GB2079112A - Circuit arrangement for modifying dynamic range - Google Patents

Abstract

Compressor and expander circuits respectively are arranged in series, the circuits each having bilinear characteristics in which there is a low level portion of substantially constant gain up to a threshold, an intermediate level portion, above the threshold, of changing gain providing maximum compression ratio or expansion ratio, and a high level portion of substantially constant gain different from the gain of the low level portion. The high and low level portions thus each have a constant compression ratio of 1:1. Over the intermediate portion, the compression ratio 10 or 12 rises to a maximum 10b or 12b such as 2:1. The intermediate level changing gain portions of the characteristics are staggered between the series circuit to the extent that the resulting overall characteristic has a maximum compression or expansion ratio substantially no greater than that of any single circuit while providing an increased difference between the gains at low and high input levels (that is, providing increased compression or expansion) than for any of the circuits individually. <IMAGE>

Description

SPECIFICATION Circuit arrangement for modifying dynamic range The present invention is concerned in general with circuit arrangements which alter the dynamic range of signals, namely compressors which compress the dynamic range and expanders which expand the dynamic range.

The invention is particularly useful for treating audio signals but is also applicable to other signals.

Compressors and expanders are normally used together (a compander system) to effect noise reduction; the signal is compressed before transmission or recording and expanded after reception or playback from the transmission channel. However compressors may be used alone to reduce the dynamic range, e.g.

to suit the capacity of a transmission channel, without subsequent expansion when the compressed signal is adequate for the end purpose. In addition, compressors alone are used in certain products, especially audio products which are intended to transmit or record compressed broadcast or pre-recorded signals. Expanders alone are used in certain products, expecially audio products which are intended only to receive or play back already compressed broadcast or pre-recorded signals. In certain products, particularly audio recording and play back products, a single device is often configured for switchable mode operation as a compressor to record signals and as an expander to play back compressed broadcast or pre-recorded signals.

The amount of compression or expansion may be expressed in dB. For example, 10 dB of compression means that an input dynamic range of N dB is compressed to an output range of (N-10) dB. In a noise reduction system 10 dB of compression followed The present invention relates in particular to a circuit arrangement for modifying the dynamic range of an input signal and comprising a first circuit with a bi-linear characteristic (where "linear" in this context denotes constant gain) composed of: 1) a low level linear portion up to a thresh old, 2) an intermediate level non-linear (chang ing gain) portion, above the threshold and up to a finishing point, providing a predetermined maximum compression ratio or expansion ratio, and 3) a high level linear portion having a gain different from the gain of the low level portion.

The characteristic is denoted a bi-linear characteristic because there are two portions of substantially constant gain.

In practice, the threshold and finishing point are not always well defined "points".

The two transition regions where the intermediate level portion merges into the low level and high level linear portions can each vary in shape from a smooth curve to a sharp curve, depending on the control characteristics of the compressor and expander.

It is also pointed out that circuit arrangements with bi-linear characteristics are distinguished from two other known classes of circuit arrangement, namely: (a) a logarithmic or non-linear circuit arrangement with either a fixed or changing slope and with no linear portion: the gain changes over the whole dynamic range.

(b) circuit arrangements with a characteristic having two or more portions of which only one portion is linear ("uni-linear").

A circuit arrangement with a bi-linear characteristic has particular advantages and is widely used. The threshold can be set above the input noise level or transmission channel noise level in order to exclude the possibility of control of the circuit by noise. The high level portion of substantially constant gain avoids non-linear treatment of high level signals which would otherwise introduce distortion, Moreover, in the case of an audio signal, for which the circuit must be syllabic, the high level portion provides a region within which to deal with the overshoots which occur with a syllabic circuit when the signal level increases abruptly. The overshoots are suppressed by clipping diodes or similar means. Only bilinear characteristics are capable of providing this combination of advantages.

The majority of the known circuits with a bilinear characteristic in use today in consumer audio products provide 10 dB of compression and expansion which is adequate for many purposes. However, this leaves some noise audible to some listeners and, for highest fidelity, more compression and expansion is desirable, say 20 dB. It is difficult to provide such a large amount of compression or expansion without encountering problems which affect the quality of the signal.

Circuits are known and commercially available which provide 20 dB of compression or expansion, and even more, but these are usually constant slope logarithmic circuit arrangements in which there is a constantly changing gain over the whole dynamic range or nearly the whole dynamic range. Such circuits suffer from higher distortion and signal tracking problems at very low and very high signal levels than the bi-linear circuits in which the change of gain is restricted to an intermediate portion of the characteristic and overshoot problems are more severe than with bi-linear characteristic arrangements. Known constant slope companders employ compression ratios in the range 1.5:1, 2:1 and 3:1, but 2:1 is most common.

Compression ratio is defined as the ratio of the incremental input dynamic range to the incremental output dynamic range. The ex pansion ratio for a complementary expander is the inverse of the compression ratio. If the compression ratio is 3:1, the expansion ratio is 1:3. It is convenient to utilize the concept of the inverse expansion ratio which, for the example just given is 3:1, i.e., it corresponds to the compression ratio. For simplicity, discussion herein will be confined largely to the compression ratio with the understanding that the same considerations apply, making appropriate changes, to the expansion ratio.

A high compression ratio has a disadvantage in that it is difficult to ensure complementarity between the compressor and the expander; in particular level errors or errors in the frequency response of the transmission or recording medium lead to correspondingly multiplied errors at the output of the expander.

It is known (e.g. US-PS 2,558,002, US-PS 4,061,874 and Japanese Patent Publication 51-20124) to increase the amount of compression available by connecting in cascade a plurality of compressor stages. These known circuits (controlled inpedance device, diodes, etc.) multiply the compression ratios of the individual stages so that a high compression ratio result, with the disadvantage explained above. For example, one circuit with a compression ratio of 2:1 and the other with a compression ratio of 3:1 will yield an overall ratio of 6:1. The resulting expansion ratio of 1:6 would place a heavy burden on the uniformity of the transmission channel.

Another consideration is the demands that are made upon whatever circuit is effecting the change of gain required to establish the compressor or expander characteristic. It is relatively easy to make a circuit effect accurate gain changes over a range of 10 dB but significantly more difficult to make the same circuit effect accurate gain changes over a range of 20 dB. It is thus difficult to establish a controlled, reproducible characteristic for use in a compander system. The 51-20124 publication concludes that a plurality of series compressors (and expanders) are not suitable as a noise reduction system for high fidelity reproduction systems.

