Recording a large orchestra can be fairly tricky, even in the most well‑equipped of studios, but trying to get a great sound on location, using only the gear you can fit into your car, adds challenges all of its own.
A 420‑mile round trip was required to get to Anglesey, North Wales, where I was to record the Cheshire Youth Orchestra. The 63‑piece orchestra was to be in residence at the site for two days of rehearsals before I arrived late on the second day (Friday) with all the recording kit. Saturday and Sunday were to be the recording days, with two sessions per day from 10am to 1pm and from 2pm to 5pm, with a break in each — a very professional approach, which I thought augured well.
I had been advised by the orchestra's manager (and the producer at the sessions), Valerie Hayward, that we were to record in the chapel on the site. I had ascertained that the chapel was relatively long and that the orchestra would fill nearly the whole length and certainly the whole width.
Orchestral layouts are very predictable, albeit subject to any restrictions imposed by the plan of a given venue. You have the strings close up to the conductor, with the first violins, second violins, violas and cellos spread out from left to right, and with double basses to the rear of the cellos on the far right. Behind the second violins and violas you usually have the woodwind — two each of flutes, clarinets, oboes and bassoons. Behind these, hopefully on risers, you find the brass section — trumpets, trombones, horns and possibly a tuba. Behind these loud instruments, the equally loud percussion section is usually found. At this site, the four players involved had their kit filling the narrower sanctuary area distant from the conductor — including timpani, glockenspiel, xylophone, cymbals and a conventional drum kit.
Lone operators like myself usually work with a main stereo pair above and behind the conductor, covering the strings and the rest of the orchestra in general. However, I knew that this particular venue and resultant layout would not be well suited to only having a pair behind the conductor, even though this method can undoubtedly work when there's an effective tiered setup.
I figured it would probably be necessary to focus on the woodwind with another pair of some sort, and although I suspected that the sheer volume of the brass would mean they'd need no separate cover, I decided that I should prepare for the possibility that I might need to adjust the percussion balance against them. I packed a figure‑of‑eight and a hypercardioid for this purpose, as they could both be set up with their pickup nulls angled to reject the brass — cardioids are not the answer in such situations.
Knowing that there was to be a movement from Rachmaninov's second piano concerto, and also an item with violin and harp soloists, I knew I would also need a moveable stereo pair available, placed in the space in front of the orchestra to help in balancing these soloists. I also included a couple of outrigger omnis in case I wanted to reinforce the left and right extremes of the soundstage.
All these mics, and a variety of cables, accompanied three tall stands with booms and one smaller set into a Cavalier saloon. Having enquired that a small separate room was available to monitor in, two stands for a pair of small BBC‑designed Rogers LS35A loudspeakers accompanied the mic stands, along with a Quad 405 amplifier. I do not like loudspeakers on tables — I use them up in the air, clear of rear walls and room corners, and having stands on which to place them usually ensures these requirements can be met.
I used one of the mixers I designed and built myself; one with eight channels in a briefcase, including talkback, Frank Fox's The Box visual soundstage monitor and two Sharp Minidisc recorders. These act as backups to the separate Sony DAT and allow instant playback without disturbing the DAT. They also provide the producer with a take‑home copy of the sessions for the edit planning.
Setting Up On Location
The hour and a half up to midnight on the Friday when I arrived at the venue was spent carrying in all the kit, setting up and checking that everything was functioning. For the main microphone I used a unique Pearl DS60, made in Sweden, with its four rectangular capsules — I have a dedicated DIY desk which means I can use it as if it were an M&S pair (see the 'Middle & Sides Mic Technique' box for details of M&S recording).
The woodwind had another M&S pair, comprising a Schoeps CCM4 cardioid and CCM8 figure of eight, held in place by a neat M&S mount. For balance reasons I favour a semicircular woodwind layout, when I can get away with it — I did so here by moving the music stands and seats around at 10 minutes to midnight, while no‑one was looking!
The percussion area was covered by a couple of Beyerdynamic ribbon mics in another M&S rig — an M160 hypercardioid coupled with an M130 figure-of-eight. I like to use ribbon mics with percussion, and the low output levels associated with most ribbon mics don't tend to cause any problems with such loud instruments. I was able easily to set up the mic nulls to ignore the brass just behind the mic stand.
