Interviews

Gillian Welch and Dave Rawlings: producing themselves

I've always been a big fan of Gillian Welch's albums. She and her musical partner, Dave Rawlings, made two records (Revival and Hell Among the Yearlings) for Almo Sounds with producer T Bone Burnett [Tape Op #67] that I enjoyed. But it was the next album, the Dave Rawlings-produced Time (The Revelator) on their own Acony Records, which put them onto my desert island list. It's quietly stunning, both musically and sonically, and Gillian's new album The Harrow & The Harvest is equally as strong. I got a chance to speak with them while they were on tour supporting Buffalo Springfield. At first they seemed a bit bored by yet another interview, but when I mentioned I wanted Dave to stay in the room so we could talk about recording, they both visibly perked up. Although Dave is credited with production, Gillian also has some deep knowledge and opinions on recording. As befits two people who have worked together for over a decade, they often finish each other's sentences and thoughts. Their strong connection as musical partners is evident.

Time (The Revelator) is one of my favorite records. I think it's a classic album.

G: We made it in the old RCA [Studio] B in Nashville that was built in the late '50s. It had no gear in there.

D: We were looking for a recording space and I had been driving around Nashville trying to find an old studio to rent, or possibly buy. One day I drove by Studio B and the door was open. I thought, "Oh my God, that's Studio B. I've never been in there." I walked in, heard my footsteps on the floor and knew that I liked the sound of the room. Bob Moore was there that day — Elvis's bass player. He just happened to stop by. I was really interested in [renting] it. I then found out that the Country Music Hall of Fame — who had been running a lot of tours through it and whatnot — were building the new Hall of Fame. In the interim they were going to be too busy to do anything with RCA B. We approached them through a friend who was on their board and they said would it be all right if we brought our gear in and rented it on a monthly basis. They treated it as a donation to the new Hall of Fame, which was real nice. We rented it out for about 14 months. When we first got in there, I spent a month or two cleaning out the troughs and I fixed the plate reverbs. The place hadn't been used much as a professional space in quite a while.

G. It had not been a functioning studio.

But they had a little bit of gear, like the plates?

D: They had the plates in the other room and they had somebody doing some karaoke sessions out of the live room. The control room was basically empty.

G: The speakers were still there.

D: Oh yeah, the old Altec 604s were still there, but they needed to be fixed.

Is that what you monitored on?

D: It was mainly [Yamaha] NS-10s and the Altecs. What we ended up bringing in was all the gear from the home studio — stuff that we've assembled over the years. Our tape machine is a [Studer] A800.

Is it 24- or 16-track?

D: 16-track. I actually bought the headstack before the machine. I found some unused 16-track heads when I was buying some other gear, and I threw those in. Then I found a 24-track machine. What else?

G: The Neve desk.

D: Yeah, we started buying [Neve] 1084s really early on. I bought a BCM-10 frame and every time we went on tour I would come back and buy a couple more modules. I found some other 1084s from the next console made, so the serial numbers were still pretty continuous. We hadn't filled the frame for ...Revelator, but we had enough — we only needed four or five.

G: That board came out of WGBH Boston. It was the old Sesame Street board.

So it had a Muppet vibe.

D: Yeah, rubber ducky. I got this other old BCM-10-style console made by Neve that has 1055 modules in it; they're the wide, black ones with three fixed bands. They basically have a high, low and a mid — you can't select the frequency — and 10 dB steps. They are very unforgiving with transients; they really don't like anything barking. There is distortion all over our records because of those modules.

Does it squash the transients or distort?

D: They break up in a weird tear-y way. If you hit them with the top of a vocal it will have a little "kkkrrrrrr" on it. I would go through those, as well as the 1084 at line level to get five dB gradiation; as a buffer stage. I had some 1084s that bypassed the fader, and those were the ones that I used before the tape machine. So the signal chain was two [Neumann] M 49s, a [Sony] C-37a on my guitar and an M 582 Neumann on Gil's guitar. There are other setups: "Dear Someone" would have been an Altec 639a, one of those birdcage mics, with a [Neumann] U 67 right on top of it. They end up perfectly out of phase and you just flip them. We were in there and we would have to break down every couple days 'cause they would run a tour, so we weren't able to leave the mics set up or anything. It was a difficult process.

G: With us, millimeters of difference in the mic setups are huge because the picture is so affected by overall phase between our four mics.

D: Everything is pretty close together.

How far apart are the two of you when recording?

G: Two and a half feet. As close as can be.

D: Some days we would set up, the phase would be great and everything would click in. Then a tour would come through and we would have to tear down. We got a little rug with everything spiked, but we would have to get within millimeters. That's the difference with this new record. Since we were finally working in our own studio, we set up and we never touched the mics.

So a lot of the same gear has made it from record to record?

D: Yeah. There are two tracks on Hell Among the Yearlings that we did at home on those same preamps. By then we also had the [Neumann] M 49s. That was the beginning of what I look at as that incarnation of duets, like "Miner's Refrain" and "Rock of Ages." "Rock" is a banjo song, so it's a little different because I used a [Neumann] U 47 on the banjo.