It is also known (US-PS 3,902,131 and US-PS 3,930,208) to cascade a plurality of compressor stages that operate in mutually exclusive frequency ranges. While such arrangements may not result in any increase in compression ratio over that of a single stage, thay do not provide any increase in compression.

In the light of all these considerations it is the object of this invention to provide an increased amount of compression or expansion without an undesirably large increase in the compression ratio and without making excessive demands upon any circuit involved in effecting a change of gain.

A further object is to provide an increased amount of audio compression or expansion without producing an undesirably large increase in the overshoots produced under transient signal conditions.

A close examination of bi-linear circuits shows that they not only have the previously enumerated advantages but a further one as well-namely, a way of solving the high compression ratio problem and. in the case of audio circuits, also a way of solving the high overshoot problem.

Note that the superposition of the linear regions does not amplify the compression ratio in these regions; the compression ratio is increased only in the limited region in which dynamic action takes place. Therefore, I have found it possible to separate the areas of dynamic action in such a way as to obtain the required overall increase in compression, while at the same time not altering the overall maximum compression or expansion ratio significantly.

A further feature of this arrangement is that the overall result is bi-linear, with all of the attendant advantages. Thus the action staggering possibility of bi-linear devices represents a further hitherto unrecognized advantage of this class of device.

The above stated objects are met in accordance with the invention which is characterized in that a first circuit, which has a bi-linear input-output characteristic, is followed by one or more further circuits which also have bilinear characteristics at any given frequency within a frequency range common to the circuits. The thresholds and dynamic regions of the circuits are set to different values so as to stagger the intermediate level portions of the characteristics of the circuits to produce a change of gain over a wider range of intermediate input levels than for any of the circuits individually, and to produce an increased difference between the gains at low and high input levels, but with a maximum compression or expansion ration which is substantially no greater than the maximum compression ratio of any single circuit, by virtue of the staggering.

In the case of audio circuits, if the circuits have overshoot suppression (limiting) elements, then it is also possible to stagger their thresholds along with the stagger of the syllabic thresholds. The overshoots of the lower level circuits, or stages, are correspondingly reduced, with minimal overall overshoot of the several stages. This is in contrast with conventional logarithmic compressors in which large overshoots are inherently produced.

Each of the circuits may introduce an alteration of the spectral content of the signal -- for example, a low level treble boost in the case of a compressor. Thus each succeding stage may be actuated by a signal of progressively hanging spectral content. In the case of complex signals, this has the virtue of spectrally spreading out the chances for error in the decoding function. In the case of a tape recorder with an uneven frequency response characteristic, for example, the spectral shifting tendency reduces the overall dynamic and frequency response errors of the decoded result.

Consideration will now be given to the amount of staggering required. For simplicity, reference is made to a series connection of two compressor circuits. The compression ratio of each of the first and second circuits will rise from unity at the respective threshold to a maximum and this will be called the rising flank of the compression ratio. The ratio will then fall back to unity and this will be called the falling flank. Strictly speaking the falling flank may approach unity asymptotically but for all practical purposes it may be considered to have reached unity when it is at some value differing only by an arbitrary small amount from unity.

The staggering of the intermediate level portions of the first and second circuits in the falling flank of one circuit overlapping the rising flank of the other circuit. At least to a first approximation, the difference between the two thresholds can be made such that the overlap of flanks results in an overall compression ratio which does not substantially exceed the maximum compression ratio of either circuit by itself.

Preferably, the threshold of the second circuit is lower than that of the first circuit (if more than two circuits are used, each further circuit preferably has a progressively lower threshold), in the case of a compressor and vice versa in the case of an expander. In principle the order can be reversed, with the first compressor circuit having the lower threshold. In the case of more than 2 circuits, the order of threshold levels among them, in principle, can be scrambled and arranged in any order so long as the intermediate level portions of the circuits are properly staggered.

Thus the ideal stagger is considered to be that which causes the falling flank of one circuit to overlap the rising flank of the other circuit in order to restrict, so far as possible, the level region in which dynamic action occurs in the overall series connected device while at the same time avoiding any substantial increase in the maximum compression or expansion ratio over that of a single device.

Then, for example, if the maximum compression ratio of each circuit is 2:1, the compression ratio of the overall circuit arrangement will rise to 2:1, maintain this value over the overlap, and then fall back to unity. Thus, ideally, there is no increase at all above the ratio of 2:1, in contrast to the prior art arrangements of cascaded compressor stages which multiply the ratios to 4:1.

In practice it may be difficult to achieve an optimum overlap at all frequencies but it can be seen that, providing a reasonable approximation is made to the ideal, the overall maximum compression ratio can be prevented from rising excessively above 2:1 in the example given. It may in a practical circuit arrangement perhaps rise to 2.5:1.

A low maximum compression ratio (e.g.

1.5:1) permits an expander to track the compressor more easily to provide good complementarity with signal channels having somewhat unreliable gains and/or frequency responses. However, a low compression ratio spreads the dynamic action over a wider range of levels, causing greater susceptibility to noise modulation for a given maximum amount of noise reduction difference in gain at low and high input levels. Hence, there is a trade off between undesirable effects caused both by high and low compression ratios.

Consequently, the ideal compression ratio will depend on the system environment and the system design goals.

The ability to stagger bi-linear stages provides the designer with an additional way in which to optimize an overall circuit. In so doing, the shapes of the compression characteristics of individual stages can be designed with staggering specifically in mind. The transient characteristics of the circuits are also taken into account and the opportunity is preferably taken to stagger the overshoot suppression thresholds in audio compressors and expanders so as to result in minimal overall overshoot.

A well known type of circuit, called "sliding band", which can be used for each of the first and second circuits, creates the specified desirable characteristic for the case of high frequency audio compression or expansion by applying high frequency boost (for compression) or cut (for expansion) by way of a high pass filter with a variable lower corner frequency. As the signal level in the high frequency band increases, the filter corner frequency slides upwardly so as to narrow the boosted or cut band and exclude the useful signal from the boost or cut. Examples of such circuits are to be found in US-PS Re 28,426, US-PS 3,757,254, US-PS 4,072,914, US PS 3,934,1 gO and Japanese Patent Application 55529/71.

Accordingly, each of the first and second circuits can be such a "sliding band" circuit.

In principle the quiescent corner frequencies of the two sliding band circuits can be different and use can be made of this to provide a degree of compression or expansion which is higher in one part of the treated frequency band than in another. However, in accordance with an important further developement of the invention the corner frequencies are made substantially identical. This leads to the advantage of sharper discriminations between the frequency region where boost or cut is being applied and the region where it is not applied and accordingly a sharper discimination between the region where noise reduction is no longer taking place, because of the appearance of a significant useful signal, and the region where noise reduction remains effective.

