circuit have been published and are well known in the art. Reference is made to the specific circuit of the U.S. patent.

RE 28 426 for convenience of presentation.

30 'Figure 5 is a block diagram showing the main elements of the compressor of Figures 2 and 4. The combination circuit 15 represents the combination resistors 14 and 18 of Figures 2 and 3.

The variable tape effect of the tape device sliding ““ ““ * ““ "~~ / 27 of real probationary tone 'from a graphics recorder, obtained from the circuit of figure 2, with incorporation of FIG. 4. The variable band effect is represented by plotting the frequency response of the compressor by means of a low level probative tone (the level of which is below the threshold of the compressor) in the presence of a signal. high level; the probationary tone is detected at the compressor output by means of a tracking filter.

The high level signal causes the compressor cir-10 to operate, the graph showing the effect on the filter transition frequency.

In a sliding band device according to the invention, the amplitude of the high level or dominant signal which causes the sliding band effect must not cause excessive sliding and the presence of other high level signals. also outside the bandwidth of the sliding band should not cause excessive slippage.

By "excessive slip" is meant here a displacement of the transition frequency of the variable filter beyond what is necessary to produce a characteristic of sliding band compressor avoiding an accentuation of the dominant signals above a level reference. The absolute value of the reference level is chosen by the designer of the device, but it is generally about 25 dB 10 dB lower than the highest levels normally used.

Figure 7 shows another family of actual graphical recorder probationary tone curves for the case of a! design similar to that of Figure 2 (including the modification of Figure 4) but with a low level gain of 8 dB and a filter rest frequency of 800 Hz. The level of the probationary tone is located 40 dB below that of the compressor. Curves are plotted for a 100 Hz signal at levels of -20, -10, 0, +10 and +20 dB, where 0 dB 35 is the reference level. We also represented a

| AT

ί v 28 curve corresponding to the absence of signal at 100 Hz. The graphical recording curves corresponding to -10, 0, +10 and +20 dB all start from around 200 Hz. The same is true for relates to Figure 8. In Figures 9 and 10, there are also shown curves for the condition of no signal.

To return to FIG. 7, theoretically, there should not be a slip in response to a signal at 100 Hz because it is located largely outside the bandwidth of the circuit at its lowest frequency (rest ). However, as the signal level at 100 Hz rises, the band slides upward. It is unnecessary for the curves at -10, 0, +10 and +20 dB to slide further than the curve for -20 dB, this in order to avoid any appreciable enhancement of the signal at 100 Hz. Unnecessary sliding for two effects : "A) an appreciable noise reduction effect disappears (during reproduction) from the leak that no accentuation occurs at frequencies at which otherwise this effect would occur and 20 b) when the signal amplitude at 100 Hz varies, it can modulate signals at higher frequencies, since the sliding band also varies under its control, which results in a possible incorrect reconstruction of the signal by the expander if 1? - recording channel or of transmission has an irregular frequency response in the vicinity of 100 Hz.

Figure 8 shows a family of actual graphical recorder probationary tone curves for the same circuit but with the addition of a modulation control circuit, as described below. Slippage practically does not occur for the same signal levels at 100 Hz as those in Figure 7- The slip band compressor is practically immune to strong signals outside its bandwidth. Its sliding band response is 3 ^^ substantially the same as its response below the threshold in the absence of dominant signals.

The effect of the modulation control for the sliding band compressors is further represented by FIGS. 9 and 10 / which are also actual graphical recorder probationary tone curves recorded using the same circuit and with the same level of probationary tone as in the case of FIGS. 7 and 8. In this case, the effect of a dominant signal at 800 Hz, frequency which is included in the desired active area of the circuit, is represented.

10 Theoretically, the sliding need only take place to an extent suitable for not emphasizing the signal at 800 Hz above the reference level of 0 dB. Thus, for example, in the response of FIG. 9, without modulation control, the slip produced by the signal at 800 ^ Hz at the levels of 15 -10, 0, +10 and +20 dB is excessive. FIG. 10 represents the response of the circuit with modulation control: the slip at 0 dB and above is greatly reduced. The effect is gradually reduced for low signal levels, but is somewhat observable at signal level 20 of -10 dB.

