Must modern metal drums rely on triggered samples? Our engineer shows how you can mic up a kit for a modern, heavy sound.
I was recently asked to record some drum tracks for an as-yet unnamed three-piece progressive metal band. I already knew their drummer, Peter Zimre, from a previous indie-rock session I’d done (the subject of SOS October 2012’s inaugural ‘Session Notes’ column, in fact), and he was able to secure the use of the same large converted loft we’d used there for tracking purposes. The recording approach, however, clearly needed to be very different in this case. Whereas indie productions can often be based around the natural sound of a four-piece band playing in the same room, metal genres take few cues from acoustic reality, laying much greater emphasis on delivering both maximum aggression and superhuman precision. With this in mind, we opted to build up the band sound via a series of overdubs, eliminating spill between the instruments and thereby allowing us to patch and edit every part freely, as required. For bands on a budget, it’s nigh on impossible to compete with the accuracy of commercial metal productions in any other way.
Mind you, I was also pretty sure we’d get more inspiring drum takes if we could maintain as much of a live performance atmosphere as possible during the tracking session, so there was clearly a compromise to be struck. In the event, I chose to have the bass and guitar players performing with the drummer in the main room. The guitarist’s amp, however, was placed in a separate room to prevent its sound from bleeding into the drum mics, and all three players were given headphone monitoring so they could still hear what he was playing. I kept the bass cab in the drum room, though, because it’s usually pretty straightforward to tackle bass spill by high-pass filtering and/or gating drum mics — at least to the extent that you can freely overdub a new bass part without audible conflicts. Besides, the physical sensation you get when you’re close to a bass cab is a crucial element of the experience of band performance, in my view, so I always like to keep it in with the players if at all possible. My impromptu ‘control room’ was set up in an adjoining bedroom, where I’d rigged my own multitrack recording system and a pair of monitor speakers.
I’m often asked for tips about recording drums, and I usually advise musicians on a budget to set up their overheads so that they capture roughly the balance the drummer’s hearing. The more of the final sound you can get out of the overheads, the more sparing you can be with your use of spot mics, which brings many benefits. For a start, you typically need less gear, because it’s often possible to get away with close-miking only a few of the more important kit components, and that accelerates the setup process as well. Fewer mics overall also means fewer problems with phase-cancellation between them, which simplifies mixdown, and in general the emphasis on the more distant overhead mics usually gives a more believable result, because close mics inherently tend to sound less representative.
The downside of the ‘mostly overheads’ tactic is that you catch quite a bit of room ambience on those main mics (they have to be at least a couple of metres from the kit to get a good overall pickup), so it’s poorly suited to delivering the super-distinct and upfront drum hits prized by metal genres. Moreover, building your drum sound around a main stereo pair provides a lot less scope for detailed tone and balance control of individual drums at mixdown, something that can be extremely desirable when working with such fast and intricate musical material. So I resolved instead to build the entire drum sound using close mics instead, without recourse to any main stereo mic pair.
This approach is already more labour-intensive by nature, but was further complicated by two factors. Firstly, Peter’s drumkit was a bit of a beast, comprising two kick drums, snare drum, two hi-hats, five toms, and about a dozen cymbals; and, secondly, the limitations of my location-recording rig meant that I could assign a maximum of 16 mic preamps and 12 recorder inputs to the drums. Straightaway that put paid to any idea of miking all the drums and cymbals individually, as well as impelling me to submix at least some of the mics into the recorder.
Planning around these kinds of restrictions is a common task for project-studio recordists, so let me explain my thought process. When there’s a limited number of mic channels available it’s sensible to start by allocating the mics you feel are indispensible. In this instance, the ‘bare essentials’ for me were individual mics for both kicks, the snare, and the toms, as well as three further mics to cover the mass of cymbals in small groups. With 11 channels spoken for, it left capacity for five slightly more discretionary mics.
