How many albums have you produced for other people?
Twenty, I've only done it for other people's music in the last five years. I engineered for some other stuff before that, but I've only been hired since about '94.
Producer/engineers tend to approach recording from two different camps. The predominant one is the technical; i.e. "If we can just get the right mics for the drums and the best gear then we'll make the perfect record." The other camp focuses more on the environment... techniques that create a space that's comfortable with an emphasis on collaboration and performance. I'm curious what percent of your approach is technically focused and what is environmentally focused.
I'd say it's half-and-half. I'm very conscious of environment and I'm always working to make it better. By that I mean, either purposefully making it comfortable or uncomfortable to the point of tricking musicians into doing stuff that they actually want. Sometimes it's technical tricks, like someone could be singing, and I can sense that they are right at the point where their voice will break. If I think they can get the pitch, then maybe I'll change the pitch of the tape a hair, without telling them. They won't sense it but then they will be able to sing it. So instead of constantly telling them, "Oh you're singing flat. Oh you're singing sharp," when it's only a hair off, I change the tape speed. Or like with the last Stereolab record, there was this one tune that both Sean O'Hagan (from the High Llamas who was also doing work on the record) and I knew really needed some kind of ridiculous "hot" guitar lead. Well, Tim Gaine(the guitarist) is just not into stuff like that. Over the course of the first two months of tracking I would constantly be playing hot licks while we were working. And Tim was always like, "Stop that! Oh No!" You know... it was very funny. Eventually, bit by bit, over time, his hands started fidgeting and after about six weeks he started doing it himself. By then it was time for him to go in and track it. So it's just little psychological things to get people to do stuff that... not trying to force them to do things that they wouldn't do...but getting them to do what they want to do. On the Superchunk record, with Mac(McCaughan) and his vocals, I was working with him, actually in a way he said he hadn't done before, where we actually went over the lyrics and talked about different ways they could be interpreted through the vocal performance. We talked about what combination of voices the songs might need? Does it need harmonies here? Should we lay off the harmonies here? My favorite example (of this) was Mary from Stereolab. She had a song on the last record where she really had to sing out because the character was supposed to be this bigot. So she's singing it like she usually sings her back-up vocals. I thought that she really needed to step-up to this. She wouldn't do it. So I ran into the live room with a stick and chased her out into the street and then back into the room and immediately pushed record. When she sang the part again she was out of breath... she sang differently and it worked. So it can be anything really that works.
Did you have to do that with the Aluminum Group at all? They seem like they talk about music in that way.
We talked a lot about the implications about the way things would be done, mostly with the interpretation of the lyrics and how the performance would color the lyrics. But I tend to talk to bands about the lyrics a lot. It seems normal to some of them and for others it seems like they've never done that before...well, with engineers. But then most engineers don't seem too interested in the interpretation of the lyrics. That's up to the singer, there's a history of singers being more difficult to record than other instruments because it's more of a personal thing than playing an instrument. Most engineers shy away from that whole area.
I think that in the whole rock realm that's a bit taboo. Either because you're not supposed to care too much about lyrics, or because it's just the vessel for this angst filled, human, catharsis. Who would dare to go in and attempt to sculpt the raw emotion? Whereas, in pop music, it's just the opposite, it's all about sculpting the catharsis.
Well, I approach everything as if it was pop music.
I can hear that.
I like pop music more, although I like rock as well. In general I don't think I've ever worked with a rock band except US Maple. But even there, interpretation of the lyrics was very important because Al has a very specific set of guidelines for the way he sings. He's worked very hard at developing the way he sings. I know him pretty well, because I've known the band since the first time they played. Luckily, I was able to understand what he was trying to do. But even in that case we worked a lot on the vocals.
We're really flying through all the questions...It's good.
I'm sorry if I'm talking too much.
Yeah, I'm interviewing you and you're talking too much. Can you stick to just yes and no answers? Think like Bill Callahan. Can you talk a little bit about the difference between the Superchunk & Aluminum Group sessions?
Did you talk with Superchunk?
Yes.
Do they hate me?
No, they seemed really happy with you. I mean they teased you, definitely.
Did they?
For being over extended. For working so much and sleeping so little...for being sleepy.
I'm always sleepy, that's how I don't sleep.
From listening to the records I assumed two different processes for the sessions. I assumed that Superchunk came into the studio with skeleton pop/rock songs and they said "We want strings or horns on these parts" and you wrote those parts. I assumed that The Aluminum Group came in with arranged songs and that you collaborated with them to come up with more parts. Although both bands when interviewed seemed to say the same thing. They both said, "We came with our songs and he helped us to do some things but for the most part they were our songs and our ideas."