On the other hand, circuits are also well known in which the frequency spectrum is split into a plurality of bands by corresponding band-pass filters and the compression or expansion is effected in each band by a gain control device (whether an automatically responsive, diode type of limiting diode type of limiting device or a controlled limiting device) in the case of a compressor, with some form of reciprocal or complementary circuitry for an expander. Examples of such circuits are to be found in US-PS 3,846,719. These split band or multi-band circuits have the advantage of independent action in the various frequency bands and, if this property is required, such circuits may be employed as the first, second, or more stages in the circuit arrangement of the invention.

In principle, one of the first and second circuits may be a multi-band circuit and the other may be a sliding band circuit. This will be of interest in a special situation in which it is, for example, desired to enhance the degree of compression or expansion in one portion of the total frequency band, the sliding band circuit and one or more of the band-splitting channels acting over this portion of the frequency band.

It is known to construct bi-linear compressors and expanders, of both sliding band and split band type, by the use of only a single signal path. However, it is generally preferred to construct such devices by providing a main signal circuit which is linear with respect to dynamic range, with a combining circuit in the main circuit, and a further circuit which derives its input from the input or output of the further circuit and has its output coupled to the combining circuit. The further circuit includes a limiter (self-acting or controlled) and the limited further signal boosts the main circuit signal in the combining circuit for the casr of compression but bucks the main circuit signal for the case of expansion. The limited further path signal is smaller than the main path signal in the upper part of the input dynamic range.The main and further circuits are preferably and most conveniently separately identifiable signal paths.

Such known compressors and expanders are particularly advantageous because they enable the desired kind of transfer characteristic to be established in a precise way without problems of high-level distortion. The low level portion of substantially constant gain is established by giving the further path a threshold above the noise level; below this threshold the further path is linear. The intermediate level portion is created by the region over which the further path limiting action becomes partially effective and the high level portion of substantially constant gain arises after the limiter has become fully effective so that the further path signal ceases to increase and become negligible compared to the main path signal.At the highest part of the input dynamic range, the output of the circuit arrangement is effectively only the signal passed by the linear main path, i.e. linear with respect to dynamic range. In dual path audio circuits the provision of overshoot suppression is particularly convenient.

Example of these known circuits are to be found in US-PS 3,846,719, US-PS 3,903,485 and US-PS Re 28,426. There are also known analogous circuits which achieve like results but wherein the further path has characteristics inverse to limiter characteristics and the further path output bucks the main path signal for compression and boosts the main path signal for expansion (US-PS 3,828,280 and US-PS 3,875,537).

Any of these known bi-linear circuits may accordingly be employed as the first and second circuits of the circuit arrangement according to the invention in order to obtain the advantages inherent therein and also to provide a good way of establishing the desired amount of staggering. This is done by setting the thresholds and dynamic regions of the two further paths appropriately.

As mentioned previously, it is not essential to create the desired form of bi-linear characteristic by such "dual path" techniques. Alternatives exist, operating with single paths, as described in US-PS 3,757,254, US-PS 3,967,219, US-PS 4,072,914, US-PS 3,909,733 and Japanese Patent Application 55529/71, for example. Although these alternative circuits usually are not capable of producing such good results as dual path circuits, or may be less convenient and thereby less economical, they can produce generally equivalent results. Accordingly, these known circuits can also be used as one or more of the circuits of a circuit arrangement according to the invention. If desired, one of the first and second circuits can be a dual path circuit and the other a single path circuit.

The invention will be described in more detail, by way of example, with reference to the accompanying drawings, in which: Figure 1 is an exemplary set of curves showing complementary bi-linear compression and expansion characteristics.

Figure 2 is a block diagram showing the present invention in general terms.

Figure 3 is an exemplary graphic depiction of the areas of dynamic action and how they can be separated in series connected compressors or expanders.

Figure 4 is a further simplifier form of Fig.

3.

Figure 5 is a series of idealized bi-linear characteristic curves illustrating a general technique for staggering the thresholds of series circuits.

Figure 6 is a schematic circuit diagram of a prior art sliding band compressor.

Figure 7 is a schematic circuit diagram of a prior art sliding band expander.

Figure 8 is a schematic circuit diagram of a modification to Figs. 6 and 7.

Figure 9 is a chart recording showing the response below the compression threshold of two series compressors and expanders according to an embodiment of the invention.

Figure 10 is a chart recording showing the response below the compression threshold of a prior art compressor and expander according to Figs. 6, 7 and 8.

Figure 11 is a chart recording of the inputoutput response as a function of frequency of a compressor having series devices according to an embodiment of the invention.

Figure 12 is a chart recording of the inputoutput response as a function of frequency of the prior art compressor having a single device.

Figures 13-15 are a series of probe tone curves illustrating the sliding band action of an embodiment of the invention and the circuit of Figs. 6 and 8.

Figure 16 shows characteristic curves below the compression threshold of a further embodiment of the invention.

Figure 1 7 shows characteristic curves similar in nature to those of Fig. 11, but for a further embodiment of the invention.

Figure 18 shows characteristic curves similar in nature to those of Figs. 11 and 17, but illustrating excessive bunching.

Exemplary bi-linear complementary compression and expansion transfer characteristics (at a particular frequency) are shown in Fig. 1, indicating (for the compression characteristic) the low level portion of substantially constant gain, the threshold, the portion where dynamic action occurs, the finishing point, and the high level portion of substantially constant gain.

Fig. 2 shows the present invention in general terms: a first bi-linear compressor 2 receives the input information and applies its output to a second bi-linear compressor 4 connected in series, which has its output applied to a noisy information carrying channel N. A pair of series connected bi-linear expanders 6 and 8 receive the input from channel N at expander 6 and provide a noise reduction system output at the output of expander 8. The areas of dynamic action of the series devices are separated or staggered with respect to each other within the frequency range that is common to the devices. Although the figure shows two devices on each side of the information channel N, two or more can be employed: the invention contemplates two or more series bi-linear compressors or expanders.When configured as a complementary noise reduction system, like numbers of series bi-linear compressors and expanders are provided.

The order of stages having particular characteristics in the compressor is reversed in the expander. For example, the last stage of the expander is complementary to the first stage of the compressor in all respects--steady state and time dependent dynamic response (frequency, phase and transient response under all signal level and dynamic conditions).

