Drum recording guide
From miking cymbals to mixing toms - all you need to know
Enough of this drum module nonsense. Get yourself a well-soundproofed room, a proper drummer and rediscover the secrets of recording acoustic drums. Mark Parsons is your guide
Do people ever record actual, real, live, played acoustic drums any more? Of course they do. Samplers may have replaced drummers in much of today's recorded music, but increasingly, in an effort to make their records stand out from the crowd, producers and artists are using sampled loops in conjunction with live playing, or combining acoustic sounds with triggered samples. And as for all your favourite drum and percussion sounds on sample CD... how do you think they got there in the first place? Yes, that's right, they were recorded using a real drummer and, er, real drums.
So knowing how to get the best sound from a drum kit is no less crucial today than it was five years ago. And, arguably, if you are recording drums with a view to creating samples, technique becomes even more crucial.
I won't pretend that the following pointers represent the only way to record drums. But if and when you find that your TR909/Akai S1000/Coldcut drum loops just aren't cutting it, here are some techniques and tips which are guaranteed to work - even if they may need a little tweaking to suit the style, mood, and groove of the session in question.
Fact: well-written, well-played drum parts record well. Bootleg tapes of great bands may have badly recorded drum sounds, but who cares? The most expensive studio in the world can't keep poorly conceived or executed tracks from sounding like demo tapes. Nothing makes a tape sound more 'demo-ish' than an overly busy drum part which detracts from the song itself. Too often, having the drummer simply copy the machine part results in a stiff, over-busy part. So when in doubt, have the drummer play a strong and simple groove.
Fact: good instruments record well. It's amazing how often people show up in the studio with badly tuned drums which have old heads, or which buzz and rattle due to loose hardware or other gremlins. Start with good drums, and you're already most of the way there.
Fact: the sound of the room you're recording in is critical. The same set of drums can sound entirely different in a different room, and, believe me, some rooms sound much, much better than others. Hard surfaces reflect sound, while soft surfaces tend to absorb and diffuse it - especially higher frequencies. Parallel walls (including floors and ceilings) can build up standing waves and produce so-called 'flutter' echoes. Each dimension of a room has a corresponding resonant frequency, and a room where two or more dimensions are the same (or multiples of each other) can cause serious problems. For example, a room measuring 16' x 16' with an 8' ceiling will have substantial 'room boom' due to multiple reinforcement of the resonant frequencies (about 70 and 140Hz in this case).
If you're absolutely stuck with a crappy room, however, all is not lost. Instead of trying to improve the sound of the room, you simply eliminate as much of it as you can from the mix. By using close-miking techniques to increase the ratio of direct to reflected sound, and foregoing ambient mics in favour of signal processing, you can actually get a very fine sound in any room. You may just have to work a little harder at it, that's all.
Once again, there is always more than one way. You'll probably have different mics from the ones suggested here. You're certainly recording in a different room (unless you've sneaked in while I wasn't looking). So don't take the following tactics as gospel. Just use them as the basis of your approach, and never forget the three unwritten rules of creative recording: experiment, experiment, and experiment.
Many people consider the snare the heart of the drum kit. If you're dealing with a drum with a battered head that looks like it hasn't been changed since Chuck Berry was a boy, don't even bother trying to tune it. Replace it with a Remo-coated Ambassador head. No, it won't be as tough or long-lasting as those multiple-ply 'rock'-type heads, but in the studio the basic white-coated Ambassador will give you better articulation, tone and bite; just listen to Kenny Aranoff's snare on any John Mellencamp record (if you can stay awake that long - Ed). While you're at it, you might also want to change the bottom (snare-side) head. It makes a bigger difference than many people realise.
To tension a drum head, you don't tighten the lugs in a circular fashion, but in a star pattern - similar to tightening the lug nuts on a wheel. The critical thing is that all the lugs are evenly tensioned. You do this by tapping the drum head (snare off) approximately one inch from each lug, listening to the pitch, and adjusting the tension until they all sound the same.
For a rock sound, tune the top head to a mid-range pitch, keeping the bottom head tuned high. I prefer the sound of a wide-open snare for most applications, but if you want to eliminate over-ring, don't use the internal muffler. Instead, cut a 1" wide doughnut out of an old head and lay it on top of the snare head. This will eliminate most of the ring without drastically altering the fundamental tone of the initial stick attack.
Place a Shure SM57 (or similar dynamic mic) above the rim of the snare, pointed down towards the head at a 45° angle with the end of the mic about 1" above the head. This placement allows you to take advantage of the proximity effect (bass boost at close distance) to add some beef to the snare; it also isolates the sound. Over at the mixing desk, try adding a few dB at 12kHz to keep things crispy, pulling out 3dB at 3kHz to get rid of the boxy element of the sound, and rolling off 3dB-5dB below 60Hz to keep the mud out.
