In October, 2002, I visited the noted composer and bioacoustician Bernie Krause in his personal studio, located in the Northern California wine country. The visit was part of my research for an article on field recording, called "Going Wild". Krause had just published his marvelous introduction to field recording, Wild Soundscapes (Wilderness Press, 2002), and I wanted to gather some background material about his work, as well as dig deeper into his recording and editing techniques. However, only a fraction of the interview was used in the feature. Because he offered so many interesting insights into the world of bioacoustics and phonography, I've decided to post our interview in its entirety.
Why did you begin recording natural sounds?
There are several reasons why I got into this. One is because I found that the natural soundscapes, particularly in North America, were disappearing very quickly. Fully 30 percent of my library from 30 years of work comes from extinct habitats. Habitats we can't record anymore, because there's no natural sound.
Another, more important reason is because listening to natural soundscapes, through this kind of kid-simple gear, has taught me how to listen. When you listen through headphones and a microphone, the first thing that you notice is how much noise there is in our environment, and how much it proliferates everything we want to hear. And the second thing is that it teaches you how far you have to go to actually get into a place where you can hear natural sound.
The third thing is that, when you listen to natural sound, it changes everything about how you perceive the world, in the sense that it changes your time frame about things: you don't get information in quick cuts. You get information over an extended period of time, because the expression of sound takes a very long time to establish. Often, bird calls are 45 seconds or longer. A theme from a humpback whale may take you 20 minutes to half an hour to get. So you have to sit there and listen to it and abide in silence, while that occurrence is happening. Because the other thing is to shut the hell up. That's what you have to learn to do.
And so it teaches you a different way of experiencing the world. Not from the visual perspective, because a visual perspective gives us very little information. We think it gives us a lot, but it doesn't. It gives us only the trunk of the elephant. Even the expressions that we use are visual. One of the reasons that it helps us learn to listen is because microphones don't discriminate like our brains do.
Those are the reasons why I record: it teaches me critically how to listen, and it gives me a chance to live a certain way, in a different kind of time frame.
Are there any well established collections of nature recordings?
The Cornell one is probably the largest library of natural sound. But they've primarily focused on individual creatures. And to me, that's a kind of 19th-century approach to science: Abstracting things from a context, where the context is able to inform people a lot more about why these creatures vocalize, how they vocalize, and how they learn to vocalize in the first place.
What happens is that we tend to deconstruct the natural world around us by taking portions out of context. For example, taking a robin or sparrow out of context. It's like taking apart a Beethoven symphony and just examining the strings lines or the horn line. And you listen to it and you say "Well, that's really nice," but it doesn't do much to tell you what the whole piece is about. And that's what we do.
You seem to be the person that came up with the term biophony. Are you the first to have begun recording in that way? So if I went to the Cornell collection, they wouldn't have a lot of recordings of biophonies?
They have more biophony recordings than ever, because of the work we were involved in. We started in the 1960s with this stuff.
Actually, it all came about because, when Paul Beaver and I got together to do our first record for Warner Bros., In A Wild Sanctuary, the mandate was to go out and try to find a way to connect the sounds of the natural world to a synthesizer and to do an album on the subject of ecology. And I'd never been outside before.
It all came about at a very propitious time because Nagra had just come out with their new stereo portable recorder. And that allowed us to go out into the field and record in ways that we couldn't have done before. Also, new mic technology had come out at that point to allow us to record in the field without a lot of humidity and wind problems, although the mics were very sensitive. At least we could go out there now and record, with a wonderful dynamic range.
So there was an explosion of technology that occurred in the late '60s that allowed us to do this kind of work. And little by little, we put together pieces of that puzzle to do it better and better.
How did you decide where to go first, or what to record first in the wild?
That had more to do with the limits of my imagination and knowledge of recording in the field than anything. To me, the wilds were Muir Woods or the San Francisco Zoo. As long as there was some separation between me and a creature, that was deemed pretty safe.
I recorded at Ocean Beach, because I could probably get some of what I called "sea gulls." There is no such thing as a "sea gull," but there I was recording "sea gulls." I didn't know much about it and was terrified.