FIG. 11 shows in general form a preferred embodiment of the modulation control according to the invention, used in a two-way bilinear sliding band device. As far as possible, we have kept the same reference numbers in FIG. 5 to designate identical or functionally analogous elements. The probation tone response curves in Figures 7-10 are taken from an essential sliding band device.

Irnent as shown in FIG. 11, the elements of the modulation control sub-circuits indicated in the dashed line frame 100 being eliminated from the circuit for the response curves without modulation control. For the purposes of the explanation, it may be mentioned here that the details of the circuit of FIG. 11 are substantially the same as those of the circuit of FIGS. 2 and 4. The circuit can be modified, as described above, without affecting operation

i / C

30 fundamental of the modulation control sub-circuit.

As shown in Figure 11, the modulation control sub-circuit draws a direct current control signal from the input of the circuit shown in this figure (or, if preferred, from the output of the combination circuit 15) by means of an amplifier 30 ', a rectifier 31' and a smoothing or filtering circuit 32a '.

A potentiometer 102 is represented to indicate that the signal coming from the filtering circuit 32a 'is .S gain, controlled. In practice, the gain is generally adjusted in advance during assembly. A combination circuit 33 subtracts the signal produced by the sub-circuit 100 from the main control signal produced by means of the amplifier 30, the rectifier 31 and the filtering circuit 32a.

1 $ The filtering circuit of figure 11 is split into "'~ 7 / * 31 rt two stages in order to minimize the cost of its components * So, for example, blocks 32a and 32a * can be identical and each of them may include only a single RCf filter section while block 32b, which ** provides additional filtering of the combined control signal, includes another RC filter section

The signals are rectified in direct current; - -, (by rectifiers 31 and 31 *) before being combined by circuit 33 in order to avoid the ambiguity of polarity which would occur if AC signals were combined and then rectified (indeed with signals of alternating current, there would be two possible stable states).

The circuit corresponding to the embodiment of FIG. 11 therefore provides a reference level for stabilization of the direct current control signal, a reference level which varies dynamically with the level of the input signal, which produces an offset. or a transposition of part of the dynamic action of the variable filter to a level region determined by the reference level. The circuit operates so as to maintain the maximum amplitude of dominant signals present on the secondary noise reduction channel equal to a constant fraction of the input signal at high signal levels, i The level relating to the output of the sub-circuit Modulation control module 100 © is chosen so as to minimize the sliding effect in response to signals outside the bandwidth of the sliding band.

Although the embodiment of Figure 11 works effectively when the input signal of the modulation control sub-circuit 100 is drawn from the broadband input (or output) signal, other arrangements give signal levels at the upper end of the dynamic range are possible. For example, we get some modulation control effects, Λ 32 even if the input signal of the lOO sub-circuit is taken from the output of the bandpass filter 20. Ideally, we use an equalization in amplifiers 30 and 30 * to optimize the overall modulation control effects (control 5 by bandwidth components / by cut-band components), taking into account the combined frequency response effects of the filters 20 f 22 and the equalization uti— - seen in the control amplifier 26.

When the invention is applied to. of the devices connected in series as described in Belgian Patent 889,428, a single modulation control sub-circuit can be used to provide a reference signal to each stage. Such a circuit advantageously draws its input signal from the output of the last stage of the compressor, when the stages in series are arranged in the preferred order, in which it is the first stage which presents / the level threshold. higher. Since the reference signal is taken from the output, the low-level stage or stages are subjected to the modulation control effect at lower signal levels, which increases the efficiency of the modulation control action.

