MAKING A DIFFERENCE

Buying The Equipment You Really Need: Part 2

Published in SOS September 2000
Printer-friendly version

Technique : Miscellaneous

Last month, Paul White looked at the vocal recording chain to identify areas in which spending more money would produce better results. This month it's the turn of outboard effects and monitoring.

If you read last month's part of this short series, you'll know that by spending your money wisely, it is possible to get good recordings of vocals and acoustic instruments for relatively little outlay. So what makes the most difference when you come to put all the elements together in the mix? You might think it's the mixer, but unless you've chosen very badly, pretty much all of today's mid-price project studio consoles are capable of doing a good job. As explained last month, you may need to think about using an external equaliser for anything critical, but other than that, the main obstacles to a clean mix are inattention to gain structure, and hum caused by inappropriate signal wiring. Both these subjects have been covered at length in the pages of SOS, so if you need to refer back to them, dig out your April '98 and August '99 issues respectively.

A greater difference is likely to be made by your choice of monitor speaker, because if what you're listening to when you're mixing is significantly altered by your speakers, you may find yourself compensating in your mix for deficiencies that don't really exist, or missing problems that do! Desktop computer monitors are usually bad news for music recording and mixing, as are hi-fi speakers designed to flatter. For example, there is a 'hump' in the frequency response of many hi-fi speakers which tends to enhance frequencies at around 100Hz. Mixing through speakers like these, you might well EQ down low-end frequencies to compensate for the apparently overwhelming bass, and then find that your mixes sound bass-light through decent monitors. Similarly, if a cheaper speaker overly attenuates certain frequencies (as many hi-fi speakers do with mid-range frequencies), you will have a tendency to produce mixes which accentuate those frequencies.

You can expect to pay around £500 for a worthwhile pair of passive monitors and twice this for active models (ie. those with amplification built in), although there are still some models below this price that do the job well. In most home studios, a two-way monitor (ie. one with a woofer and a tweeter) with a reasonable, but not hugely extended, bass response is ideal. Very small nearfields will leave you guessing at what the bottom octave is up to, while monitors that cover too wide a frequency range at the bass end may, in an acoustically untreated room, stir up more problems than they solve. I find that models whose frequency response starts to roll off below 50Hz are about right for home-studio applications. If you want me to name names, you must realise that I can only comment on monitors I've actually reviewed. With this in mind, I'd say you should check out the Alesis Monitor Ones and Twos, Event 20/20s, Harbeth Xpressions and HHB Circle 5s (which are made by the same designer), AVI's NuNeutrons, Tannoy's Reveals and Dynaudio's BM5s, BM10s and BM15s.

Active monitors, though usually more expensive than equivalent passive speaker/power amp combinations, tend to produce a tighter sound and take the guesswork out of buying suitable power amplifiers. Again, check out Alesis, whose M1 Actives are surprisingly good for the price, Event's Active 20/20s, Dynaudio's BM6As and BM15As and Harbeth/HHB's Xpression/Circle 5 Actives, as well as the Tannoy Reveal Actives, but add KRK to the list as they also make active monitors that fall into this category, such as the E7, V8 and V6. If your budget can be made to stretch a little further, look at Genelec's 1030As and Mackie's HR824 actives — and if you really have money to spend, check out ATC's studio range, which I feel gives truly accurate results but with a hefty price tag.

Before proclaiming dissatisfaction with your present monitors, however, ensure that you have them set up in a sensible position and on rigid stands. Sand-filled steel hi-fi stands are perfectly adequate (cardboard boxes are not!) and you should use heavy-duty cable with passive monitors. Different cables do affect the sound slightly (see the box, left), but providing you're using heavy-duty, low-impedance cable and keeping the lengths as short as possible, you should be fine. Monitors are best used the right way up — not on their sides, as this affects their horizontal dispersion — and they should be positioned a little away from walls and definitely not too close to corners. Putting a speaker close to a corner produces an increase in low-frequency energy, but this increase varies with frequency, making it impossible get an accurate picture of what you're listening to. It's also better to place speakers on stands behind the mixer rather than on the mixer meterbridge, as reflections from the desk surface can alter the sound significantly.