As I explained earlier, I wasn't raised around animals and never spent very much time outside: I was a city boy. If there wasn't a window between me and the world outside, I didn't feel particularly safe. And I think most people feel that way.
So, In A Wild Sanctuary did several things for us. One, it introduced us to technology that was portable and we found we could carry our recorders around with us rather than just have them in the studio. And that got me outside. From there on, it just exploded for me. I was hooked: I wanted to learn more about what I was doing, and I ranged further and further away from cities.
And now I'm to the point where I live here surrounded by 10 acres. We've been here 10 years and I don't think I've been to the city a half dozen times. I have no interest in it anymore.
In Wild Soundscapes, you talk about how quiet it was here after 9/11.
We're directly under the descent route to San Francisco International Airport for planes that come in from Europe and the north (such as Seattle, Portland, and Boise). They fly over at 15,000 feet and make a lot of noise.
Then there are the tourist flights. On any given weekend, and often during week days, there are light planes flying up and down the valley. You can't record for a minute or more without a plane flying by. It's very quiet today because of the weather, but after 9/11, there wasn't a thing in the sky. Not even a helicopter.
We would sit outside and be amazed in September about how much wonderful sound there was that wasn't masked or obliterated by all this mechanical, human noise around us.
I followed your book's examples about listening through mics to the environment where I live, and I was shocked at how noisy it really is, even late at night.
What's fascinating about what you're saying about there being so much noise and we're completely unaware of it — we shut most of it out — is that people like me are afraid of technology, so they won't use it, particularly when it comes to recorders. They have no problem using a video camera or still camera, or any of the new digital cameras or anything like that. But put an audio recorder in their hand and somehow they're stymied, because we're a visual culture. They have an affinity to anything visual, really quickly. But put a recorder in their hand, which is just as simple, and they block: they cannot figure out how to push the Record and Play button at the same time to record.
But when they have been out and done exactly what you did, their lives have changed. It profoundly affects them because it gives them something to think about that they haven't thought about before.
One of the things that I do with kids in schools is to give them an assignment to go out and record a robin. Okay, simple: everybody knows what a robin is. And they have to record it in their own backyard. I tell them they can get a little lavalier microphone from Radio Shack for $14. They can put all the equipment together for under $50 — really inexpensive. And then just go out and work as a team, and share it among five or six people. Go out and record a robin.
And sure enough they hear the robin chirping and they record it, and they bring the tape back to me and they say "yeah well, there's a robin in there, but there's also a bus, and a helicopter, and an airplane. There are people talking, and I hear a television, and a radio, and a booming car; and 808's going by." I said, "I want a robin. That's what I want to hear. Now go and record a robin for me." And it changes their mind. "There's too much noise. I can't get the robin." Well maybe you have to go someplace else to get the robin. Maybe you have to go to the park, or to the woods.
When I started in 1968, it used to take me 14 or 15 hours of recording to get an hour of usable material. Now it takes me a year to get that same amount of material.
When you get back to the studio, do you use the takes that don't have man-made noise, then crossfade that into another section that doesn't have man-made noise? You would have to match the time of the day and the biophony.
Sure, and I have to record over a period of many days.
Is that where the art form comes in? In matching those scenes?
Yes, and it has to be seamless. The problem is that there are recordists out there that claim they're purists. My answer to them is that, of course, if you choose a microphone system, you've done some editing. If you choose a recorder, you've done some editing. And the time of day, location, where you point your mics — all editing.
Then you've only got 74 minutes on a CD. Which 74 are you going to choose? So these guys and gals have done all the editing. Where's the pure?
Krause in his studio, October, 2002.
What do you think of the style of nature recording that was popular in the '70s?
It doesn't sound very good: it's technically not done very well. The stereo imaging is not very good. The depth and dimension is not very good. There's no indication that the recordist knows anything about the habitat they recorded in. There's no information in the notes about what creatures you're listening to.
