Read Next:

David Lewiston: The Father of World Music

He’s spent the last 40 years recording in Indonesia, Morocco, Peru, India, Mexico, Istanbul, Nepal, Georgia, Pakistan and more. He’s recorded everything from Balinese gamelan to Tibetan monks. Greg Simmons talks to the father of world music…


9 April 2006

With a career spanning 40 years of recording and releasing the music of other cultures, David Lewiston could be described as the father of world music. His pioneering field recordings of Balinese music from the 1960s gave many Westerners their first taste of the music of other cultures.

But what makes Lewiston worthy of such a lofty title (and one which he would probably dismiss as ‘pretentious’)? Field recording legends like folklorist John A. Lomax and his son Alan had been archiving the music of other cultures from the moment they were able to lift a record cutting lathe into a pick-up truck. Lewiston’s work is unique because it was the first to be mastered and released as entertainment for discerning listeners, rather than archives for anthropologists. Those before him mapped the relatively accessible musical terrains of Europe and the Caribbean, gently easing the West into the music of other cultures and opening the way for Lewiston to venture into the peculiar sounds and musical structures of Asia and the Himalayas – music that had evolved with little or no Western influence. His earliest Balinese and Tibetan releases provided a timely soundtrack for a Western world that was dabbling in Eastern philosophies, transcendental meditation and other consciousness-altering endeavours.

No discussion of Lewiston would be complete without acknowledging the contributions of the late Teresa ‘Tracey’ Sterne, the visionary head of Nonesuch Records who saw the potential of Lewiston’s first recordings and pushed to have them released, launching the famous ‘Explorer Series’ in 1966. In a working relationship that lasted over a decade, Lewiston and Sterne created and defined the genre we now call ‘world music’.

Greg Simmons: 2006 marks your 40th year as a field recordist. What’s the most important thing you’ve learnt during that time?

David Lewiston: The most important thing is not to bore or scare the musicians. You need a rig that allows you to set up quickly and unpretentiously, so you can maintain a casual attitude and create the impression that what you’re doing is simple and no big deal. Also, rather than hiding behind the musicians, you want to be among them, expressing your pleasure at the music. This encourages them to give a better performance. If a take is unsatisfactory, just keep going, and tell the musicians their music is really great. Then, later on in the session, tell them you really enjoyed the piece that was poorly recorded and that you’d love to hear them play it again.

You have to create a relaxed atmosphere conducive to music-making. I always ask a local informant what the musicians traditionally expect to have while playing, so they feel taken care of and comfortable. In some cultures it’s traditional to have booze on hand, but don’t offer too much – I’ve had musicians pass out and ruin the session!

GS: Let’s talk about your equipment. What microphones do you use?

DL: I have three microphone rigs, depending on the application. For street work I am enamored of the MS configuration, mounted in a Rycote windscreen. I use that rig for 80 to 90 percent of my recording. It’s great for recording anything outdoors and it works very well for indoor recording; say, a singer with harmonium, tabla and flute. I simply point the business end at the soloist, arrange the other musicians in a semicircle, and say ‘Ready’!

Lewiston at Tashi Jong monastery, 1998.

GS: What’s in your MS rig?

DL: You need a decent small-capsule cardioid mic for the M signal; I use a Neumann KM84. For the S signal I use a Sennheiser MKH30.

GS: Many people would’ve chosen M and S mics from the same manufacturer. Why the mixture?

DL: I wanted a bidirectional to use with the KM84, but one problem of bidirectional mics is that they have an inherent low-frequency roll-off. When I was putting together my MS rig I asked [microphone designer] David Josephson for his opinion and he recommended the MKH30. It uses internal EQ to deal with the inherent roll-off, producing a fuller and flatter response. I really like the results that the Neumann/Sennheiser combination delivers.

GS: …and your other microphone rigs?

DL: For formal recordings in Tibetan temples I use a pair of KM84s in the classic ‘crossed’ configuration, with the angle between the capsules either 90 or 130 degrees, depending on preference.

Both those rigs use condensers, and sometimes they don’t suit the music and circumstances. If the music is very loud it can overload condenser mic electronics (I’ve done it!), resulting in horrendously distorted and unusable sound. Also, in highly humid places like Bali, condenser microphones are problematic. So I keep a pair of Electrovoice RE50s handy for such situations.

