DOD R-825 Compressor/De-esser
Regular contributor Dave Simpson, who runs a Fostex B16-based commercial studio, recently added the DOD R-825 to his extensive armoury of effects. Having compressed, limited and de-essed bis way through many sessions since, here he presents his wholly subjective User Report.
Over the past few months, this magazine has contained various articles dealing with the techniques of compression, as well as reviews of several machines (not to mention the compressor/limiter project in the January issue). This all accurately reflects the ever growing importance of the compressor in any modern studio and home set-up. Correctly used, they can reduce noise (by squeezing higher signal levels on to tape), compensate for a lack of musical technique (particularly vocals, by smoothing out irregular vocal levels) and can even become an extension of the instruments themselves, as in the case of the modern rock and disco bass guitar sound which relies heavily on compression.
Owing to the recent articles (which I assume you will all have read!) I will be brief in the explanation of compression and limiting. Basically, a compressor/limiter is a dynamic device in which gain reduction is automatically controlled by programme level, increasing attenuation progressively as the level rises above a pre-determined (threshold) point. Compression is used to describe a process of gain reduction that is more or less continuous, with the original input signal dynamics being compressed producing a resultant increase in overall loudness. Limiting is where the signal level is restricted at some specific point, preventing transients (peaks of short duration) from exceeding the pre-determined peak recording level.
The DOD R-825 is unlike any of the other compressor/limiters we have reviewed in two major respects. Firstly it is a mono unit, unlike the more commonly found stereo devices such as the Vesta Fire, MXR, Drawmer and Ashly. There is, however, a mono jack socket on the rear panel by means of which two units can be linked together to produce stereo compression with no image shifting. Secondly, it contains a de-essing device which can be used independently of, or in conjunction with, the compressor. More of that later however.
As you can see by the photograph, the front panel of the unit is divided into two sections. To the right are the compression controls, activated by an in-out switch. To the left of the group is the Ratio control, variable from 1:1 to 30:1. This sets the slope dictating the amount of compression. What happens is this: up to an input threshold level of 0dB the device acts like a normal amplifier with unity gain. This means that for every 1dB of level put in, you get 1dB out!
Above the threshold though, output levels are reduced according to the slope you have chosen. Incidentally, in case you are unsure of the difference between compression and limiting, it is this. Compression is the term used where gain reduction is more or less continuous. This usually involves a slope of between 1.5:1 and 5:1, thus reducing the original dynamics of the signal by varying degrees (I will give you some settings later). Limiting occurs when, a hard slope is used - say 20:1 - where, in effect, the output level remains unchanged no matter what the input level.
This can be used in two ways. You can set the device so that the VU meters of your tape recorder flick immediately to 0VU and stay there, whatever the music content. AC/DC seem to use compressors in this way, as did many records in the 1960s - for instance, Telstar by the Tornadoes. This means that you obtain perhaps the optimum signal-to-noise ratio. Unfortunately, it also means that you end up with practically no dynamic range.
The other way of using limiting, is to prevent the occasional peak from pushing your meter needles into the red, by reducing the input level so that only the very loud passages trigger the compressor. This needs to be judged carefully however; if the unit cuts in too sharply the audible volume drop can be annoying. I attended a Rolling Stones concert once, where the whole mix was poorly limited - every time the band hit a loud note, the entire PA volume actually dropped, resulting in a ludicrous reversal of emphasis.
Returning to the controls on the DOD, the fact that the compression ratio is variable up to 30:1 is in itself slightly unusual, since most other compressors, rather than offering hard limiting in terms of a specific ratio, offer it in terms of infinity:1. In practice though, 20:1 seems to cope with most eventualities.
Moving to the right, the Attack time control is variable from 1ms to 40ms. This determines the speed at which the compressor cuts in. With short attack times the unit responds almost instantaneously whilst longer attack times allow a smoother transition to compression. Unfortunately, if slow attack times are used with vocals, sibilants tend to be emphasised, since it is just those peaks which are allowed to pass unattenuated. I will return to this point later (for obvious reasons).
I would have liked to see a slightly faster attack time; the majority of modern units offer 0.5ms or better. The difference is small, and in practice probably makes little or no difference, since a very fast attack time can result in a harsh sound. However, if Fostex can do it, I am sure DOD can.
To an extent, the same criticism applies to the next control - the Release knob. Release time, perhaps better thought of as recovery time, is the time taken for the compressor to return to its original setting. Fast release times are useful simply for limiting peaks, where the original signal dynamic is restored immediately the transient has passed. Longer release times suit compression applications - to smooth out a vocal for instance. The release times on the DOD can be varied between 170ms and 1 second. Again, comparable units have release times up to 5 seconds. Although I must in all honesty admit that the MXR unit I also own has never been set beyond 1.5 seconds, I still find it hard to see why DOD could not incorporate a release time of even 2 seconds. Ah well...
The other two controls relating to the compressor section are fairly self explanatory. The Input control dictates the amount of signal that is fed through the compressor and is used in conjunction with the Ratio control to determine the amount of compression. For instance, a high input with a low ratio results in a touch of compression over much of the signal. A low input with a high ratio leads to parts of the signal (the peaks) being heavily compressed, resulting in limiting. The Output control does exactly what it says; it controls the output from the unit.