All of our albums, and those of Lang Elliot and Ruth Happel, both of whom are fine recordists, are produced with the understanding that there is something we need to know about, beyond just the fact that it's an alpine wood. What's there? What are we listening to? Consequently, listeners have more information on the new titles that we're putting out.
And it's really important that recordists pay some attention to that kind of information. They'll record at one site in a national park. Well, in Yellow Stone alone, there are probably 70 different types of habitat. And they've recorded one, and they call it "Yellow Stone National Park."
We've just come back from Sequoia National Park and Kings Canyon.
How much tape does that translate to: one, two...eight boxes of DATs?
It's a year's worth of recording at four sites, four different times of day for an hour at each site, over the course of a year. Let me tell you, we haven't got anything!
I was talking to one group that was doing a study (a group that we were commissioned to work with): they've got 26,000 hours from one place. Now you're getting the point.
What's represented there [Points to boxes of DATs.] is 80 hours worth of material from a Sequoia National Park project that we did last year for the National Park Service. For a lifetime of work — almost 35 years now — I have 3,500 hours of material. Compare that with 26,000 hours from one park. There's a lot of work to be done out there.
What is your current setup?
I record M-S, and then I transfer it to stereo through a matrix. I use an analog matrix because I like its continuously variable capabilities. Also, I like analog sound: I don't like 44.1 kHz. So I immediately convert my stuff out of 44.1 kHz on the DAT through some kind of analog process. I have to convert back to digital to get the CD, obviously.
You'll save that entire archive of tapes, even though there may be only a couple hours of useful stuff on there?
Sure. Because it's an indicator of the dynamic of that particular habitat.
Let's say you record a dawn chorus. The loudest part of the day during the spring, summer, and fall is, typically, during the dawn chorus. As soon as the sunlight hits, the creatures start to warm up their voices and sing. So there's this huge envelope of creature-sound density that occurs in the morning. And it falls off during the daytime. There's another one maybe an hour after sunrise that comes up, which is a secondary dawn chorus, and then it levels off for the rest of the day, depending on what the temperature gradient is. There is a dynamic there.
Then there's an evening chorus and a night-time chorus. So there are four different types of dynamic that you want to get during each session. It's a lot of work. And you only have one day's sample. Then you have to go back over several days to see what the layer and texturing is of those particular sites. We had four people working these four sites.
The base noise level, or noise floor, seems high on some nature recordings.
We don't use noise reduction on our recordings. And mics don't discriminate like our ears do (at least most mics don't discriminate). When you go outside, and your ear tells you you're in a relatively quite place, the mics will pick up ambient sound. They'll pick up wind through the leaves and the boughs of trees. They'll pick up a stream from a 1/4 mile or 1/2 mile away, even though it's a tiny stream. They'll pick up all kinds of stuff that you're not sensitive to. Even when you put on your earphones, you don't really hear it. It's not until you get back to the studio, and it's really quiet there, and all you hear is this damn hiss in the background. And you try and figure out where it came from.
We are trying to use equipment that has the lowest signal-to-noise ratio as possible. And for that reason we only use Sennheiser mics because, when the AES did a study on the noise floors of different mics, Sennheiser came out the best at that time. We've not used anything else since.
Another reason we use Sennheiser mics is because they've got a really nice dynamic range. And we can take them into habitats that are very humid, and they don't fail like some of the other mics do, like the ones from AKG and Schoeps.
We have all of those, but we don't use them very often because, unless the conditions are perfect outside — humidity under 70 percent, for example — they don't work very well. Also, the Sennheisers are a little bit less sensitive to wind noise then the others. But we still have to use wind screens and other equipment to quiet it down.
Which Sennheisers are they?
For most of my systems, I'm using the figure 8 MKH 30 and the cardioid MKH 40. They're mounted on top of one another. There are other mics out there: Sanken has a really good M-S system. Other people like to use other technologies, such as the Crown SASS-P system. I'm not too fond of that, and I'm not too fond of binaural because you've got to put on earphones in order to hear it.
The other reason I use M-S is because I get several options out of one recording. First, because of the mid mic, I'm able to get some directionality out of it. So, I'm able to get the single creature I'm aiming at.
