Barry Diament is a veteran recording and mastering engineer, the head of Barry Diament Audio, and founder of Soundkeeper Recordings, whose philosophy is to create recordings that are natural-sounding and capture the spontaneity of musicians playing together in great-sounding acoustic spaces. Barry got his start in a small but noteworthy recording studio in New Jersey, moved to Atlantic Records in the 1970s, and ultimately decided that he wanted to achieve a level of sonic quality that could only happen if he followed his own path. I spoke with him about his career, recording techniques, and a lot more, which we’ll cover in a two-part interview beginning here.
Frank Doris: How'd you get started in audio? And what made you decide to become a mastering engineer?
Barry Diament: I got started when I was probably around 11, recording weekly jams I had with a friend on my brother's tape recorder. He had a Concord, maybe it was a [Model] 440. And I realized that you could engage Record in each channel individually. So I was able to play a guitar part and then overdub a drum part or vice versa. I kind of found out about overdubbing before I knew what overdubbing was.
When I was in college, they didn't have a studio like many schools have now. The audio production class was basically learning how to edit tape with a razor blade. That became very useful in time. But I decided I wanted to be a recording engineer and make records. I always loved records. I answered a job ad in The New York Times one day and took a job as an assistant at an R & B studio in New Jersey that ultimately became one of the homes of hip hop.
Barry Diament.
FD: What studio was that?
BD: It was called All Platinum when I was there. Their big band was the Moments. They also had Sylvia, who had a hit with “Pillow Talk.” I learned a lot there. They had a little mastering room with a Scully lathe, so I learned how to cut vinyl there as well.
I would set up the mics in the studio for the chief engineer per his instructions, and then I'd come into the control room and listen to what we had. I'd walk back into the studio to make a little change if he asked me to, and come back into the control room. Ultimately I asked him, “How come in here it doesn't sound anything like what it sounds like out there?” He got pissed off and said, “it's not supposed to sound like it does out there!” I never really found out what it was supposed to sound like. I just knew that it wasn't really what I was hoping for.
So, I [finally] got a shot at sitting behind the console. There was a big string overdub session scheduled, and the string players came in from New York. The chief engineer had been on vacation and this was supposed to be his first day back. He didn’t show up. So they said, okay, Barry, you get behind the console. And I went from assistant to chief engineer.
I had the opportunity to experiment all I wanted. So I got to try every mic in the mic closet. I think one of the first things I did was, when we were cutting rhythm tracks, instead of setting up seven mics on the drums, I decided to set up three. And I thought things got a lot better. Nowadays. I no longer “mic instruments.” I mic the event instead.
FD: We’ll get to that. When did you actually start in the recording studio?
BD: It was right out of school, so maybe 1976. I was there in ’76 and ’77. And then in ’78 I heard that there was an opening for an editor at Atlantic Records. I applied and got that position. My job was to make long songs shorter and cut [them for] singles, and make short songs longer for the clubs. I’d take the album track and cut it down to three or three and a half minutes for the radio. Then I'd also make an 11-minute version for the clubs by doubling solos, having the hook come back again, and maybe repeat one of the verses. I just did lots of editing sessions those first few years and learned a lot. I guess it was 1983 when Atlantic bought one of the [first] Sony digital systems and we got into digital editing. It felt like I could get in and out of an edit in the width of a razor blade.
FD: But then there was the early digital sound. It’s pretty well accepted now that early digital didn’t sound as good as what we have today. How much of that do you think was the early technology, and how much was that engineers had to learn to work with the new medium?
BD: In most ways? I mean, the stuff was painful. I remember a colleague saying, “There isn't any noise! And I would respond, “Well, listen to the music,” because the price we're paying for losing the tape hiss, which never really bothered me that much, was way too high musically.
I was telling people, well, we've had vinyl for a hundred years, so in another a hundred years, I expect CD is going to sound as good. Happily, it didn't take a hundred years. I think a lot of it was in the players. [But back then] we ended up putting, I think, Apogee filters in the Sony system. We replaced the stock filters. That helped a little bit, but I think the CD players took a long time to get better.
FD: How did you decide to change your miking method?
BD: Sometime around when I was at Atlantic I started experimenting with the idea of “miking the event” rather than using a lot of microphones. It started to occur to me that I didn't want multiple time arrivals to mix together in a given playback speaker. I decided, since I'm listening with two speakers, I want to capture the sound with just two microphones. That was the beginning of the road I'm on.
