Recording A Choir On Location, Part 1

Recording a choir on location presents a number of unique technical and artistic challenges. Hugh Robjohns describes how he approached a recent project. This is the first article in a two‑part series.

Earlier this year a colleague and I were asked to record the choir of Cheltenham College, an independent school in my home town. Whilst this particular project may not seem directly relevant to the kind of music many SOS readers are involved with, the wide range of technical and organisational issues involved are common, in one form or other, to most live recording situations.

The school's idea was to produce professional‑quality CDs of their very accomplished amateur choir, which could then be sold to raise funds for the school, as well as acting as a promotional tool in attracting new students. We were asked to handle the complete project, from the first recording right through to handing over the finished CDs, all within a time scale of about 10 weeks. The initial recording was made at the start of the summer term and the CDs had to be available for sale in the last week of term.

Planning

The choir receive directions. In the foreground are the talkback speaker and the home keyboard used to ensure consistent tuning between the recordings made in the new chapel (with the organ) and those made in the old chapel.

The first thing to do when faced with a project like this is to discuss the whole thing in detail with everyone involved, making absolutely sure there is a clear understanding about what is going to be done, by whom, and when. I find it useful to ask the 'client' to provide some examples of the kind of sound they are looking for, too, as this avoids any misinterpretation over the style of the recording. Any confusion at this early stage will result in complete chaos later on, so it is essential to make certain everyone knows what is going on. Equally, it is important not to bite off more than you can chew yourself — a challenging recording is a good way to develop skills, but don't make promises you can't keep if you want to build any kind of worthwhile professional reputation!

After agreeing precisely what is required, the next stage is to reconnoitre the proposed recording venues. From our discussions we had already learned that the choir consisted of around 40 students (plus a few adults), and that the recordings were to be made in two different halls at the school: the 'old' chapel and the 'new' chapel. The former, we discovered, was not designed with acoustics in mind, being a big rectangular box with lots of wood and hard reflective surfaces, and featured an enormous reverb time, especially at low frequencies. In fact, speech was completely unintelligible for anyone more than about 10 feet away from the person speaking — a serious problem which had eventually led to a second chapel being constructed some years later! The 'new' chapel is much more along the lines of a conventional large church, with fine acoustics to match, and houses a large pipe organ.

The old chapel's acoustics, although impossible for speech, were actually well suited to some of the slower, unaccompanied choral works which were to be included on the CD, so we agreed that this hall would be the first recording venue. The new chapel, with its Romantic‑sounding organ and shorter, brighter reverberation, was to be the venue for the more elaborate, accompanied material. The school's proximity to a main road junction and the A&E department of Cheltenham General Hospital was a major concern, and to minimise the likelihood of traffic noise problems we decided to record on a Sunday, planning to record in both halls in the same day — the old chapel in the morning and the new one in the afternoon.

Recording in two different venues on the same day would only be feasible if we could establish a fixed control room somewhere between the two halls, so that we would not have to waste time relocating the control room — it would take long enough to re‑rig the microphones and cables, talkback and cue lights in the second chapel. Fortunately, there was a convenient lecture theatre more or less mid‑way between the two halls, which was ideal for the job and within the reach of our multicores.

When you are looking over a potential recording venue, the kinds of things to look out for are the locations of power sockets and what kinds of plugs they need — some churches and old buildings don't use modern 13A plugs! You also need to find out if there are likely to be any noise problems from equipmet (heaters, chillers, lifts and so on), or from other people, traffic or planes. Are there any safety considerations, such as having to avoid laying cables in front of fire exits? How long do the cable runs need to be, and who will be letting you into the building to rig? A good recce will save a great deal of time on the recording day, which means you will have more ability to deal with the completely unexpected bombshell which always turns up at the last minute!