Second, I get the figure-8 perspective, which gives me the instantaneous surround. That's a separate mono channel that I can use. It gives me the ambiance of the recording, minus the front and back.
The third thing I get is a wonderful stereo combination without the 3 dB level drop in the center channel, which happens when using an x/y or an ORTF stereo miking, and with most binaural systems. With the M-S system, the mid mic fills in the gap and gives you a much more robust stereo image than x/y, binaural, or ORTF.
The fourth thing is that I'm able to encode the mix signal into surround, if I don't do any processing to it, with any number of systems. We encode ours with surround data which works in almost any surround decode format: 5.1, Dolby Digital, and any of the others.
The last is if you don't do any signal processing, such as compression or heavy EQ, you can convert the stereo signal back into M-S again.
By running it back through a matrix?
Yes, you get the same results. Let me add that M-S is the least intuitive of the systems to use for recording. You have to learn to use it because the M-S signal is very confusing when you hear it in your headphones, unless you have the right kind of preamp (a preamp with an output on it that allows you to listen to things in stereo). Otherwise, you get the mid signal in your left ear and the side in your right. And it's very confusing when you're seeing things go by and you don't hear the spatial modulation that you would in a stereo recording.
But some M-S mics have a built-in decoder.
Some do. I have one here, the Sony ECM-MS 957. It's an M-S microphone that gives you a stereo result. It's converted within the tube to a stereo signal, and the stereo signal is what you record. If you wanted to reconvert it back into M-S you can. The stereo result from an M-S mic will sound significantly different than an x/y recording, and it can always be reprocessed and returned to an M-S signal.
There are preamps, like the Schoeps (although it's too damn noisy) that gives you an option of going right to stereo. Again, you can take your signal and put it back into M-S again after you've recorded it.
I use this now — it's from Sound Devices — and it is a great preamp. Now they have a preamp that has a headphone output that you can listen to things in stereo. But this offers just straight M-S.
You've glued down some of the switches?
Yes, I hot glue them. What happens is that when they're in the packs or in your pocket, the switches often move. I wouldn't know if they change. So I just glue them into position. You can pull the dried glue out quickly if you need to.
What else do you pack besides the Sony M1 DAT machine and the Sound Devices M-S preamp?
A gel cell, which is a battery that lasts me 18 hours of recording, powering both of these guys. This is a 48V version, or course, for the Sennheiser mics. I can run 18 hours with one battery charge. Plus my mics, 10 meters of coaxial cable, a Rycote zeppelin for the mics, and an aluminum tripod, which doesn't weigh more than a pound.
That gives me a month of recording. I can hike back and spend days out there.
[At this point, Krause takes me outside for a listen.]
I hear a chainsaw. The white noise you hear is the stream. That would be the noise floor you would hear in a recording. When I did the recordings in Sumatra, for instance, I must have hiked 15 miles. I could never get away from the stream noise. That's part of the geophony: it's part of the signature of that place.
I thought it was the sound of the mic and recorder.
The noise floor of those Sennheisers is down 85 dB, and you're not going to hear them. And the Sound Devices preamp is really quiet. It's down there with the noise floor of the Sennheisers.
When I use my Schoeps mics, the noise floor comes way up, because you hear the electronics. When I use my Neumanns, it comes way up. They're wonderful mics, and when you have enough signal you don't notice it, because you're comparing signal to noise.
But if there were birds singing right now, it would drive you crazy to hear that stream, because that's what the mics would pick up. If there was a breeze right now, you'd hear the wind in all of these oak trees blowing through the limbs or the leaves. Again, it's part of the ambience of the space, and there isn't any way to get away from it.
You get those perfect moments, if you live long enough and you stay long enough, and you're patient enough. You get those perfect moments where everything conspires to work for you. But 80 percent of the time, that doesn't happen.
[We return to the studio, where we listen to recordings of some of his projects.]
What are we listening to?