I got permission to use Atlantic Studios late at night; Studio A, which was wonderful. So many records we love were done there. It’s a historic room. They had some B&K microphones. I loved those compared to the Neumanns and AKGs and Electro-Voices and all the others, which I thought were, it's hard to describe, but I found them to be very colored. They didn't sound like what happened in front of them. They have a sound, which is why a lot of engineers like a certain mic for vocals or certain instruments; they like the color. I was looking for an absence of color, and the B&Ks were the first things I heard that suggested that to me.
I did my first two-mic recording. I was thinking of how far apart my speakers were at the time in the room I was in, about six or seven feet apart. So I thought, that's how far apart my mics should be. I was thinking there might be some reciprocity between the time it took a signal to get from my left speaker to my right speaker and getting from my left mic to my right mic. So I tried that, and one thing I noticed was that using only two mics had a lot more focus than I was used to with the typical studio multi-mike situation.
What I found was that you have to be a lot more careful about where you place all the instruments. The thing was if an instrument or a vocalist in the center moved just slightly off-center their image went all the way to the speaker. I attributed this to the spacing between the mics. I started to experiment. I made several recordings in my room with different mic spacings, and capturing myself speaking in different positions in the room. Somewhere around [the mics being spaced] 15 or 16 inches [apart], things seemed to gel, and I wasn't getting that pull to the extreme anymore.
I started reading about the types of cues our brains use to locate sounds. A sound that's on your left is louder at your left ear than your right ear. But it also arrives at your left ear before it arrives at your right ear. There's also a frequency difference because your head blocks some of the highs from getting to the opposite ear. And what I thought was, nature’s very efficient, and if nature could have gotten by fewer than three types of cues, I think it would have. So that made me want to supply all three types in my recordings.
So I made a baffle, something like a Jecklin disc, but where the Jecklin disc locks the mics in at ear spacing, I wanted to keep my 15- to 16-inch spacing. I made a disc out of two types of foam, a dense core and a lighter outer layer, a roughly 12-inch circle two inches thick. I put that between my microphones. Once I did that, the images on the soundstage seemed to solidify more, they had more meat to them, and they were less “phantom.” So that became my way to record.
FD: You were eliminating the crosstalk between channels?
BD: Well, I don't know if it's totally eliminating it. I'm just supplying some frequency discrimination between the mics, rather than just volume and time discrimination.
Soundkeeper Recordings' first release, 2006's Lift by Work of Art.
FD: What kind of mics and configuration and pickup pattern are you using?
BD: Well, those B&Ks were omnidirectional. I tried a pair of Earthworks, the QTC-40s, and I like those even more. They seem faster than the B&Ks, wider band without the upper-bass boost, lower bass rolloff, lower-treble boost, and upper-treble roll off that most typical studio mics have, even the very expensive ones. These have a frequency response that looks more like an amplifier frequency response, and I think these particular mics are rated from four cycles to 40 kHz, which is nice. I like recording at high resolution. They're really good at giving you something that sounds like what happens in front of them.
FD: And even if you think you can't hear that high, having that high a frequency reproduction can have an impact.
BD: Yeah, I think there's probably an impact on the transients lower down. I'm sure I can't hear 96 kHz, but when I switch from a 192 kHz recording to the 44.1 kHz version, it's harder to tell where the room boundaries are in the recording and much harder to tell what those boundaries are made of, where with the high-res, suddenly it's just focused on very, very fine detail. At least that's been my experience.
FD: What recording system are you using? DSD or PCM?
BD: I record in PCM at 192 kHz and 24 bits. Even if I'm going to end up on a CD, I find that if I start at high-res, which of course I'll keep for the high-res version, the CD sounds better than if I started at 44.1 kHz. Some of that information is somehow preserved. I don't know how.
FD: I know a lot of people who really like DSD, and they feel that it's very natural sounding.
BD: I know a lot of folks that love it also. It sounds strange, but I think as you go up in frequency, it feels less and less dynamic to me. I do think the bottom end and the lower midrange are very nice. With PCM, I feel like the dynamic range at the top stays where it was. When I hear DSD, I don't know if I always get that impression. It goes back to the mic thing also, because I have some DSD recordings and, while I'm listening to them in PCM, I think they're great recordings, and I think they're great because what came off the microphones sounded great.
FD: I really think it all comes back to the quality of the source. It seems obvious, but I mean, if you don't capture it at the beginning…and it’s frustrating because once the sound hits the diaphragm, the diaphragm’s got inertia and you inevitably lose something. Nobody's come up with a microphone with no mass. Not yet.
BD: Although those Earthworks are pretty fast, I must say. They have very tiny diaphragms. Maybe that has something to do with it.