GS: The RE50? But isn’t that an interviewer’s microphone…

DL: Right, it’s an omnidirectional dynamic microphone designed for hand-held outdoor use. Being dynamic, it doesn’t have electronics to overload when the music is very loud, and it’s not affected by humidity.

GS: Oh, of course! Also, omnis have excellent low frequency response, no proximity effect, and very low handling noise. The RE50 looks like a great solution…

DL: …and it’s easy to get a good sound using a pair, especially outdoors, holding one in each hand, fairly far apart. The only problem is how to hold the two mics in the appropriate positions and adjust the recorder at the same time!

I made all of my successful 1987 Balinese recordings outdoors, direct-to-stereo, with just a pair of RE50s. If you have access to Nonesuch’s Explorer Series album Bali: Gamelan & Kecak or the Bridge album Kecak – a Balinese Music Drama you’ll get an idea of their utility. For environments where volume and/or humidity cause problems with condenser microphones, the RE50s deliver reliably glorious sound.

…what could be better than travelling to incredible places, meeting amazing musicians, and recording them?

Dalhousie, Northern India, 1972: Lewiston, local informant, musician and monk setting up for recording.
Hemis Monastery, Ladakh, 1977: Recording Thukse Rinpoche, a highly respected incarnate lama, as he explained the significance of one of the monastery’s rites.


The dynamic range of these rituals can be extremely high, up to 50dB, so switch the recorder’s preamp sensitivity to low, ask the instrumental ensemble to play as loudly as possible so that you can set levels, then set the pots so that the modulometers read no more than –10dB. Watch out for distortion, in the mic preamps and in the microphone’s internal electronics, during the loudest patches.

A potential problem that needs discussing with the Khenpo (abbot) is that of unwanted sounds. Monks are accustomed to clearing their throats noisily during rituals, which ruins the recording. Naturally, the Khenpos want the recordings of their rituals to be as perfect as possible, so they understand the need for banning throat-clearing. Before recording starts, I make sure that all present are reminded about the ban. As the recording gets under way, someone will forget and hawk, spit, or cough noisily, at which point I stop recording. I let the monks know that their chanting was beautiful, but we have to start again because the cough spoilt it. They usually decide to leave a door open so that anyone with a tickle in his throat can tiptoe outside, cough as much as he needs, and then rejoin the session.

Monasteries often have several dogs, and their barking can ruin a recording. Also, novice monks are always curious about recording sessions, peeking in the windows and making noisy comments, which also ruin the recording. I discuss both of these potential problems with the Khenpo. Typically he will decide to have the novices drive the dogs away from the temple, and then stay away from the temple themselves, too. – David Lewiston

GS: I expect you’ve used many different recording devices. How about a brief history?

DL: Okay… There were no battery-powered stereo tape recorders to be found in New York before I headed out to Bali in ’66, so I packed a mono Uher 4000. Stopping over in Singapore I found, to my delight, a Concertone 727 battery-powered stereo recorder. It was well within my budget, so I took it to Bali with the rest of my gear. I used the Uher for a few recordings, but everything else was recorded in stereo on the Concertone.

GS: Those are the recordings on your first release, Bali: Music from the Morning of the World?

DL: Yes. That album was very successful. It was a sampler of what you could hear in Bali at the time.

When I left Indonesia to head West I sold the Concertone to a Javanese friend, who told me it broke down a few weeks later. Cheap crap! Then, in ’67, I was walking along Nathan Road in Kowloon [Hong Kong] and spotted a battery-powered stereo recorder from Akai. It was bulky and heavy but it was stereo, so I lugged it around South America for 18 months.

Lusting after a real pro deck, I saved my pennies and, in 1972, bought a Stellavox SP7 with a matching pair of T-powered Neumann KM74 cardioid condenser mics, plus my first pair of Electrovoice RE50s. I took that rig to India in May ’72, along with a stereo Uher 4200 for backup. By this time my so-called ‘portable rig’ added up to 13 pieces of baggage; I was forever counting it – in and out of hotel lobbies, in and out of taxis, etc. – to make sure nothing was missing! Returning to New York in February 1973, I added some bits and pieces to the Stellavox (the 10.5 inch reel extension plate and a custom-made clip-on Dolby A box), and used it until 1986.

