With a band citing a particularly broad range of influences, our engineer came up with a plan for capturing a bit of them all. Find out how he fared...
Like any starry-eyed muso, I like nothing more than bands mashing up different stylistic influences, but recording them doesn't half test your engineering mettle! The main difficulty, I find, is in trying to meet the traditional sonic requirements of more than one genre simultaneously. So in this month's column I'm going to look at a recent recording project that presented just such a challenge, and explain how I tackled the trade-offs in practice.
Impossible Colours
The band in question was four-piece psychedelic rock group Impossible Colours (www.impossiblecolours.com), comprising David (guitar, vocals), Chacha (bass, sax), Miko (drums), and Romeo (keyboards). I'd agreed to record a three-track EP for them, and their cited references for the session included a range of classic rock tracks from Led Zeppelin, Jethro Tull and The Doors; more modern indie/alternative numbers from Tool, Panzerballet and Weezer; vintage funk from Larry Ellis & The Black Hammer; and jazz fusion from Joshua Redman! To make matters trickier, the detailed interplay of the performers was an important part of the band's character, so I was loath to rely much on dynamics processing to shape the timbre.
This month's Session Notes article features the psychedelic rock band Impossible Colours.
As a consequence, I decided to hedge my bets by using directional close-mics to achieve a very focused and up-front sound, akin to Joshua Redman's, my reasoning being that it would be easier to add room sound to a drier capture than to remove unwanted printed-in ambience. I figured that this approach would also leave me more freedom to layer in additional samples at mixdown if necessary. A dry sound turned out to be the most pragmatic choice too, because the project studio available for the sessions featured only a small carpeted live room (measuring roughly 4x5m) that didn't boast the kind of open-sounding acoustic typically required for normal rock-style mic techniques.
Part and parcel of this approach was minimising room ambience and spill from other instruments on the drum mics, but I didn't want this to prevent the band from performing as a unit. For example, I was convinced that I should seat Miko and Chacha together so that they could hear each other directly, rather than separating them and relying on headphone monitoring. Yes, that would mean some spill between them, but I felt that this would be a small price to pay for a more closely locked rhythm performance. (Besides, in practice, bass guitar usually spills mostly onto the low frequencies of a drum kit's overhead/room mics, which I usually high-pass filter at the mix anyway, to keep the kick-drum sound clear.)
Where To Put The Guitarist?
Romeo wasn't available to put in keyboards for the session, so we didn't have to factor him into the setup, but where to put David was more of a quandary. Fortunately, the band were happy to leave vocals to the overdubbing stage, so there was no need to separate David from the others physically, although I did initially consider bringing him into the control room with his amp so that I'd have the option to record a 'keeper' guitar take alongside the bass and drums. The main disadvantage was that I'd have had to monitor on headphones out in the corridor, and, even leaving aside concerns of listening accuracy, I knew that would impede my access to the live room and my communication with the performers, slowing down workflow considerably. If I'd recorded the band in that studio before, perhaps I might have risked these additional technical hurdles, but in the event we all decided that nailing the drum and bass tracks was the overriding concern for the day, so the control room was kept for engineering and monitoring purposes.
Without any additional isolation room in which to mike up a guitar amp, we had to find another way for Miko and Chacha to hear David's playing — something that everyone agreed was important from a performance perspective. My suggestion was to plug the output of David's effects pedalboard into my battered old Behringer V-Amp 2, feeding its digital amp simulation to all the performers via the studio's headphone foldback system. None of us were thrilled at the thought of creating any final guitar tracks this way, so we were effectively committing ourselves to overdubbing all the guitar parts after the fact, but overall this seemed like the best compromise available to us. (I did take the precaution of recording the V-Amp output, though, and we did eventually use it for a couple of improvised sections where it was impossible to match the live take's inspiration during later overdubbing.)
Here's a picture of the inside of the 'drum tent' we used to eliminate the room sound from our recording, taken from the drummer's seating position. A big lighting stand was used to provide a ceiling-level support for four duvets, which were then draped over two freestanding rockwool-based gobos. As you can see, the whole edifice was held together primarily with clip-straps and bungee cords, using carabiners fixed to the duvet edges. One of the sand-filled petrol cans I use as weights is also visible hanging behind the stand, counterbalancing the fully extended boom arms.