An exemplary graphic depiction of the separation or staggering for two bi-linear devices is shown in Fig. 3, which plots compression ratio versus input amplitude level (horizontal axis) for a compressor or expander operating at a particular frequency. For clarity, the curves are shown in idealized form; as a practical matter the curves are somewhat asymmetrical in practical embodiments of applicant's A-type and B-type noise reduction systems, (US-PS 3,846,719 and US-PS Re 28,426, respectively). Curve 1 2 refers to the dynamic action of one compressor or expander (the high level stage). Curve 10 is that of a further compressor or expander (the low level stage), with a separated area of dynamic action.If the high level stage is first in the series of compressors (second in the series of expanders), the curve 1 2 represents the variations of compression ratio of the first (compressor) stage as a function of the input level to the first stage and curve 10 is the variation of the compression ratio of the second (compressor) stage as a function of the input level to the first stage. The top curves are those of compressors, the bottom curves those of expanders. In this example, the areas of action in response to input amplitude level are separaed such that the product of the two curves results in an overall characteristic having a compression ratio or expansion ratio which does not exceed 2:1 (1:2) between the two maximum compression points 1 0b and 1 2b (1 Oa and 12a) of the two devices.

Thus, even with two devices in series, the end regions of operation still remain fixed, the maximum compression and maximum expansion ratios are not increased beyond those of single devices and the advantages of single bilinear devices are retained. Consequently, any errors occurring within the range of dynamic action caused by the devices in series should not exceed those of a single device.

Most bi-linear devices determine the fixed end regions of constant gain by means of fixed, pre-set circuit elements, such as resistors and capacitors, which are inherently stable and cannot introduce dynamic errors, waveform distortions and the like. Consequently, only in a transitional area of operation, between the linear, constant gain regions, can any dynamically active portions of the circuits introduce signal errors.

Note that in the representation of Fig. 2, the dynamic action of a conventional logarithmic compressor or expander becomes a horizontal line; line 11, for example, is the characteristic of a 2:1 compressor, line 13 is that of a 1:2 expander. It is clear, in this analysis, that there is no opportunity for separating or staggering the actions of such devices.

For purposes of analysis and to obtain a first order approximation of necessary thresh oed levels to provide optimum staggering in accordance with the invention, it is useful to idealize Fig. 3 even further. Assume therefore that each compressor (and expander) immediately reaches its maximum compression ratio at a threshold level and holds that ratio until it reaches a finishing point at a higher level where its dynamic action abruptly stops. Then a series of compressors and expanders depicted in the manner of Fig. 3 appear as contiguous rectangular curves such as shown in Fig. 4. As an example, three bi-linear characteristic compressors and expanders are connected in the series.The low level device, which is preferably the third compressor (first expander), has the lowest threshold (T3) shown at - 62 dB, with its finish point (F3) at - 46 dB which is the threshold (T2) of the medium level stage. The medium level stage has its finishing point (F2) at - 30 dB which is the threshold (T1) of the high level stage.

The high level stage has its finishing point (F1) at - 14 dB. All levels are with reference to the overall input. It is further assumed that each stage has a gain of 8 dB and a maximum compression ratio of 2:1.

Fig. 5 depicts idealized characteristic curves (overall input versus output) for compression based on the example of Fig. 4 (the mirror image expansion curves are omitted for clarity). The drawing shows how the dynamic action of each stage occurs next to that of its adjacent stage resulting in an overall compression ratio of 2:1 while obtaining 24 dB of compression.

Based on observations of Figs. 4 and 5, a single equation sets forth the relationship between threshold levels (T), finishing point (F), maximum compression ratio (c) and gain (G) of any particular stage: CG T= F C- 1 Using this equation the threshold levels for each stage can be determined to a reasonably close approximation by an iterative process.

For example, if an overall finishing point (Ft) of - 14 dB is desired with a stage gain of 8 dB and a maximum compression ratio of 2, the equation shows that the high level threshold (Tt) is - 30 dB. This value is then used as the finishing point (F2) of the middle level stage to determine that its threshold should be - 46 dB, and so on. Thus each stage is referred back to the result of the previous stage in this analysis. However, the calculated threshold is the overall threshold referred to the input of the series. To obtain the threshold of a particular circuit referred to its own input, the cumulative signal gain up to that point is taken into account. For example, the threshold of the lowest level stage of Fig. 5 is - 46 dB when referred to the input of that stage.

The equation can also be solved for the finishing point F, the compression ratio C, or the gain, G. Thus, the designer can determine his circuit parameters based on his design goals. Such goals may include requirements that the lowest level threshold be above the noise floor, that the highest level finishing point be low enough to permit the use of overshoot protection and that the overall maximum compression ratio not exceed a particular value.

In practical circuits, the threshold and finishing point are not always well defined points as they are in this analysis. As discussed in the introduction, the regions in which the intermediate level portion of the characteristic merge with the low level linear portions may be smooth or sharp depending on the characteristics of the circuits that control the dynamic acition. Thus, in practice, the threshold region of one circuit will overlap with the finishing point region of another circuit.

Consideration of the above equation and Fig. 5 shows that for the particular case of a 2:1 compression ratio, half of the threshold staggering is provided by the signal gains of the stages and that the other half must be provided by an altered bias on the control element and/or an altered control amplifier gain (increased gain for lower threshold). Similarly, for 1.5:1 and 3:1 compression ratios, respectively, 1/3 and 2/3 of the staggering, respectively, is provided by the stage gains and 2/3 and 1/3 of the staggering, respectively,must be provided by the control circuitry.

In both Figs. 1 and 5, 0 dB is a nominal maximum or reference level. In practice, a headroom of some 10 to 20 dB is provided above the 0 dB level.

As mentioned earlier, it is usually preferable for the high level stage to be first in a compressor series and the low-level stage to be last. However, a reversed arrangement is also possible. In the reversed case the control amplifier of the first stage needs a high gain in order to achieve the required low threshold.

This low threshold than applies even in the presence of high level signals, which in the case of sliding band systems known in the prior art usually leads to poor noise modulation performance of the overall system. In this reversed arrangement each stage must provide sufficient control amplifier gain to achieve the threshold required of that stage.

Moreover, each threshold is essentially fixed and independent of the operation of the other stages. This is a consequence of the fact that the signal gain of each earlier stage has fallen substantially to unity when the threshold is reached for the corresponding succeeding stage. The calculation of the thresholds required for optimal staggering in the reversed case is the same as that in the preferred case.

However, the threshold of each stage referred to its own input becomes the same as the overall threshold.