Digital recording had arrived by then so I went looking for a new rig, settling on a Sony PCM F1 feeding a pair of Panasonic 8420 portable VHS recorders. I also bought a pair of Neumann KM83 omnis, a pair of KM84 cardioids, and a fresh pair of RE50s. The F1 didn’t provide phantom power, so I used a Stellavox preamp. I took this rig to Karnataka in Southern India to record the rites of the great Tibetan monasteries, then on to Bali, where it had endless problems. The F1 would have level drops in one channel from time to time… I soon discovered that a hearty slap would restore correct functioning. Sessions were nervous affairs; I never knew which bits of equipment would function properly.

By 1994 I had converted to DAT, using an HHB Aiwa Pro – excellent recordings, but at 48k sampling rate only. For field trips in the late ’90s I used an HHB Portadat. Amazingly rugged: recording in Tbilisi, Georgia, in 1998 I dropped it five feet on to concrete. I was expecting the worst, but it continued working without complaint. Its only drawback? 16 bits…

A year ago I plunked down the cash for a Sound Devices 744T portable four-track hard disk recorder. Brilliant design (it’ll do just about anything you want except pick out a really cute hooker), but the ergonomics suck. Most of the important functions are stored in menus, but the menu structure has 84 items (!) and, to my tired eyes, the LCD is nearly impossible to read. It took me forever to get comfortable with it.

GS: The 744T is amazingly portable, with many functions and some clever ideas. I particularly like the idea of powering it from camcorder batteries…

DL: Camcorder batteries are commodity items, of course, which is preferable to using exorbitantly priced custom batteries. But the recess for the clip-on battery is only sufficient to match relatively low-capacity batteries, rather than the 6AH batteries we need for field work. I get the impression they were thinking of sound recordists for film and TV rather than those of us who do field work. Oh well…

GS: How does it sound?

DL: Very good. A month ago I recorded a pair of Bengali musicians with it. I used the KM84s as a crossed pair just over the musicians’ heads; much closer to the stringed instrument (Hindustani slide guitar!) than the tabla to get the right balance. I ran them directly into the 744T’s internal mic preamps. I also placed two KM83 omnis close-up as ‘detail’ mics, using a Mackie 1402 as a preamp into the 744T’s line inputs. A very pleasing recording.

With the passage of time field work seems progressively harder and much more time-consuming.

Ubud, Bali, 1966: Lewiston and the Chokorde Gde Agung Sukawati, a very great prince and patron of the arts, after a recording session in his palace.


Recording in Bali calls for very unusual recording techniques. For everything there I use a spaced pair of Electrovoice RE50s, and record outdoors. The hardest part is finding the right space for recording – if it isn’t quiet, away from packs of howling dogs and noisy two-stroke motorbikes, the recording will be ruined.

The way the Gamelan [Balinese percussion orchestra] sets up for a live performance isn’t suitable for recording. After many unusable sessions, my collaborator, Cliff de Arment, and I came up with a setup that worked amazingly well. Think of the recording soundstage as an inverted ‘V’, with the widest part of the V nearest the mics and the narrowest at the back. We positioned the big gongs at the back of the soundstage. Then we arranged all the gangsa (single-mallet metallophones) on the right of the soundstage with the highest-pitched kantilan (‘flower parts’) nearest the mics. The left leg of the V was occupied by the reong (frame containing gong pots), with the gender (softer two-mallet metallophones) close behind them.

There are two things you don’t want to hear much of: the time-beater, and the chengcheng (cymbals). Both are placed behind the gangsa, so the mics don’t pick up too much of them. The two drummers are placed in the open part of the V, facing forward so that the mics pick up both sides of each drum – it’s difficult for them because they’re accustomed to playing facing one another, but if you explain its importance for recording, they’ll adjust. If there’s a suling (flute) or rebab (two-stringed instrument) player, he goes in this central space with the drummers. In such a setup the mics are separated by 18 to 25 feet, roughly six feet above the ground. The widest part of the V extends several feet beyond the mics.

