geoffsample.com

ecology music landscape mythology

home | news | projects | wildsong | beyond | contact


dawn chorus :: marcus coates

field & studio work


Phase 1: Planning | Phase 2: What can go wrong?
Phase 3: Sound ascendant | Phase 4: Extracting the score

PHASE 1: PLANNING

We had decided on a two week intensive period at the beginning of May to get the field recordings. Well, maybe we felt that was all the time we could afford to devote to the recording side of the project. I can’t quite remember now. But I do remember that, after a year thinking through the practicalities and working out the plan, as the date approached I was feeling very nervous.

Timing was a matter of choice that would affect composition. If we’d gone for an earlier period, then the chorus would be more restricted to resident birds and the weather tends to be more turbulent, sometimes still hanging on to winter. May is really the peak month for birdsong, though as the month progresses, the output of some of the resident males declines as they have to provide for nestlings. So the first fortnight of May it was - with the first wave of summer migrants settling in and the resident males still in good voice. And hopefully we would get a few days of calm, dry weather - microphones just don’t work very well in wind.

Choice of sites
We’d spent a lot of time discussing sites. I felt more confident on this. We had permission to work in three separate sites, all with slightly different qualities, two of which I knew quite well and had used on a regular basis for my own recording work. In this aspect I was at least on familiar ground: I’m used to arriving at a site, living with it round the clock for several days and really getting to know the bird community. Marcus and I had visited the sites during February and March, but there wasn’t much to be gained from mapping out the singers and song-posts at these early dates, since it was all likely to change as spring came on.

We wanted somewhere with a reasonably diverse community of birds, so there would be good variation in the sonic themes. But we also wanted several species with more than one individual male singing, so there would be some counterpoint in the final score. Depending on the site, this should not be too difficult in the case of wrens and robins, or maybe even dunnock, blue and great tit, and possibly one or other of the finches or warblers, but was more difficult in the case of the thrush family, which tend to slightly larger individual territories. The practicalities of rigging out a site with up to 16 cable runs meant that we could cover a roughly circular area, from the central point of the recording device, with a diameter of about 100m, though if needed we could extend an individual line further by doubling up leads.

Marcus’ requirements in deriving clean individual signals from within the sum of the chorus meant that, on the other hand, we would not want the individual birds bunched up together too much. The ideal would consist of a diverse community, not too heavily-populated, with individual males evenly spread through an area within about 100m of a central point. As much as possible we needed individual birds to sing close to individual microphones, without any other singers close-by. This requirement suggested that woodland or any other site with a surfeit of tall trees would give us problems: it would be too difficult to consider getting mics up trees to cover birds singing from tree tops or ranging through the canopy. So we tended to sites with scattered scrub and saplings; open habitat with little cover (rank vegetation, scrub and bushes) would not be likely to provide us with enough individual birds within our potential work area.

mic in windsock In theory male birds of this kind of habitat have favourite song-posts that they sing from; a bird will sing a while from one, then move on to another, its song showing the area that this bird is claiming as its territory. Some birds (wrens, tits, some warblers) are more mobile: they may sing while foraging for food - they stop, deliver a song, then continue moving through the foliage. Other than fitting individual birds with tiny radio mics, I didn’t see how we could deal with this; some day soon this may be feasible. It seemed to me that we had to choose our mic locations carefully and spread them through the area. It wasn’t simply a matter of micing an individual’s song-post, but putting mics in spots where mobile birds might sing for a while in passing - isolated bushes, wood edges, the corner of an old hut and such-like.

Actually we did consider another possible approach - multiple mobile recordists with parabolic reflectors. A parabolic reflector with a microphone fixed at the focal point looks like a satellite dish and is a fairly standard part of the wildlife recordist’s toolkit, that acts like a telephoto lens for a photographer. This was a feasible approach, apart from the impact of having 12 or more people moving about in our fairly limited work area. By impact I mean the effect of this human presence on the birds and the disruption to their singing behaviour.