As described above, it is also possible to provide modulation control of 9-band circuits by means other than deriving a reference of the control signal from the input signal (or exit). One or more control signals can be taken from the output of the variable filter and be limited so as to ensure results similar to those obtained with the embodiment comprising an attenuation of FIG. 11; the essential result remains the same, namely to desensitize the action of dynamic modification of the circuit at high signal levels? located in the cut strip. Figures 12, 13 and 14 are aimed at such embodiments using a limitation *

L

33

In the embodiment of FIG. 12, the means generating control signals (blocks 30, 31 and 33 of FIG. 5) is subdivided by amplifiers 30, 116 and 124 and by filters 110, 118 and 126 into three channels, 5 Know a high-frequency channel, a medium-frequency channel and a low-frequency channel. Each channel includes a limiter (112, 120, 128) which has a threshold established at 1 advance. The limiters can be back-to-back diodes such as the diodes 28 in FIG. 2. For a high frequency acoustic compressor 10 ensuring performance substantially as shown in FIGS. 7 to 10, the frequencies of the filters can be, for example, the following s filter 126, 200 Hz (low pass); filter 118, 2C0 at 800 Hz (bandpass) and filter 110, 800 Hz (high pass). The respective output signals of the three limiters are rectified by respective rectifiers 114, 122 and 130f then these signals are combined (or are subject to a selection of maximum value) and applied to a filtering network 32. According to a alternatively, the limiting functions may intervene after the rectification. In operation: the low-frequency and medium-frequency band limiters are adjusted so as to minimize the slip effects exerted by signals outside the bandwidth. The high-frequency channel may require only a small degree of limitation, or even require no limitation, and the control effected by this channel can be improved by associating with the amplifier 30 an accentuation of the frequencies. high frequencies, as shown by block 52.

FIG. 13 shows another embodiment of the control circuit split into several channels. In the example shown, two channels are used, namely a high-frequency channel and a low-frequency channel. The high-frequency path is substantially the same as in the embodiment of FIG. 12, except that the limit—

AT

34 tor 112 is omitted. The low-frequency channel comprises an amplifier 132 provided with a network 134 for attenuation of the high frequencies. The output of this amplifier is connected to a low-pass filter 136 and to a limiter 138. The threshold of the 5 limiter is chosen concurrently with the characteristics of the various filters and amplifiers, so as to ensure the best possible immunity against control by the band slip by cut band signals. The signals present on the two channels are rectified by rectifiers 10 114 and 140 and combined at the input of the filtering circuit 32.

A simplified variant of the embodiment of FIG. 13 is shown in FIG. 14. The high-pass filter 110 ^ the low-pass filter 136 and the high frequency attenuation network 134 of the amplifier 132 15 are deleted * The high frequency pre-emphasis network 52 * of the amplifier 30 is modified with respect to the network 52 so that the high frequency boost becomes effective at a higher frequency.

As a result, only the broadband channel containing the am-20 amplifier 132 transmits low-frequency signals (in conjunction with high-frequency signals). The threshold of the limiter 138 is adjusted concurrently with the boosting characteristics of the high frequencies of the network 52 * in order to minimize the effect on the slip of cut-band signals 25.

* Figure 15 shows an embodiment comprising a single-channel control circuit, which comprises an amplifier 141 sensitive to frequency, provided with a network l42 for boosting low frequencies, am-30 amplifier which is followed by 'A limiter 144 and another amplifier 146 provided with a network 148 for boosting high frequencies. In operation, the low-frequency part of the spectrum, which tends to cause undesirable slippage, is first emphasized, then limited. The l44 limiter

L

35 is preferably syllabic and has its own closed loop circuit comprising an amplifier, a rectifier, a filtering circuit and a gain-controlled element (such as blocks 276, 280, 282, and 270 of Figure 17) · L 'am-5 amplifier 1 ^ 6, which is provided with a network 148 of * high frequency enhancement, restores any pre-emphasis of the high frequencies if necessary · The output signal of 1 * amplifier 1 ¾6 is then rectified and filtered, respectively, by blocks 11¾ and 32 · In this single-channel control circuit 10, the signal components of high cut-band level are notably reduced at the rectification point 11¾. ''

For convenience and for the sake of simplicity, the sliding band embodiments have been described in relation to a particular configuration of sliding band compressor. The invention is also applicable to expanders without modification of the additional noise reduction channel control circuits shown in the embodiments of FIGS. 11 to 1¾. In the noise reduction devices using compressors and expanders, it is preferable to apply the modulation control according to the invention to these two types of device. to ensure complementarity. The invention is also also applicable to low frequency sliding band circuits, in which the compression and expansion action is designed so as to occur in the low frequency region, "

FIG. 16 represents a functional diagram of a configuration comprising a bilinear compressor and expander 30 with two fixed-band channels. The basic aspects of this arrangement are described in U.S. patents 38 ^ 6,719 and 3,903 ¾85 and in the Journal of the Audio Engineering Society, Vol, 15 "No. k, October 19 & 7" pages 383-388.