Yeah, sure.
I was thinking that the Aluminum Group was much more of collaboration, and Superchunk was more about selective collaboration. "We need a part for this chorus here." "Make this string part and put it in here." Etc. Because that's how the records sound.
With Superchunk they had never really tried anything like this before so in some ways they were unfamiliar with what it takes to make a record like that, and they wanted to do a two- week record in a week. I mean, you start bringing strings and horns, which aren't something that you just throw on. So in that case, of course I wasn't sleeping because I was home at night and I'd be writing the arrangements. The Aluminum Group was much more familiar with what goes into making a(arranged) record like this because as a band they do it already. That said, with the Aluminum Group, pretty much all of the arrangements changed in the studio.
They said that it changed a lot in the mix.
They're were also new parts. They had me replace almost all of the piano.
And the drums?
The drums were replaced. A lot of the new parts came from the original part on tape, triggering this stuff (Jim gestures at a wall of effects) and that's actually what was used. On that record the arrangements changed an awful lot. The keyboard player did the strings, and I brought in some people to play the horns and I wrote stuff for that. It was always collaboration. Over the course of the record all the songs changed from the basics. The basics were actually full performances with most of the arrangements in there. Sometimes the keyboard equivalent sound would be replaced by another instrument. In general, all the arrangements were there in the beginning but by the end it was very different.
I noticed that there is a lot of play between acoustic and electronic sounds in your mixes. For example, the AG's "Two Bit Faux Construction" begins with acoustic drums and ends with electric drums. It's almost like Paul McCartney's work where he uses his voice to bridge two completely different song environments.
I think that that's one thing that we really agreed on. One song doesn't have to be this group of chord changes. It can be what someone else might split into 3 or 4 other songs. In general, I don't want the complexity to clunk people over the head. It's there for people to find it. I don't want to foreground songs with the complexity of the arrangement but in general the stuff I do is fairly complex.
I noticed that you'd have two separate songs in one song and by the end you'd collapsed the parts together.
Some of my favorite pop music is McCartney & Sparks. That's really what The Aluminum Group are about construction-wise. When we'd get to the arrangements we were usually on the same page about, why this melody should move from this instrument to that instrument. I would always push it because I really like it.
I like it too. There's a lot of 70s Wings stuff happening.
Oh yeah.
So let's talk tech. Tell me some of the technical choices you made in the different sessions and tell me why...
Oh boy, I don't know where to start. In general, I'm very picky about what mics I use and I spend a lot of time with the placement. (See recipe) I set them up, I hear them and if they aren't right, I change them...because you know each instrument...snare drums... will be different. With the Superchunk record I spent a lot of time changing mics. Especially when it came to the bass. Laura's bass has no bottom to it. It's all mid. That was a little frustrating because, of course, they wanted bottom. I had to use what was coming out of the speaker...so I spent a lot of time trying to get a bass sound that fulfilled what they wanted. I'm still not sure that I did. I think I maybe should have changed the cabinet, but it was her cabinet. I tend to spend a lot of time on all technical aspects. I generally don't use anything digital in mixing unless it's overtly digital. I like using ProTools; I have no problem with that, but for mixing and such, I don't use any reverbs. I don't think I used any reverbs at all on the Superchunk record. I mean if they need it, I'll only use a spring or tape delay instead of using digital reverbs or something like that.
And your reason for that is?
They sound like crap. I think they sound terrible and they are so horribly culturally loaded at this point that they are of no interest to me. They take away any character from the singing. It relates the singing to the reverb instead of vice versa. There isn't this equal relationship between the treatment of the vocal sound and the vocal sound. That's generally why I shy away from them. A lot of the ambience on the vocals and the drums was just from using mics in different rooms and blending them in and such. Which, I think is much more authentic and interesting sounding approach because it actually has a direct relationship with the vocal sound. It's not just a relationship with some digital algorithm. I used this stuff a lot. (gesture to wall of effects.)
What's that stuff?
It's all modular voltage control filters and amps. Like on "Tiny Bombs" on the Superchunk record. The drums at the beginning, a lot of people have asked if they are played backwards, but they're not. It's actually each individual drum channel going through a voltage control amp. The bass drum triggered an LFO[Low Frequency Oscillator] and the LFO was sending out 5 different shapes and that shape was amplifying each of the drums. So it would trigger on the beat so the snare drum would have one envelope, and the high hat would have this other envelope, etc. Again, I think that's much more interesting because there is a direct relationship between the sound and how it's being processed. I tend to use this stuff[gesture] a lot. This was all over the keyboards on the Aluminum Group. All their keyboards are like mono, Yamaha, $200 keyboards. In order to make the sound at least a little bit more detailed they were all processed through this stuff. Especially Frank's keyboards, they were going through the Serge to "stereo-fi" it. If it's just mono, you're stuck with, basically, one option of how to mix it. You can put it off to the side, you can put it up the middle, but it sort of has a gravitational pull to it that you can't really play with. Except for in relation to the other instruments. This way, I could put it in stereo and then create the depth that I wanted for the keyboard as opposed to the depth being in relation to the other instruments in the mix. The keyboard could actually have it's own depth which the other instruments could then work with.