Drums In A Padded Cell
Given the restricted live-room space, the setup process had to be tackled in stages. First, I asked Miko to set up his drums towards the back of the room, facing the control-room window, but with a couple of feet of clearance from the wall. While he did this, I busied myself constructing something of a padded cell around him to shut out the room sound. (It's not a bad idea to keep drummers in padded enclosures as a matter of course, in fact...) There were no fixing points on the ceiling, so I extended the arms of a portable lighting stand over the kit and laid duvets over those to create a roof. Kevin Fellows (who featured in this column with his band Dunning Kruger, back in SOS October 2012) was helping with the production, and helped to position a couple of large free-standing Rockwool-based acoustic panels to create walls on either side of the kit. The duvets draped down to fill the gaps between the panels — except at the front, where some space was left for sight lines between the band members.
Although this sounds fairly uncomplicated, the practicalities were a little more involved. For a start, getting duvets to stay put is often a challenge. Although gaffer tape is the traditional tool for immobilising them, I appear to have been cursed by the Gods Of Adhesive, such that some corner always comes loose mid-take, even if I use rolls of the stuff! That's why, a few years ago, I bought myself a bucketload of cheap spring-loaded carabiners and fixed those securely along the edges of my 'recording duvets' with wire, so they're now a lot quicker and more secure to rig up.
This picture shows the final miking setup for the drum kit, including both overhead mic pairs.
That said, you can't always find somewhere to clip a carabiner, so I keep a selection of bungee cords and spare clip-ended straps on hand to increase my options, and these proved a godsend here, because the weight of the duvets kept threatening to topple the Rockwool panels onto the drummer! Connecting a chain of straps between the frame panels and some heavy equipment racks behind them made the whole edifice much more stable.
Although I'd already piled around 25kg of extra ballast onto the base of the lighting stand to prevent it from overbalancing, the weight of the duvets was also causing the fully extended boom arms to sag noticeably, reducing the clearance between the booth's roof and the drummer. Tackling this required a further 8kg counterweight hung from the opposite end of the central boom arm using some 'S' hooks and lengths of chain — not exactly pretty, but it did the job!
Twice As Nice
Listening to Miko drumming in this makeshift tent, I was pleasantly surprised at how well we'd managed to eliminate the room sound from proceedings, but I guessed that this was going to cause me some problems with the cymbal sound. In most venues that are well-suited to recording drums, the early reflections from the room effectively bounce a blend of all the dispersed cymbal frequencies into your overhead mics, which does a great deal to enrich the sound. No such assistance would be forthcoming from our DIY booth. This is one reason why slightly quirky and/or anaemic cymbals are actually quite commonplace on records with tight, funky drum sounds, but such records typically also have enough space in the arrangement for these sounds to poke through easily — indeed, the band's Joshua Redman reference track ('Put It In Your Pocket') was a good example of this. Our cymbals, on the other hand, had to be able to hold their own against heavy guitars in the rockier sections of the music too.
My approach to squaring this circle was to put up two pairs of overheads. My idea was to try to capture a more holistic picture of the cymbals by miking the kit from two different angles at once, and thereby compensate for the lack of early-reflections reinforcement. The first mic pair were Shure KSM141 small-diaphragm condensers operating in cardioid mode, positioned above the drummer's shoulder and as close as possible to the kit. The main placement restriction here was that I didn't want the microphones to be hit, so I asked Miko to help me set them up just out of harm's way. While I was at it, I got him to hold the end of an XLR cable in the centre of the snare head while I used a length of the cable to match the distances of each mic from the drum. This is something I now do on most of my sessions, to lock the snare into the centre of the image in stereo, as well as to minimise phase-cancellation in mono.
I stuck the second pair of overheads out in front of the kit, pointing back towards the drummer. Because spaced overhead pairs usually end up with a bit of a 'hole in the middle' of their stereo picture, I decided to adopt a coincident configuration for the second pair, and then used the hypercardioid capsules of my Avantone CK1 condensers to minimise pickup from the room. (Remember, we hadn't boothed off the front of the kit.) You have to be a bit careful with crossed hypercardioids because they give a wider image than you'd expect of cardioids for a given mutual angle — splay them any more than a right angle and everything quickly starts squishing up against the left and right extremes. In this case, I was already quite close to the kit, so even though I went with a mutual angle of around 70 degrees, the cymbal image was still more than wide enough for me, with the crash cymbals well panned, but not right out at the extremes.
As a by-product, the toms and snare came across as quite narrow, because they were significantly further from the mics, but this didn't bother me because it meant that the crossed pair would serve the 'centre-fill' role that I'd hoped it might. In addition, the cymbals were louder in the crossed-pair signal (the mics were much closer to them, after all), so when I mixed the two sets of overheads together, the wider snare and toms of the spaced pair dominated. The result was a surprisingly creditable dry balance of the kit which nonetheless seemed to offer enough cymbal complexity to carry through a rock arrangement.
Getting the tight kick-drum sound involved removing the resonant head and padding the inside head with some old curtains. The mic was then placed inside the drum with the fabric interrupting its line of sight to the beater, to take the edge off the clicky high frequencies. Note also the gaffer tape on the unused lugs, stopping them from rattling.
Not that the sound was by any means perfect: all the drums needed to be a bit more clearly defined, for a start; the kick drum lacked weight (as it invariably does when heard through the overheads alone); and the hi-hat presented a rather indistinct image in both stereo pairs. Fortunately, the remedy for all those things was next on the agenda: time to put up some close mics!
Kick & Snare
A Shure SM7B dynamic mic was used for the snare drum, with both its frequency selector switches set flat. An offcut of acoustic foam was used to reduce spill from the hi-hat. Notice also the damping on the top head, comprising small wads of tissue paper stuck on with gaffer tape.
The kick drum was my first port of call, and listening to it in the room I felt that it was probably too resonant to achieve the focused sound we were seeking. Miko was happy to experiment with removing the drum's resonant head completely, and that certainly improved matters. (We gaffered the loose lugs to stop them rattling.) An old drape inside the drum refined things further, and once I'd adjusted the lie of the material against the batter head while Miko was (cautiously!) playing, I felt ready to pop an ADK S7 large-diaphragm condenser mic about a foot away from the head, with its 18dB pad switch engaged, and nip back into the control room for a listen.
The sound certainly had more than enough low-end 'woof', but otherwise felt rather soft and diffuse, no matter where we moved the mic. (Kevin did some repositioning tests while I listened over the speakers.) Swapping the S7 out in favour of a Blue Kickball immediately set us more onto the right track, though, and I was able to find a decent spot for that mic pretty easily. As a final tweak, I rearranged the drape to obstruct the 'line of sight' between the mic and the kick pedal, because the Kickball's characteristic 6kHz peak seemed to be making the beater click a touch too aggressive — a misjudgement on my part, as it turned out, which meant that I had to reinstate some bite at mixdown, in order to keep the kick pattern clear.
The snare was also initially unpromising — too spiky and resonant, where we wanted something tight and punchy. Reaching a more promising sound was another collaborative effort: Miko and Kevin worked together on the tuning; I swapped out my initial mic choice (AKG's C414B XLS) for a Shure SM7B, which rounded the transient, added a bit of beef, and reduced spill from the hat; and we all experimented with different tactics for damping the drum's heads. The final arrangement involved little pads of tissue paper gaffered to the head in half a dozen places. As an engineer, it's easy to get into the habit of trying to solve every problem by yourself, but it's often so much easier to home in on a good sound if you get the musicians to help you!
Hi-hat & Ride Cymbal
The full studio setup for the Impossible Colours tracking session is shown in this diagram. The drums and bass were miked in the room, a DI feed was taken from the bass guitar as well, and a guide guitar part was recorded through a digital amp modeller.
Having wrapped an offcut of acoustic foam around the SM7B to reduce the hi-hat spill a little further, I threw a couple more Avantone CK1s on the hi-hat and ride cymbals in case I needed to give those a clearer image at any point. I used hypercardioid capsules for both of them, to achieve reasonable separation without having to go in too close, and I also pointed the hat mic outwards away from the rest of the kit to reduce the bleed.
However, leakage from the kit onto the hi-hat mic wasn't nearly as much of a concern as too much hi-hat spill appearing on all the other mics! This can be an issue at the best of times when you're working in a small room, because of the way hi-hat energy bounces off nearby walls, but even when reflections aren't part of the equation, it's quite common for a hi-hat to overpower a heavily damped snare, as in this case. Although I know lots of ways of dealing with this at the mix, it seemed silly to make a rod for my own back unnecessarily, so I brought Miko into the control room to listen to the raw overheads signal for himself, and he was happy to adjust his playing a little to make the hi-hat more evenly balanced against the snare at source.
The bass was recorded in the same room with the drums, using a large-diaphragm condenser mic a few inches away from the amplifier grille. The drums are off to the left, and you can see the corner of one of the drum-tent duvets in the background.
Final Preparations
For each of the three toms, I put up a Pro 228A mic taken from a budget Superlux drum-mic set. The supercardioid pattern again allowed me to keep the mics 4-5 inches from the skin without incurring too much spill, and it therefore wasn't too hard to get a reasonably believable tone out of each one. The top tom did throw a couple of spanners in the works, though. First of all, it seemed to be a lot more resonant than the other toms, so that tom fills were coming across top-heavy, and its sympathetic vibrations were also interfering with the snare tone. Soloing the overheads signal indicated that this wasn't a miking issue, so we did a bit more work with gaffa and tissue paper, and managed to improve things enough that I felt the result would be workable in the mix.