AT

'; -! 36

In the known embodiment of FIG. 16, the networks 2 ^ 0 of the additional channel provide four bands. Bands 1, 3 and 4 have conventional 12 dB / octave input filters: a low pass filter 5 at 80 Hz, 252 at the input of band 1, a high pass filter at 3 Hz, 254 at the input of band 3 and a high-pass filter at 9 kHz, 256 at the input of band 4. Each of these filters is followed by an isolation stage with follower emitter 258, Band 2 has a frequency response which is complementary to that of bands 1 and 3 · This response is obtained by adding (in the adder 260) the output signals of the follower transmitter stages 258 of bands 1 and 3 and subtracting the signal- sum thus obtained of the total input signal (in the subtractor 202), the output signal from the stage 15 with follower transmitter 258 of each band and the output signal from the subtractor 262 are applied to respective limiters 264 and 264 ' , The limiters 264 and 264 * are identical except that the limiters 264 * of bands 1 and 2 are time constants double those of the limiters s 264 of bands 20 3 and 4, The output signals of bands 1 to 4 are combined with the signal of the main channel in a corabiner 266, The output signal of the compressor is applied to a noisy channel for transmission to the expander complementary, in which the output signals of identical additional channel networks are subtracted from the input signal to ensure that the complementary expansion characteristic is obtained.

Figure 17 shows further details of the 1 ± --mitters 264 and 264 *, each of which includes a 27Ο T.E.C. attenuator, which operates in response to a control signal. The output signal of this attenuator is amplified by a signal amplifier 272f, the gain of which is adjusted so as to ensure the desired low level signal gain. The output signals of all bands are combined with the main signal to produce a signal of

AT

/ 37 I low-level compressor output * signal which is uniformly 10 dB higher than the input signal up to 5 kHz [approximately, frequency above which the level rises gradually to reach 15 dB at 15 kHz.

5 The T.E.C. is controlled by a control signal generator sub-circuit, which provides a compression threshold of 40 dB lower than the peak operating level. The control subcircuit includes a control signal amplifier 270, followed by a phase divider 278, which energizes a two-wave rectifier 280. The resulting direct current is applied to a filter network 282, at the output of which the control signal appears. The network 282 comprises an RC pre-integrator, a follower emitter stage and a final RC integrator 15 which operate in conjunction with diodes, so that the pre-integrator and the final integrator both exhibit non-specific characteristics. linear produced by the diodes. Large and rapid variations in signal amplitude are transferred quickly, while small variations are transferred slowly. This dynamic filtering action produces optimal results with respect to modulation effects, low frequency distortion and the distortion components generated by the control signal. The circuit provides both rapid recovery and low signal distortion.

FIG. 18 shows a response plot corresponding to an actual graph recording, below the compression threshold of a fixed band compressor having a low level gain of 8 dB and a band filter frequency 30 800 Hz (high pass). Emphasis is provided in the active frequency zone of the device (determined by the angle frequency of 800 Hz) up to levels of approximately -10 dB (compared to a reference level of O dB).

Figure 19 shows the effect on compression when a high level signal (+ 10 dB) is h 38 present at 100 Hz, which is much lower than the filter angle frequency of 800 Hz , The strong 100 Hz signal in the cut band effectively blocks the compressor and prevents compression in the bandwidth · 5 As a result, the desired noise reduction in the bandwidth is not achieved. In addition, if the 100 Hz signal is intermittent, sporadic compression occurs in the bandwidth, while the 100 Hz signal, which controls the onset and disappearance of this compression, causes modulation of the noise and / or a modulation of the signal.

