One of engineer and producer Frank Filipetti's singular talents is that he's comfortable behind the board for literally any kind of music. His credits range from James Taylor's Hourglass — an album recorded on a Yamaha O2R in a cottage on Martha's Vineyard that won Grammys for Best Engineered Album and Best Pop Album in 1998 — to hit albums for contemporary bands such as Korn and Fuel. He has also worked with perennial stars such as Elton John, Barbra Streisand, and Rod Stewart. He's worked on Grammy- and Tony Award-winning Broadway cast albums, done live sessions with Luciano Pavarotti in Italy, and recorded Hole for the CD Celebrity Skin. Then there's that remote he did recording the songs of Bushmen tribes in Botswana's Kalahari Desert — well, you get the picture.
Filipetti is quite comfortable working on the cutting edge. As one of the first engineers to champion digital consoles, in the past few years he's become the “go-to” guy for numerous manufacturers seeking trusted ears to evaluate their gear. When they go to him, they have to be ready for honest assessment; he isn't shy about expressing his opinions, whether positive or negative. He doesn't fall for hype, and has been known to personally phone manufacturers to discuss what he thinks can be improved about certain products of theirs that he's purchased.
Filipetti, who lives in New York State, had a prior career as a moderately successful singer, songwriter, and drummer with his own publishing deal and recording contracts. But after nine years in the songwriting trenches, he reevaluated and made a left turn into engineering and production and never looked back.
These days he's busier than ever; pinning him down for this story took over two months. I finally caught up with him by phone one Sunday a few minutes before midnight. He was at a hotel not far from Manhattan's Right Track Recording (of which he is part owner,) trying to catch a few hours of sleep between the day-long setup he'd just completed and Monday's 7:30 a.m. session call. I tried to make our conversation quick, but, as is usual for Filipetti, once we were talking, he gave it his all.
What were you working on today?
Setting up for the cast album for the Broadway musical The Boy from Oz.
A big setup?
An orchestra: woodwinds, horns, strings, full percussion, and a rhythm section of keyboards, guitars, bass, and drums — along with 4 principals and a chorus of 18. Broadway shows are some of the toughest gigs because you have only one day to record everything. This one has 27 numbers, with a variety of musical formats, including small jazz band, big band, gospel, and Latin.
You've recorded plenty of orchestras. You must have a routine.
You can't really use that routine. With an orchestra, you generally don't have live vocals or a chorus and a rip-roaring rhythm section in the same room. Also, for a normal orchestral setup you'd have 20 to 25 strings to balance the horns and woodwinds. For Broadway shows the pits are incredibly small, and the costs are very high. So the string section has four to six violins and a couple of cellos. It's tough for such a small string section to compete with live drums, guitars, bass, woodwinds, and brass.
How do you work with that problem?
We're using “bugs” on the fiddles: little omni mics you fit on the bridge of the violins that give you an acceptable direct-sound-to-leakage ratio. Unless you're prepared to screen everybody else off, standard string miking brings up drums, horns, rhythm, and a lot of room. With the bug mics and some judicious equalization, you can make it work.
For this show we're using DPA bugs, but there are a lot of others. Schoeps makes a really nice one, and there are lots of homegrown varieties. Guys take little lavaliere omni directional mics and make clips for them. Some fit inside the hole, some sit on the bridge, some sit on the side of the instrument.
They give you a more direct string sound you can treat with artificial ambience?
Any other tricks?
You try to use microphone patterns as best you can, bearing in mind the size and the sound of the room and where the reflections are. It's all about what you screen, how to use mic polar patterns to maximize on-axis signals and minimize what's off-axis. Or at least to make sure the leakage off-axis doesn't interfere with what you're trying to capture.
Those are useful techniques for recording in general.
It crops up all the time. You want to minimize the bleed and make sure that whatever bleed you do get isn't ugly due to the frequency anomalies in the polar pattern. As you move toward the rear of a cardioid mic, moving away from the hot side of its polar pattern, it can exhibit an aberrant frequency response. In some circumstances, you might actually want to go to an omnidirectional mic, because you'd rather have good-sounding leakage than bad-sounding leakage. Generally the omni's polar pattern exhibits a much more uniform response all around the mic. The bottom line is to know the polar pattern as well as the frequency response of your mics.
So you're always thinking spatially as well as sonically.
