How do you begin to review a unit such as this? Symetrix describe it as a 'multifunction controller', and it certainly is that! It compresses, limits, expands, gates and ducks either in mono or, by means of a linking switch on the front panel, in stereo. Is it though as versatile as it appears on the surface?

Before I answer this question, it would seem a useful exercise to undertake a brief description of each individual function. If you know all about compression, expansion and ducking, skip the next bit!

Compressor/Limiter

This is probably the most well known of the four functions, and perhaps the most useful. To refresh your memory, compressors and limiters are dynamic devices in which the gain is automatically controlled by programme level, the attenuation progressively increasing as the input level rises above a pre-determined point. The signal passes through an amplifier which maintains a fixed gain relationship between input and output at all signal levels below a chosen threshold. Beyond this point, for every dB increase at the input, the output rises by an amount determined by the ratio or slope selected.

This can prove advantageous in many recording applications. Instruments which play both very quietly and very loudly, have always been difficult to record. For instance, imagine a vocalist during the course of a song. At the start, he or she might sing fairly quietly, in some cases almost at a whisper. As the song builds to its climax though, in order to increase the feeling in the voice, the full power of the vocal is used.

This presents a problem to the studio engineer. If he records the vocal so that the quiet sections are recorded with the VU meters peaking correctly, the loud sections will almost certainly introduce distortion onto the tape. On the other hand, if he anticipates the loud vocal and sets the recording levels accordingly, the quiet passages will be recorded at very low level onto the tape, resulting in a possibly inaudible vocal when it comes to the mix, not to mention the increased tape hiss due to the large gap between signal level and tape headroom.

There are three ways around this problem. The first concerns microphone technique. Quite simply, every time the vocalist hits a loud note, he or she backs away from the microphone. This is difficult to judge in a studio situation though, since monitoring usually takes place through headphones whilst singing.

The second method is for the engineer to 'ride' the faders as the vocal is being recorded, whilst trying to anticipate the dynamics of the vocal and move the fader up and down as required. This can also be done on mixdown, but the signal to noise ratio is worsened. Both methods are less than satisfactory.

The third method involves the use of a compressor. Since the compressor can be set so that the louder the vocalist sings, the more the gain is reduced, the dynamic range (the difference between the quiet passages and loud passages) is reduced. More level is therefore 'squeezed' onto tape resulting in an audible signal at low levels, no distortion at high levels and less tape noise since the tape headroom is more effectively utilised.

The difference between a compressor and a limiter is determined by the ratio of compression. A ratio of 2:1 means that as soon as the signal reaches a pre-selected threshold, for every 2dB of input, the output rises by 1dB. A ratio of 20:1 means that the output continues in a linear relationship until the threshold level is reached, and from then on, every 20dB of input raises the output by 1dB. A 2:1 ratio or a shallow slope is referred to as compression, which preserves more of the original signal dynamics. A high ratio such as 15:1 or greater is called limiting, the function of which is to prevent transients (peaks of short duration) from exceeding the pre-determined peak recording level.

Expander

If a compressor/limiter is a device which causes automatic gain reduction above a certain level, then an expander can be thought of as a device which causes automatic gain reduction below a selected signal level. In other words, an expander is really the opposite of a compressor/limiter. It is used to increase the dynamic range of an audio signal, exhibiting unity gain above threshold, as opposed to a compressor, which exhibits unity gain below the threshold.

Expanders usually operate at very low ratios. An expansion ratio of 2:1 indicates that for every 2dB output increase the input has increased by 1dB.

This facility has two important uses. Firstly, it means that instruments or signals with a low dynamic range can be given a larger one. After my previous comments regarding compressors being almost indispensable in contemporary recording situations you might be surprised that this might ever be necessary but it does happen. Modern keyboards for example, can be given 'feel' by the use of low ratio expansion. There might even be cases where a signal has been compressed, and the original dynamic range needs to be restored. DBX and related systems of noise reduction are based on this principle.

Another use of an expander is as a noise reduction system in its own right. Very low signal levels often represent unwanted noise - for instance, mains hum or buzz after the signal has ceased.

The threshold on the expander can be set so that any signal below the quietest signal level produced by the instruments is reduced at a selected ratio.