I'm feeding a Mac laptop to capture the audio and the software and hardware between the mics and the Mac are made by a company called Metric Halo, which I wish one of the magazines would review for the audiophiles in the world. The stuff is amazing. The designer is one of those geniuses I mentioned that I've had the good fortune to sit and talk with. You spend a half an hour with the guy and learn so much it’s like you aced a three-credit class in the subject.
I use a unit of theirs called the ULN-8. It has eight channels, but [the units] can be stacked and have as many channels as you want. Then the software that's part of it is the most transparent recording device I've ever experienced. You [can] design the console [software interface] yourself. There's a monitoring section, and you can control the level of the monitoring for zero for studio use or minus 10 if you're interfacing with typical home gear, which is how I have it set. It's got a great headphone amp and it does RIAA [phono equalization]. Their EQ plugins are the most transparent I've ever experienced. Their EQ makes a lot of the others sound really coarse.
A similar thing happened with the cables [I use]. At one time I was using really good pro mic cables, but then I got some Nordost balanced cable, and the old cables got put away. There was just no comparison.
The mics feed the Metric Halo through the Nordost cables. The Metric Halo feeds the Macintosh through an Ethernet connection and the Mac captures it on the Metric Halo software at 24-bit, 192 kHz. I also use software called RX from a company named iZotope. I use it for sample rate conversion, dither, and noise shaping.
FD: Hence the name RX. (The software is used for audio restoration, mixing, mastering and other applications.)
BD: Yeah; that's really for fixing [problems]. I've been able to remove spurious noises from projects that came in for me to master. I've actually used it to kind of un-mix and then mix again. It separates [the recorded tracks]. I got a project from a band that said they could not hear their bass player or the lead vocals very well, and I used RX to get the vocals and the bass on separate tracks, then I could just lift them a little bit and bring them right into the mix. Some of the software today is just unbelievable.
FD: I never thought we’d have software to separate individual instruments and vocals from a mix. Look at that Beatles song that was put together, “Now and Then.” You can even get stem separation software for free, like Moises.
BD: The Soundkeeper recordings don’t really use that, because happily, I've been fortunate and not needed to. What hits the [hard] drive is pretty much done.
I record with a lot of headroom. I like to leave a lot of room for peaks to happen, and I found that it's good for digital recording. A lot of people think you should get up to 0 dB when you record digitally, but I found that you'll get a much better recording if you don't let the maximum peak rise above – 6 dB on that initial conversion to digital. Once it's [in the] digital [domain], you can bring it up a little more.
FD: Conventional wisdom would say you want to use as many bits as possible.
BD: Yeah. [But] it seems that most A-to-D converters, I think, exhibit their lowest distortion when you're below that top level, and I never go up to 0 dB anyway. I consider 0 dB the equivalent of an overload, because there are inter-sample peaks on the reconstruction of the analog signal. So I usually leave a part of a dB [for headroom]. I don't go all the way up to 0 dB in the final mastering, but the mastering of a Soundkeeper Recordings project is typically just setting the final levels, deciding on the spaces between tracks, and the track sequence. There really isn't any sonic processing going on.
FD: So the musicians really have to be on top of their playing also.
BD: They do, but sometimes circumstances get in the way. On one song from Lift, our first release, as the last note of the song was dying out, the mics very clearly picked up a fly all the way at the back of the church. The last note was an E and he was [buzzing] in E-flat! It made it sound like the last note kind of sagged, so they had to do [the track] again, even though they had nailed it.
FD: What speakers do you like for monitoring?
BD: I've been using different Magnepans over the years. (See this article's header image.)
FD: I noticed that on your website. That’s an unconventional choice.
BD: For record making, yeah. I’ve been using the Magnepan 3.7i for a while now. I also have Linkwitz LX521s in my room right now. Each does something the other doesn’t.
FD: I’ve heard the Linkwitz LX521s at shows and really like them. They’re dynamic and open and clear and natural. It's interesting that you are using those for mastering.
BD: They truly are amazing. I’ve been using them to master new Soundkeeper releases and am so glad to have another set of monitors I can trust. I haven’t heard many speakers that can, as I put it, “get out of the way” as well as these can. I also love the Maggies; there's something about line-source dipoles that is just magical. Dipoles load a room differently than monopoles do.
FD: And the tech keeps advancing. I’ve said this many times but I think there’s no limit. If the goal is to have a system sound like real life, who's to say we might not get there someday.
BD: It's getting better. I mean, everything is getting better. I think I'm the weak link in my recordings right now, not the gear. I can't blame the gear anymore. At one time. I could have done that. I think the gear has surpassed me, and so I'm doing my best to do it justice.
FD: And so many musicians I’ve talked to are thrilled to be able to hear themselves with this quality of playback.
BD: Yes, absolutely.