This is something we had thought about in our planning. Birds in urban parks, suburban gardens and regular rural walks become accustomed to the presence of humans and would be less likely to be disturbed by our activities. Recording in such sites offered clear advantages from this point of view, but such sites are likely to be quite noisy and the patterns of noise can be unpredictable; a rubbish van might happen to work through the adjacent area on the most important period of our one calm morning. Then there would be the worry of having many thousand pounds worth of mics carefully laid out overnight in a site frequented by other people.

So we needed sites where we would have some privacy to do our stuff. We wondered about renting an isolated cottage with an extensive garden and spent some time looking through the possibilities. Or even moving in on some unsuspecting friend. We wondered about travelling further afield - into the Borders, the Scottish Highlands, somewhere in Wales. It would certainly have been more comfortable to work from such a situation, with the recording centre set up under a solid roof and the big bonus of mains power. But maybe the more solid barrier between us and our subject matter out in the open would be detrimental to the process of the field work. And I think we had a feeling that this was a bit like putting all our eggs in one basket - there wasn’t much scope for adaptability.

In the end we went for a campervan as our shelter and recording centre and planned to work the sites we knew in our area of north Northumberland that allowed us access for the van to a central point in a recording arena. The fact that we could get the campervan to that point would suggest that the local bird community would be used to the occasional presence of vehicles. This gave us more flexibility, depending on what we learned as the sessions proceeded and how the weather conditions developed. And was quite luxurious compared to my usual field work conditions. I tend to rough it, Marcus likes his basic comforts - and a decent sleep! But it was surprising how this spacious camper became clogged once the recording equipment was rigged up. Try bringing 16 heavy duty cables discreetly through the window of a vehicle. Flightcases, heavy duty batteries, mic boxes, multicores from pre-amps to recorder, rigging improvisatory kit and so on.

We had three scrub/wood-edge sites in mind to work, with the tempting possibility of trying a moorland site, if time or conditions allowed; we couldn’t rely on such open habitat, since so many of the birds’ territorial displays involve aerial song and advertisement and a lot of motion. Difficult. The three main sites had the added advantage that they were sheltered from different wind directions, which proved to be useful.

Site 1
A disused quarry near the coast. The high edge of the worked-out rock face formed a crescent on the south-west side, whereas the arc forming the circle to the north-east side had scattered scrub and saplings. Just outside this circle to the east was another area of rough grass and scattered bushes bordered on the south by a young wood on a steep slope.

Site 2
An open area behind some disused farm buildings, bordering a mixed woodland and surrounded on the other side by some scrub and mature oaks. The photo doesn't show the area we were recording; this was from a walk along the side of the woodland.

Site 3
A hollow with extensive hawthorn scrub and gorse patches, bordered on the north by a scots pine plantation (behind the campervan in the photo), with another further off to the south.

Equipment
We still didn’t have a working recording system, but we had decided on the set-up to go with. This hadn’t been at all straight forward. We wanted a 16 track system and had originally thought that recording to laptop would be the best solution. Research into others’ experience of mobile multitrack recording suggested we’d be taking a laptop-based set-up to its full capacity and this might not be a robust enough system. We needed the system to record continuously for up to an hour to 16 tracks, running at a minimum of 16bit and 44.1k f/s. Reliability was more important to us than sheer quality: it could well turn out that we only had one or two opportunities when weather conditions and the birds’ movements were right for us, and we couldn’t afford for our recording system to jam at the crucial moment. So we hired a dedicated 24-track hard disc recorder that we were advised was rock-solid - an Alesis HDR24 with two focusrite Octopres (each with eight microphone pre-amps). This did prove to be a reliable system.