You had 14 months to make Time (The Revelator). However, it wasn't really 14 months because you were constantly interrupted?

D: We made that record in five weeks. Most of the album was probably created within three weeks, and then there was a little bit of time on either side. I also produced part of the first Old Crow Medicine Show record in that time period. We just happened to be renting the studio for that long.

How long did Harrow & The Harvest take to record?

G: Four weeks. That's about how long our records take.

Is everything recorded live?

G: Totally.

D: Yeah, everything is live. It is pretty much all from takes one, two or three. Very few mixes. This is the first record we've done that Stephen Marcussen [our mastering engineer] listened to and said, "Okay, Let's transfer it." We didn't compress or EQ anything. Just transferred it from a machine of his that we really like, through the nice converters and a clean signal chain.

When you are two-feet away from each other there is no way you are going to punch in and fix a part.

G: We have never done that.

D: About half the songs on the album are complete takes. Five of them are composites of adjacent takes. G: Edits between takes.

Edits on the 2-inch master tape?

D: Yeah, I do a lot of 2-inch editing. I've always done that.

G: Dave's really good at editing. I'd put him up against anybody at this point, because he's not even getting to cut on drums. [Drums make it easier to find the edit points. -ed.]

D: I know where the edits are on this record, but they are pretty hard to find. I bet you can find a few on ...Revelator.

G: But I don't even mind. I like the sound of a tape edit.

Sometimes they're cool. It can totally change the ambience in an unexpected way.

D: Yeah, as long as it's musical.

What speed, 15 ips?

D: 30 ips. I think Soul Journey was at 15, but everything else we have done was at 30.

Why 15 for Soul Journey?

D: Drums. There was more of that vibe. There might be a couple [of songs] at 30, but I just wanted to try it. But that was a very different rig. That was mostly [Shure] SM57s and API preamps.

Was that still mostly tracked live with the band?

D: Yeah. The only thing I should say is that I overdubbed some organ on a few things. I'm a terrible [Hammond] B3 organ player, but if I get one pass at something I usually do a really good job. So I go in, do one pass and that's it.

G: One band song went down without any singing and I had to go back in and sing.

D: We were jamming with the chords of it and it sounded good, but then I think we used your scratch vocal to see if it worked.

Do you have an engineer helping you?

D: We have worked pretty closely with Matt Andrews in Nashville for a while now, and our methodology has developed around the three of us. I'm not in the control room while we are tracking, so we rely on Matt, to some degree. We have some sense of whether or not we are getting there, but it's always good to have another set of ears. If we are going to be editing between takes, it's generally good to get parts from adjacent takes. I suppose I handle most of the responsibilities that you'd associate with a producer.

G: We all listen and weigh in on what the good takes are. Happily, we pretty much agree. It's pretty evident.

Do you do the mixing?

D: I do a lot of mixing. On The Harrow & The Harvest, we weren't really moving faders very much — we never really ride stuff. For most mixes we set the faders and let them run. It's very rare for there to be fader moves within a song. Matt did a lot of live mixing, where he would get the picture a particular way. If we liked it when we came in, we might only tweak things slightly. He did a lot of riding the preamps and then we would adjust from there as far as color and compression.

G: We did a lot of printing tracking mixes — this is very common for us. If we like what we have — even if we can narrow it down to one, two or three takes — we will print them that night.

D: We did that when we were at Studio B and we needed to bring songs back to the other room. We don't have automation, but I have a system of recalling mixes that is crazy accurate by using voltage to get faders in exactly the right spot.

Do you measure it with a voltmeter?

D: I measure the fader levels with a voltmeter. It's actually more accurate than any of those moving fader systems. I mean, you can be off a quarter or a half [dB], which, in our world, a quarter and a half is like...

MORE INTERVIEWS

Recording Nona Invie’s <i>Self-soothing</I>
INTERVIEWS · ISSUE #167 · Apr 2026

Recording Nona Invie’s Self-soothing

By John Baccigaluppi

On Friday mornings I go to the new releases page on Tidal and wade through the new music released each week. I'll check out records from artists I know, but what I really enjoy is finding new music from an artist I'm not familiar with that resonates with me. On the last day of...

M. Ward
INTERVIEWS · ISSUE #167 · Apr 2026

M. Ward Leaving the Door Open to Chaos

By Geoff Stanfield

Geoff Stanfield spoke with M. Ward for an episode of the Tape OpPodcast in August of 2023, around the time of his album supernatural thing was released. Here they dig into his love of collaborations, his analog approach to recording, and more.

Daniel Tashian
INTERVIEWS · ISSUE #166 · Apr 2026

Daniel Tashian Having Fun

By Larry Crane

In 2017, one of my best friends, Craig Alvin [Tape Op#137], kept texting me about a record he was engineering. He was saying how amazing the process was, and how awesome the results were. The album turned out to be Kacey Musgraves' Golden Hour, which went on to be a platinum...