You can begin to see now why this device incorporates an expander alongside a compressor/limiter. One unfortunate side effect of a compressor is that it increases unwanted noise. With a straight compressor, it is best to operate on the direct signal before it is recorded on to tape. When the compressor recovers, the gain in the system increases, and any low level programme content is brought to a higher level relative to the unprocessed signal. Thus any source noise is increased, and tape hiss in particular can become exaggerated. If however an expander is inserted in the signal chain after the compressor, this noise can be attenuated by setting the threshold of the expander at a point just above the noise ceiling. As soon as the signal level falls below the threshold it is reduced sharply, thus attenuating the noise.

Noise Gates

I have mentioned that expanders typically operate at low ratios. A noise gate is simply an expander which operates at a high ratio, (for example 40:1). In other words, whenever the signal moves below the pre-set threshold level, it is attenuated by 40dB for every 1dB drop in input level. In practical terms, this has the effect of muting the channel instantaneously. I should qualify this though by mentioning that noise gates usually have a number of other controls relating to attack, decay, and the degree of attenuation (not the ratio, but the level to which the signal drops once threshold is reached, usually referred to as the range control).

It is no exaggeration to say that noise gates can be enormously useful, both in remedial and creative contexts. As a 'gate' for shutting out unwanted noise, it is used in much the same way as an expander, setting the threshold level below which all signals are attenuated. The gate opens to let the signal through, then closes when it has ended, thus ensuring that any noise content is eliminated. The noise of course, is still present when the signal is above threshold level, but most signals have a sufficient signal to noise ratio to enable the threshold to be set somewhere in between.

One problem with a noise gate, however, is that although they open quickly, clicks sometimes occur due to the square edge imparted to the leading edge of the input signal as the gain suddenly increases to full volume as the gate opens. The answer to this is to set the gate so that the attack time (ie. the time taken for the gate to open) makes the level rise more gradually after the threshold is reached. This, though, might mean that part of the signal will be lost if the gate opens too slowly. Decay time (ie. the time taken for the gate to close) is also usually user controllable, and this is set according to the decay characteristics of the instrument being recorded.

You might begin to wonder why an expander is used when a simple noise gate would appear to do the job more efficiently. Because gates operate at higher ratios, the audible effects sometimes tend to be drastic, and unwanted modulation can be produced. It's as well to be cautious with such a device when gating a signal on to tape, as one fractional error in setting the threshold or release levels can result in the loss of part of a valuable take. Gates are often used therefore during a mix-down when the multitrack can be re-run until the gate is set correctly. Some desks even have gates built in to every channel (notably SSL).

Expanders on the other hand can be set so that modulation effects are minimal. Admittedly, the amount of noise reduction is less, but use of an expander permits some noise reduction without jeopardising the signal during the recording process. Ultimately, the final decision as to whether or not to use an expander or a gate rests with your ears!

Gates have other uses apart from noise reduction, though. A signal can be processed such as a bass guitar, but the gate keyed (ie. made to open and close) by another signal, such as a bass drum. This has the effect of ensuring that the bass guitar is heard at precisely the same time as the bass drum, so resulting in a tight, punchy sound. Noise gates are also useful in eliminating crosstalk or leakage, most notably when recording drums. Putting a noise gate on the bass drum for example, eliminates any other extraneous sounds from being recorded on the bass drum track, and therefore allows scope for flexible equalisation without affecting other drums.

Ducker

A few months ago I described a de-esser as the most onomatopoeic name for an effects processor. I think you would have to go a long way to find a more descriptive name than 'ducker', considering that this is exactly what the unit does. Basically, it causes gain reduction upon command from an external audio signal, rather like a noise gate but with two important differences. Firstly, the signal being reduced in gain (ie. ducked) is not usually the signal which triggers the ducking. The second and most important difference is that in the case of the noise gate, when the trigger signal exceeds threshold, the gate opens. In the case of a ducker, when the trigger signal exceeds threshold, the ducker closes, or attenuates the processed signal. It's also important that unlike a compressor, when the external signal exceeds threshold, the ducker responds with a fixed amount of gain reduction - the gain reduction not being dependent upon the amplitude of the external signal once threshold has been exceeded.

There are two main applications for this facility, one being in a voiceover situation. In this case a music track is made to 'duck' or decrease in gain whenever the announcers voice begins (exceeds the threshold level), a technique widely used on radio and television. The second application involves a multitrack mix-down situation, where a rhythm or lead instrument is reduced in volume every time the lead vocal or instrument appears, in order to reduce clutter on the final tape. (I expect that other variations on this theme may spring to mind.)