Big reverb can sound great on snare, but that same reverb sounds rotten on hi-hats and cymbals, so we're going to gate the snare before we send it to the reverb. (I generally recommend saving the gating and reverbing until the mix, but for those who aren't going to put snare on its own track or are limited to one reverb or gate: do it now.)
Set the gate to its fastest attack time with the maximum attenuation. Set the threshold so that the gate just barely opens on the quietest snare hits of the song. If doing this still gives you too much bleed from the rest of the kit, here's a tip: buy a cheap piezo buzzer (about £1.79 at Tandy) and solder on a cable and plug. Tape the buzzer to the snare shell. Plug this into the trigger input on the gate, and the noise gate will open only when the snare itself is hit. If you have extra tracks available, you can record the spike from the piezo buzzer onto its own track, which can later be used for extra-accurate gate triggering on mixdown. Now adjust the 'time open' function so that the gate stays open long enough to capture the hit, but no longer (usually 300mS or less). It's this clean and dry snare-only sound that we're going to send to the reverb.
For a powerful rock-snare sound, try a fairly grainy room reverb with an RT60 of one to 1.5 seconds. Program 15 on the Alesis MIDIverb II is a good example of this sort of room.
If possible, return the reverb through an extra input channel or outboard EQ rather than just an aux return, so that you can EQ the reverb return. The secret to getting a big, fat snare without it becoming muddy is keeping the source sound (dry snare) clean and sharp, while EQing the heck out of the reverb return to get the 'boom' after the 'crack'. So try adding some top (5dB at 6kHz) to the return, pulling out the same amount at around 3kHz, and boosting at 120Hz by six or more dB.
This should give that killer 'bigroom' sound, yet retain all the attack, clarity and articulation of the original drum sound - a resounding 'smack!' to help give your recordings the drive they need.
The operative word for an effective kick drum is 'punch'. Let's define this subjective word. First, a strong low-end fundamental without a lot of extraneous sound below that fundamental to make it boomy; the sound should be tight and have a relatively brief duration so that each note is clearly defined. To this add a certain amount of attack (that brief high-frequency spike that accompanies any percussion instrument) to add articulation to the beat and allow it to cut through the mix without resorting to excessive gain. So how do we get this sound?
Again, let's start with the instrument itself. Here, too, new heads are in order. Some people like to take the front head off (this gives easy access for muffling materials and mic placement), but I like to keep the front head on with a 10" hole cut in it. This way you can still muffle the drum easily, but also get resonance and sound control from the ventilated front head. Get a piece of 2" foam wide enough to fit snugly between the front and back heads and long enough to cover the bottom third of the inside of the drum (about 14" x 22"). Roll it up, insert it through the hole, and press it into place. This will give you a fairly natural yet controlled sound.
When tuning a kick, work your way from a 'flap' sound (too loose) up to a 'bong' (too tight). Somewhere in between there's a nice solid 'thud.' That's the sound you want, and you'll find it a lot quicker if you tune with the muffling already installed.
Any sturdy dynamic mic (SM57 or similar) will work; my personal favourite for a good, punchy sound happens to be the AKG D112.
Location is important. The closer a mic is placed to the spot where the beater contacts the head, the more attack you'll get, and the further toward the front the more 'air' it'll pick up. A good starting place is a few inches from the front head, aimed at the beater contact spot.
To keep it from sounding boomy, we're going to roll off everything below about 50Hz, and to increase the attack definition we're going to add a few dB at 6kHz. For the final EQ tweak, it helps to have an EQ with a sweepable mid-range (semi-parametric) section, although a graphic equaliser will also work. Crank the gain of the mid EQ way up and start sweeping the spectrum, but instead of looking for an improvement, decide which frequency is the most annoying. When you find this boxy, obnoxious, unmusical sound (on kicks it's usually around 1.5kHz-2kHz), simply attenuate it, and what's left is the tight, musical, punchy sound we've been looking for.
The comments made in the snare section regarding heads and tuning also apply to toms, with a few variations. For the most sustain, tune the top and bottom heads to the same pitch. Some drummers tune the bottom head slightly lower than the top, giving a 'falling pitch' sound as the drum is struck. I prefer to avoid mufflers - the role of the tom within the kit is more melodic than the snare, so the notes should have a clearly-defined pitch and a longer duration. To fatten up the sound somewhat without destroying the tone, you may wish to use heavier heads or ones with sound-control features such as integral rings, black dots, or oil filling.