The Rig

The old chapel's acoustics, although impossible for speech, were actually well suited to some of the slower, unaccompanied choral works which were to be included on the CD, so we agreed that this hall would be the first recording venue. The new chapel, with its Romantic‑sounding organ and shorter, brighter reverberation, was to be the venue for the more elaborate, accompanied material. The school's proximity to a main road junction and the A&E department of Cheltenham General Hospital was a major concern, and to minimise the likelihood of traffic noise problems we decided to record on a Sunday, planning to record in both halls in the same day — the old chapel in the morning and the new one in the afternoon. Recording in two different venues on the same day would only be feasible if we could establish a fixed control room somewhere between the two halls, so that we would not have to waste time relocating the control room — it would take long enough to re‑rig the microphones and cables, talkback and cue lights in the second chapel. Fortunately, there was a convenient lecture theatre more or less mid‑way between the two halls, which was ideal for the job and within the reach of our multicores. When you are looking over a potential recording venue, the kinds of things to look out for are the locations of power sockets and what kinds of plugs they need — some churches and old buildings don't use modern 13A plugs! You also need to find out if there are likely to be any noise problems from equipment (heaters, chillers, lifts and so on), or from other people, traffic or planes. Are there any safety considerations, such as having to avoid laying cables in front of fire exits? How long do the cable runs need to be, and who will be letting you into the building to rig? A good recce will save a great deal of time on the recording day, which means you will have more ability to deal with the completely unexpected bombshell which always turns up at the last minute! The Rig For this project we had agreed to start the rig at nine o'clock on the Sunday morning, and the choir were scheduled to arrive at about ten‑thirty, so we had an hour and a half to get everything rigged and ready. The lecture room and chapel had already been unlocked for us so we were off to a good start. In classical music recording you rarely touch the desk again once the right balance has been achieved, because any alterations to the sound between takes can make editing impossible. Instead, the emphasis is very much on careful listening to the recorded performances and making accurate editing notes. Since it won't see much action during the recording itself, the mixer can be rigged out of the way to one side, rather than in the usual position in front of the speakers. I like to set up the monitoring first and work backwards through the rig towards the microphones; in this case, the speakers were set Left: The Soundfield ST250 stereo microphone needed to be much higher than an ordinary mic stand could reach — here the Ambient Jumbo Pole (!) comes to the rescue.

For this project we had agreed to start the rig at nine o'clock on the Sunday morning, and the choir were scheduled to arrive at about ten‑thirty, so we had an hour and a half to get everything rigged and ready. The lecture room and chapel had already been unlocked for us so we were off to a good start.

In classical music recording you rarely touch the desk again once the right balance has been achieved, because any alterations to the sound between takes can make editing impossible. Instead, the emphasis is very much on careful listening to the recorded performances and making accurate editing notes. Since it won't see much action during the recording itself, the mixer can be rigged out of the way to one side, rather than in the usual position in front of the speakers. I like to set up the monitoring first and work backwards through the rig towards the microphones; in this case, the speakers were set up near the front wall in the middle of the room, with the mixer and the recorders off to the right. Although the lecture theatre was a large room, its acoustics were surprisingly good, and the raked floor meant that we could sit 'line‑astern' and still all hear a good stereo image. I use PMC TB1 nearfield monitors for my location‑recording work, because they produce an excellent sound stage with good bass extension and are unfussy about positioning, yet easy to transport. They need a good amplifier, though; I used a Bryston 4B, although it is a bit of a heavy beast!

Sat on top of each loudspeaker was a metering unit: a traditional BBC‑style PPM unit on one, and a home‑made meter on the other. The PPM was a dual meter showing left and right on one twin‑needle meter, with mid and side on the other. The home‑made meter was a modified and re‑housed 'The Box', along with a custom LED bargraph meter. The Box (which is, unfortunately, no longer manufactured) consists of a diamond array of 100 LEDs which is effectively an audio vectorscope. I find it invaluable when recording in an unfamiliar monitoring acoustic, and it provides an immense amount of information about the stereo imaging of a recording. This was the first outing for the unit since its reconstruction, though, hence the additional PPM meter as a backup — on subsequent sessions I have been happy to rely on The Box alone.