Yes. The same principles can be used when recording a drum kit. In that case, the off-axis leakage isn't coming from 10 or 20 feet away but rather from 6 or 7 inches away. Still, the techniques and application are very similar. Take the overhead mics — some engineers like to use them to capture only the cymbals. The snare, tom, and hi-hat mics provide the sound for their respective instruments in the final mix. As an ex-drummer, I've never been fond of that clean, dry sound. For me, much of the power of the snare and toms comes from the overhead mics. I place them (and their polar pattern) in such a way as to maximize the snare and tom leakage. Consequently, I mic from behind the kit rather than from in front (see Fig. 1). I make sure that the distance from the snare to both overhead capsules is identical. I don't want the final overhead balance, which is quite hot, to pull the snare from the center of the mix. Again, sometimes I use an omni if the snare and tom leakage isn't sounding quite right. Whereas some engineers will face the capsules over the cymbals, which gets 80 percent metal to 20 percent of the rest of the kit, I move them farther back over the drummer's head. I look for about 50 percent metal and 50 percent kit.
What about the other drum mics?
When miking a tom — in addition to just placing the mic where it will pick up the best sound on the tom itself — you have to focus on keeping the leakage of the cymbals out of the tom mics. Although the sound may seem okay on the recording date, by mix time, after you've added a fair amount of top-end EQ to your toms, you may find that your cymbals sound excruciatingly bright due to the cymbal leakage in the tom mics. Some people deal with this by gating, but that's never been my choice. Just take an extra couple of minutes on the recording date and play with some hypercardioid, figure-8, or even omni patterns before settling on your standard cardioid mainstay. For example, say you have to mic a rack tom with a ride cymbal slightly to the right of it. If you use a cardioid mic on the tom, you'll end up with a lot of rack tom and a lot of off-axis cymbal. Instead, you may want to use a hypercardioid mic or even a figure-8. That way, the cymbal coming in on the right actually hits the capsule at its null point, which is along the sides instead of the rear (see Fig. 2). You're not just miking to get the sound of the individual drum, you must always be aware that there are other tonalities in close proximity that, by the time you get to the mix, may cause you problems.
There's the sound of the mic, the polar pattern of the mic, and then there's phase.
Phase is tough on a drum kit because you have so many mics in close proximity. You can try to apply the “rule of threes,” which is less a rule than a guideline; you want the direct microphone's capsule at least three times closer to the source than any other mic. The idea being, at that distance, phase cancellation (or more importantly, comb filtering) is less of a problem. With a drum kit, there's only so much you can do.
An example, please.
Say you have a mic on the snare. It's probably about six inches from where the actual attack of the drum is at the center of the head. As I said earlier, I don't like just the sound of that one mic; I like a bigger sound that comes from above the drummer, which I get from my overhead mics. But I don't want those overheads so close to the original source that they cause comb filtering on the snare. So I'll have the mics a couple of feet above the drummer's head. That adds weight to the snare drum as both capsules are temporally far enough apart to enhance rather than cancel frequencies out of it.
Also, your snare and hi-hat mics are always going to be in close proximity. There again, you do the best you can. Put your snare mic on the snare and try to maneuver it so that the hi-hat is facing the null point as much as possible. I usually try to put the hi-hat between its mic and the snare drum, using the cymbal as a kind of screen to the snare.
You recorded and mixed the upcoming Korn recordTake a Look in the Mirror[Epic/Immortal, 2003] at lead singer Jonathan Davis's personal studio [see Fig. 3]. How did you mic David Silviera's kit?
I used an Audio-Technica AE2500 — the dual capsule mic — for the bass drum; an old mic, an AKG D36; and a Neumann U 47 FET. I also had a tube Neumann U 47 in the room aimed at the bass drum. The drum had no front cover, and the AE2500 was slightly inside the rim of the drum right on the beater. The AKG D36 and 47 FET were slightly off-center and back a couple of inches, with the U 47 about two feet away for the extreme low end (see Fig. 4).
The snare mic was an Audio-Technica AE3000, with an Audio-Technica AE5100 underneath. The hi-hat was a Neumann KM 84, the overheads were Sanken CU44s, the ride cymbal was a Schoeps, and the toms were Sennheiser 421s, except for his “gong drum,” which was mostly an Audio-Technica AT4047/SV. And then a couple of Neumann CMV 3s for the room.
Is using that many mics on a bass drum normal for you?
It's typical for a rock session. On a jazz date, I use two mics. On a Broadway date, I'd probably use just the AE2500.