All the compressor controls are in/out switchable from the front panel, allowing A/B comparison with the original signal. However, the front panel LEDs, which show the amount of compression, keep functioning even when the compressor itself is switched out, enabling visual monitoring to continue. To the right of the controls, a single LED indicates clipping, remedied by lowering the input so that distortion does not occur.
To the left of the front panel LEDs which indicate the degree of compression are two pots marked De-ess Level and De-ess Output. Although this is probably the most onomatopoeic word in modern music technology, it is perhaps worth explaining what a de-esser does, and why it is required.
The problem of sibilants has been with the film industry in particular, and latterly with the recording industry, for many years. Several factors can cause the 'ess' sounds of a vocal to become objectionably prominent; close-miking techniques, cheap microphones, and the boosting of those frequencies which can appear to give a vocal presence - say around 8kHz. To cap it all, compression, even with a fast attack, tends to let sibilant peaks pass before modulating the remaining portion of the signal, which serves to emphasise them even more.
Traditionally, there have been two main ways of tackling this problem. One way has been the use of a programme control filter, used for some time in disc cutting. The other, most popular, has been to make the sidechain of a limiter frequency conscious, so that the threshold is more sensitive to high frequencies than low ones. As the threshold is reached, the sibilants are modulated to an acceptable level.
Many modern compressors offer sidechain access, making them potential de-essers by patching in a graphic equaliser and boosting frequencies around 8kHz. Unfortunately, this method suffers from two deficiencies. Firstly, it ties up the use of a graphic or parametric equaliser. If your mixing desk has limited EQ then any outboard equalisation may well be needed elsewhere (for instance, on the vocals themselves). Secondly, even if you are fortunate to possess a spare graphic equaliser, using this method of de-essing, modulation still affects the whole bandwidth. What this means is that every time a sibilant occurs, the volume of the whole signal will drop. This problem can become acute on multiple signal applications, or when more than one vocal is involved. If de-essing is heavy enough, every time the vocalist triggers the effect - the orchestra fades into the background!
The DOD also offers sidechain access. However, the unit already has a built-in de-esser, operated from the front panel by the aforementioned controls. It functions in the same way as for sidechain access, except that it has fixed attack and release times - 4.7ms and 200ms respectively, and presumably over a fixed high frequency band. I say presumably because this information is not included in the owner's manual supplied with the unit. I will return to this mighty tome presently!
What it does reveal, however, is that in order to combat whole bandwidth modulation, the de-essing circuit of the unit separates the high frequencies (all those above 1.8kHz), compresses them independently from the low frequencies and then sums the two outputs. Thus, low frequency information is not lost when excessive low frequency signals trigger the voltage controlled amplifier. The De-ess Level knob controls the amount of compression in the high frequency portion of the input signal. Suppression of sibilants occurs as this is turned clockwise. The Output control matches the output level with the compressor output.
Since the de-essing circuit can be switched in and out independently of the compressor, allied to the fact that both compressor and de-esser function well, the possibilities of the unit are many and varied. It can be used as a de-esser only, as a compressor only, or as both. It can even be used as a kind of dual compressor, with different compression ratios for high and low frequency signals.
So far, so good. Now for a few typical settings you might find useful. I must also point out that you will need them, because the owner's manual as it is laughably referred to, is the worst I have yet come across. I admit, it grudgingly gives you most of the specifications, and indeed includes a brief line on each of the control functions. But that's as far as it goes. If some manufacturers can produce a good sized booklet giving possible settings and ideas for application (I had to file the previous Ashly ones along with the dictionaries), I am sure that DOD, given this versatile little unit, could have pointed the potential user in the right direction. Anyway, they didn't - so try these.
Example Control Settings
BASS GUITAR — Ratio 10:1, Attack 15ms, Release 500ms.
PIANO — Ratio 8:1, Attack 30ms, Release 170ms.
STRINGS — Ratio 5:1, Attack 5ms, Release 170ms.
GUITAR — Ratio 12:1, Attack 5ms, Release 250ms.
SNARE — Ratio 9:1, Attack 1ms, Release 300ms.
CYMBALS — Ratio 10:1, Attack 1ms, Release 200ms.
VOCALS — Ratio 8:1, Attack 8ms, Release 190ms.
Basically, the factors to be weighed up when considering a product of this nature are that it is a mono unit, as against most of its competitors which are stereo/dual units. In addition, it retails at a price which although competitive, is still £70 more than half the price of a low cost dual unit (which would provide you with two compressors).
Balanced against this though, is the extremely useful de-essing circuitry. You may have noticed that I did not supply settings for the de-esser. This is because every vocal will require a different amount of sibilance control and the only way to find out how much is to experiment. Basically, the technique is to introduce the de-esser until it becomes noticeable, and then back off a bit. If you consider the cost of de-essing using a normal compressor and using a graphic equaliser in the sidechain, then the DOD looks relatively good, with more sophisticated de-essing into the bargain.
Given that in a home studio set-up only one compressor at a time is really needed (using multitrack techniques), and given also the highly objectionable nature of sibilants in a vocal line; if you are considering compression and do not feel that stereo compression is a must, I would think carefully about this DOD unit.
The de-esser is excellent, and the compressor, given the minor limitations of its attack and release specification, copes well. I would also have liked to see, and would have been prepared to pay more for, an expander/gate function incorporated into the unit. Since I bought mine I have used it on every vocal track I have recorded, with a noticeable improvement in recording quality. I only wish it had the same effect on the vocalists!
User Report by David Simpson
Previous article in this issue:
Next article in this issue: