Recording Steel Drums: Our engineer rises to the unfamiliar challenge of capturing a loud steel band in their octagonal rehearsal building.
Amongst the many things I enjoy about being a recording engineer is the variety of people you meet and the unusual situations you can sometimes find yourself in. This past month alone, I’ve recorded a few different guitar bands and singer-songwriters, a brass section, audio for an English exam paper, and a hurdy-gurdy player. For this month’s session write-up, though, a personal first: recording a steel band.
When I was initially contacted by the Cambridge University Steel Pan Society, to see if I was able to accommodate their 14-piece steel band at my studio, I had to politely decline on the grounds that they wouldn’t all be able to fit in my studio’s live room at the same time — and I couldn’t see much merit in attempting to record a band like this in a modular overdubbed fashion. But I did ask them if they had a good room that they rehearsed in. If they did and it was suitable, I reasoned, I could take my little mobile rig along and record them at their place. They seemed to quite like the idea, but before committing to the gig, I thought it sensible just to pop out and take a look at the venue, to listen, and figure out more broadly what I’d be dealing with.
The band rehearse in a music room in the grounds on one of Cambridge University’s colleges. In essence, the room is like a large wooden garden outbuilding, but it was purpose-built, had pretty good soundproofing and an acoustic that was immediately quite pleasing on the ear. The recording space was octagonal, had an all-wood interior and had a fairly steeply pitched roof that rose about 15 feet from the edges to the very middle of the room. It would be great to record a little ensemble in a space like this, but when I walked through the door in the middle of their rehearsal I was faced with 14 steel drums being played at full pelt — all the while being kept in time by a full drum kit!
I sat and listened to a few of their songs, from a few different points in the room, and then we had a chat about the options available to us. The band filled the room in both senses of the word — sonically and physically — but, happily, the room still had a balanced, controlled sound to my ears; it was very comfortable to listen to. Trying my best not to be patronising — it’s impossible to gauge how well 14 different people get some of the concepts of recording — I explained that I’d be quite happy to record them in that room, but that it would be impossible to achieve any useful separation between the different elements of the band.
Our other option was to decamp to a larger hall somewhere else on the campus. That would potentially allow us to get some physical distance between the different elements of the band. We talked this idea through for a while, but the band was very happy to go with what I recommended — which was to capture as balanced a recording as we could in their rehearsal space. I’m always thinking about how to get the best quality recording in a situation like this, but when faced with having to make a quick decision that will have a huge effect on the outcome, you often have to take a pragmatic approach. The project had a very modest budget, and the priority for the band was to get a good-sounding record of what the group was sounding like at that point in time. As a University Society band, they had a constantly revolving cast of members, and we wouldn’t have the luxury of decamping to the larger hall to do any kind of test session — the logistics of moving a band like this are quite significant! Furthermore, the larger space might pose some challenges. For instance, I might have had to contend with the sound of a large steel band swirling around a big reverberant sports hall! I also sensed that they were hoping I’d want to do it in their regular space, and the final deciding factor for me was the sense that recording them in familiar surroundings would probably encourage the best performances.
I still very much rely on my little RME Babyface interface for my occasional adventures in location recording, and despite only having two onboard mic preamps it can connect to eight external preamps via its ADAT connection, so it can offer me up to 10 simultaneous mic or line inputs if necessary. I was able to borrow a Focusrite ISA 828 eight-channel mic preamp for the recording session. These two bits of kit, combined with my Mac Mini and a pair of headphones, formed my recording rig for the session.
Microphone-wise, my strategy was to presume that the vast majority of the final sound would be based on a single stereo array of some sort, but I wanted to give myself a couple of different stereo options to choose from, while leaving the remaining channels to hook up a few spot mics, in case they gave me a small amount of control come mixdown. Things got much more interesting, however, when after chatting to SOS’s Sam Inglis about the project he offered to lend me the Soundfield TSL ST450 MkII microphone he’d recently acquired, and also kindly volunteered to help me set up the session. Having never had the chance to record with a Soundfield before, I readily accepted: it’s an ingenious invention (see box) that allows you to decide on different stereo and surround options after recording.
My experience of working with steel bands was nonexistent, so it was a real treat to experience the wonderful sound they produce and also to gain a better understanding of how such a band might be made up. It seemed to my uninitiated ears that the band comprised three sections, but it was explained to me that we actually had five different types of drum contributing to the overall sound for our session. Looking into the room from the doorway, on the immediate left there was a set of six tenors, and at the far left the bass (three lower-pitched drums played by one person). Then, back right, was a conventional drum kit, and to my immediate right was a section that consisted of two ‘double guitars’ and two ‘double second’ steel drums. Whatever they are called, the whole ensemble playing together produces a wonderfully upbeat sound — it’s hard not to listen to a band in full flight without having a stupid grin on your face!
If you stood in the middle of the room, there seemed a natural logic to how the band would be presented within a stereo field. My only slight concern was how much it would matter if the drums were slightly off to one side of the image. With the basic recording rig I described earlier, my only means of monitoring what I’d be recording from a loud band such as this was to record a section and then listen back through my headphones. Even this was a real challenge given the number of people in a relatively small room. So, despite me (politely!) kicking them all out to give me a few precious minutes to check that everything was OK, I was flying blind to some extent.
Our approach, then, became about having a basic plan of where things would sit in the stereo field, using instinct, and making the most of the 10 available recording channels. I’d already earmarked two channels for a spaced pair of omni-directional mics. For this, we used MBHO preamp bodies with KA100DK capsules, positioned about three feet apart and fairly central in the room.
The B-format Soundfield setup, which we would not be able to audition until after the session, would occupy another four channels. Sam’s Soundfield mic also has the option of giving you a conventional X-Y stereo output. This seemed essential to me, as, despite the flexibility the system would give us, we still needed something to audition, just to make sure the mic was placed in a sensible position. Realistically, there was a fairly small space in the centre of the room that we had to play with, positioning-wise, and we tried three different placements before settling on a position closer to the drum kit and bass section. We also tried adjusting the height of the mic before settling on our final position of about six feet above the floor.
If I chose to record the X-Y output of the Soundfield, this would only leave me with two channels for spot mics, so the big question was: what did I want, or hope, the spot microphones might give me? I was confident that we’d achieve a nice sense of the space with our stereo options, but I felt the key to making the recording more than just a well-captured rehearsal was to have some means of enhancing the bottom end of the band and some degree of definition. I was certain that I wanted a close mic on the kit’s bass drum, and I positioned an AKG D224E just outside the front beater of the drum (there was no hole in the drum’s resonant head). We then played around, hoping for even a small amount of control over the bass section. We tried a ribbon mic, with it’s figure-of-eight pattern, about three or four feet above the centre of the bass section, but it was immediately obvious that this wouldn’t be much help — it just picked up everything as it swirled around the room. Another idea was to place the mic below the drums, but it was impossible to represent all three drums without getting further away, where the mic was again greeted by a wash of sound. Feeling slightly resigned, we settled on another AKG D224E dynamic mic focused towards the middle of the section, about three feet above the middle drum.
By the time we’d played with the mic placement options, we’d blown all our available setup time, and needed to get recording. The band were hoping to get through quite a number of songs in a few hours, and we settled into a rhythm where they’d rehearse part of a song, and we’d then record two or sometimes three takes, as necessary. Recording-wise, there wasn’t a huge amount to report, except to stress the need to keep your concentration on sessions like this. After each take, I made a point of listening to a little snippet to make sure everything was gremlin free. It was a good job I did, too, as a random digital noise appeared on just one song. Luckily I spotted it, and was able to ask for a retake; I was careful to check it didn’t happen again!
Back in the studio, I could audition the recordings. Our chosen stereo option was doing it’s job well, but I was disappointed with the lack of options for giving the sound a fuller, more ‘hi-fi’ feel. Being able to beef up the kick drum helped a bit, as this lent the drums and the recording as a whole a little more weight. The close mic on the bass drums was of limited use though; it only succeeded in providing a little more low-end rumble that I decided, on balance, was better included than left out. Once Sam had decoded the recorded Soundfield signals into a stereo file (see box), I was able to compare the available stereo options. Listening to the spaced omni pair first, I really liked the natural tone that these mics captured, but the stereo image wasn’t fantastically defined and, despite our best efforts at the recording stage, the drum kit was quite noticeably placed to one side, which I found distracting. With the decoded Soundfield option, I didn’t like the overall tone quite as much, but it was OK, and we were able to position the drums firmly in the centre of the stereo field, and also choose the best way of getting a sense of stereo spread that represented the band’s layout in the room.
Processing-wise, I did very little to the stereo channels apart from apply some EQ, to smooth out a little ‘peakiness’ around 2kHz and to perform a few other little nips and tucks. On the mix bus, I ran everything through a small amount of compression, to help smooth things out a touch, and to the same end added just a smidgeon of top and bottom boost via a Pultec-style EQ plug-in.
This was a really interesting challenge for me and, happily, we managed to get a good result in the end. Most importantly, the band were really pleased. I’m writing mostly about the technical aspects of the session here, of course, but it’s important to remember the whole point of it. The band felt relaxed playing in their rehearsal space, and they obviously felt that we had captured them playing to their potential. From this point of view, it was the right decision to record in that room. However, would I do things differently if I was recording them again? Quite honestly, yes. Using the Soundfield mic was an unknown to me, so I didn’t feel comfortable relying on decoding its signals after the event. Thus, I felt I needed to record the mic’s standard X-Y output too. Hindsight tells me that this really wasn’t necessary, and that the two channels this used up would have been much better used to capture a close mic on the drum kit’s snare, or by close-miking all the individual drums of the bass section, to give me some control at mixdown. I’ll take those lessons away for another time, but we still managed a good result, and enjoyed a challenging new experience to boot!
Like hovercraft and tilting trains, the Soundfield microphone is one of those great British inventions that was miles ahead of its time. Developed by Michael Gerzon and Peter Craven in the late ’70s, the basis of the mic is a tetrahedral array of sub-cardioid mic capsules, from which a four-channel ‘B-format’ signal is derived. The four B-format channels correspond to three figure-of-8 mics at 90 degrees to one another, and an omni mic. In turn, this B-format signal can be used to synthesize any coincident stereo or surround pickup. In other words, if you record the B-format output of a Soundfield mic, you can decide after the fact whether you want to hear an X-Y cardioid array, a Blumlein array, and so on. What’s more, you can also ‘steer’ that array to point in any direction in 3D space.
This is obviously very useful in the sort of session we’re describing here. When you’re short of time and working in an unfamiliar acoustic, with no separation from a very loud band, the chances of finding the optimal configuration with a conventional stereo pair are pretty low, as we discovered with the spaced omnis. With the Soundfield, the only variable you really need to get absolutely right on the day is the physical position of the mic in relation to the sources and the room. Everything else — the stereo width, polar patterns, direction and so forth — can be put off until you get back to base.
The original Soundfield MkIII and MkIV mics came with a large rackmounting box that, as well as powering the mic and incorporating four gain-matched preamps, sported a full B-format decoder. However, the ST450 MkII variant we were using here is designed for portability, and its much smaller and lighter box incorporates a simplified decoder offering only Width and Polar Pattern controls. To access the full range of B-format decoding tools, including direction, up/down tilt and ‘zoom’, requires Soundfield’s SurroundZone 2 plug-in, set up in quad-to-stereo format. Neil had access to two DAWs, Ableton Live and the native version of Pro Tools, neither of which supports quad channels, so he sent the recordings to me to decode in Pro Tools HD, which does.
As we’d set up the mic in the middle of the band, with instruments all around it, I knew I needed to decode to a stereo array with a 360-degree pickup pattern. This pretty much limited the choice to cardioid mics at 180 degrees, and although pin-sharp stereo imaging was never really on the cards in such a small and reflective space, this did produce a greater sense of width and fullness than our spaced omni pair. The SurroundZone plug-in gave me the ability to rotate this back-to-back cardioid array in any direction with a simple mouse move, and doing so made clear that our on-the-spot decision to place the low drums in the centre of the stereo image with the Omni mics had been wrong; everything gelled much better with the drum kit in the middle. By tweaking the zoom and front/back tilt settings I was also able to make some adjustments to the relative balance of the drum kit, the steel drums and the room ambience, although the overpowering nature of the last meant there wasn’t a great deal of flexibility on offer. Sam Inglis