I think the keyboards are really important with the Aluminum Group. They run through almost every song. While the songs on that album tend to be pretty eclectic, accessing different genres that are instantly referenced, you've got this reassuring constant of the singing and keyboards that make it all of a piece so you don't feel like you're listening to a mix tape.
Right, related to that, the other thing that I like about this is that it doesn't make this sound like processed keyboards, because it's actually making a keyboard sound with an actual direct relationship. Again, you could put it through digital reverb and make it stereo, but it's a processed keyboard then and it carries that weight. That's why I like using the analog voltage control better.
I notice you often mix an "authentic" sound into a "processed sound". Many times it's the same instrument, you'll have real strings coming in at times with fake strings, or real drums substituting for the fake drums.
It has a narrative value. That's the only thing I like about it. By pitting these sounds together...they aren't just sounds, they have a narrative value moving from "authentic" to "non-authentic" which means that that is another thing you can use. You can work that with the song and the lyrics so that it has something to do with that.
Give me an example of what you mean by narrative value.
Well, something as simple as the song "Rrose Selavy's Valise" on the Aluminum Groups album. When the vocals change, when the singer changes, the whole mix also changes too. I mean, this is an obvious example, in that section, the drums go to electronic drums. Sometimes it's not a narrative question and it's just a question of which performance is better. I mean, almost all of the songs for the Aluminum Group record were recorded first with the electronic drums. I didn't think they were appropriate for about 60 % of it so I asked them to re-record the drums on about half of the tunes. On some songs the electronic drums worked perfectly but on other songs they just drew too much attention away from the other parts. But back to "Rrose..." that first song is a real scene changer. It has real epic feel with several sort of different scenes. We tried to use the switch between "authentic" and "inauthentic" to exaggerate this idea of scene changing. I think the other song that we were referencing the most when working on "Rrose..." was "Uncle Albert." You can use the transitions to really push this idea of changing slides. Something as simple as changing from acoustic drums to electronic drums can emphasize that scene change.
What about in a song where it happens more subtly? Like on "A Blur in Your Vision" there is an extended instrumental beginning where every two or four bars another instrument is added. Still, it happens in such a balanced way that it masks the complexity and gives the casual listener more of a sense of maybe a volume increase but not so much an increase in the number of instruments.
(We listen to the track so he can jog his memory). Yes, you see I took out the twelve-string here so there would be room for this other slide guitar here. Also, with this cross-rhythm entering here, I thought it needed to be emptier. And there...the twelve-string comes back in there so that the effect of this rhythm moving the rhythm sort of that way (gestures from right to left), would be more effective. It just seemed more effective with the twelve-string out so that when the twelve-string came back in it had less of a sense of things just piling on.
It gets wider.
Yeah, well also it gets wider because the string keyboard was running through the Serge and here the original keyboard is on the left. So over the course of this whole section... it's going slowly out of phase with the right channel so it is actually getting wider. I'm really heavy on mixing stereo. This sounds goofy, but every record that I've mixed, if you sit and listen to it right in the stereo spot you'll hear stuff you won't hear otherwise. I'm really heavy on exactly where to place things in the stereo and the depth. John Macintire and I had our inevitable jokey fight about the Stereolab record. You know... "You're track sounds better!" No, "Your track sounds better." We mix very differently. He mixes with that great in your face plateau.
Explain that.
He's got that great compressed mix sound where everything is up front, but everything has it's space but on a plane across your face. I think it sounds great, especially for the type of stuff that Stereolab do. Particularly the songs that he worked on- on the new album. I haven't personally had the opportunity to mix a band that does that, because the songs that I mixed for Stereolab were just entirely different. I love the way he mixes that material, it sounds great. To me, my mixes sounded like crap and his feelings were vice versa.
So what are your mixes like?
My mixes are all about depth. They are a triangle. They are all back there like this... (gestures like a diminishing point on a horizon). I'm not saying that one approach is better than the other is. Certain material needs that, I don't know if you heard the last Smog record, but the big "Queen- like" rock song on there, that's real flat and in your face, because that's what the song needed. Even so, there is still some depth in the drums. Almost all of my mixes are like a triangle, but I haven't really had the opportunity to do it a different way. I mean I could've with the Superchunk record , but in a way it seemed more appropriate not to do it, because that's what their other records sound like. I wanted this one to have more air in it. There is a lot of space.