If you choose speakers to suit your room and position them sensibly, you should be able to get good results but, in any event, you should always get used to the sound of your system by playing commercial CDs through it before doing any serious mixing. If you have a commercial studio, you may want to spend thousands on full-size, super-accurate, foundation-crumbling monitors, but unless these are used in a properly designed room (from the acoustic point of view), you're unlikely to get satisfactory results. If, as suggested earlier, you use monitors whose frequency response begins to roll off below 50Hz, any undetected errors below this should be fixable when you have your material mastered.

At one time, digital reverb was an expensive luxury, but now effects units costing less than £100 come with an assortment of reverb treatments. This being the case, does it make sense to spend more?

Actually, reverb is the one effect that has to be got right — it's a natural effect that we all hear every day, so if the electronic simulation doesn't cut it, the ear/brain combination won't fully accept it. And if the brain isn't convinced, it doesn't enjoy the music as much. A good reverb doesn't need hundreds of adjustable user parameters — once it's in a mix, the fine distinctions between room algorithm 5 and room algorithm 6 soon blur into meaninglessness. What you actually need as a minimum is a small but well-chosen selection of plate, room, hall and ambience settings, with the option of adding pre-delay, adjusting the decay time and changing the brightness, either by variable controls or by having a range of suitable presets to use as a starting point. Having some way to tweak the damping and reverb density is also useful, but when it gets to dialling in the number of room sides or the colour of the carpet, you're in the realms of the ridiculous. Remember virtually all the classic recordings up to the end of the '70s were made using plate reverbs which had one control for damping and a bit of basic outboard EQ. If you wanted pre-delay, you patched a Revox open-reel recorder in front of the plate and used the tape speed switch plus varispeed to set the delay between the record head and the playback head.

How convincing a reverb sounds depends on how well the algorithms are designed. Obviously there should be very low noise and adequate headroom but, in subjective terms, the best test is to adjust the wet/dry mix control, close your eyes and listen. If the dry sound fuses with the reverb to produce the illusion of a real sound in a real space, that's a good thing. If, on the other hand, the reverb seems separate from the dry sound, almost as though it's stuck on as an afterthought, it's probably not going to work very well. Another test of a good reverb is that you can add much more of it without swamping the original sound than you can with a cheap unit.

Cables     Digital equipment should always be connected using purpose-designed digital cable (especially S/PDIF connections), as this has a different impedance from audio cable — cheap phono leads might appear to work, but the error rate may be surprisingly high, and you're quite likely to suffer from the occasional digital click. True digital cable has a 70(omega) impedance and will result in more accurate data transfer. Speaker cable needs to be heavy in order to exhibit a low impedance but don't feel you need to buy anything too esoteric. Just keep both sides the same length and don't use more cable than you need to. You also need to ensure that good connectors are used and that they're kept clean. I use Caig Labs Deoxit to clean my plugs, patchbays and computer connections, which probably makes more difference than anything else, and it costs less than £15 a can.

The final test should be to try shorter reverb settings (small room and ambience programs with a decay time of under one second) to see if these are believable, and again check that they fuse with the dry sound in a convincing manner. Listen to the illusion of space — does the reverb seem one-dimensional and lifeless, or does it appear to have depth as well as width? Reverb is all about creating the illusion of space, so its effect on stereo imaging is important.

In the context of a mix, a good reverb unit will provide space without clutter and it will help to knit together the different components of a mix. Top of the tree in my opinion are definitely Lexicon and TC Electronic, with high marks for realism also going to Sony for their extraordinary DRE S777 Convolving Reverb, although like all technological breakthroughs, it's not cheap (see SOS December 1999 for more details). Further down the scale, the Alesis Q20 and some of the models made by Roland (such as the SRV3030), and Yamaha (such as the REV500) are also very good. Best buy at the moment has to be the sub-£200 Lexicon MPX100 (or if you can afford a little more, the MPX500 sounds noticeably classier), but if you want the very best without paying silly money, then check out the Lexicon PCM90 or PCM91, which is as close as you'll get to the industry-standard Lexicon 480L without paying serious used car prices (the 90 and 91 are essentially the same machine, though the 91 has 20-bit A-D converters, a digital output and more presets). TC Electronic also offer an upper-mid-price box, the M2000, that works exceptionally well and has the advantage of being relatively straightforward to program.