In contrast to the reversed situation, in the preferred arrangement (in which the high level stage is first in the compressor chain, and the low-level stage is last), there is a useful interaction between the stage gains and the thresholds. The thresholds of the downstream stages are partly determined by the signal gains of the preceding stages. Thus in a 2 stage system with 10 dB of low level gain per stage, the control amplifier gain requirement of the second stage is reduced by 10 dB, by virtue of the low-level signal gain of the first stage. When a high level signal appears, the 10 dB gain of the first stage is eliminated and the threshold of the low level stage is effectively raised by 10 dB. With sliding band companders this improves the noise modulation performance of the noise reduction action.

In the preferred arrangement the gains of all preceding stages are fully effective up to the threshold of any particular succeeding stage. Thus, in contrast with the reversed order system described above, the preferred arrangement takes best advantage of the prevailing signal gains of the individual stages.

Namely: 1. Under very low level (sub-threshold) sig nal conditions the control amplifier gain requirement of each stage is reduced by an amount equal to the cumulative signal gains of all preceding stages. In the exam ple of Fig. 5, the control amplifier gain required of the lowest level stage, to achieve a threshold of - 62 dB, is thus reduced by 1 6 dB relative to that required if that stage were operating independently or in the reversed configuration described previously. Similarly, the control amplifier gain of the mid-level stage is reduced by 8 dB, thus leading to the most economical circuitry.

2. A signal dependent variable threshold effect is achieved, whereby with sliding band stages, noise modulation effects are reduced. The effective thresholds of the low level stages are progressively raised with increasing signal level at a particular fre quency. At high signal levels (on the high level linear portion of the transfer character istic) the effective threshold of the lowest level stage is raised by a level equal to all the low-level (sub-threshold) stage gains up to the point. In the example of Fig. 5 the threshold of the lowest level stage, normally - 62 dB under low level signal conditions, is thus raised by 1 6 dB, to - 46 dB, under high level signal conditions. Similarly, the threshold of the mid-level stage is raised to - 38 dB.

In a first practical embodiment of the invention, employing series sliding band devices, the compressor 2 and expander 8 of Fig. 2 are essentially standard B-type sliding band devices as set forth in US-PS Re 28,426, while compressor 4 and expander 6 have modified response characteristics. I have found that, with the noise generated by cassette tapes, a workable result is obtained when the second device (in the compression mode) not only has a staggered input amplitude level response but also has a cut off frequency some two to three octaves lower than that of a standard B-type device. More specifically, the threshold levels of the second device are lowered, both of the syllabic filter/ limiters and of the overshoot supression limiter, to effect staggering, and the corner frequency of the fixed filter is lowered by two to three octaves.

Details of the B-type circuit are set forth in Figs. 6, 7 and 8 which are the same as Fig.

4, 5 and 10 respectively of US-PS Re 28,426 and further details of said circuits, their operation and theory are set forth therein. The following description of Figs. 6, 7 and 8 is taken from US-PS Re 28,426.

The circuit of Fig. 6 is specifically designed for incorporation in the record channel of a consumer tape recorder, two such circuits being required for a stereo recorder. The input signal is applied at terminal 10 to an emitter follower stage 1 2 which provides a low impedance signal. This signal is applied firstly through a main, straight-through path constituted by a resistor 14 to an output terminal 1 6 and secondly through a further path the last element of which is a resistor 1 8 also connected to the terminal 1 6. The resistors 1 4 and 1 8 add the outputs of the main and further paths to provide the required compression law.

The further path consists of a fixed filter 20, a variable cut-off filter 22 including a FET 24 (these constituting the filter-limiter), and an amplifier 26 the output of which is coupled to a double diode limiter or clipper 28 and to the resistor 1 8. The non-linear limiter suppresses overshoots of the output signal with abruptly increasing input signals. The amplifier 26 increases the signal in the further path to a level such that the knee in the characteristic of the limiter or overshoot suppressor 28, comprising silicon diodes, is effective at the appropriate signal level under transient conditions. The effective threshold of the overshoot suppressor is somewhat above that of the syllabic filter/limiter.The resistors 14 and 1 8 are so proportioned that the required compensating degree of attenuation is then provided for the signal in the further path.

The output of the amplifier 26 is also coupled to an amplifier 30 the output of which is rectified by a germanium diode 31 and integrated by a smoothing filter 32 to provide the control voltage for the FET 24.

Two simple RC filters are used, though equivalent LC or LCR filters could be used.

The fixed filter 20 provides a cut-off frequency of 1 700 Hz, below which diminishing compression take place. The filter 22 comprises a series capacitor 34 and shunt resistor 36 follows by a series resistor 38 and the FET 24, with its source-drain path connected as a shunt resistor. Under quiescent conditions with zero signal on the gate of the FET 24, the FET is pinched off and presents substantially infinite impedance; the presence of the resistor 38 can then be ignored. The cut-off frequency of the filter 22 is thus 800 Hz, which it will be noted is substantially below the cut-off frequency of the fixed filter 20.

When the signal on the gate increases sufficiently for the resistance of the FET to fall to less than say 1 K, the resistor 38 effectively shunts the resistor 36 and the cut-off frequency rises, markedly narrowing the pass band of the filter. The rise in cut-off frequency is of course a progressive action.

The use of a FET is convenient because, within a suitable restricted range of signal amplitudes, such a device acts substantially as a linear resistor (for either polarity signal), the value of which is determined by the control voltage on the gate.

The resistor 36 and FET are returned to an adjustable tap 46 in a potential divider which includes a temperature compensating germanium diode 48. The tap 46 enables the compression threshold of the filter 22 to be adjusted.

The amplifier 26 comprises complementary transistors giving high input impedance and low output impedance. Since the amplifier drives the diode limiter 28, a finite output impedance is required and is provided by a coupling resistor 50. The diodes 28 are, as already noted, silicon diodes and have a sharp knee around 1/2 volt.

The signal on the limiter and hence on the resistor 1 8 can be shorted to ground by a switch 53 when it is required to switch the compressor out of action.

The amplifier 30 is an NPN resistor with an emitter time constant network 52 giving increased gain at high frequencies. Strong high frequencies (e.g. a cymbal crash) will therefore lead to rapid narrowing of the band in which compression takes place, so as to avoid signal distortion.

The amplifier is coupled to the smoothing filter 32 through the rectifying diode 31. The filter comprises a series resistor 54 and shunt capacitor 56. The resistor 54 is shunted by a silicon diode 58 which allows rapid charging of the capacitor 56 for fast attack, coupled with good smoothing under steady-state conditions. The voltage on capacitor 56 is applied directly to the gate of the FET 24.