What do you think about listening?
Well, how do I talk about that? Listening is the whole thing.
Though it seems to me that a lot of producers fall into a trap of mimicking.
Oh, yeah...I wouldn't want to say what other producers do because it would sound like I was insulting them. But yeah, for me it's all about listening. It's hard for me to dissect the way I do things. Not to sound like a hippie, but when it comes down to it, if the vibe isn't right, it's just not right. I mean, something can sound technically great but still not be right. The best record I've heard in years is this record by a woman called Stena Nordinstom from Sweden. Technically it's a mess but it's brilliant because the vibe is right. She's got these really badly recorded basic tracks of her singing. They are almost purposefully badly recorded (tracks of) piano and singing against fantastically recorded strings and horns, and the way the clash is what gives the record its character. I mean, generally, when I'm in the studio with a band, the only time I put my foot down is when someone wants to go over a part a million times. The more you play a part, the more you get it perfect, the more character you take out of the performance. When you think about your favorite records and you think about your favorite moments on those records...it's never the perfectly played part it's the anomaly of the performance.
It would benefit many bands to be less perfect.
Oh yes.
And for some bands it's even dangerous. Like for example if the Aluminum Group were perfect...
It wouldn't work. See that was part of my problem with some of their other recordings. They sounded too perfect... at times it sounds too programmed. That takes away from the character of Frankie & Johnny when they are being accompanied by what basically sounds like robots. There isn't an interaction between the voices and the programming. I don't mind sloppy performances, if the character is there, that's more important. I've worked with some people who will go over a performance a million fucking times to get it perfect and by that point it's completely uninteresting. And that's not just because I've heard the other performances. It's just not as interesting. Some people are interested in technical perfection.
Or when they exhaust all the spontaneity out of something (Eyes Wide Shut).
I really couldn't give a rats ass about technical ability because you know...it might take someone a week to learn a part perfectly, it might take them a year, but anyone can play something perfectly. It's not interesting to me. What's interesting to me is hearing the person play it. Hearing the character of the person playing, that's the overriding thing for me at all times, if I don't hear the character of them playing in there then it's not there. Sometimes it does require a few attempts to get that, other times the first take is perfect. So as far as listening, that's kind of the thing I'm always listening for...the sense that it's an actual performance as opposed to a perfect take. This is especially true with vocals. I don't care if the pitch is off a hair if the performance is there. Other times though, if it's not about being a performance, then the pitch is more important because that's the aspect you're trying to get across.
Like background vocals...
Backgrounds...or like in some of the Stereolab stuff where it's not meant to be a performance. I mean with the stuff I did with Laetitia [Sadier] we did actually go for more of a performance than she usually does. But, at other times that's not what you want...you don't want one performance blocking out or distracting from another performance, which was the case with the backing vocals on The Aluminum Group records. Often times we did go over the parts a lot because they have to remain in the background. Even with Edith Frost, the choice of having a voice double tracked has a lot of implications. She isn't singing just to you anymore.
Yeah, I don't much like double tracking. That's why I liked where you placed the harmonies and double tracks on the Superchunk. I mean it wouldn't really sound like Mac if it wasn't doubled...but...
A lot on the new record isn't double tracked.
Yes, but in those cases there is usually a harmony or something else, which strengthens the part without doubling it.
Right. I actually like double tracking but I like the implication of double tracking. Again, it has to work with the lyrics. You can remove the idea of a person singing to you. In the tiny bit of time that I've been doing that on my own records that question has been of major importance to me. Is it one voice, two voices or twenty voices?
Well that gets back to the pop versus rock dynamic again. It's strange how certain environments just traditionally allow for more ornamental approaches.
Well yeah, that contrast is of interest to me. There's sort of, well not a "rape" song, but something like that on this thing (Half-way to a Three-way EP) I just finished. There were so many ways that I could do it. The first time I did it super-cheerful so that you didn't notice the subject matter so it would have more of an effect. But you know, I tried different approaches to it, both with how many voices and what the tone was, because it effects the way the lyrics are interpreted so much. They aren't just words and music you know. I think a lot of people just think of it as words and music. I don't understand how you can.
I get a strong sense from your body of work that recording is fun for you. Not to imply that other engineers don't have fun recording, but that fun for you isn't just an effect, but an essential part of your recording process.
Yes, if it isn't fun, it's of no use. That's why I play a lot of pranks in the studio. If people aren't having fun then I don't think they make good records.