Miking a tom is similar to miking a snare. The mic should be over the rim of the drum and the diaphragm within a few inches of the top head. Sometimes you can get by with one mic for every two toms on a large kit with four or more rack toms, centring it between the drums above the point where the rims are closest together.
A great mic for toms is that allround 'Wondermic', the Shure SM57. Another popular choice is the Sennheiser 421; there are also times when you might want to use a condenser mic for the increased high-end response. An AKG D112 works very well on large toms - I have an 18" x 18" floor tom that sounds like The Voice Of God when tuned low and close-miked with a D112. Again, these are mics that have worked for me; there are many other great mics out there.
For increased articulation I like to add a few dB at 6kHz, and to help out with the 'well-rounded' aspect, try adding a little bit at 120Hz - this'll add some 'whoom' after the attack and give a feeling of solid sustain. To increase musicality and intelligibility while reducing gain, try the above-mentioned technique: fully boost the parametric EQ and home in on the annoying frequencies. This usually results in a cut somewhere around 3kHz and an overall smoothing of the sound.
Before we get into recording cymbals, is there anything we can do to improve their sound? There certainly is: we can clean them. The thin layer of dirt, grime and tarnish that collects on cymbals over time can dampen the vibrations of the instrument and darken the tone. Cleaning cymbals can be a time-consuming chore, so here's a tip: go to your local music shop and get a bottle of cymbal cleaner. Then head for the garden with a bucket, a hose and some rags. In no time, cymbals will look and sound like new - and no drummer will resent you for it.
How about stick selection? 2B or not 2B? Sorry about that, but different sticks do sound different. A 2B stick across the bell of a ride cymbal will give much more volume than a smaller 7A, for example; but more important to the tone is the material used on the stick's tip. A wooden tip gives a slightly mellower 'jazz' feel, while a nylon tip produces a brighter 'ping'. Selection depends on the type of material being recorded and personal preference.
Recording cymbals (including hi-hats) is a job tailor-made for condenser mics, due to their open, 'airy' reproduction of high-frequency sounds. Mics I've tried as overheads above a drum kit include AKG C414s, Neumann KM84s and U87s, Calrec CB20-Cs, and a variety of Audio Technica condensers. The good news is that they all sound fine (even the less expensive ones), so you don't have to bankrupt yourself to get a decent sound.
Most major mic manufacturers offer a reasonably priced 'studio-quality' small-diaphragm condenser, and a couple of these can do wonders in opening up the sound of acoustic drums.
As far as placement of overheads goes, there are a couple of ways to proceed. If you're only going to use one mic, it's pretty simple - place the mic on a boom several feet over the drum kit, pointing down. Using two mics, the best way is probably the coincident-pair method (also known as XY), which uses two directional (cardioid) mics arranged so that their diaphragms are almost touching and they coincide at a 135° angle over the kit, with the signal from the mics panned hard right and hard left in the mix. This setup is simple, gives a true stereo image, has excellent mono compatibility, and avoids phase problems because the two diaphragms occupy virtually the same position. For hi-hats, place the mic so it looks down on the hats at an angle, about 6"-12" from the top cymbal.
Most of us don't have access to unlimited quantities of world-class vintage valve microphones, but we can optimise our recordings by making the most of what we have. If you only have one decent condenser mic, forget about the hi-hats and place it over the kit - it'll do a lot more good there, and the hats will get picked up just fine along with the cymbals. If you have two condensers, you may still be better off using them as stereo overheads rather than putting one above the kit and one on the hi-hats. If your total mic complement is limited to a couple of dynamics and one or two condensers, forget about close-miking the toms. Mic the kick and snare with the dynamics, and run the condenser(s) overhead.
You compensate for the way you're using the condensers by the amount of low end you roll off the signal. For example, if mics are being used strictly for picking up cymbals (overhead and/or hats) you can pull out quite a bit of bottom. This will greatly attenuate the tom levels in the signal without doing anything to the cymbals. However, if the overheads are responsible for picking up the body of the kit, you'll probably want to run them flat or even give them a boost in the bottom.
Finally, if you've got the mics and inputs available and you're recording in a good-sounding room, you should run a couple of ambient mics. It's difficult to give hard and fast rules (due to the individuality of every room), but here are a few guidelines.
Walk around the room while someone is playing the kit and listen for spots where things sound good. These locations may not necessarily be at ear-level - pro engineers sometimes place room mics high up in the corners. Remember that our ears are somewhat omnidirectional; what may make certain spots in the room sound good is the combination of direct and reflected sound, and it may take an omni or bi-directional mic to pick up the same effect.