Figure 20 shows the effect of adding a modulation control sub-circuit, described below, to a fixed band circuit. Compression is re-established in the bandwidth area even in the presence of the strong signal (+ 10 dB) at 100 Hz. The modulation control sub-circuit effectively immunizes the fixed-band circuit against the strong cut-band signal. .

FIG. 21 shows in general form 20 the preferred embodiment of the invention in its application to one of the bands of a bilinear two-channel compressor with fixed band of the type described above with reference to FIG. 16. Two additions were. make to the circuit to ensure modulation control. A modulation control sub-circuit 198, similar to that of the sliding band embodiment of Figure 11 is provided; it includes a rectifier 208 'and a first filtering stage 210a *. Modulation control can optionally be provided from the compressor output. The elements 203, 208 ′ 30 can be identical (but separate) as well as the elements 210a, 210a *. The level of the modulation control signal coming from the filtering circuit 210a 'is adjusted by the attenuator 212 or any other suitable means and is combined by the circuit 214 in opposition of polarity with

U

lw- 39 the cut-off control DC signal from the filter circuit 210a ”In addition, the output signal from the series assembly including amplifier VCA 204 and amplifier 206 is applied to a filter 5 216 preferably having the same angle frequency as the filter 202, although this is not essential: the comparative graphs of FIGS. 19 and 20 were produced with a simple low-pass filter 216 at 3 hHz, at 6 dB / octave. However, the filter 216 should ideally have a relatively steep cutoff characteristic, for example of 12 dB or 18 dB per octave (which would be the case for example with a bipolar or three-pole filter), with substantially the same cutoff frequency as the filter 202. The output signal of filter 216 is rectified and smoothed by blocks 218 and 220 ", to form the control signal located in the passband. The filtering provided by the blocks 210a, 210a * and 210a "can be a preliminary stage filtering followed by an additional filtering in the circuit 210b. The output signal of the band filter channel! 20 e3t applied to a selector maximum 222, which receives at its other input, from the output of the combiner 214, the cut-band control signal, itself with modulation control. In its simplest form, the maximum selector comprises two diodes which pass the stronger of the two input signals; in more sophisticated circuits, operational amplifiers are used to eliminate voltage drops across the diodes and to increase accuracy.

In operation, the signals located in the cut band are subjected to the action of the sub-circuit 198 if there are no dominant signals within the pass band in which it is desired to occur. compression action. Thus, although a strong signal, such as the aforesaid signal of + 10 dB at 100 Hz, causes the generation of a strong

AT

4ο control signal by blocks 2C8 and 210a (as well as 210b), this control signal is attenuated by the signal of the modulation control sub-circuit, so that the gain of the amplifier VCA 204 is not lowered , which would suppress compression in the bandwidth. On the other hand, if a 100 Hz signal appears in the region of -20 dB levels, the attenuation effect is considerably reduced and the cut-band control signal then appropriately controls the action of the compressor every 10 times that the signal conditions are such that the bandwidth control signal does not control the compressor.

If strong signals are present in the bandwidth of the active area, at the output of the sharp filter channel, the bandwidth control circuit will control the maximum selector 15 and allow the VCA amplifier to react accordingly.

The level of the modulation control sub-circuit relative to the input or output is adjusted so as to provide a dynamic reference signal (relative to the input) of a level sufficient to ensure substantial immunity the action of the compressor against signals located largely outside the bandwidth.

The comments which have been made above with respect to the equalized control and modulation control amplifiers with reference to sliding band circuits are also applicable to embodiments V · 'of fixed band. For example, if desired, filter / equalizer assemblies 224 and 226 can be inserted in the respective paths leading to the rectifiers 208 ′ and 208. However, the opportunities to make a characteristic function advantageously depend on the frequency given to the against another are less numerous in the case of fixed bands than in that of sliding bands; in fact, it is for this reason that an additional control circuit is necessary

AT

4ι in the case of fixed bands (three circuits against two).