The second problem we had was that the Superlux mic gave out on us just before we began doing takes, and I had to swap it for a Shure SM57. This wasn't a disaster from a tonality point of view, but the wider polar pattern did let in more spill. However, by this point everyone was geared up to start playing, so I wasn't going to hold up the whole session for 10 minutes messing around with different mics while the band went off the boil.
The band's guitarist played along with the rest of the band in the live room, recording a live guide part via a Behringer V-Amp 2 modelling preamp. The output of this unit was fed to all the players via the studio headphone system.
Setting up the bass was quite quick. Angling Chacha's amp up more towards his ears with one of my K&M amp stands enabled him to play at a fairly low volume while still hearing himself clearly, and this, combined with the padding around the drums, reduced the instrument's spill on the drum mics to the point where we could easily have punched in on the bass part as an overdub if we'd wanted to — although that turned out to be unnecessary. I did take the precaution of recording a DI feed from the bass, though, so that I'd be able to isolate the bass from the drums. This was important because Miko wanted to record with a click track, but Chacha wanted to respond to Miko's performance rather than the click. That meant that Miko had to knock his sticks together during gaps in the drum part so that Chacha would keep in sync. These 'metronome' parts naturally bled onto the bass amp recording, of course, so the DI signal allowed me to mute them at the mix.
Wot, No Processing?
Although this session involved a lot of careful planning and setup work, our commitment to designing the right drum sound while tracking really paid off at mixdown: even though I'd recorded the mic signals completely without processing, I didn't need to trigger any samples at all (as I'd feared I might), and EQ and dynamics requirements at the mix were also minimal: across all my drum channels, I used a total of four bands of EQ, one compressor (for more click on the kick), a limited-range gate (to reduce hi-hat spill on the snare), and some parallel compression of the whole kit.
So if you routinely find yourself stacking up dozens of plug-ins when you're mixing your own recordings, next time you're tracking try to save yourself a load of mixing work (and CPU power!) by spending some extra time sorting out the sounds at source.
Canned Sand
There are lots of times when a heavy weight can come in handy in the studio, perhaps to stabilise a mic stand that's at full reach, to damp vibrations in a floor or riser, or just to club the singer. Accordingly, I've frequently taken sandbags with me on past location sessions. Unfortunately, these had an annoying habit of scattering sand everywhere, and weren't easy to rig where I wanted them.
Then, a couple of years back, when I was mooching around the local DIY shop, I came across rather a neat alternative that some readers might be interested in: five-litre plastic petrol cans. These are cheap, well-nigh indestructible, and can be filled with whatever you have to hand. (Sand weighs around 8kg and water about 5kg. Beer also weighs around 5kg, but somehow doesn't retain its weight as reliably...) They also have a good solid handle you can use for fixing them where you need them. I put together four of these for about £25, and use them for pretty much every recording job I do. The Velcro-strapped wrist/ankle weights used for fitness training can also be handy in the studio, especially for counterweighting mic stands, but they do tend to work out two or three times more expensive than the sand-filled cans for a given weight.
More Reasons To Kick Myself
All in all, I was pleased with the way this session turned out, especially considering that I was working with a new band in an unknown venue. As usual, though, there were a few things to kick myself about after the fact. I should have resisted the urge to soften the kick-mic sound, for instance, because I know from experience how important a good 'click' is when you're trying to get the kick to cut through a rock mix. Some more low end on that mic would have been sensible too (perhaps via the Kickball's frequency tailoring switch), as that was where I still needed some EQ at mixdown. To be fair, though, the subs weren't that easy to judge via the studio's monitoring.
From the perspective of keeping the session running smoothly, I think we should also have worked up a detailed cue sheet for each song, with the names and bar-lengths of all the sections pre-agreed. As it was, there was a certain amount of confusion while navigating and discussing takes, in that everyone was referring to certain sections by different names! In retrospect, I'd have liked a bit more recording space on the hardware multitrack recorder we were using as well, because we didn't have room to do quite as many takes of the last song as we'd have liked to. Having discussed with the band prior to the session how many takes they'd roughly want to do, I'd thought an hour and a half of recording time would be more than enough, but I clearly under-estimated. What can I say? I'm an optimist!
Hear The Results!
Accompanying this month's column is a dedicated resources page on the SOS web site. This includes a variety of captioned audio demonstration files from this session, as well as rough mixes from the tracking session and one of the final full mixes so you can hear how the recordings sounded in their eventual context. There's also a link on that page where you can download the full multitrack session for one of the songs as unprocessed 24-bit/44.1kHz WAV files, so you can examine the recordings further or try remixing them for yourself.