A complete circuit diagram of the complementary expander is provided in Fig. 7, but a full description is not required as substantially as the circuit is identical to Fig. 6, component values, are therefore not for the most part shown in Fig. 7.

The differences between Figs. 6 and 7 are as follows: In Fig. 7, the further path derives its input from the output terminal 16a, the amplifier 26a is inverting, and the signals combined by the resistors 1 4 and 1 8 are applied to the input (base) of the emitter follower 12, the output (emitter) of which is coupled to the terminal 16a. To ensure low driving impedance, the input terminal 1 0a is coupled to the resistor 14 through an emitter follower 60. Suitable measures must be taken to prevent bias getting in the expander.

The amplifier 26a is rendered inverting by taking the output from the emitter, instead of the collector, of the second (pnp) transistor.

This alteration involves shifting the 10 K resistor 62 (Fig. 6) from the collector to the emitter (Fig. 6), which automatically gives a suitable output impedance for driving the limiter. The resistor 50 is therefore omitted in Fig.

7.

It should be noted that it is important in aligning a complete noise reduction system to have equal signal levels on the emitters of the transistors 1 2 in both compressor and expander. Metering terminals M are shown connected to these emitters.

Fig. 8 shows a preferred circuit, for replacing the circuit between points A, B and C in Figs. 6 and 7. When the FET 24 is pinched off, the second RC network 22 is inoperative, and the first RC network 20 then determines the response of the further path. The improved circuit combines the phase advantages of having only a single RC section under quiescent conditions with the 1 2 dB per octave attenuation characteristics of a two-section RC filter under signal conditions.

In the practical circuit, using MPF 104 FET's, the 39 K resistor 36a is necessary in order to provide a finite source impedance to work into the FET. In this way the compression ratio at all frequencies and levels is held to a maximum of about 2. The 39 K resistor 36a serves the same compression ratio limiting function in the improved circuit as the resistor 36 in the circuit of Fig. 6 or Fig. 7. In addition, this resistor provides a low frequency path for the signal.

Modifications to Figs. 6, 7 and 8 As mentioned previously, in the first practical embodiment of the present invention, compressor 4 and expander 6 of Fig. 2 employ devices of the type shown in Figs. 6, 7 and 8 with modified characteristics. The altered cut off frequency and lowered threshold are accomplished, respectively, by modifying the characteristics of the fixed filter (fixed filter 20 of Fig. 6) and also the control amplifier gain by altering its pre-emphasis characteristics (emitter time constant network 52 of amplifier 30 in said Fig. 6). The threshold of the overshoot suppressor is lowered by the application of suitable DC biases (in the forward direction) to the diodes 28.The impedances of the variable filter (variable filter 22 in Figs. 6 and 8) are left generally unaltered in order to retain a suitable match to the characteristics of available voltage controllable variable circuits elements. Suitable modifications of the B-Type sliding band circuit shown in Figs. 6, 7 and 8 are to change the value of the 3.3 K resistor in fixed filter 20 to a value of 1 8 K in order to lower its cutoff frequency two to three octaves. For an increase in the control amplifier gain, the value of the capacitor in the amplifier 30 emitter time constant network 52 is increased from 0.15 to 0.60 yf (or from 0.1 to 0.4 4 if the suggested 0.1 4 value is used).Biases of about plus and minus 1/4 volt in the forward direction are applied to silicon diodes 28, thereby reducing the overshoot suppression level by several decibels.

Variable filter 22 has an all pass frequency characteristic in response to the quiescent control voltage and thus the overall filter cut off is lowered two to three octaves. Increasing the capacitor value in the emitter network of control amplifier 30 increases the amplifier gain at any given frequency. As explained above and and in US-PS Re 28,426, as the control voltage (from amplifier 30, rectifier 31 and smoothing filter 32) increases, the cut off frequency of the variable RC Filter 22 rises.

Thus, with larger values of capacitance in network 52, the variable filter responds by moving up in frequency from its quiescent value in response to lower level signals, thus staggering the level response or threshold from that in the unmodified B-Type circuit.

The level response can be staggered in numerous ways in addition to changing the emitter network of the control amplifier. Other possibilities include changing the bias on the control element, otherwise altering the gain of the control amplifier, altering the relative signal levels between the filter path and the control signal derivation path, and so on.

Certain details of the circuit of Figs. 6, 7 and 8 have evolved over the years and more modern forms of the circuit have been published and are well known in the art. Reference to the specific circuit in US-PS Te 28,426 is made for convenience in presentaton.

Fig. 9 shows an actual chart recording plot of response below the compression threshold of the two series connected compressors, the first being modified as described above; the expander response is also shown. Compare to Fig. 10 (which is Fig. 12 of US-PS Re 28,426) showing an actual chart recording plot of response below the compression threshold of a single compressor or expander according to Figs. 6, 7 and 8.

Fig. 11 is a chart recording of the inputoutput response of the series compressors as a function of frequency. Inspection of the response plots shows the two dynamic areas for the curves indicating the two staggered areas of action. While the observability of the dynamic areas in these curves is useful to demonstrate the staggered action of the devices, in practice it is preferred that the curves are as smooth as possible, without discernible dynamic areas or "bumps". Parallel lines A and B are drawn through the threshold regions: Line A referring to the standard circuit and line B to the modified circuit. Compare these curves to Fig. 1 2 (which is US-PS Re 28,426) which plots similar response curves for a single unmodified B-Type sliding band compressor.Fig. 11 shows that the compressor comprising series devices provides essentially twice as much compression which is distributed over a greater frequency and level area.

The variable band action of the staggered action series devices can be seen in Fig. 13 and 14, showing a chart recorder probe tone response of said series connected compressors. Compare to Fig. 1 5 (which is Fig. 1 5 of US-PS Re 28, 426) which is an actual chart recording obtained from the circuit of Fig. 6 incorporating Fig. 8. The variable band action is shown by plotting the compressor frequency response by means of a low-level probe tone (the level of which is below the compressor threshold) in the presence of a high-level signal; the probe is detected at the compressor output by means of a tracking filter. The high-level signal causes the compressor circuitry to operate, the graph showing the effect on the turnover frequency of the filter.

Fig. 1 3 shows the response for a probe tone at - 65 dB and 200 Hz signal tones at levels ranging from - 28 dB and below to + 10 dB. Fig. 14 is for a 500 Hz signal tone at levels ranging from - 34 dB and below to + 10 dB.

In a further practical embodiment of the invention, giving improved performance, the compressor 2 and expander 8 of Fig. 2 are both modifications of standard B-type devices.