Many metal productions make extensive use of sample-triggering, which takes the pressure off your recording skills (the close mics don’t need to sound that good as long as they provide reliable trigger signals) and gives you maximum control over the drum sounds at mixdown. However, both Peter and I were distinctly unenthusiastic about taking this approach here: Peter was happy with the sound of his kit the way it was, and was keen to retain the organic quality of his live performance as much as possible, despite our sonic concessions to the musical style; and I didn’t relish trying to achieve natural-sounding triggering for Peter’s many nuanced rolls, flams, and ghost notes!
It was to avoid triggering samples that I decided to allocate a second mic to the snare drum. I knew I’d have to move in very close to the snare to keep spill and room ambience tightly under control, and I’ve never found any close snare mic that sounds very good on its own. By combining two mics in different positions, I hoped to capture more of the character of the drum, rather than just the ‘donk’ of the stick hitting batter head. I didn’t feel the need to do the same with the other drums, though. The kick drum in metal is such an unnatural sound, so I wasn’t too concerned about multi-miking that, and I was happy to mic the toms from slightly further away to reduce their need for additional mics, calculating that a bit of extra spill on those tracks would be fairly easy to tackle with mix processing given that those drums are usually hit relatively infrequently. Instead, I decided to spot-mike each of the two hi-hats in case we needed a drier, more rhythmic sound for those, even though I figured they’d likely be plenty loud enough in the cymbal mics most of the time.
My final addition was a pair of room mics. Partly, this was so that I could give an increased sense of space around the kit for certain arrangement sections, but that wasn’t the whole story. I also wanted the option to add some extra richness to the snare and toms at mixdown, if required, again to avoid having to rely on sample-triggering. You see, in pursuing such up-front drum sounds, I knew we’d inevitably sacrifice some of the kit’s essential timbre — close mics simply can’t pick up the full frequency dispersion of instruments they’re only a few inches away from. Room mics, by contrast, capture a much denser, more well-rounded kit sound, and can therefore make a real sonic improvement to a mix that’s otherwise reliant on extreme close miking, even if you only mix them in a touch. (Incidentally, artificial reverb is a poor substitute in this scenario: because it doesn’t add any new frequency content, it won’t give you any more drum tone, but will just make the kit sound further away!)
To fit all these mics through my 12 recorder inputs, I planned to submix the three overheads and five tom mics to stereo pairs, so that all the other mics could have their own dedicated inputs. Most of the time you don’t get more than one cymbal or tom being hit at once, and I reckoned I wouldn’t lose too much mixdown flexibility by submixing in that way, because I’d still be able to significantly rebalance hits and fills with moment-to-moment fader automation. What I wouldn’t retain, though, was independent control over the tone and sustain of the individual instruments in those submixes, so I was aware that we’d have to take special care over those during the tracking session.
So much for the planning: how did it pan out in practice? Well, once on the session, I started work (as I normally do) with the mics which were likely to have the most spill on them, which in this case were the overheads. These were all Shure KSM-series small-diaphragm condenser mics, set to their cardioid pickup patterns for a drier sound. If I’d used a more traditional overhead pair, it would have been tricky to get a representative balance of all those cymbals from close up, because the ones closest to the mics would have been favoured in the mix. By adding a third mic I was able to bring the overheads closer, while mitigating this balance problem.
Which cymbals to group with which mic was an issue of stereo positioning as much as anything. I’d already decided to pan the left and right overheads to the extremes to give a fairly wide kit image overall (as per metal convention) and have the third mic panned down the middle as a ‘centre fill’ to avoid problems with mono-compatibility. Following discussions with Peter, we decided to slightly rotate the overheads ‘triangle’ towards the drummer’s right so that his primary crash cymbals would occupy complementary positions on each side of the stereo panorama. This meant that one china cymbal on the drummer’s far left became recessed in the balance, but this didn’t strike either of us as a huge loss, because that cymbal was used rarely. (Besides, I find that chinas usually come across a little too loud through overhead mics anyway!)