This is Alpine Meadow (Miramar Recordings, 2002). You hear some ambient noise, but there's no way to get around that. That's the wind in the trees and the pines, but this was also recorded with a different kind of M-S system. This was recorded with a Neumann RSM 190 i, which tends to be noisy as opposed to this. [He plays Midsummer Nights (Miramar, 2003).] There, it's just a normal ambience of the environment. [Then, he plays Douglas Quin's Madagascar: Fragile Land (Miramar, 1998).] It's very low noise: there are no planes or boats.
They all have different levels of ambient noise, and the ambient noise is very much a part of the signature. If you're using gear at the level of quality as these Sennheiser mics, you're going to get very low-noise recordings. We take a lot of care with this stuff. Each one of these albums took me 100 hours to mix.
When people go out there to record, they don't realize they're recording in one habitat, and it isn't representative of the environment, because it's so diverse. When Ruth Happel and I were doing the desert recording in the panhandle of New Mexico, it was a wonderful place. There's a ranch of 500 square miles. No traffic. We heard one plane in two weeks.
When you go out to record in that kind of habitat, you quickly realize that even in the high desert, there are probably 50 or 60 different kinds of microhabitats within, each represented by a different sound signature.
Krause recording a gorilla.
How do you pick the ones you're going to record?
If I'm thinking CD or if I'm thinking I want an expression of that habitat, I go for the location with the densest mixture of creatures that are vocal. I'm recording animal music. So I want something that's going to sound.
There are places within the jewel national parks, where you can go to hear natural sound unimpeded. As a matter of fact, the National Park Service is the only federal agency and the only group in the country that has designated natural soundscape as a resource. Because they're a federal agency, it means it must be protected like any other resource.
How do they make that happen?
Part of it is restricting fly overs, snow mobiles, and jet skis. These are rules that they implemented over the course of ten years.
One of the next steps is to establish noise-free zones. What that would mean is no human-induced noise can intrude for 15 minutes at a time. There's a manual from the National Parks called The Nature of Sound, that all of the superintendents have to read and enforce. It's all about soundscape.
Chainsaw noise travels 20 miles through forests. They're pretty loud. Motorboat noise, those diesels with the straight pipe, is a low-frequency sound that carries a long way.
How do you get rid of it?
I usually put a 160 Hz cut on the preamp, which the Sound Devices has. It has an 80 and 160 Hz low-cut filter. I usually put that in and it cuts out some of the longer waveform material. Also, most of the natural sound is above that, ranging from 200 Hz to 20 kHz or more.
Do you use parametric EQ for boosting or to remove audio artifacts?
Mostly to remove artifacts, such as low-frequency wind noise. There's often a low-frequency component to the material in jungles because of distant streams and stuff like that. Low frequencies carry a much longer distance through the jungle and it's just too noisy and too distracting.
When you're transforming this material to a CD, you're transforming it from one medium to another: from the natural world into another medium. And you better make sure it's listenable. That's the art of this. That's why so many recordings aren't very good, because people haven't taken into account that it's being transformed into another medium. They haven't done due diligence to consider how it's being perceived by the other person.
You may, as the recordist, be able to imagine the spot that you recorded it and it evokes an image in your mind. But the listener has no such capability of doing that because they weren't there when you recorded it.
Do you test your recordings on somebody?
I test my recordings on somebody all the time. Typically my wife, to see if she gets it.
Does she often give you the thumbs down on things?
Often. More often than you'd like to know. That's why I spend so much time in this damn studio and wear out so much equipment. She's good because she has a good sense of that stuff.
Tell me about your homemade hydrophone.
It has a frequency range of 3 Hz to 20 kHz. It's designed to work with water, and it matches the impedance immediately.
Sometimes when you drop it into water, it gets little air bubbles around it, which keeps it from matching the impedance, so it doesn't transmit the sound very well. To get rid of the bubbles, you put a little grease on it, like some olive oil. Go online and do a search on hydrophones. Hydrophones go down very low, and this one goes down to 3 Hz.