Power
Marcus had arranged for the loan of a heavy-duty 12-volt ‘leisure’ battery with the hire of the campervan. The plan was to run the whole rig (240-volt recorder plus two 8-channel mic pre-amps) via an inverter from this battery. It was a case of fingers-crossed: I knew that inverters are inefficient in their use of power and, from the figures I could find for the power consumption of the equipment, I felt it should just work. If it didn’t we would have to rethink things. At least at Pawston (site 2) there was the possibilty of laying an extension lead from the camper to one of the farm out buildings and running from mains power. There was some rational comfort in having a plan B, but it didn’t really soothe our anticipatory nerves on the power issue.

PHASE 2: WHAT CAN GO WRONG?

We took delivery of the kit on bank holiday monday and spent the afternoon testing it out and going through all our gear. We decided for our first night, at site 1, we’d just listen and observe, to get an idea for mic placement. It was rather a cold dawn with a light breeze blowing; the dawn chorus was hardly a fanfare and left us slightly apprehensive. The rest of the day was spent preparing kit, testing mics and improvising wind covers. (Bird peanut feeders - remove the ends and cover the mesh with a couple of socks. Suspend the mic inside using elastic bands.)

Tuesday evening was our first rig-out. It took much longer than expected and we didn’t finish until around 11pm, after working into the dark and getting a fair come-uppance from brambles. Scrub might be the ideal habitat for various reasons, but it proved to be a difficult habitat to rig in practise through the abundance of brambles, nettles, spear-like dead stems of bracken and willow-herb, (haw)thorns and such-like. Working into the dark also meant that we would probably have driven out any birds trying to roost in the area, though I hoped that they might return when we finished and certainly at dawn if they were used to singing here.

So, the moment of truth - powering up the whole rig and going through each channel to check the signal. Naturally one or two weren’t working and a couple of mics didn’t appear to be in the right channels; but it really wasn’t very easy to tell when there was no sound to pick up except the soft hiss of the breeze, not that different from the soft hiss of system noise or a dead cable. More tramping through the brambles and up the steep brackened slope in the dark to check connections. Marcus got busy cooking us up some tea. Particularly delicious with midnight hunger pangs. It was about 1am when we were finally ready to get some sleep, with the alarm set for 3:50am. The last thing I remember was the breeze playing round the campervan and the patter of drizzle starting. Oh hell. But we still hadn’t bottomed out yet.

The plan was to keep all lighting and noise to a minimum, with the camper’s curtains drawn and all windows shut but the one with the dozen snaking cables entering. We didn’t want to do anything to disturb the birds, though this meant we wouldn’t be able to observe what was going on (even though it was still too dark to see). Coffee is an essential luxury for me on dawn sessions; even more so on a couple of hours’ sleep. Powering up was amazing; in the cold gloom of dawn, technology’s multi-coloured lights sparkled like a christmas tree. Well, it was all a bit pointless anyway since it was still drizzling; but at least the mics were still working - except channels 5 and 8.

By 4:45am the rain was just about finished and we went into record. The first wave of the dawn chorus, such as it was this grey morning, was about over now, but we were desperate for a test run to see how the power held out. 65 minutes later it was still running OK, so we stopped it there; we’d set our goal at one hour of continuous recording. The weather may be against us, but this gave us a definite lift. Now we needed to get more channels in operation.

On the other hand, the birds weren’t exactly co-operating - we had very few birds singing directly into the mics. There was a very obliging whitethroat that was a sheer joy to me at least. Not only was he singing his heart out, he was moving from one mic to another on predicted songposts. But that was about it; we needed another dozen or so to perform like him. There was also a lesser whitethroat about this morning, quite a scarce bird in these parts with an unusual song - a complex warble leading into a mechanical rattle. But he never reappeared on subsequent mornings.

After a short period of half-baked euphoria, tiredness set in. And with it a draining depression - in spirits and barometric pressure. The wind built up and it was pretty cold, certainly not conditions to enthuse the birds’ singing nor lift our mood. The bottom line: several mics weren’t working properly, including an expensive B&K 4006; the birds hadn’t appeared where I expected them to; the forecast was poor until the middle of next week. Time was slipping by, each day costing a small fortune for hire of mics, recording system and campervan.