Beyond these considerations of ambience pickup and stereo picture, it was the balance between the different cymbals in the mix that dictated the exact positioning and level of each overhead mic. In other words, the goal was to try to get the cymbals feeling fairly evenly matched across the kit. For the most part this wasn’t too tricky, because Peter, like all good drummers, instinctively balanced the instruments by virtue of the way he played. In one case, though, we did swap out a cymbal for a different model where it seemed to zing out too prominently — I’d have had to mike up every cymbal individually to do anything about that in the control room. A quick polarity check confirmed that the sound was best without any inversions, and then I dialled in the built-in high-pass filters of my Audient ASP008 preamps to cut away unwanted low-end ambience from the kick and toms. Incidentally, one concern that frequently arises when pointing a cardioid condenser directly at cymbals from close up is that the mic’s inherent on-axis high-frequency boost overbrightens the tone. In metal, however, the mix tonality of commercial productions is typically so mid-scooped and sizzly that this effect actually played into our hands!
Next stop for me was the snare, because I wanted to experiment with different mics and mic positions. Over the top we tried a Shure SM57 and a Samson C02 small-diaphragm condenser, and the Shure won out with its meatier-sounding attack. The physical placement of the toms and cymbals restricted the mic-stand access a good deal here, and the most practical miking angle (approaching the snare from between the two top toms) overemphasised a 460Hz pitched resonance from the instrument. Sometimes I just use a bit of gaffer on the drum head to tackle resonances like this, but you always lose a little high end into the bargain, so in this instance I decided to deal with the resonance at mixdown instead, given how simple I knew it would be to dial it out with a narrow EQ notch.
Under the drum, another SM57 lost out to one of the cymbal condensers from a cheap-as-chips Superlux drum-miking kit, the latter doing a better job of capturing the high frequencies of the snare wires. Miking from the side of the drum shell often does nice things for me as well, so I experimented with that for a while too, but whatever I got from that mic seemed to make the drum feel bloated somehow, and in the end I ditched it completely. It can sometimes seem a bit galling throwing away something you’ve been working on for 10 minutes, but I console myself with the thought that eliminating rubbish-sounding mics is also a vital part of a recording engineer’s job description!
Kick patterns are blisteringly fast in modern metal, so there’s really no way you can allow the kick to have any kind of sustain tail — otherwise, the whole low end of the mix quickly turns to sludge. With this in mind, I agreed with Peter that we should remove the resonant heads from both his kicks, which is a very effective way of deadening the sound. An AKG D112 was then positioned inside each one, pointing right at the beater contact point, in order to achieve that high mid-range slap that’s so important for cutting through metal’s obligatory phalanx of distorted guitars. Some discussion amongst the band about how much of this beater noise we needed led us to reposition the mics a couple of times. (I suspect, though, that all our ears were already beginning to tire by that point in the evening, because the final kick signal turned out to be a good deal more aggressive at the high end than we eventually needed in the mix!) When we’d committed to our mic positions, we draped a couple of duvets over the front of each drum to take the edge off any cymbal spill — it’s so easy to do, and really does make a difference.
The toms were miked using five Avantone CK1 cardioid small-diaphragm condensers. I didn’t process these mic signals at all, but I did expend considerable effort on their pan, fader, and polarity settings in search of a suitable stereo picture, a reasonably even level balance, and a sense of tonal commonality across the five drums. Remember, we were submixing all these mics to stereo for recording, so there’d be no way to remix them after the fact. Fortunately, Peter’s a bit of a tuning demon, so he’d headed off many potential problems before the mics even got a look-in, but we did re-tweak his tunings a little to bring the sustain characteristics of the centre tom more into line with the others, and small strips of gaffer tape helped rein in overprominent pitched overtones here and there.
People are often overzealous about minimising cymbal spill when placing tom mics, even in styles like this that require an up-front sound. If your tom mics are an inch away from the drum heads, you’ll likely get just a nasty combination of slap and boom, and it’s only once you back off two or three inches more that you start to hear something resembling a tom! A bit too much cymbal spill is a hell of lot easier to deal with (using gating, say) than a horrible tom sound, so I’d always rather err on the side of too much spill here. That said, I prefer not to make a rod for my own back unnecessarily, and in this instance I chose to switch the capsules of the top three CK1s to give them more directional hypercardioid polar patterns, because those mics seemed to be suffering most from spill.