On this occasion, the mixing console was a Yamaha 03D, fitted with an AES‑EBU buss output card providing independent digital feeds to two DATs and a CD recorder (see Digital Ticks And Analogue Hiss box). Sony PCM7000 DATs made the main and backup recordings, and an old Marantz CD‑R machine recorded audition copies for the playbacks. These discs were subsequently given to the choirmaster to take home and listen to at his leisure, to select preferred takes for editing. As a form of security, the two metering units were separately driven from the analogue outputs of the two DAT recorders so that any problems with one or other machine would (hopefully) be spotted quickly! I avoid replaying material from the DATs, as far as possible, to minimise the risk of over‑recording something vital, and I always rewind the tape to the leader at one end before ejecting it, just in case the tape gets chewed as the drawer opens!

Once the monitoring, mixer and recorders were rigged, we had the ability to play some reference tracks (from the CD‑R) to help familiarise ourselves with the acoustic foibles of our temporary monitoring environment. All that was then left to do was to organise some talkback and cue lights, and to run out the multicore to the hall. A switched mic, a spare channel and an aux output of the 03D became the talkback system in the control room, which was routed through to a small powered speaker in the hall. Reverse talkback was care of the main recording mics and monitoring — a crude system which had to be used with care to avoid howlrounds. In a pukka studio the monitoring feed is reduced in level when the talkback mic is activated but, unfortunately, the 03D does not have that facility.

I avoid replaying material from the DATs, as far as possible, to minimise the risk of over‑recording something vital, and I always rewind the tape to the leader at one end before ejecting it, just in case the tape gets chewed as the drawer opens!

The cue‑light system was another home‑grown affair. The 'Producer Box' contains red and green lever keys with associated LEDs, a PP3 battery and a female XLR connector. The 'Studio Box' carries a male XLR connector, a mic‑stand adapter and another pair of red and green high‑intensity LEDs. The system uses a spare mic circuit in the multicore and, because there is usually always one circuit in every multi with a broken screen connection which can't be used for powered mics, the cue system is arranged so as to do all of its signalling over the balanced pair without using the screen connection. The LEDs are incredibly bright and can be seen clearly even in broad daylight, yet an alkaline PP3 battery lasts for several days even if left switched on continuously. The lever switches are momentary in one direction and latch in the other, so it is simple to provide flicking greens and steady reds, or whatever other signalling arrangement is devised.

Choosing The Mics

The recording and monitoring arrangements in the lecture theatre. Against the front wall of the theatre (left) are the PMC TB1 monitors (the one in shot has the modified The Box meter unit on top); in the corner, Marantz CD‑R and Sony PCM7000 DAT recorders; right of shot, Yamaha O3D digital desk.

Based on the original brief, and having seen and heard the halls, the microphone configuration was planned around a Soundfield ST250 as the main stereo pair, with Neumann KM100s or AKG C414s as outriggers. With choral music, I like to capture a good impression of where the individual voices are, and a coincident pair works better in this respect than a spaced‑pair system — hence the choice of the Soundfield, which has to be the very best coincident‑mic system available. In the reverberant old hall we assumed the mics would need to be fairly close, with a crossed cardioid pattern for the main pair to keep the choir within the stereo sound stage. A common problem with this arrangement, though, is thacentral voices can end up much closer to the mic than those at the edges and thus tend to be far louder — hence the need for outriggers to redress the balance.

The ST250 mic was designed as a semi‑portable system with a small control box to adjust the stereo width and the polar patterns. This can take a little getting used to because it actually operates as an M‑S system: the polar pattern control determines the directivity of the mid component, and the width control sets the level of the figure‑of‑eight side component. However, the output signals are (optionally) converted into standard left and right signals for recording — albeit at a strange voltage which is lower than line level, but higher than mic level! Previous experience with this particular microphone made me doubt its ability to drive well over 50 metres of multicore, and I elected to put a four‑channel GML mic preamp in the recording hall and use it as a line driver, buffering the Soundfield mic as well as amplifying the two outriggers.