Description

So much for the principles behind the design of the Symetrix 522. The question is; do they all work and how well? Beginning again with the compressor, there are four controls: threshold, attack, release and ratio. It is perhaps worth mentioning at the outset that these same four controls apply also to the expander, gate and ducker functions, and change their function depending on which mode is selected. Returning to the compressor/limiter, the threshold control selects the signal level at which compression begins. The attack control determines how quickly the unit responds to a triggered signal. Fast attack times suit sharp percussive sounds such as snare or slapped bass, whilst with a high compression ratio they can limit transients. Slower attack times with shallower ratios are useful for controlling more gentle signals such as vocals.

The release control determines the time the device takes to return to unity gain. If the 522 is being used as a limiter, then fast release times will return the signal to its original level after distortion has been effectively prevented by limiting the transient. With sustained sounds however, a sudden return to unity gain might introduce unwanted modulation, so a slow release time should be used.

An unusual feature of this control is its 'program dependency' mode. As the control is adjusted towards the centre position, the nature of the signal itself helps to determine the release time. For example, long sustained sounds will tend to automatically create long release times, while short staccato material will create fast release times, thus aiding situations where the nature of the program material is constantly changing by automatically determining the optimum release time. This feature, incidentally, applies also to the other three modes.

Finally, the ratio control does just that - adjusts the ratio, from 1:1 (unity gain) to infinity to one (hard limiting). Ratios of between 1:1 and 7 or 8:1 are usually referred to as 'compression'. Higher ratios are known as 'limiting'.

Expander Mode

The threshold control operates in the same way in this mode as the compressor except that, as previously stated, the device cuts in when the signal level passes below threshold rather than above it. The attack and release controls function in much the same way as they do on the compressor. With most signals, fairly fast attack and release settings seem to work well, but the release setting is particularly dependent upon the ratio, with high ratios calling for longer release times. The ratio control adjusts the amount of downward expansion once the signal falls below threshold. Using higher ratios, the result is an increase in dynamic range, or noise reduction (if that is the intention). Lower ratios are best for subtle increases in dynamic range. Higher ratios can be used for special effects. It is worth noting that the higher the ratio, the more the device acts as a noise gate, which can lead to unwanted side effects (such as modulation).

Gate mode

In this mode, once again the threshold control acts as in the case of the previous two examples and as with the expander, the unit is triggered when the signal passes below the threshold. Attack and release knobs perform in the same fashion as the expander with 'attack' meaning the amount of time it takes for the gate to open, and release the amount of time it takes the gate to close. In the case of the fourth control that was previously designated ratio, this now becomes 'range'. It adjusts the range of attenuation or the amount of gain reduction which takes place once the input signal falls below threshold. Calibrated in dB steps, this is variable from 0dB (no effect) to 60dB (full gating). Although the maximum setting will usually be the one selected, in certain cases full range gating will produce unnatural responses. This sometimes happens with drums, and by adjusting the range control, the signal is not completely attenuated thus leaving some signal as background producing a more natural sound. It is worth bearing in mind though, that a sharply gated signal might sound strange when listened to in isolation, but within a multitrack mix, these characteristics may become less noticeable. It is definitely worth experimenting.

Ducker Mode

As far as the ducker goes, the threshold control works in the same way as the compressor, triggering the device when the trigger signal passes above threshold. Unlike a compressor, however, which reduces the rate at which gain takes place, the ducker reduces the actual gain itself, the attack and release controls determining the time taken for the signal to reduce in gain and recover to unity gain. This will tend to depend on both the voice over signal and the background or music signal - it can sound a little strange sharply reducing a soft violin piece in volume!

As in the case of the noise gate the last control adjusts the range of attenuation, or the amount of ducking. As the control is adjusted clockwise the signal is increasingly reduced in volume. The amount of ducking is again dependent on the individual situation. For most applications between 3 and 10dB of attenuation will suffice.

Layout

So far I have detailed the theory behind each function and the controls relating to each. At this point, I should explain how the unit is laid out, and a few other additional features not so far mentioned.

As you have probably realised, the fact that the four main controls are multi-functional indicates that at any one time only one of the modes can be used. However, there are two complete sets of controls, and they can either be linked in stereo (for instance stereo compression or gating, ensuring that no image-shifting occurs), or linked so that any one mode can be used in conjunction with any other mode in the second channel. In other words, you can select the same signal to a noise gate on the second channel.

