Last week I found myself on the crest of the Moytura ridge, on the east side of Lough Arrow in County Sligo. It had been raining for hours, on and off, sometimes heavier, sometimes only light drizzle blown on steadily by a wind from the west. The ground was almost swampy in parts, so the twenty minute walk up-hill from the roadside had soaked my trousers six inches from the soles of my shoes. My hands were numb from setting up microphone stands, adjusting microphone positions, fiddling with digital recorder levels, and listening, always listening, to the sound of the wind on everything it touched.
This was my second trip to the same spot, the first being a reconnaissance the previous day. I had been led on that occasion by Kaspar Aus, a dance and video artist from Estonia. He's in the neighbourhood choreographing a new work. And he'd found a strange sculptural metallic form of that was singing in the ever-present wind. So he'd called me up from Limerick for three days to do some field recordings, something I generally only do for my own compositions. But I am also happy to oblige other intriguing projects. After all, every new recording is a learning experience.
Wind is normally the enemy of field recordists, right up there with cars and passing aircraft. The problem is a simple one: we want to record a sound of interest, sound waves being longitudinal compressions of the air. But we do not wish to record wind noise, this being the bulk movement of air itself. Unfortunately, wind commonly has lots of energy relative to the sounds we desire, and so tends to overload the microphone. The solution is to employ wind-shields of various types, the simplest being a foam fitted covering, really only sufficient for an interior setting. The next line of defence is some sort of a faux fur cover, which I employed on this occasion.
Since we are generally trying to muffle wind noise, it's a tricky proposition to actually record the wind itself. A bit of a paradox! However, what we usually mean by "the sound of wind" is the sound of the wind's effect on some other object, like long grass, or the leaves and branches of a tree. In this case, the wind was fluting down through hollow metal tubes that made up the ambiguous structure in the middle of the boggy field. This had a definitive pitch, with one bold overtone and likely others I couldn't hear as well. The sound one experienced when sitting inside the structure was quite hypnotic. Kaspar wanted this recorded so some experimental musicians could improvise/compose around it.
It would have been relatively easy to filter out a good deal of the wind, using a few methods. First, the microphones could have been put inside a larger superstructure, commonly called a blimp. This creates a region of dead air into which the wind has difficulty penetrating. But the more you put in the way of the microphone to cut wind, the more you lose other frequencies as well. Another problem was the result of my recording technique, using a spaced pair of omnidirectional microphones. This would have made for a rather large and bulky blimp setup, which is not my style. Second, one can apply rather severe equalisation in post-production, especially if you know which frequencies you really want to keep.
But I find such recordings unnatural. The strength of the wind itself as it gusts against the metal is an important part of the visceral experience. So instead I used three simple and obvious techniques. First, I employed omnidirectional microphones which, by their very nature as pressure-sensitive devices, are less wind sensitive. Second, I engaged the low-frequency cut-off filter, since most of what exists in that range is wind energy. And third, I positioned the microphones so the structure itself protected them from the direct force of the wind rising up over the ridge.
On my first visit I made two recordings and on the second three, one utilising a different (inferior) microphone approach. The very last recording did the trick, which at least saved me from another visit to the same site. I could have saved more time except for a rather foolish error. Given the rain, I decided to wrap the recorder in a plastic bag, the end of which I secured around the XLR cables exiting to the microphones. Unfortunately, each time the wind gusted, I unwittingly record a very prominent "Oh look! there's a plastic bag in the wind" sound. Next time I'll bring a heavier bag for the task. And also a flannel to wipe down the gear after each recording attempt. There's no point letting moisture build up on expensive electronics!
I let each recording go for about half an hour, knowing that I'd need to edit off the first and last few minutes as I moved in the vicinity of the mics. I like to get a good long sampling of any ambience, so that the stochastic fluctuations in sound events can play out over a naturally extended period. Looping a shorter recording does not have the same effect (though certainly it can present a convincing illusion).
The first thing I did on returning to my basecamp was ensure I had a safe backup of all files. This entailed plugging the right sort of USB cable into my laptop and copying all the files from the digital recorders, a process that takes far too long. Next I copied these from the hard drive to a USB stick. Then I erased the recorder for the next session, knowing I now have two safe copies. (If I had plentiful flash disks I'd keep a third copy safe that way, as well.)
Back in the studio I did some simple equalisation to reduce the very low frequencies and give a slight boost to the regions where the tones of interest occurred. I down-sampled from 96 KHz to 44.1 KHz and likewise reduced the bit depth from 24 to 16, resulting in a typical CD-quality file for upload to SoundCloud. Of course I also keep the full original files for applications where storage space and transfer speeds are not issues.
Here is an edited version of the result. I hope it's of interest, and that this article has explicated some of my working methods. I trust it has also demonstrated some of the considerations I go through in order to preserve essential qualities of a given sound source.