It is also possible to provide modulation control of fixed band circuits by means other than that which consists in drawing a control signal reference from the input (or output) signal of the compressor or of the expander. One or more control signals can be taken from the output of the controllable element (attenuator or VCA amplifier) and can then be limited so as to provide results similar to those obtained in the attenuation embodiment of Figure 21. Figure 22 is aimed at such embodiments -with limitation.

In the embodiment of FIG. 22, the means generating control signals (blocks 276, 278, 280 15 and 282 of FIG. T7) is divided into two channels, one of which comprises an amplifier 228, a filter of sudden cut 216 (as in the embodiment of FIG. 21) and a rectifier 218, while the other comprises an amplifier 230, a limiter 232 and a rectifier 218 ·. The threshold of the limiter 232 (which may consist, for example, of back-to-back diodes) is chosen such that the limiting effect begins at relatively high levels, at. know substantially at the same levels as those at which the output signal r of the combiner 214 begins to become predominant in the embodiment of FIG. 21. The respective output signals of the rectifiers 218 and 218 * can be combined and applied to the filter circuit 210, the output signal of which is applied, as a control signal, to the amplifier VCA 204, or alternatively outputs of said 30 rectifiers can be connected to a maximum selector circuit (such as block 222 of FIG. 21), or even constitute such a circuit themselves, the output signal thereof being applied to the filtering network 210.

/ U

} k2

In service, the embodiment of FIG. 22 operates in substantially the same way as that of FIG. 21.

"

Claims (8)