To select any particular effect, a mode switch is provided, depression of which moves an LED indicator along a row of four. At the side of each is labelled one of the four functions. You depress the button until the LED is opposite the mode required. (This is one of those things that is almost impossible to explain but obvious when you see it!) In addition to the four LEDs, channel two has a separate additional LED. By depressing the mode switch until the LED labelled 'slave' is alight, channel two is 'slaved' to channel one, whose controls are overridden. Stereo operation is then possible using the controls of channel one as the master. To process a signal using both channels in different modes — i.e. expanded compression - you take a lead to the input of channel one, link the output of channel one and the input of channel two using a patch lead and take the output from channel two. In other words the unit can be used as two independent signal processors, two processors acting upon one signal or a linked stereo unit.

In addition to the mode select switch, each channel has an in/out switch allowing an A/B comparison between affected and unaffected signals, and a switch labelled Int/Ext. The normal operating position of this switch is in the Int (in) position. This allows the incoming audio signal to activate the control voltage processor circuitry (CVP). In other words, the unit is triggered by the signal being processed. In the Ext (out) position, the CVP is activated by the signal present at the return/ext jack on the rear of the processor. This means that the processor is triggered by a signal other than that being processed, (for instance, the example already given of a gated bass guitar triggered by a bass drum). This switch would also be used in the ducker mode, and the triggering signal (the voice for instance) is input via the socket on the rear panel.

Other features include an on-off switch on the front panel (nice to see one for a change) and sidechain access allowing the patching in of external processors such as equalisers (allowing facilities like frequency conscious gating, or vocal de-essing whilst in the compressor mode). Finally, a LED ladder display adjacent to each channel indicates the amount of gain reduction taking place at any one time.

Conclusion

When reviewing a budget unit, the conclusion often consists of saying something like 'the unit does do what it is supposed to but it's a bit noisy,' or 'the unit performs an adequate job considering the price'. When reviewing a unit of this calibre, however, one assumes that all the functions will perform to a professional standard. The conclusion is not so much based upon whether or not the features work, but rather, how useful is the device when used in a wide variety of situations.

Certainly, in this case the assumption that all the features offer professional performance is a valid one. In addition, despite the multifunction system of controls, the Symetrix offers most of the features available on dedicated units. In the case of the compressor/limiter, dedicated units such as DBX and MXR tend to offer slightly wider parameters such as a longer release time, and dedicated noise gates will usually offer features such as a 'hold' control (determining the length of time before the fade begins), or in the case of the Drawmer 201, frequency conscious gating controllable from the front panel. However, using the Symetrix in a wide variety of applications I found it up to scratch every time and I must also point out that the expander and ducker facilities offer all the features available on dedicated units.

I liked the concept of a multi-function processor, but I found it difficult to decide on the most appropriate environment for its use. I finally came to the conclusion (perhaps surprising in view of the professional specification of the unit) that the more basic the recording set-up, the more use that this unit would be. In fact, using it with a four track cassette machine such as a Portastudio or Fostex would utilise its potential to the full. To me, working in a professional sixteen track studio, although the unit performed faultlessly, I would rather pay the extra money and have a pair of compressors, a pair of gates and a pair of expanders, with all the flexibility that this configuration would allow. In that situation, six different signals could, if necessary, be processed at any one time. Having said that, there are times, for instance when a band is laying down a basic track, that I do need several gates and several compressors. If you trace this argument back, you realise that the fewer instruments you record at any one time, the more useful the Symetrix becomes. Used in conjunction with a basic 4-track, where the tracks are often recorded one at a time, I would imagine that a user would soon wonder how he ever managed without it. Tracks could be recorded with gated (or expanded) compression, squeezing every last dB of level onto tape without compromising in terms of noise. Any track bouncing could be done (even in stereo) using the expander as a noise reduction unit, and the final mix could be recorded on to the master cassette machine (or whatever) using a touch of compression and expansion to reduce any unwanted tape hiss. You begin to see the possibilities!

I am fairly sure however that Symetrix did not have the budget four track studio in mind when they designed this unit, but it seems to me that this is the market which might prove most profitable. As far as I am aware, there is nothing comparable available to small studio owners, and although expensive when viewed as a single unit, it becomes economic when seen as four stereo professional quality processors.

If you can only afford one dynamics processor, then this is the one to go for. It would not be as flexible as separate dedicated units, but then neither is its price as high. You pay your money and you take your choice!