I usually make sure that I leave some spare mic cable under spot pairs, so that they can be easily moved to suit different works being recorded. In this case, I knew that there was to be a James Bond theme selection, and that I'd have to move the positioning to suit the drum kit for this.
The soloists, near the conductor, were covered with another M&S pair of Schoeps mics, this time with a soft cardioid CCM21 as the middle.
Efficient Cabling
I had packed a 16‑way 75m Klotz multicore, but with the short distances involved I preferred to use separate cables for each pair of mics — if I'd used the multicore, I'd probably have had to saw the centre corners from the doors to the playing area in order to prevent damage to the cable! I exclusively use dedicated four‑core 25m cables, with five‑pin XLRs, for each of my M&S rigs, each mic feed using opposite wires of the screened quad. The desks have five‑pin XLR inputs ensuring each pair is correctly connected each time. I had packed eight reels, making 200m in total.
An important regime to follow when running cables through an orchestra to your control room is to tape down the cables with white gaffer at sensible intervals. I also use the gaffer tape to protect the cables wherever they pass through doorways which need to be shut so as not to compromise the loudspeaker monitoring.
Creating The Best Balance
The first thing I do, during the warm‑up at the start of the first session, is to set the gains and the faders for the main pair. For setting up the Middle channel, the conductor will always oblige by rehearsing a loud bit. I try to allow a cautionary headroom of about 6dB on the DAT, as the players' enthusiasm tends to rise as sessions proceed.
The stereo width then needs to be set by the amount of Sides mic in the mix. Apart from loudspeaker and Sennheiser HD600 headphone judgment, I also rely on The Box's display, which helps reassure me that things are not too wide or too narrow. You can see its diamond‑shaped grid of LEDs at the top of my suitcase mixer in the picture, and a full diamond display is the aim.
With the main pair muted I then individually set the gains and the faders for each of the other pairs in the rig. Each also has its width provisionally set. Having got the ball‑park settings, the main pair is brought in again and the balance and perspective influence of the other pairs is judged, with levels and widths being trimmed. There is an argument for including time delays in the feeds of 'spot' mics, but this recordist remains to be convinced of this when relatively distant stereo spot pairs are used.
The dominant source of peak levels will generally remain the main pair and, overall, the balance should give the impression of strings nearest, woodwind behind, brass and percussion behind them. It will be the main mic which will give the more ambient pick‑up of the woodwind, brass and percussion — the spot pairs adding focus and clarity in the mix.
The producer's ears also assisted here. Valerie had the scores of the works and advised me on the balance, as this particular engineer can't (and doesn't want to!) read musical scores — I take the view that an alternative judgement from a listener's perspective is a good thing. Finally, once we had a balance, the conductor, Tim Redmond, was invited in to hear a test take. We must have got the sound pretty good, as he felt he could make any further changes merely by working with the orchestra.
Homing‑in On A Soloist
When it came to recording the piano concerto movement, we had to find the best way to balance the soloist, Fergal O'Mahony, with the orchestra. This involved getting the piano sensibly positioned and employing the soloist M&S pair. There were a number of factors to take into account here, not least that Tim Redmond had to be able to see Fergal's hands at some crucial moments. In the end they decided that the baby grand piano should be to the right of the conductor, in front of the cellos, which meant that Tim didn't need to turn too far to see Fergal's fingers.
The piano lid was fully raised and the spot pair was placed 2m from the centre of the strings at about 2.5m above the ground. The level balance and perspective, along with the apparent width of the piano image, was quickly established as they played through the whole of the eleven minute movement. I favoured a 'slightly to the left' stereo position for the piano, with a width related to that of the orchestra. The spot pair was angled to get the slightly left‑hand image, this having the advantage that spill from the cellos behind and to the right of the piano was captured to the right of centre, coinciding with the main pair's perception. Conflicting spill can create imaging problems, as well as balance difficulties.
Taking, Editing & Mastering
The four sessions went by without any technical problems. There was some subtle rebalancing and mic repositioning, especially for some of the film scores. By the end, we had a total of 124 takes recorded on three DATs, with four Minidisc backups. The backups were given to the producer Valerie, who had to listen through in excess of four hours of music in order to plan the editing. She was assisted in this by Simon Roach, who has a degree in recording and who was then to come down to North Bucks with the scores in order to guide me in getting the edits right.