43 1, - Circuit making it possible to modify the dynamic range of an input signal, characterized in that it comprises: a frequency selective circuit determining a frequency passband in which the modification of the dynamic range takes place; and a dynamic modification means for carrying out a progressive dynamic modification of signal components contained in this pass band, or for producing a progressive slip of said frequency pass band, by means of. Whereby the dynamic range is modified, the dynamic action of said dynamic modification means being sensitive to increasing levels of the linear additive combination of the components of signs. bandwidth and cut band signal components within the fixture and the dynamic action of said dynamic modifying means becoming less sensitive to cut band signal components for high level input signals. 2, - Assembly according to claim T, characterized in that the dynamic action modification means 20 comprises a variable gain or variable loss device, or a variable filter, whose variable dynamic action is controlled by a circuit control. 3. Assembly according to claim 2, character-. rised in that the control circuit comprises antagonistic means capable, in response to cut-band signal components, of counteracting, respectively, the progressive dynamic modification, or the progressive sliding, at the levels of the high input signal. 4. An arrangement according to claim 3, characterized in that said antagonistic means comprise means for non-linear processing of said components of y cut-band signal. h kk 5. Arrangement according to claim 3, characterized in that said antagonistic means comprise a sub-circuit for generating a reference attenuation signal as a component of the signals to which the control circuit is sensitive, said reference signal providing a information at least at high input signal levels, said information being such that the effect of cut band signal components present in the control circuit is reduced at high input signal levels. 6. Circuit arrangement according to claim 3, characterized in that the control circuit comprises at least one limiter and one frequency selective circuit element, the arrangement being such that the limiter (s) act preferentially * on signals of cut band, such that the effect of cut band signal components is reduced in the control circuit at high input signal levels. 7. Assembly according to any one of claims 1 to 6, characterized in that it has a bilinear characteristic composed of a low-level part with substantially constant gain up to a single one, of a part of intermediate level, greater than the threshold, with variable gain, ensuring a maximum compression ratio or expansion ratio, and of a high part * 'substantially constant gain level different from the gain of the low level part, or a unilinear characteristic, in which the low-level part is also of variable gain. 30 8, - Assembly according to any one of claims 1 to 7, characterized in that it is bilinear and constitutes a two-track circuit, comprising a main linear track by ranoort at the dynamic range, an A 'h5 circuit combination channel in said main channel, and an additional channel, the input of which is connected to the input or output of the main channel and the output of which is connected to the combination circuit, said additional channel producing a signal who, at least in a par- | , /. upper part of the frequency band, boosts or attenuates the main channel signal by means of the y / combination signal, but which is limited in such a way that, x in the upper part of the dynamic input range, 10 le signal of the additional channel is weaker than the signal of the main channel. 9. A circuit according to claim 8, characterized in that the additional channel also has a bilinear characteristic such that at high input signal levels the signal level of the additional channel is not less than a fraction. the signal level of the main channel. 10, - An assembly according to claim 2, for acoustic signals, characterized in that the means for modifying the dynamic range comprises a variable filter ion, which ensures an accentuation or a cut in a region of high or low frequency of the band of signals and which is capable, in response to dominant signals, of causing its angle frequency 25 to slip in a direction tending to narrow the accentuated or cut region, and in that the control circuit comprises rectifying means, filtering and amplification applying a control signal taken from the output of the filter to a device with controlled impedance of said filter, to cause the angle frequency of the latter to slip, said control circuit comprising a sub- circuit, which produces a reference signal to attenuate the control signal, reference signal which is drawn / lr \ yu i S 46 from the input or the output of the assembly, said reference signal presents being a level which varies dynamically with the level of the input or output signal, the sub-circuit comprising means for adjusting the gain of the reference signal such that the effects of the cut-band signal components contained in the control signal are reduced to high input signal levels. 11. A circuit according to claim 2, for 10 acoustic signals, characterized in that the dynamic action modification means comprises a variable filter, which ensures an accentuation or a cut in a region of high or low frequency of the band of signals, and which is capable, in response to dominant signals, of causing its angle frequency to slip in a direction tending to narrow the accentuated or cut region, and in that the control circuit includes rectifying means , filtering, and amplification, applying a control signal drawn from the output of the filter to a device with controlled impedance of said filter to ensure the sliding of the angle frequency of the latter, said control circuit comprising at least There is a limiter and a frequency selective circuit element, the arrangement being such that the limiter (s) act pre-ferentially on cut-band signals, so that the effects of com cut signal signal levels are reduced to high input signal levels. 12, - Assembly according to claim 2, for 30 acoustic signals, characterized in that the dynamic action modification means comprises a variable gain or variable loss device, which provides dynamic nodification of bandwidth signals, and which is capable, in response to dominant signals, of causing a progressive dynamic modification, and in that the control circuit comprises means applying a control signal to a control input of said device, to ensure the progressive dynamic modification said control circuit comprising, on the one hand, a first sub-circuit which comprises at least one pointed filter offering frequency characteristics similar to those of the aforementioned frequency selective circuit element, and rectifying means, filtering, and amplification, to generate a first signal; and, on the other hand, a "second sub-circuit, which draws a second / signal from the output of the aforementioned variable gain or variable loss device, which comprises means for rectification, filtering, and amplification, which draws a third signal from the input or output of the assembly, and which comprises additional rectification, filter and amplification means, said third signal constituting a reference signal for attenuating the second signal so as to producing a fourth signal, said reference signal having a level which varies dynamically with the level of the input or output signal, and said second sub-circuit further comprising means for adjusting the gain of the reference signal in such a way that the effects of cut-band signal components present in the third signal are reduced to the high input signal levels, and means for applying the strongest of the first and fourth signals, as a signal of command, at the command input of said; 50 device. i 13.- Assembly according to claim 2, for acoustic signals, characterized in that the dynamic action modification means comprises a / A device | 48 ♦ variable gain or variable loss, which ensures a dynamic modification of signals situated in the passband, and which is capable, in response to dominant signals, of causing a progressive dynamic modification, and in that the control circuit comprises means applying a control signal, taken from the output of said variable gain or loss device, to a control input thereof, to ensure progressive dynamic modification, said control circuit comprising, on the one hand, A first sub-circuit, which comprises at least one pointed filter offering frequency characteristics analogous to those of the aforementioned frequency-selective circuit element and rectifying, filtering and amplifying means; and, on the other hand, a second sub-circuit 15 characterized by at least one limiter and one frequency-selective circuit element, the arrangement being such that the limiter (s) act preferentially on cut band signals of such that the effects of signal components, cut band are reduced at high input signal levels. : Uaaxu