Then to really dig it in, a rumble and clattering started up just round the corner: workmen had arrived to begin maintenance work on drainage nearby. They were going to be here for a few days at least and planned to use the quarry for storage of some of their kit.

The next few days somehow got lost in the weather, trying to improve the kit, get replacement microphones, think through our methodology and look for alternative plans. Have you ever tried to enjoy something when you feel in your heart it’s a waste of time, making the best of elaborate plans that appear to be disintegrating into complete failure? Can there ever be pleasure in futility?

The other difficulty we faced at this site was the fact that it was rather public, with a road, admittedly very minor, running alongside. In fact several of the mics were in roadside hawthorns. This all meant we couldn’t leave the site with the mics out: one of us had to be here all the time it was rigged. On the Thursday when we both had to get off for provisions, meetings etc., we took all the mics in and weatherproofed the ends of cables (sandwich bags and elastic bands) and left the site for a few hours.

That night was my real low point. Fortunately Marcus had the resilience to resist my mood and maintain a constructive outlook. Whatever happened, there would be a show. There WAS light at the far end of the tunnel, I just couldn’t see it at that point. And since the field work was my area of responsibility, I felt I was letting Marcus down. Another late night and early morning: beware of the demon tiredness. I know from hard experience of many dawn sessions that I need to get a few hours sleep during the day to recharge my batteries. Marcus tried to skip on this, but suffered at dawn, when he would struggle to keep awake. Anyone can do one dawn, the difficulty is doing five in a row. Topping up your sleep in the early afternoon is essential - it’s a boring bit of the day anyway. You wake up pretty groggy, but after an hour or two you can get working for the evening.

No recording Thursday morning. I bought a heavy duty battery charger - but it needed a good 24 hours to charge up one of these leisure batteries, so in the meantime I put it in to the local garage for a quick high-rate charge. Sleep - ‘knitting up the ravelled sleeve of care’. Copy files to another hard disc in the studio and check out playback with my software (Logic Audio). No problems - I’m thinking I must be missing something. Back-up to DVD. Picking up the battery from the garage, I realise it’s election day in some other world, so get out to vote and back on site (15 miles away) soon after 7pm.

It was a little worrying that the battery indicator was still showing low, but it’s just a rough indicator. We put out the mics again, with better wind protection, and it was quite dark when we finished. Marcus got on with cooking some pasta, while I started testing the lines. Problems. And the battery was low: shutdown. I never did trust these quick chargers. Back to assessing the options - can we run it from the main campervan battery?

Friday morning it ran for just over a minute then the system shut down - not enough power. The forces of entropy were clawing back again. Trying to make the best of a reasonable morning (there was some sun, though chilly with a gusty breeze trying to get going) we ran 8 channels through a single octopre from the main campervan battery and recorded 45 minutes. Just to feel that we were making some progress.

Back to filming and photography, making the most of a heartening dawn. Then later, since it was proving to be a reasonable morning with a rather good birdsong chorus, I did some stereo recording to DAT. The workmen returned at 8am, so we packed in and assessed the situation. The wind was forecast to swing round to the north-west, so we decided to pack up at Brada for the time being, take the night off and try Pawston (site 2) from Saturday. We were pretty run-down at this point, after a week’s hard slog with little to show for it; although things hadn’t gone our way here, we had a pretty good feel for the site. We just needed a break. The north-west wind would get in at Brada, whereas Pawston should be a little more sheltered. Give the battery a good 24-hour charge and get a second one, so we could alternate them - putting one on charge each night.

There’s a real sense of security comes from having a solid plan that makes sense.