Some software-based reverb plug-ins are also pretty impressive; TC Works, Lexicon and Waves are probably the leaders here, both for VST-compatible plug-ins and their TDM counterparts for Pro Tools systems. When you take into account the processing overhead needed to provide good reverb, however, a hardware outboard unit still makes more sense if you have the means to patch it into your system. Alternatively, desktop studio users could opt for a soundcard that includes high-quality onboard chip-based reverb, such as those made by Yamaha (SW1000XG and DSP Factory) and Lexicon (Studio systems with Core 2 and MP100, or Core 32 and PC90).

Reverb is the only effect where accuracy is a real consideration, because it's trying to recreate a natural phenomenon. Most other effects, by contrast, have no direct natural counterpart: obviously you still want the things to be quiet and clean, but the subjective quality of the effects is the only thing that really matters after that. Most multi-effects boxes offer the usual delay and modulation treatments, but it's now fairly common for things like vocoders and ring modulators to be included too. More money usually buys you a better technical spec, more editability and more real-time MIDI control, but you can still get great results from the cheaper units, such as the Alesis Midiverb IV, Zoom 1204, ART Quadra/FX and Digitech S100 and S200 units. If you want to move up from there, check out the Alesis Q20, the Lexicon MPX1 and MPX500, the TC M*One, and the Lexicon PCM80 and 81. Both Roland and Yamaha also build some extremely good mid-price effects units, Roland mainly under their Boss imprint, with such units as the SE70 and 50 and the SX700, and Yamaha with their REV500, which is primarily a reverb, but includes some multi-effects like echo, tremolo, chorus, and flanging.

You can, of course, spend a lot more money on high-end studio effects, and it's true to say that they can sometimes do things that budget and mid-priced units can't. Specifically, pitch-shifting is likely to be smoother-sounding and more natural. Nevertheless, for the majority of tasks where you're looking to combine delay, reverb and modulation effects, the majority of mid-price units work perfectly well. If you need something a little more out of the ordinary, then look at a TC Fireworx or one of the Electrix boxes.

With the exception of reverb (which I still feel is too processor-intensive to trust to software plug-ins), most conventional effects can be handled perfectly well by software plug-ins, and in many cases the graphical user interface makes them easier to adjust. Delays use relatively little processing power with modulation effects being only a little hungrier, and often plug-ins offer features that hardware units don't have.

How important multi-effects are to you depends a lot on the type of music you're making. Dance music thrives on attention-grabbing effects, whereas more traditional musical styles use mainly reverb, delay and chorus. Most units can handle these basic effects adequately, but pitch-shifting will always be disappointing unless you spend serious money.

Monophonic intonation correction is a relatively new science, and the only hardware devices which I'm aware of in the field are those made by Antares and TC Electronic. I've owned an Antares ATR1 box, the hardware version of the ubiquitous Auto-Tune plug-in, since it first came out (there are also various software plug-in versions, including a VST-compatible one), and I can honestly say that it does exactly as claimed, and makes a huge difference if you're working with anyone who has slight vocal pitching problems. Obviously, there are singing problems that no amount of technology can fix, but the ATR1 and/or Auto-Tune can turn an 'almost there' vocal performance into one that's spot on pitch-wise. It's also easy to use and, when properly set up, artifact free. I've also tried the ATR1 on fretless bass, slide guitar, flute, electric guitar (monophonic only of course) and a few other instruments and found it to be both effective and (when correctly adjusted) free from side-effects. Considering that the Auto-Tune VST plug-in is under £100, it's more an essential than a luxury. I found TC Electronic's hardware equivalent, the Intonator, took slightly more setting up, but it also includes additional vocal processing, such as de-essing, and has digital I/O, which the ATR1 doesn't.

I pointed out last month that your choice of microphone is crucial to the quality of your finished work, because it is the first stage in the recording signal chain. If you get a bad result with a cheap mic at this point, it's unlikely you'll be able to rescue it later. If you do any of your recording, mixing, processing or mastering in the digital domain, then the same quality-control-related point applies equally to analogue-to-digital converters; no matter how much care you take with your recording up to the point where you convert to digital, if you then put everything through a set of budget A-D converters, you may seriously affect the quality of your work from that point on. You might think that A-D/D-A units are a rather dull way of spending money, but in my experience, the quality of the converters you use makes a big subjective difference, not only to the amount of background noise and distortion in your material, but also to the smoothness of the overall sound and the stereo imaging. If your recording medium is digital but you intend to record a lot of acoustic instruments via quality mic preamps, you'd be crazy not to invest in decent converters; likewise if you plan to do your own mastering in the digital domain from an analogue source.