For the bass drum on rock sessions, you want lots of impact and extreme amounts of low end — attack and depth. I use miking and phase to get the initial attack and the low impact that occurs microseconds later. I'll have one mic, possibly two, that are slightly out of phase with each other, and I'll add them together in a way that combs out 300 to 700 Hz. You can also do that with EQ, but I like it better using phase.
The same with rock guitars. You can put one or two mics right on the cone, then add a couple more slightly out of phase with them by moving them back a few inches. It's taking that rule of threes and standing it on its head — putting two mics at slightly different distances from the source to get a comb-filtering effect. That sometimes works well on guitars; you move mics around until you get the sound you're looking for.
Do you listen on headphones while you're moving the mics?
No. While we set up, even before the drummer gets there, I have someone hitting the bass drum. I record a bit, then listen, and also look at the signal path in the waveforms on the Euphonix R1, which is what I record to. Then I play with the phase. It's a process of experimentation.
That's a rock date, when you don't necessarily want the bass drum to sound the way it does in the room. You're accentuating some of the fundamentals and removing some of the midbass harmonics. Unlike a jazz date, when it's important to get the natural sound of the drum — the ring of the head, the natural dynamic. On a jazz date, I don't generally play with phase. In the past, I'd use a dynamic and a condenser and try to get the capsules matched as closely as I could. With the AE2500, a very clever mic that has two capsules — one a dynamic, and the other a condenser, which are totally phase aligned — A-T has done that for me.
You prefer to use mic placement and phase manipulation to create your sound rather than filters and EQ?
Yes. I start with that. I want to get the best sound I can before I reach for the processing. I have no problem going -30 or +20 on an equalizer if I need to. But my instinct, training, and experience have led me to place a greater value on mic placement than on the equalizer and compressor.
Do you use the same techniques on a rock snare?
It varies. I use two to four mics. There's also the cross-stick issue; some guys are good at it and some aren't, and you have to mic with that in mind. I usually place a mic under the snare with the polarity reversed. Some snare drums have a lot of sound coming out of the air hole just off the center of the drum. Sometimes I mic that for impact.
You mentioned putting high end on toms for snap. What frequencies?
Somewhere in the 3 to 4 kHz range. You're looking for the sound of the stick on the head. That brings out the metal in the cymbals, which brings us back to what we were talking about earlier regarding minimizing cymbal leakage.
Do you compress drums?
Usually I don't compress drums on the record side, because you can't undo it. Rock kits often end up being compressed by default, anyway, due to the overall bus compression you're generally using in the mix. I like that better because it compresses more uniformly with the rest of the track. Sometimes I'll put the toms into a compressor because it brings up sustain and adds a certain type of bloom that's good. But compressing the bass drum or snare drum during recording can lose the crack of the initial attack. And once it's gone, it's gone.
Do you often record guitar amps in stereo?
No. For one thing, mono usually has more impact. With Korn, of course, there are two guitar players, so I'll place them left and right in the mix. If I'm looking for a particular effect, or for a solo, I'll mic two different cabinets, put them on separate tracks, and maybe spread them left and right. But that isn't stereo — it's two mono sources. For actual stereo miking, I'd probably use a coincident technique so that the sound would be captured in stereo but would still have a powerful, phase-coherent center.
What mics do you like on the guitar amps?
In the past, my main mics were a Shure SM57 and a [Neumann] U 67. I'd use both and match the capsule distance. But since A-T came out with the AT4047[/SV], that's become my favorite guitar mic. I used it with Korn on three different cabinets. Each cab had a 4047 and one additional mic: a Royer ribbon, a 57, and a U 67. Generally, I put the mics close to the speaker cones.
That gives you a similar sound from the 4047s to help blend, and different sounds from the other mics for character.
Korn's bass player, Fieldy [Reginald Arvizu], has a very distinctive bass sound. How do you work with that?
Fieldy likes lots of click on the top and doesn't like mid bass. His bass sits above and below the guitars, which lets them stake out that middle ground.
Do you use direct and amp sounds on him?
Yes, but 80 to 90 percent amp. I use three or four mics: the Sanken CU44X and the AT4047 on one speaker, and an AKG D36 on the other. He also has a tweeter in his cabinet for high end, and I threw a 57 on that.
Is there a DI you like for bass?
The Vipre, by Groove Tubes — GT. It's a great mic pre and DI, especially for bass.
What did you use for Jonathan's vocals?
On the previous record, we used a Neumann M 49. On this one, we used a Sanken CU44X into a Tube-Tech MP 1A mic pre, then into the line-in of a Neve 1073 mic pre, for a little EQ tweak. Out of there, it went into an 1176 and a Massenberg EQ. From there into a dB [now Lavery Engineering] converter ahead of the R1.