PHASE 3: SOUND ASCENDANT

We arrived at our second site on Saturday soon after 5pm. Very nice - a few birds were singing and there was a feeling of calm, though the rumble of distant thunder was a bit foreboding. This was the most wooded site with a surrounding of tall, mature trees and we were aware that this could be a problem in terms of sound separation and how close we could get mics to singing birds. But it looked good for sheer numbers of birds (Brada had been a little less populous than expected), it was a secluded, private site and we had plenty of battery power, with the option of using mains if we ran into problems.

We were feeling a positive apprehension at the possibilities here when the distant thunder became a heavy local downpour. Well, at least we didn’t have the mics rigged out. But the shower was short-lived. We waited for a while to let things dry out and check that another wasn’t on the way, while we drew up a plan for mic positions. Rigging out took over 3 hours, the mics needing decent weather protection with the threat of showers, and it was well into darkness by the time we were finished; but we were better prepared now with a choice of torches and a better idea of the lengths of cable runs. Testing showed that it was mostly working fine: the low output of a few of the mics was a niggle.

During the night a light breeze picked up, which strengthened after dawn. Nevertheless we went into record at about 4:25am, though really the wind noise was too much. Since we didn’t want to disturb the birds by switching on lights or drawing the curtains, much less going out and walking around in the half-light, listening down the mics was our only way of monitoring the bird activity. This proved to be a densely-populated site and revealed the difficulties of monitoring mobile individual singers down a multichannel system. I had a small mixer which enabled me to set up a rough stereo mix and solo to channels. Keeping track of individual songs was a matter of watching the led levels on the recorders (if the bird was close enough to a mic) and a lot of switching on the mixer, soloing channels.

The basic practicalities of keeping a check that the system was running OK, noting any problems to be sorted out later and trying to get an idea where individuals were tending to sing, meant that it was impossible to also note down much observation on the activity. It really needs several people, each with their own monitor mixer, keeping track of a few channels each and maybe dictating notes to a scribe; but it would still be very difficult without being able to back this up with viual observation of activity and movement.

The mature trees at this site did prove to be a problem and it wasn’t just in their height. The large volume of space available to birds within the network of branches and the outer foliage, offers a large area for movement and also works as a reverberation chamber to ‘degrade’ the direct sound-signal from the singer. Although there was a good volume of sound coming from quite a large band of singers here, making up a full chorus, there was very little clear separation of singers to individual mics; most of the birds were a little distance from any mic and being picked up by several mics. And looking at it from the other side, most mics were picking up more than one singer.

We did have the satisfaction of recording a big chorus on a decent number of channels across a large sound stage. After a while we stopped the recording and, now that it was light, went for a walkabout to get a better picture of the local bird community. This was the one site that was new to me, though Marcus had visited it earlier in the year. The east side of our recording area was bordered by a lovely mixed wood in the flush of spring, with a carpet of flowering ramsons and bluebells - full of the fresh, damp, musty scent of a northern spring dawn. And a surprising number of male blackcaps.

Further recording passages were interspersed by farm activity, photography sessions and a walk round the wood with the landowner. Back home to put put this morning’s battery on charge, get some sleep, copy and back-up the audio files, deal with any pressing correspondence, then return to site at tea-time. And this became my daily routine for the rest of the week.

After a walk up to visit a wooded lake with Tom, the landowner, it was back to base. The evening was sunny, very calm and still, if a little cool in the shade of the oaks. We wanted to change some of the mic positions, but before that recorded 15 minutes of the evening chorus. We re-arranged the mics to try and encompass a small paddock that extended down one edge of the wood; and we changed some of the improvised mic weather-proofing (styrofoam packing and cling-film) that was making unwanted noises when brushed even by a light breeze. All working nicely at 10:45pm; what’s more, it’s calm and clear. Just maybe ....

Monday 4am it was still calm, though cold, as the under-lying airflow was swinging round to the north. 4:05am and we’re into record mode, running a continuous 65 minutes to 14 tracks. It’s a decent chorus as well, though going a little quiet later on because of the chill. By the time the finches and warblers we’re going, with the sun now warming things, farm activities were starting up. Another walk round the wood on a glorious May morning; then we made a few more recordings between 6:45 and 9am, but the breeze was gradually picking up.