Before getting into the subject of different makes of converter, what subjective difference is there between 16-bit and 24-bit converters, or a 48kHz sampling rate and 96kHz? I've attended a number of controlled listening tests, and the consensus seems to be that the difference between using 48kHz and 96kHz sampling rates is very, very small, and that unless you're using the best of professional audio chains with superb monitors in a good acoustic environment, you probably won't hear any difference at all. In my experience, even when you can hear the difference, you still can't say which sounds best — there's just a very small difference.

Enhancers     We hear a lot about enhancers that can make things sound bigger, brighter and better, but are they really any better than EQ? In reality, there's little they can do that a good EQ can't, though some models can synthesize high-end harmonics to brighten up an inherently dull sound. However, the 'good EQ' that could take their place might be very expensive and complicated to use, and an enhancer will often provide a cost-effective and simple way to add transparency and detail to a mix. As you probably know from my previous articles, one of my favourite toys is my SPL Vitalizer and, if you can afford it, go for the 1U gold-coloured Vitalizer Tube — I think that's the best model they make. The result is part mastering EQ, part high-end enhancer and, used sparingly, mixes can be made to sound bigger, more immediate and more open. The original Aphex Aural Exciter comes into its own when restoring the top end to a dull mix or track, and for those on a tight budget both the Aphex Exciter and the low-cost SPL Stereo Vitalizer can be picked up new for well under £200 each. In my book, one good enhancer is worth having, and used carefully it will make a difference.

So what about 16-bit converters versus 24-bit ones — do the extra eight bits make any difference? For classical music recording, where you need plenty of dynamic range and you also need a few dBs of headroom, 24-bit recording certainly improves the low-level signal resolution, but the only way to appreciate this fully is to deliberately under-record the signal by 50 or 60dB, then compare the 16- and 24-bit results. With material recorded at sensible levels, it's very difficult to tell which is which, even with acoustic music. And when you get to pop music, with its very restricted dynamic range, I'd venture to say there was no audible difference at all.

I think it's safe to say that 96kHz is pretty irrelevant for pop music production, at least for all but the very top level, however, there is a case for tracking, mixing and editing at 20- or 24-bit resolution. Digital mixing and processing can result in a lowering of the audio resolution, so if you start out with 16-bit signals, you'll probably lose two or three bits of resolution by the time you get to your master tape or CD. On the other hand, if you start out at 20 bits or more, you can do all your processing then dither down to 16 bits while still maintaining the maximum dynamic range and low-level resolution. Whether this is worth doing for pop music depends on whether you can afford to lose a third of your recording time and playback tracks: in most situations, the bit depth isn't the weakest link in the chain, so there'll be little or no subjective difference.

With that out of the way, what about the converters themselves? Some of the better soundcards do have good converters, such as the AKM ones Martin Walker has praised in several of his recent soundcard reviews (think Creamware Powersampler, Midiman Delta 1010, Terratec EWS88MT, Aardvark Direct Pro 2496, and SEKD Siena, although these aren't the only ones). Nevertheless, it's worth pointing out that you can now buy fairly high-quality converters in stand-alone boxes for a fraction of the price they would have cost just a few years ago. Various models are available within the typical SOS readers' budget, ranging from two-channel A-D/D-A affairs like Midiman's Flying Cow and RME's ADI-1, which are available at around the £200 mark, up to eight-channel units, which either come as physically separate A-D and D-A units like Swissonic's AD24 and DA24 (£499 and £399) or as one A-D/D-A unit like RME's AD8 Pro (at just over £1000). There are also oddities like the Fostex VC8 8-channel analogue-to-ADAT optical digital converter, which was designed to be used with Fostex's stand-alone digital recorders, but is a steal at £199, provided you have an optical connection to use it with.

Pragmatically, you'll notice a much greater difference in quality between a games soundcard and a respectable audio card costing a few hundred pounds than you will between the latter and an esoteric high-end converter costing upwards of a thousand pounds per channel.

I've touched on the subject of digital recording this month by mentioning A-D converters, and next month I'll revisit the topic. Computers and software have developed to the point where they now play a fundamental role in many studios, and such is the pace of progress that upgrade options in this field can be particularly confusing. They'll be the focus of next month's third and final part of this series.