How did that particular chain come about?
On the Untouchables album, I worked for a couple of days with Jonathan and [producer] Michael Beinhorn trying a variety of preamps and mics. Generally, I don't get that complicated. But working with Michael you get to experiment. We started with that setup but found that the Sanken worked better with Jonathan this time out. It was actually [producer] Peter Asher who introduced me to that mic.
More often than not, I'll start with a U 47 or a 269, put it through an 1176 and begin recording. Especially when I'm working with a new singer, I tend to go minimal. I want to know what their voice sounds like, and I don't like to do a lot of things until I'm confident I'm making the right decisions. I'd rather do something later than undo something later. We have so many tools now that we can use after the fact; a good, clean signal is the most important thing.
But Jonathan needs to hear a particular sound on his voice, and he likes to hear a finished product. We were trying to maximize the width of his voice — the low end and the high end — and to minimize a bit of his midrange. This chain worked well for that.
Speaking of doing things later, what about ambience on rock vocals?
With a rock vocal — especially with a band like Korn, where you've got such a huge sound behind the singer — it's very important that the vocals are right in your face. Jonathan likes to have a lot of things on his voice, but he doesn't want it to sound like there's a lot. You'd actually be amazed at how much there is going on. We've got vocoders, delays, and choruses on most of the songs, but all are subtly applied to enhance the presence.
What's the most important piece of equipment that today's engineers need but don't have?
We desperately need a fail-safe, bulletproof storage mechanism. That's something that's really going to bite us in the ass in the not-too-distant future. I started talking with some of the hardware manufacturers and OEMs, and found that a lot of people in the know are saying that if you let a hard disk sit idle for three to five years, there's a good chance it won't play back and that it will freeze up. You have to run it periodically. And I can guarantee you that in storage houses and at record companies, nobody's in there spinning these disks. People are putting all their stuff on Pro Tools hard disks, and then tucking them away. It's going to be horrifying if all this material goes down the drain.
Any new analog gear you like?
I love the new Tube-Tech multichannel compressor, the SMC 2B. It's my favorite new piece of analog gear. It has three bands — low, mid, and high — and you can set the crossover point so you can compress each band individually. I also like the new tunable JBL 6300 series speakers. I used them on the mix of the new Korn album and was very impressed. I also have to say that I'm very enamored with the TC-Helicon VoiceOne. It's digital, but it can do things you can't do with anything else.
The Grammy you won forHourglass[Best Engineered Album — Non-Classical] was the first to be awarded for a project recorded in a home. What did you learn from that?
I certainly don't want to do that all the time! I wholeheartedly support the great commercial studios, and I want them to survive. They're really important to the art of music. But I learned that if you know what you're doing and you're in a good-sounding space, you can make a recording of very high quality for a relatively small amount of money. I used to get very uptight about situations I couldn't control, and it made me much more aware of the fact that you can just go with it. You can use the knowledge that you have to make it work.
Maureen Droneyis the Los Angeles editor for Mix magazine. Mix Masters, her book of interviews with top recording engineers, is available from Berklee Press.
FRANK FILIPETTI: SELECTED DISCOGRAPHY
Luciano Pavarotti, Pavarotti and Friends: For War Child (Polygram, 1996)
Barbra Streisand, Higher Ground (Columbia, 1997)
Billy Joel, The Stranger (Columbia, 1998); surround remix
Luciano Pavarotti, Pavarotti and Friends: For the Children of Liberia (Polygram, 1998)
James Taylor, Hourglass (Columbia Records, 1998); CD and 5.1-surround mixes
Luciano Pavarotti, Pavarotti and Friends: For Guatemala and Kosovo (Polygram, 1999)
Meatloaf, Bat out of Hell (Sony, 2000); SACD surround remix
Korn, Untouchables (Epic Records, 2002)
Rod Stewart, It Had to Be You: The Great American Songbook (J Records, 2002)
Fuel, Natural Selection (Epic Records, 2003)
Elton John, One Night Only (Universal, 2003); concert, CD, DVD, video
Rod Stewart, As Time Goes By: The Great American Songbook, Volume III (J Records, 2003)
Wicked, Original Broadway Cast Album (Decca Broadway, 2003)
Korn, Take a Look in the Mirror (Epic Records, 2004)
Paul Simon, Graceland (not yet released); surround remix