Recording any acoustic source in a nice‑sounding space requires some careful listening and a lot of experimentation. In this particular case the aim was to have the choir fill the sound stage, with a pleasing reverberant acoustic and with an appropriate perspective. Although it sounds trivial, the last point is the tricky bit because it requires a precise balance between clear, detailed diction from the choristers, and a homogenous sound in a warm and reverberant hall acoustic. The choir were physically arranged in the old chapel hall in an arc on a low platform at one end, and the Soundfield mic needed to be suspended well above and slightly in front of them — a position which is well beyond the capabilities of ordinary studio mic stands. The normal solutions are either to suspend the microphone from a catenary wire rigged across the hall from points high on the side walls, or to use a 'Cathedral stand'. The former idea, although popular in churches, was not possible or appropriate on this occasion because there were no suitable fixing points and it would have taken too long to rig and de‑rig. Consequently, we elected to support the main microphone, in its shockmount cradle, at the end of an Ambient Jumbo Pole, which is the modern lightweight equivalent to the traditional Cathedral stand. This carbon‑fibre pole can be extended to over nine metres in length, but is light to carry, easy to adjust, and very fast to rig, move or de‑rig by just one person. A heavy‑duty tripod with an adjustable clamp holds the pole at any desired angle with complete stability, and the microphone is counterbalanced by a sandbag attached to the opposite end of the pole.

With the mics roughly positioned and plugged into the multi, we performed a stereo line‑up to calibrate the gains of the paired mic channels on the desk. Having checked that everything was working properly, we then listened from our makeshift control room to the choir warming up and rehearsing through the Soundfield mic, which had been rigged quite close to the choir and set up as crossed cardioids, as per our original plan. The stereo acceptance angle of the cardioids was about 180 degrees, and the choir filled the stereo image nicely but the sound was much too dry, with a little too much presence and clarity in the voices — it sounded far too clinical, in other words.

The obvious solution was to pull the mic back further away from the choir, to increase the amount of reverberation relative to the voices and give a slightly less clinical sound. You don't usually have to move the mic far to make a big difference in perspective, and we moved it back about 20cm. However, moving the pair back meant that the choir filled less of the total acceptance angle, so the polar pattern was also altered towards a hyper‑cardioid pair to narrow the stereo acceptance angle and keep the choir filling the full image width. These corrections brought the sound much closer to the required goal, with good clarity of diction and plenty of presence, but with a nice reverberant bloom and a lovely homogenous quality to the choir as a whole.

In classical music recording you rarely touch the desk again once the right balance has been achieved, because any alterations to the sound between takes can make editing impossible.

Although it was close, we all felt it was still not quite right, so we tried moving the mic just five centimetres back in towards the choir. This achieved the exact sound balance we were after... but the stereo image became overly wide. Opening out the polar pattern towards cardioids narrowed the image nicely but also altered the perspective too much, so we settled back on hypercardioid and repositioned some of the members of the choir to reduce the physical width of the group a little. The balance and width aspects were now perfect... but these tweaks had had the side effect of making the voices at the back of the choir a little too distant compared to those at the front. I like to create a good depth perspective in my recordings — far too many modern records present everything on a straight line between the speakers — but this was too deep! The solution was to simply raise the mic about 20cm higher and fractionally closer to the choir again. The extra height brought a more uniform balance to all the members of the choir, and moving it in slightly compensated for the extra reverberation gained by increasing the height — all without messing up the stereo width.

Although we had rigged a pair of KM100s with the idea of reinforcing the edges of the choir, we had managed to find a position for the Soundfield mic which allowed it to do an admirable job all on its own. Fading up the outriggers didn't add anything useful to the balance, so we happily abandoned them. The fewer mics there are, the easier life is, and we were all very happy with the sound we were hearing in the control room. This whole tweaking exercise probably lasted about quarter of an hour or so and fitted in well with the choir's warm‑up. It is important not to take too long in finding the right sound in this kind of situation, because the available recording time is always too short anyway, and a bored or frustrated choir (or any other group of musicians) won't perform very well!

With the recorders, mixer and monitoring set up, the mics in position and the choir warmed up and ready to go, the effort invested in preliminary checking and reconnaissance seemed to be paying off. As with any location project, however, there was still plenty that could go wrong before we had the finished recordings in the can — and next month, I'll describe the second half of the project, from the first press of the Record button to editing and producing the finished CDs.