The Kecak is based on an excerpt from the Ramayana, and is usually performed in the local Temple of the Dead. It consists of the cak chorus (typically 80 to 100 men) seated in concentric circles in an open courtyard, with the performers (a narrator and several singer-dancers) in the middle. The powerful rhythms of the cak chorus are derived from an ancient trance exorcism and provide the ‘motor’ for the performance. It’s important to get the right balance between the caks and soloists, so I string a length of 1/8th inch diameter nylon twine across the courtyard and hang the mics off it. They’re positioned about 10 to 12 feet above the performers, and separated by about 20 feet. – David Lewiston

GS: Judging by your methods for recording Balinese gamelan and similar difficult ensembles you’ve obviously had to figure out some unique recording solutions…

DL: I’m a firm believer in learning by doing. Even after doing this for 40 years I still screw up. For instance, I’m not at all happy with the recordings made in Georgia in ’98, and the recordings made in Jammu in the same year can only be charitably described as a disaster, unbearably distorted. After discussing this with Klaus Heyne, we concluded that the mics’ FETs had overloaded. To guard against similar problems in future, Klaus tweaked my KM84s and replaced the FETs.

In Bali in ’87 it took five months of work, lots of sessions, to figure out how to record big gamelans effectively; that’s when I realised that condenser mics weren’t suitable, and switched to the RE50s. Same applied to the Kecak: it took four tries to get something usable. What has this taught me? Patience. I like to allow enough time in the field so that I can screw up and then redo the work until the recordings are satisfactory. Jammu was a problem because I had only a week there. And Kashmir in ’98 was another disaster; 10 days in a city that had been turned into an armed camp!

It’s interesting, though… When I first started doing this sort of thing I would make very short visits to these locations, just one week in Kashmir in ’72, for example, and less than three weeks in Bali in ’66, and wind up with remarkable recordings. With the passage of time field work seems progressively harder and much more time-consuming.

GS: …and I thought technology was supposed to make our lives easier! Speaking of technology, do you use much processing in mastering, or do you prefer to leave the recordings in their pure form?

DL: I want the released versions of the recordings to convey the excitement that was present in the live performance, and that often means some modification is necessary. Recordings that sound perfectly OK through headphones often sound tame when played back over speakers, so, after doing the editing in Samplitude, I use a bunch of Waves plug-ins to ‘sex up’ the recordings and return the excitement.

An example: A decade ago I was present at a huge festive occasion in Ubud, Bali, where a very large gamelan played the traditional Barong repertoire while two famous Barongs danced. It was an exceptional event; the loudness and richness of the gamelan, the enthusiasm of the huge crowd, and the power of the dancing. It sounded amazing and, by good fortune, the two mics suspended overhead captured the music very well. But when I returned home and played the recording, it seemed flat. I eventually realised that the very large dynamic range of the recording robbed it of excitement. I restored it with some hefty limiting from Waves L2.

The other thing I add to my field recordings is a very discreet amount of reverb to provide a sense of ‘place’. Samplitude’s Room Simulator has a Nature category, and there are a couple of setups there (‘forest’ and ‘valley’) I very much enjoy. I use just a touch, typically setting dry at 0dB and wet at -18 to -24dB. For outdoor recordings I remove most of the early reflections to create the illusion of limitless space.

By the way… Seva, a good friend who worked at Waves for many years, advises me that when using a plug-in or other process, it’s prudent to set the output level within the process to just under 0dB so there are no ‘overs’ from one process to the next. I thought that 32-bit float made this unnecessary, but he says it isn’t so. Now I do as he suggested, and the sound is much sweeter.

GS: That sounds like good advice… Do you have any closing comments, David?

DL: Some people think that what I do is difficult, but do you know what the hardest thing is for me? Writing the liner notes! I feel that if people are reading the notes then they aren’t listening to the music. From time to time a journalist will ask me something like “As an ethnomusicologist…” to which I reply “No, no, ‘musical tourist’ is a better description”. To me, an ethnomusicologist is a boring person who takes wonderful music and analyses it until all the joy is lost. I can’t stand such pretension. I just want people to enjoy the music; that’s the reason for my wanderings. After all, what could be better than travelling to incredible places, meeting amazing musicians, and recording them?


Leave a Reply

Your email address will not be published. Required fields are marked *

More for you