The neat thing about some of the digital Sony recorders, particularly the D10, is that they can capture infrasound down to 3 Hz. Sometimes you can use the hydrophone as a microphone. I used this in the sand dunes and I also used it to record elephant sounds, because they vocalize around 14 Hz. Giraffes, hippos, and elephants transmit sound through the air at very low frequencies, which allows them to communicate over great distances. Hippos do it under water and in the air. So this is what we use for that.
Do you just hold it when you record with it?
Yes. You have to hold it very still. And it's very sensitive, so you'll hear your muscle fibers firing.
This is one we used to record a tree. Every heard a tree singing? It's 70 kHz. This is a B&K 4103, which has a range of 0.1 Hz to 100 kHz. It's used to record dolphins. The dolphin is the highest frequency creature on the planet. Whales are the lowest on the planet. Their lowest sounds go down to 3 or 4 Hz, because they want to transmit over long distances. And the Ganges river dolphin goes up to 356 kHz.
Do you ever normalize recordings to bring up the level?
Yes. I sure do. After editing and just before I'm ready to cut a master.
How did Douglas Quin approach his underwater recording in stereo?
The Weddell seals that he did are all in stereo. He used two mics set about 10 meters apart. Where we would record in stereo with about 7 or 8 inches between mics, you have to place the mics about 10 meters apart to get a stereo image, or at least a spatial image, underwater due to the impedance and density of water, and the fact that sound travels five times as fast underwater.
How did you know to record the tree?
We were recording bats, and we had a bat detector, which is, essentially, a frequency divider. We were listening for the sounds of bats, which are up in the 47 kHz range. And we heard a steady signal, very unbiological in the sense of being a creature. As we moved closer to this cottonwood tree, the signal level increased.
We drilled a little hole in the tree and put this hydrophone in. We had an instrumentation device with us that could record a frequency that high, and we got a signal coming from the trunk of the tree. We couldn't figure out what it was. Then we slowed it down by a factor of seven, to get it down within our hearing range.
What we discovered was that, during a drought, the cells in the xylem of the tree usually maintain a certain pressure from the water that comes into the trunk during normally wet seasons. When that pressure drops during a drought, the cells automatically fill with air to try to maintain the osmotic pressure. And when they get too dry and they're pumping in air, they pop. When they pop, they die, and the dead cells form the tree's rings.
So, when they pop, they make a noise: we can't hear it, but insects can. And when insects hear multiple cells popping, they're drawn to the tree because certain ones are programmed to expect sap. And when the insects are drawn to the tree, the birds are drawn to the tree to eat. it's all a microhabitat formed by sound: The sound of popping cells.
You've also recorded ants.
People would never think that ants make this kind of noise. If you block an ant hole with a lavalier microphone, the ants will gather under it. And there will be one ant in particular that's directing all the others to dig the damn thing out, and get the obstacle out of that entrance. They're rubbing their little legs on their belly: they're not vocalizing. They're communicating with one another through this kind of stridulation, and it is sound. They're not communicating using pheromones, as most people think, but through sound.
There are different species of ants, and many have a unique sound. If you put your ear close enough, you can actually hear the sound. With that little $14 lavalier microphone, you can hear ants singing in your backyard.
How did you record the singing sand dunes at Kelso?
You kick the sand into motion — it's not the wind that creates the sound — and the dunes begin to resonate in a frequency range between 4 or 5 Hz and 120 Hz. It sounds like a freight train, and you can hear it 7 or 8 miles away. The reason they make sound has to do with the moisture content in the sand.
Do you have any words of advice for budding recordists?
Begin to explore the acoustic component of our world to see what it's really like out there. To the extent that people are willing to take the time to do that, they're going to discover a whole new part of themselves that they never thought of before, because we're so visual. We're not really tactile. We don't really use our olfactory senses at all. We're something like 99 percent visual.
That's largely because the technology for preserving visual stuff has been with us for a long time: we've been drawing on caves for 40,000 years. We haven't been able to capture sound until a little over a hundred years ago. And when we first began to do that, it came very late in our human development of civilization, particularly in the west. We've lost that ability to hear, because it's been put on the backburner, and it's been overcome by so much noise.