The editing program we used was Minnetonka Audio Software's Fast EdDit, which I find quick and tactile to use. The appropriate takes or parts of takes were loaded in as straight digital feeds via S/PDIF from a Tascam DAP1 DAT. After some hours of copy‑and‑paste editing, involving over 50 edits within the music, the 16 works were assembled in a provisional final order.
It was not too difficult to come up with an opening number, the music from Star Wars, and the final track was to be Ron Goodwin's 633 Squadron, with which the orchestra always ends their concerts. The piano concerto movement and the other piece with soloists (John Williams' music from the film Schindler's List) were spaced apart in the middle, with the other works between. Each piece was trimmed to about two seconds at the start and about three seconds at the end.
Fade‑ins and fade‑outs were then made, and an appropriate number of seconds of digital silence placed in between — classical music needs around six to 10 seconds between works. Then we sorted out the track IDs, and these were placed at the start of the ambience fade‑ins (rather than right up against the music) as is the norm for classical CDs.
The mastering technique we used was to output the entire playlist from the S/PDIF output of the editor to an Audio & Design DMM1 digital mastering desk. From there it went to two master DAT machines in parallel.
This setup allowed the levels of the quieter items to be raised manually, making them much more interesting to listen to. Increases of around 6dB were common, with one or two particularly low‑level deliveries going up as much as 9dB. By watching the cursor move along the editor's waveform display we could gently get the level back to 0dB where necessary, thus preserving the impact of the orchestral dynamic range. This 'musical' approach is something which cannot be done with compressors or limiters!
I had noticed that we might need to add reverberation while on location, but I usually do this after the editing to leave my options open. In this case, I mixed in reverb from a TC Electronic M2000, though the send was first passed through a TC Electronic Finalizer in order to add some low‑frequency lift and high‑frequency cut as a means of warming up the sound. I also increased the stereo width of the reverb send in the Finalizer.
Finally, the tapes from the editing session were passed to fellow engineer and producer, Patrick Allen (of Opera Omnia Productions) who discussed it with the conductor, and put finishing touches to the track ordering, adding a touch more reverb using Patrick's TC Electronic M5000 system.
The Cheshire Youth Orchestra's A Brief Encounter was recorded in aid of the Sargent Cancer Care For Children fund, following the recent early death of Joe Anson, who had been a trombonist with the orchestra for five years. The CD is available for £10 (including UK p&p) direct from: Friends of CYO, The Croft, Chapel Lane, Lower Whithington, Macclesfield, SK11 9DE.
Middle & Sides (M&S) Mic Technique
M&S stands for Middle and Sides, which is a powerful stereo miking technique. Two mics are required. The Middle mic faces the centre of the soundstage, and can be an omni, cardioid, figure-of-eight, or any pattern in between. The Sides mic faces sideways to the soundstage, and must be a figure of eight. The mics should be as close together as physically possible and vertically coincident.
When the signals from these two mics reach your mixing desk, the Middle mic is fed at equal levels and in phase to the left and right of your master stereo buss — connecting it to one channel of a desk and panning it centrally takes care of this. The Sides mic is also fed to the left and right of the stereo buss at equal levels, but with the phase of the right channel inverted. This can be done by splitting the mic feed to two channels of a desk, panning one full left and the other full right, and having a phase inversion in the right‑hand feed. If the desk can't do the phase inversion, it can be incorporated in the splitter lead.
The level of the Middle mic is decided according to the headroom of your master recorder in the usual way, depending on what is being recorded. On the other hand, the level of the Sides mic determines the width of the stereo image — it shouldn't really exceed that of the Middle mic, and should usually be a maximum of around three decibels below, due to its out‑of‑phase nature.
I have been sold on the use of M&S microphone rigs for years. For a start, you get instant control of stereo width on your desk, as just mentioned. In addition, the stereo image starts between the loudspeakers, as the centre of the soundstage is focused on axis by the Middle mic — compare that with any crossed or spaced pair — and the imaging is very clean, clear and coherent, with a depth of perspective. Mono compatibility is second to none, as the Sides mic contribution is simply cancelled when summing the stereo to mono, and if you have suspended a pair upside down on location, as is usually convenient, flipping the sides of the stereo is as simple as changing the phase‑inversion settings on the desk. Another reason I like M&S is because I can use all sorts of combinations and makes of mic from my armoury — even the BBC‑designed long‑ribbon figure of eight microphone, the Coles STC4038.