With the lack of individual separation at this site, an over-abundance of high song-posts and the wind swinging round to the north, we decided we’d be better moving on to try out our third site at Ford Moss, which was more sheltered from the north. Sod’s law: we’d put two mics out by the entrance gateway where several birds had been singing the previous morning, but not this morning - until we started derigging: first a dunnock came to one, then a wren came to the other and delivered just where we’d hoped. Oh well, new site, new opportunities. But first, food, sleep, backing-up, the daily chores.

I’d tried to keep some diary notes on the project, but they’d been growing thin. They run out entirely at this point. There just wasn’t time to do everything. And all along any time that as going spare was well-spent observing the goings on of the bird community at whatever site we were working - this is how you work out where to put the mics and is all the difference between getting serious recordings and just recording. Usually it also leads to a deeper understanding of what the recordings represent. I call it ‘tuning in’ to the place. The awkward irony is that, if the weather isn’t conducive to recording or wildlife display, there’s usually time to reflect and write up the diary; when the weather comes good, there’s too much to do.

Now we were in much better spirits since we had one good dawn chorus in the bag, even if it wasn’t ideal for extracting the models Marcus needed. Monday was a day of glorious sunshine and it always feels a bit strange and wasteful going to bed in the middle of a sunny day. Ford Moss is a favourite spot of mine: the topography, the scattering of gorse, heather, bracken and hawthorn, the patch of mature woodland with shapely mature oaks and scots pines all give a sumptuous sweep to the eye. It’s set back from a very quiet road, with access up a track which we could use. As I drove out in the late afternoon sunshine, I felt very optimistic for the first time.

We were unsure about mic placement because of the many potential song-posts in the sweep of hawthorns and gorse scrub and the plantation edge behind us. We were inclined just to observe the singing behaviour of the bird community the first morning; but, as the wind died to calm through the evening, I was desperate to try some recording, so rigged out a spread of our six best mics to take in the ambience.

We thoroughly enjoyed the evening activity here, listening to cuckoos, then roding woodcocks, barn owl, tawny owls and finally a female long-eared owl somewhere off in the plantation. As I recorded several sequences between 4 and 10am the next morning, we tried to get a grasp on the use of the area by the bird community. I was really pleased when Marcus came back beaming at one point after a close encounter with a couple of roe deer grazing among the bluebells and bracken shoots. Stuff like this is the bonus of being out working quietly in the early morning.

We set out the full rig of mics that evening and enjoyed it for a change; the weather was good, we had the place to ourselves and the cable runs didn’t involve too many encounters with thorns. With a clear sky, as last night, the temperature quickly dropped once the sun had set and there was a light frost the next morning. It was just about dark when we finished, as usual, but we recorded for a little while, to try to capture something of the evening. Searchlight beams sweeping around off to the east had puzzled us. Was it someone out poaching or was it the estate keepers lamping for foxes? Later, when headlights appeared from further up the track, we found out it was the keepers and we had a bit chat before they went on their way.

Going into record in the pre-dawn darkness at 3:50am I could hear a distant redstart from the old oaks and a skylark somewhere out on the moor; a cuckoo called out on the Moss, then a robin song, another, a blackbird somewhere in the distance and a song thrush and the chorus was gradually building. We recorded for 65 minutes continuously then stopped. We were a little disappointed that the several local blackbirds hadn’t put on much of a show. A song thrush had sung almost continuously by our front pair of mics, but the garden warbler which had used those bushes the previous morning had moved a little further off, outside our area.

We had a break and a wander for some filming and photography, then we entertained visitors. Pippa and the new director from the Baltic dropped in to see what we were about. We recorded some further sequences between 6:30 and 10am, while the sun was warming up; this tends to be a good time for birdsong, not strictly the ‘dawn’ chorus, slightly more spacious and relaxed than the surge from dawn, more charcterised by liquid meanderings of the warblers and finches.

As the air warmed, the day’s breeze gained in strength. It had been a lovely morning with a rich variety of birdsong developing in phases throughout the larger area, but yet again we were a little disappointed with our recorded sequences. Although we had some nice passages with a variety of birds cleanly recorded on individual mics, there didn’t seem to be any passage where there was a good number of individuals singing at the same time, recorded cleanly and discretely; and a few of the birds that we’d hoped for, and had been present the previous morning, had not performed well or had not been close to mics - blackbird, garden warbler and linnet.

The wind was continuing to swing round from the north to the north-east; we’d recorded a pretty good chorus from this site, but probably not ideal for what we wanted to provide a workable score for Marcus. This site was more open to the east, so we faced a difficult decision - whether to continue for at least another morning here or move back to our first site at Brada Quarry. As well as Ford Moss being open to the east, we felt that we faced an additional difficulty here in the extensive, scattered scrub at the site with a surfeit of songposts, which meant the birds were very mobile in their singing behaviour (also an intrinsic characteristic of many warbler species). The lack of a good blackbird chorus this morning was puzzling and we felt that blackbirds were an essential ingredient in a chorus that would be generally representative of Britain. In principle Brada still offered the best prospects for what we were after. Back to Brada it is.

My memories of the rest of this week are dominated by the sunshine of spring coming on strongly. That evening Jaime arrived from Bristol with his cameras to do some filming of the project. It was a fine morning with a clear dawn leading into a warm, sunny day. Although it was fairly calm, the air flow was still from the north-east, drifting to east, so there was a soft roar coming over from the sea a mile or two away. Our old friend the whitethroat was still at full throttle. In the unpredictibility of the natural world, a sedge warbler spent the morning wandering hidden in the scrub in the centre of our area, pouring out its erratic rhythms and mimicry. This wasn’t the usual habitat of sedge warblers and sure enough he’d moved on the next morning. And we had the base of blackbirds, robins, a song thrush, wrens, linnet, yellowhammers and swallow now, in an abandoned hut, plus other incidental vocalists.

Things were flowing quite smoothly now: alternating the two batteries was providing us with reliable power, the weather was good for recording and encouraging the birds to sing and we had 14 channels in operation. It seemed that now we had a pretty good chorus recorded from Brada, though we could still do with more individuals cleanly recorded. Proceedings this morning had been slightly disrupted by considerations for the filming. After the mid-day maintenance and recovery period, we assessed the situation in the late afternoon sun.

We adjusted some of the mic locations and even decided to experiment a little. The mobility of the birds was proving to be our major problem: very few individuals of any species were singing for very long at any one spot. So, once it was reasonably light, (so as not to disturb the start of the chorus) Marcus would go up into the wood and man the highly directional MKH70 mic, while I would slowly move along the roadside bushes recording to DAT with a parabolic reflector, having synced up the start time with the main recorder. This way we could make sure that these two tracks were focussed on individual birds and even follow them, if feasible. We recorded for a continuous 70 minutes from 3:50am; then another 65 minutes in several passages between 6 and 10am.

This was a great morning - calm, quiet, sunny and full of bird song. In fact the activity was so intense at some points that it was impossible to keep up with what was going on; there had seemed to be three male blackbirds singing really quite close together at one point. The song thrush seemed to be flitting about and popping up to deliver a verse or two all over the place. But, unlike yesterday, none of the local blackcap males sang in our arena (as we’d come to think of it) until later in the morning. We could hear a willow warbler singing for much of the time, but never close to a mic. Later we realised the female was busy building her nest just a few metres in front of the campervan, very trusting of us watching so close.

