Last month Paul White covered gates and compressors; this month it's up to enhancers to justify their existence...
When Aphex invented the Aural Exciter, some people latched onto it right away, while others thought they could see right through the emperor's new clothes and that conventional equalisers could produce exactly the same results. Now we're all a lot wiser and the principles of sound enhancement are more widely appreciated, with numerous different companies making enhancers that all work on a slightly different basis. There's a tendency for people to use the term 'exciter' as a generic description for any type of enhancer that brightens the mix, but Aphex would be the first to point out that the trademark belongs to them — at least when it has a capital E. In fact they're probably putting down their Thermos flasks and Hoovers right now and looking for a Biro to write us a stiff letter!
Over the past couple of years, SOS has examined the most popular enhancement systems, and while each manufacturer understandably claims that their system has some advantage over all the others, they're all reasonably effective in making sounds seem brighter, better focused, and more detailed. They do sound different when compared directly, but the purpose of this series is not to pick winners — it's to look at which situations benefit from signal processing and which don't.
Enhancers, on the whole, tend to emphasise transient detail, either by the use of dynamic equalisation, frequency‑dependent phase shift, or, in the case of Aphex, by a mixture of compression, filtering, phase shift and harmonic synthesis. The Aphex approach is particularly useful in adding a plausible high end to source material that has a limited bandwidth, but unless the programme material is badly lacking in high‑frequency content, all the popular systems will bring about some subjective improvement in clarity.
The benefits of enhancement are the ability to make a sound stand out in a mix, to make individual instruments and voices seem better separated, and to improve vocal intelligibility; the negative aspects, though, are less well known.
Like compressors, enhancers can be built using the quietest circuitry in the world, but if the input material is noisy, the process itself will still make that noise more obtrusive. Most enhancers attempt to emphasise low‑level transient detail, and whether that's done by harmonic synthesis or by compressing the output from a high‑pass filter, it inevitably results in some of the high‑frequency components of the original waveform being increased in level. Even if the process doesn't significantly increase the peak level of the signal, low‑level signals may be made louder, and it's during periods of low‑level signal that noise is less likely to be masked. As a significant proportion of background noise is high‑frequency hiss, it follows that this too will be made more noticeable by the process. After all, hiss gets enhanced just as effectively as any other high‑frequency component of the source signal!
Don't use any more processing than is needed, regardless of how much you paid for the box!
Some manufacturers claim that their enhancement processes don't have a significant effect on background noise; ultimately, though, if the process is successful in lifting low‑level, high‑frequency detail, it must also lift noise. Often a single‑ended noise reduction device working on the dynamic filtering principle will help remove noise during quiet passages, but the only fully satisfactory approach is to make sure that your source material is as quiet as possible, and only use noise removal processes as a last resort.
Sibilance is a problem that every recording engineer dreads — the singer whose every S and T is accompanied by a burst of high‑frequency noise. This isn't anybody's fault — it's all down to how an individual's mouth works — but it seems that the better the microphone, the more sibilance is captured. This is especially true of some capacitor mics, but unfortunately some people tend to equate a very bright vocal sound as being 'more CD‑ish' or better produced. What's more, adding effects such as reverb or using heavy compression can make sibilance noticeably worse.
As sibilance is a high‑frequency problem, and enhancers are designed to emphasise high‑frequency detail, it's hardly surprising that using an enhancer tends to exaggerate sibilance even more. The best place to tackle this problem is back at source, and if you have a mic that's less susceptible to the offending frequencies, try that instead. Don't worry if it's less bright than the original mic: you can use the enhancer to help compensate for that. Changing the position of the singer relative to the mic may help, but in serious cases, you may need to resort to using a de‑esser.
I'm not a great fan of de‑essing using a standard full‑band compressor, as the result of gain reduction is often to make it sound as if the singer has a lisp. However, split‑band models, or dedicated de‑essers that affect only the required part of the spectrum, can work extremely well. If you need to use a de‑essing process, put this before the enhancer, and don't use more enhancement than you really need.
The effect of adding an enhancer is to make the treated sound seem closer and louder, even though its peak level may hardly have changed. The reason this happens is closely associated with the mechanics and psychoacoustics of the human hearing process. There's always a temptation to take the easy option and always put the enhancer on a finished stereo mix, just to add that extra sparkle, but this can be problematic on at least two counts.