Both series devices have their corner frequency lowered by two octaves in order to provide a sharply rising low-level response characteristic. Dynamic action staggering is provided by descreasing the thresholds (both syllabic and overshoot suppression) of the second (in the compressor mode) device.

A feature and useful advantage of the invention is that the frequency responses of the individual circuits are compounded. If the most sharply rising noise reduction characteristic is desired, this is accomplished by the use of circuits having the same low level (quiescent) frequency response characteristics.

Accordingly, in the improved embodiment, the choice of identical filter characteristics at about 2 octaves below that of the standard Btype device results in a characteristic that rises rapidly above about 300 Hz. Thus, the system becomes capable of achieving substantial noise reduction in the critical 300 Hz to 2 kHz range, a region in which tape noise is discernable once noises above 2 kHz have been reduced. There is negligible audible noise contribution from the tape below about 300 Hz. By providing only minimal noise reduction action below 300 Hz, the system avoids the manipulation of fundamental signal frequencies and improves the complementarity of the system in paractical tape recorders, which, for example, may have frequency response errors due to head bumps and the like.Moreover, by avoiding the compression of low frequency signals, system compatibility is improved because the boosting of low frequency signals would result in annoying rumble and bass enhancement when encoded tapes are played on systems not having complementary expanders.

Referring once again to Figs. 6 and 8, the two series devices in the practical embodiment under discussion both have the resistor in fixed filter 20 changed from 3.3 K to 1 3 K, which causes the overall lower cut off frequency of the filters 20 and 22 to shift about 2 octaves lower to about 375 Hz. In the second device, the capacitor in the emitter network 52 of control amplifier 30 is increased in value by a factor of about 4 as in the previously discussed embodiment. This results is a staggering of threshold levels of roughly 10 to 1 5 dB (depending on signal level and signal frequency). Appropriate biasing is introduced in the diode limiter circuit 28 to lower the overshoot suppression level.

In a modification of the last described practical embodiment the capacitor 34 in filter 22 can be increased in value of 0.01 4 in order to promote consistency in characteristics from unit to unit and to improve the noise modulation characteristics. In that case, due to the substantially equal time constants of fixed filter 20 and variable filter 22, the arrangement is equivalent to a single pole variable filter and the fixed filter can be eliminated.

In that case the resistor 36a (which has value of 47 K in modern forms of the B-type circuit) is placed in shunt with the source-drain path of FET 24 so as to provide a quiescent corner frequency of about 375 Hz. It is however, desirable to retain the fixed filter in the high level circuit so that the circuit can be switched to operate by itself as a standard B-type circuit.

As a practical matter, a consumer product embodying the improved systems just described will be compatible with existing nonencoded and B-type encoded software (e.g., tapes and FM broadcasts). The improved systems incorporate a standard B-type device and, hence, can be switched to operate as a B-type device for full compatability. On the other hand, when recorded tapes become available encoded with the improved system, existing B-type consumer systems may observe excessive high frequency information or "brightness" which can be treated by adjust- ing the high frequency tone control in the same way that non-equipped consumer systems presently treat B-type encoded software.

The standard B-type circuit described in US PS Re 28,426 has a maximum compression ratio of about 2:1. This compression ratio has proven to be a good practical choice for consumer cassette tape compander systems.

In the series connected circuits of the embodiments described above, each circuit retains a maximum compression ratio of about 2:1 and the maximum compression ratio of the overall combination of series circuits is about 2:1 at most input signal levels and frequencies.

However, in practical embodiments, it is difficult to avoid somewhat greater ratios, e.g.

2.5:1, in a small range of levels and frequencies. This can be tolerated if the compression ratio is not more than about 2.5:1 (or about 1-1 /4 times that of each circuit) and if the range of levels and frequencies in which it occurs is not large.

Another specific embodiment of the invention depicted generally in Fig. 2 is to configure one compressor and expander as a sliding band device, such as described in US-PS 3,846,719 and US-PS 3,903,485, and the other compressor and expander as a sliding band device. One suitable split band or multiband device is described in the Journal of the Audio Engineering Society, Vol, 15, No. 4, October, 1967, pp. 383-388. Split band devices in accordance with the parameters in that paper have become well known as A-type devices.

In a practical embodiment, an A-type compressor receives a flat input signal and provides its output to a specially tailored sliding band device. It is most advantageous to locate the A-type device so as to receive an unprocessed input signal because it is designed to process a flat input signal. Placing the sliding band device first would have the disadvantage of changing the flat input signal to a form less suited for the A-type device input. On the reproducing side, the sliding band expander receives the channel N signal, processes it and applies it to the A-type expander.

Fig. 1 6 shows curves similar to Fig. 9 for the low signal level response of an A-type compressor alone, the sliding band compressor alone and the combined compressor response. The expansion reponse curves are complementary in the manner of Fig. 9. The A-type device provides 10 dB compression up to about 5 kHz, above which the increase in level rises smoothly to 1 5 dB at 1 5 kHz.

Advantage is taken of this rising response of the A-type characteristic to desensitize the sliding band characteristic at high frequencies (see the high frequency portion of the "sliding band" curve in Fig. 16); this is advantageous in reducing the effects of high frequency channel-response uncertainties, hereinafter described in greater detail. Thus, the combined response curve climbs smoothly to 20 dB where it remains essentially level to about 14 kHz where it drops off. The sliding band device is designed to have operating thresholds and resulting areas of dynamic action which are well clear of those of the Atype circuit.

Fig. 1 7 shows a series of response curves at different levels for the series A-type and sliding band compressors. These curves present the same type of information as Fig. 11.

The shaded area C indicates generally the dynamic regions resulting from action of the A-type device; shaded area D from action of the sliding band device. This arrangement results in a maximum compression ratio which at any level or frequency does not exceed about 2:1 and therefore is relatively free of error amplifcation effects in practical tape recording channels.

It will be understood that for exemplary purposes a standard A-type device is connected in series with a special sliding band deivce. However, in principle, the A-type device can be modified to shift its areas of dynamic action so as to provide the best fit with the action areas of the sliding band device.

The precise amount of staggering or shifting necessary in this and other configurations set forth herein will depend on the parameters of the signal processing devices used. The goal in staggering the areas of dynamic action is to minimize bunching effects in the response curves. Bunching is an indication of large compression or expansion ratios. For example, see Fig. 1 8 which shows excessive bunching; that is, at some frequencies and levels a 10 dB change in input level results in a 2-1 /2 dB change in output -- a 4:1 ratio. Optimally, with appropriate staggering, a 2:1 ratio is never substantially exceeded in a cassette compander system throughout most of the range of levels and frequencies. In other types of transmission systems higher compression ratios may be acceptable.

Claims (26)

1. A circuit arrangement for modifying the range of an input signal, comprising a first circuit with a bi-linear characteristic composed of a low level portion of substantially constant gain up to a threshold, an intermeddiate level portion, above the threshold, of changing gain providing a maximum compression ratio or expansion ratio, and a high level portion of substantially constant gain different from the gain of the low level portion, and at least one second circuit, which also has a bi-linear characteristic within a frequency range common to the circuits, in series with the first circuit, the intermediate level portions of the characteristics of the circuits being staggered within a frequency range common to the circuits such as to provide a change of gain over a wider range of intermediate input levels than for any of the circuits individually, and an increased difference between the gains at low and high input levels, but with a maximum compression or expansion ratio which is substantially no greater than that of any single circuit, by virtue of the staggering.

2. A circuit arrangement according to claim 1, wherein the circuits are compressors and the thresholds thereof are set to different values in order to stagger the intermediate level portions of the characteristics of the circuits.

3. A circuit arrangement according to claim 1, wherein the circuits are expanders and the thresholds thereof are set to different values in order to stagger the intermediate level portions of the characteristics of the circuits.

4. A circuit arrangement according to claim 1, 2 or 3, wherein the falling flank of the compression or expansion ratio of each of the circuits overlaps the rising flank of the compression or expansion ratio of another of the circuits to the extent that the overall compression or expansion ratio in the overlapping region does not substantially exceed the maximum compression or expansion ration of the adjacent circuits.

5. A circuit arrangement according to claim 4, wherein each of the circuits has a maximum compression or inverse expansion ratio of substantially 2:1 and the maximum ratio for the complete circuit arrangement does not substantially exceed 2:1.

6. A circuit arrangement according to any of claims 1 to 5, wherein at least one of the circuits comprises a variable filter providing a boost or cut in a high or low frequency region of the signal band, and responsive to signals in that region to cause the filter corner frequency to slide in the sense narrowing the boosted or cut region.

7. A circuit arrangement according to claim 6, for audio signals, wherein the or each variable filter has a rectifying smoothing and amplifying control circuit providing a con trol signal to a controlled impedance device of the filter to effect the sliding of the filter corner frequency.

8. A circuit arrangement according to claim 7, wherein each circuit comprises a variable filter.

9. A circuit arrangement according to claim 8, wherein the control circuits of the variable filters have different gains such as to establish the different thresholds of the circuits.

10. A circuit arrangement according to claim 8 or 9, wherein the quisecent corner frequencies of the variable filters are substantially the same.

11. A circuit arrangement according to claim 10, wherein the boost or cut is provided in a high frequency region and the quiescent corner frequencies of the variable filters are in the region of 300 to 400 Hz.

1 2. A circuit arrangement according to claim 8, 9, 10 or 11, wherein in at least one circuit the variable filter is in series with a fixed filter having a narrower pass band than the variable filters in the quiescent state.

1 3. A circuit arrangement according to claim 12, wherein at least two circuits have a variable filter in series with a fixed filter and the corner frequencies of the fixed filters are substantially the same.

1 4. A circuit arrangement according to claim 13, wherein the boost or cut is provided in a high frequency region and the fixed filters have a corner frequency in the region of 300 to 400 Hz.

1 5. A noise reduction arrangement comprising a compressor arrangement according to claim 2 or any of claims 4 to 1 4 insofar as dependent thereon and an expander arrangement according to claim 3 and any of claims 4 to 14 insofar as dependent thereon, wherein the threshold of the previous circuit and the threshold of each successive expander circuit is higher than the threshold of the previous circuit.

1 6. A noise reduction arrangement comprising a compressor arrangement according to claim 2 or any of claims 4 to 14 insofar as dependent thereon and an expander arrangement according to claim 3 and any of claims 4 to 14 insofar as dependent thereon, wherein the threshold of each successive compressor circuit is higher than the threshold of the previous circuit and the threshold of each successive expander circuit is lower than the threshold of the previous circuit.

1 7. An arrangement according to any of claims 1 to 16, wherein at least one of the circuits is a dual-path circuit comprising a main path which is linear with respect to dynamic range, a combining circuit in the main path, and a further path-which has its input connected to the input or output of the further path and its output connected to the combining circuit, the further path providing a signal which, at least in an upper part of the frequency band, boosts or bucks the main path signal by way of the combining circuit, but which is so limited that, in the upper part of the input dynamic range, the further path signal is smaller than the main path signal.

1 8. An arrangement according to any of claims 6 to 1 7 and claim 17, wherein at least one of the circuits is a dual path circuit with the variable filter in the further path thereof.

1 9. A circuit arrangement according to claim 1 wherein the approximately threshold level T of a given stage is determined by the relationship CG T = F- C- 1 where F is the finishing point of the stage, whereat the intermediate level portion ends and the high level portion starts, C is the maximum compression ratio of the stage and is G is the gain of the stage.

20. A circuit arrangement according to any of claims 1 to 14, wherein the circuit arrangement is for modifying the dynamic range of audio signals and wherein each circuit includes an overshoot suppressor having a threshold level, the threshold levels being staggered among the circuits such as to provide a reduction in the overshoots of the overall circuit arrangement.

21. An arrangement according to claims 1 5 and 20, wherein the threshold of the overshoot suppressor of each successive compressor circuit is lower than that of the previous circuit and that the threshold of the overshoot suppressor of each successive expander circuit is higher than that of the previous circuit.

22. An arrangement to claims 1 6 to 20, wherein the threshold of the overshoot suppressor of each successive compressor circuit is higher than that of the previous circuit and that the threshold of the overshoot suppressor of each successive expander circuit is lower than that of the previous circuit.

23. A circuit arrangement according to claim 11, wherein there are two bi-linear circuits, each having a low level gain of substantially 10 dB.

24. A circuit arrangement according to any of claims 1, 2, 3, 4, 1 5 and 16, wherein there are three bi-linear circuits, each having a low level gain of substantially 8 dB.

25. A circuit arrangement according to claim 6, wherein one of the circuits comprises said variable filter, and wherein there is one further circuit comprising a plurality of band pass filters each in combination with a limiting device.

26. A circuit arrangement according to any of claims 1 to 5, wherein the maximum compression or inverse expansion ratio for the complete circuit arrangement does not exceed about 1.25 times the maximum ratio of any one of the circuits.