On my walkabout, I realised that a young peregrine falcon had spent the night on the crags of the main quarry, just round the corner from us. The whitethroat was less obvious this morning and disappeared for a few periods, possibly up the road to the far end of the quarry site. My theory was that he’d just arrived the week before and was singing strongly on the core of his territory; but maybe he’d failed to find a mate and, though still attached to his main patch, was having to range further afield to try to pick up a female. A blue tit took a liking to my expensive Sennheiser wind-sock and returned several times to try to pull out lumps of the synthetic hair, presumably to line a nest.

We’d come a long way from the angst of the first few days here at Brada - it was actually fun this morning. But the round-the-clock work had taken its toll and we were pretty low on energy reserves. We were unsure about trying to record at Brada over the week-end, since there are a couple of large caravan sites nearby and the quarry was sometimes being used for unofficial raves, general partying and messing with motorbikes. We thought we’d got a pretty good recording this morning; after talking things through, we felt we’d got about as far as we could and decided to wind up the sessions at this point.

WIth relief I was looking forward to a proper night’s sleep, comfortable in the misguided notion that the hardest part had been completed reasonably successfully. We had over 12 hours of recordings, varying from 6 to 15 tracks between sessions. Saturday was spent tidying up the equipment and setting up all the files for playback, checking that there were no technical problems while we still had the recorder (which, incidentally, had worked flawlessly). The weather was still calm and on the Sunday morning I was drawn back to Brada to do some more recording on my own; I recorded some of the individuals with my reflector, in case we needed to do any patching up, and some stereo ambiences, just because.

PHASE 4: EXTRACTING THE SCORE

Other projects took up my time for a while and it was July before I could continue with Marcus’ stuff. After setting up a rough mix, with the stereo spread corresponding approximately to the mic placement on site, the choruses sounded very nice - when listened to as choruses. Our next task was to select a passage with as many individuals (from a good variety of species) recorded cleanly and discretely on different channels: after a week of listening and noting timings and individuals on different channels, I appreciated the enormity of this task. It’s a very slow process: you listen to short passages ensemble, then go back through the same half minute or so soloing different channels. With music this sort of thing is relatively easy: you know the hi-hat is on a particular channel corresponding to the mic placed close to the source. With mobile birds, often singing a little way off from any particular mic and spilling through to several channels, this was much more involved.

We tried some experiments slowing down recordings of different species. Some species such as robin and wren needed to be slowed at least 16 times to bring their songs into the range of a human voice; this meant that any individual singing for a full 10 minutes would produce a 160-minute model for a human singer - a tall order. It was also clear that, although any slight overlap between two birds at normal speed didn’t cause too much difficulty in following one individual’s pattern, at the slower speed it was generally impossible.

We quickly rejected the chorus from the more wooded site at Pawston since there was just too much spillage and overlapping between tracks. The choice was between Ford Moss and Brada: Ford was more spacious with reasonably good separation, but not much blackbird song, whereas Brada had a really good representative chorus, though it was difficult to assess the separation without actually going through and working on the tracks.

In the end we went for Brada mainly because of the quality of the ‘turdine’ chorus; the turdidae make up the thrush family, including blackbird, song thrush and robin, which are the main singers in the real dawn chorus. At the peak of the first wave of the last Brada chorus there had been three individual blackbirds singing in our arena. We also decided that, rather than simply use one 10 minute section, which would show limited dynamic development, we’d use several shorter sections from different times through the full morning chorus beginning at dawn, which would show more variety and reveal the changing phases of the full dawn chorus.

It might have been easier if we’d used Ford Moss, since it proved very difficult to derive clean tracks from individuals (the wrens get everywhere!). We had to resort to our bottom-line safety net and substitute cleaner songs in some places, trying to match the original with a similar verse of the same bird from elsewhere in our recordings. I’d known all along that what we’d aimed for was an ideal and we’d have to do some substitution in the end. But what an ideal - to reveal the full live interaction of a community of birds singing counterpoint, within and between species - the rich interplay of shared phrases, mimicry and assimilation producing the subtle, dynamic evolution of themes that characterises a real dawn chorus.