As expected, the hi-hats were coming through the overhead mics pretty nicely, but without much in the way of body. In my view, a lot of the rhythmic emphasis of a hi-hat isn’t about the very high frequencies, but more about the mid-range, so I decided to use dynamic close mics for both of the hi-hats here — a couple of tom mics from the same Superlux drum-miking set, as it happened. An additional benefit of the dynamic mics was that their slower response to transients prevented the stick noise from getting too spindly, which can be a danger when you’re miking any cymbal close up.
By this point, we had a respectable drum sound coming out of our 14 close mics, so it was time to give Peter a well-deserved break while we set up the other musicians. I put the bass cab in front of the kit, pointing off to the side to weaken high-frequency spill into the drum mics, and miked it up with a Sennheiser MD421, a mic which not only has decent low-end extension, but also a hefty mid-range boost to help the bass compete with the guitars. For the live guitar we tried out a couple of different cabs in the main live room, and then took our favourite (an Orange 2x12) to its own room, miking the clearest-sounding speaker cone with a Superlux R102 ribbon and a Shure SM57 dynamic so we’d have a few tonal options in the control room — handy when you’re moving quickly between different songs, as we were scheduled to do.
The final step before we could start takes was to sort out the room mics, another two Superlux R102 ribbons set up as a spaced pair. I’d deliberately held off mixing those in until after the bass amp was set up, because I wanted to aim the rejection nulls of these mics to reduce bass spill. The rejection characteristics of ribbon mics never cease to amaze me: although the bass cab was twice as close to the mics as the drums were, it might as well have been in a different room for all the spill level it generated!
Of all the instruments frequently recorded in project studios, the drum kit arguably presents the greatest engineering challenges, and it’s tempting to lay the blame for uncompetitive end results on the lack of an acoustically-designed recording space or the shortcomings of affordable gear. In this month’s featured session, however, we recorded in an untreated domestic room using mics averaging $200 apiece, but still managed to get a commercially viable sound without recourse to sample-triggering or heavy-duty mixdown processing. So if you’re dissatisfied with your own drum recordings, don’t immediately reach for your wallet, because spending a little more time planning and setting up your tracking session may provide all the improvement you need.
The practical reality in project studios is that you’re seldom able to mic a drum kit as comprehensively as you might like (as in this month’s featured session), and this forces you to make some irrevocable sonic decisions while tracking. While I’m all in favour of such early commitment in principle, because it’s usually the quickest route to decent end results, I do still like to hedge my bets slightly by recording a couple of hits of each drum and cymbal through the whole multi-mic rig at the end of every drum session I do. By triggering these samples at mixdown, I can deal with exceptional balance problems that level automation can’t remedy. Unlike traditional mixdown sample-triggering, however, these samples aren’t meant to change the essential sonic character of the recording, and tend to be used very sporadically — ideally not at all, in fact!
For example, let’s say Peter hadn’t hit the second rack tom hard enough on a certain fill. My first port of call would be to boost the level of the toms submix briefly to emphasise that specific drum hit, as explained in the main article. However, if there happened to be exceptionally strong cymbal spill on the toms submix just at that moment, it might prevent me boosting that tom’s level as much as I’d like without audible pumping of the cymbal levels. This is where carefully time-aligning the sampled hit underneath the live part and mixing it in would be able to bail me out, lifting the individual tom’s level without boosting the cymbal spill. Clearly, it isn’t a stunt you want to pull that often (if only because it’s rather time-consuming), but as a kind of ‘get out of jail free card’, it does make the prospect of committing to submixes during the tracking process a lot easier on the nerves!
I’ve placed a number of audio examples on the SOS web site so you can hear for yourself the sounds that were captured on this session.