The first objection to routinely enhancing stereo mixes is that you lose the enhancer's ability to create front‑to‑back perspective in a mix. If you process everything, the whole mix seems to move forward, which might sound louder and brighter, but doesn't help you improve the contrast between sounds — and in home‑produced music, maintaining clarity and contrast is often one of the most difficult things to achieve.
Hiss gets enhanced just as effectively as any other high‑frequency component of the source signal!
A better alternative is to use the enhancer while you're mixing. By patching it into a stereo subgroup, you can route all sounds destined for enhancement via that group and leave alone any sounds that don't need treating. For example, you may feel that enhancing the lead vocal and acoustic guitar makes the mix sound more immediate, and this contrasts nicely with the pad synth parts and fretless bass you've left in the background. Enhancing drums can help clarify a muddled sound, but unless the music is strongly rhythm‑based, you have to be careful not to pull the drums too close to the front of the mix, otherwise the lead vocals can get upstaged.
The other problem in enhancing a full mix after you've finished mixing is that the process may change the subjective balance of some of the sounds. If you feel the need to enhance everything, it's best to have the enhancer in circuit as you mix so you can compensate for any apparent balance changes. Working this way can also help you make more appropriate EQ settings, because if a recorded sound is already perfectly equalised, enhancing it will make it sound too bright. Once you're more experienced, you may also be able to combine enhancement and conventional EQ to recreate depth perspective, by making sure that the sounds destined to be at the back are not so toppy as those you want to place near the front. Having said that, the worst thing you can do is stick an enhancer on a perfectly good mix just because you have one.
The human hearing system is a marvellous thing — it's capable of resolving sounds against noisy backgrounds, it can compensate for different room acoustics, and it can even make dreadful transistor radios seem listenable. In other words, the hearing system is always adapting what we really hear to make it easier for the brain to extract the information it needs. So if you decide to enhance a track, you'll notice an immediate difference, but after a relatively short time your brain will have compensated to make it sound normal again. Now, turn off the enhancer, and the sound you originally thought was OK will sound dreadfully dull. The danger is that as the session progresses, you'll be tempted to add more and more enhancement, just to keep your brain thinking the device is working, and by the end of the day the original sound might be over‑enhanced to an alarming degree. Over time, you've got used to the new, enhanced sound, but the next day, when your ears have recovered, you'll wonder how you ever produced such a dreadful mix.
Another problem inherent in using enhancers is that it's very easy to come up with sounds that are impressive when you first hear them, but once you've listened for a few minutes, you feel irritated or fatigued. I find a lot of modern CDs suffer from being mixed fatiguingly bright; after you've played a few tracks, you just want to turn them off. The trick is to keep bypassing the enhancer to see how much effect you've added, and also to make comparisons against other records that you know are well mixed. Don't use any more processing than is needed, regardless of how much you paid for the box! Fatiguing sound is a particular problem if the original source material includes much in the way of of high‑frequency distortion, because most enhancement processes emphasise this further. In some cases, the distortion might not even be audible at all until you switch in the enhancer, after which it becomes almost unbearable.
Though they're all designed to meet roughly the same needs, every type of enhancer sounds a little different to its competitors, and in some situations one type will work better than another. For example, some of the models use dynamic filtering so that boost is only applied when the input signal exceeds a specific threshold; these are likely to be less prone to noise problems than the type that boost the lowest‑level signals the most. It's also true that, while systems based on dynamic equalisation can give very smooth results, harmonic reconstruction as provided by the Aphex Exciter range is more successful when the input signal contains little or no information in the frequency range you wish to emphasise.
The watchword when using enhancers has to be moderation. Don't take it as a sign of defeat if a mix or track sounds better without it — sometimes it will. You have to know when to recognise that the material you're processing is fine as it is, which is always a problem when a client brings in a tape and says they want it brighter, more punchy and with more body. What they're really asking for is more of everything, and if you give them it, you'll end up with exactly the same mix but a few decibels louder!
It may also be a mistake to reach for the enhancer before you've first tried a good‑quality outboard parametric EQ. Quite often, adding a wide boost centred between 15 and 20kHz will add the necessary shine without affecting the noise level in a very significant way, and, once you've brightened the top, you may also need to balance that with some bottom‑end EQ. Contemporary enhancers are sometimes fitted with bass boost facilities of varying sophistication, but often the simplicity of a really good equaliser will sound better. At the end of the day, an enhancer is a powerful and useful audio tool, but it's only one of many tools in the box.