Keep in mind that for every foot of space between the microphone and the sound source you're going to get approximately one millisecond of delay. This discrete delay won't be apparent to you when you're standing across the room, but the delay between the close- and distantly-miked signals will be noticeable if the distances are great enough, and the gain on the room mic is high enough. This can be a positive effect if used creatively.
Ambient miking is one area where experimentation can really pay off. I once placed a PZM on a large piece of plate glass facing a drum kit from about ten feet away. The signal from this mic, when solo'd at the desk, was awful - metallic, harsh, and bouncy. But when I added a small amount of this track to the mix, it added just the right amount of edge to the tune.
Whether you're doing a percussion submix or a final mixdown, it's helpful to proceed in a logical order. First, zero the desk. Bring up the kick, panned dead centre. It should punch. If it sounds boxy, pull out some mids.
If it sounds poofy, it could use more high end to emphasise the attack. If it booms, you need to roll off some bottom. If possible, listen to it on another pair of speakers. The sound will be different, but it shouldn't change character too much on differently-sized monitors.
Next, listen to the consistency of the playing. Except for obvious accents, each kick-drum hit should be at the same level. To tighten this up - and give the impression of power without over-saturating the tape - run the signal through a compressor. Don't squash it, just keep the peaks to a manageable level.
Bring up the bass guitar (or bass synth, or whatever) track to see how it sits with the kick. Do they complement each other, or do they compete for the same spot in the audio spectrum? Considered as a single instrument, the kick should provide the deep, powerful (but brief) attack, while the bass should contribute the sustain to the note.
Time for the snare. EQ as previously recommended, if you haven't already done so, and set up a gate (again, fast attack, maximum attenuation, 300mS decay, threshold just below quietest snare hit) before sending the signal to the reverb. Process the reverb return separately from the dry-snare signal to keep the 'big-room' sound from muddying up the initial snare hit. If the snare loses some of its clean, crisp articulation, you can give the reverb some pre-delay to separate the attack from the reverb tail. (Don't overdo this, or it'll sound as though you recorded in a gym.)
Remember that reverb is additive - a little bit on every instrument adds up to a lot. You want to be selective about what you put it on and how much. Sparse and/or slower recordings can generally accommodate more reverb than faster, busier ones (as well as more pre-delay and longer reverb times).
The snare should be panned to the centre and, as far as gain goes, it should really kick posterior. Nothing makes a tape sound more like an amateurish demo than hearing the cymbals, toms, kick and snare all at the same level (and usually with so much mid-range that they have to be pulled way back in the mix so they don't bury the vocal). Don't be shy about the snare (and to a lesser extent, the rest of the kit). If the drums are recorded to sound musical and EQ'd so they don't compete with the vocals and guitars, you can really crank them in the mix and they'll still be relatively transparent.
Next, bring up the toms. I like them panned from left to right, from the drummer's point of view. When EQ'd properly, you can hear the stick attack and a nice, round sustain following it. In the mix they should have the same relative volume to the snare that they have in the real world - a strong presence but not as loud. The reverb decision regarding toms depends on whether or not room mics were used during tracking and how live the room is. In a close-miking/dead-room situation, a little reverb is appropriate, but you can't get away with as much as on the snare. If ambient mics were used, wait until they're in the mix to decide.
The overheads are pretty easy - they almost mix themselves. If you used an XY pair, pan them hard left and hard right. If you haven't done so, EQ them according to their intended purpose as explained in the section on overheads. If you sit in a relatively small room with someone playing live drums, you'll notice that the cymbals are a very loud part of the kit. They shouldn't be as prominent in the mix as they are live because they have a lot of cutting power and will be clearly present at maybe half their live level.
If there is a separate hi-hat signal, bring it up and roll off the bottom, or you'll hear a 'thump' every time the drummer steps on the pedal. My preference is to pan them slightly to one side and place them somewhat below their relative live volume, keeping in mind that the overheads will also contribute to their level.
If there are ambient mic tracks, bring them up. They may sound horrible by themselves, but no one's going to hear them solo. Add them to the mix a little at a time, and stop when the mix has the right amount of 'liveness'. You may want to try different EQ, the most common being a reduction in the 2kHz-4kHz region to eliminate some harshness.
Now your drum tracks are really happening. Bring up the rest of the mix. We're looking for a drum track that sounds musical and 'vocal-transparent'. This means that even when you crank it up, you can hear the vocal because you've left the latter plenty of room within the audio spectrum.
When you get to this point, reach over with your right arm and clap yourself soundly on the left shoulder. Congratulations! You've succeeded in capturing probably the most difficult instrument to record well - and the one that can do most to bring a recording to life.
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!