Finding Their Way: Making Kowloon Walled City’s Piecework

So, I listened to your record.

Thanks!

And I listened to some of the previous Kowloon records again. There's a solid theme going on. The new record feels like an expansion on previous themes. How much of that is conscious?

It’s completely conscious. When we started writing we had a few changes in mind, based on Grievances. So, it’s naturally evolutionary. Despite the running joke in this band that my original idea was that we would never evolve.

You can’t make the same record over and over again. [laughs]

Oh, I could! But I find it okay to make creative rules and sometimes break them. Maybe make a rule like, "No cymbals on this album." Then, at some point, it’s like, "We do need a cymbal here." It’s fine. The rule is not religion.

Sure. Wait, are there actual jjrules?

Oh, my god; there are so many rules. Along with that, we’ve figured out what we’re good at. Container Ships was the first record where it was like, "Oh, there’s a thing here." The Grievances record was us doubling down on that, and maybe me being a better recordist.

This one sounds the best of all of them. The sonics are very much grounded in reality. You don’t give into the larger-than-life overinflation of elements, like a boomy sub-bass kick drum. I assume that’s deliberate?

Yes. Super deliberate. I’ve realized that’s my favorite type of recording to make.

Realistic?

Maybe hyper-realistic. You know this – if we gave someone a recording that really sounds like what’s in the room, it would be totally underwhelming. I’m shooting more for what it feels like when you’re in the room. I’m trying to accentuate that.

Listening to the whole record, I almost get the feeling that it’s made for musicians or engineers. There’s all this negative space, and you’re celebrating the texture of everything. It’s not like, “Check out this riff.”

Yes! I love hearing that. There are parts that feel like when you look at a minimal painting for a long time and start zoning out. We all love that.

Along with the kick drum not being sub-bassed out, the bass guitar isn’t either.

My bass mic choices have always been pretty mid-rangey. The [Electro-Voice] RE20 and a [Shure SM]57, most of the time.

No DI?

Nah.

Oh. Offended?

No, but I always feel like the big winner if I can avoid a DI. You get a lot of tone out of your DI. I didn’t even know that was possible. I grew up playing bass, so I love dialing in bass tones. I don’t find it that hard. In our case, Ian [Miller] is a fantastic bass player. He could be playing through a $25 practice amp, and it would sound great. His rig has been consistent for years. It’s very simple: a Washburn Vulture, which is [like] a hyped P-bass, [Tronographic] Rusty Box [bass preamp pedal], and a [Gallien-Krueger] 800 [bass amp]. It sounds great.

Would you say that new strings are the fifth member of the band? It sure sounds like it.

We don’t actually change strings that often. [laughs] But I’m playing a [Fender] Telecaster, and Jon [Howell] plays a [Gibson] Les Paul with very spanky P-90s. My guitar cabinet is punishingly bright, and I’m usually working to get Jon’s guitar sound into that zone. On this record we were playing Hi-Tone amps, which are amazing Hiwatt clones with infinite headroom and very sharp. Unforgiving. Jon played two boost [pedals] into the Hi-Tone, and that’s it.

Two boosts? No actual distortion or overdrive pedals?

Yep. Then stacking them when needed. It’s weird, but it works for him.

And you’re using the [Ben Adrian Audio] KWB [kowloon walled bunny, overdrive] pedal?

Not much on this record. I used the Hudson Electronics Broadcast [pre-amp pedal] and Electronic Audio Experiments Mu Blaster [boost pedal]. Both are fantastic pedals. For a couple of songs, I got some clank from the Big Game Pedals F800B, which is like an old Peavey preamp in a pedal.

Was this the record where it ended up being mostly tracked with ribbon mics?

Yes. For guitar I had two or three mics up on each cab, but, in the end, I only used one for each guitar – both Extinct Audio BM9 ribbons. And crazily enough, that’s what I used for vocals too.

That means that in real life, the guitars were really bright.

Take-your-fucking-head-off bright. There’s very little EQ on my guitar. There’s a little bit of Pultec lift on Jon’s. Each guitar track has a couple of EQ notches to take out some resonant ring around the pain zone. But not much else.

You know that I’m perpetually hypercritical of the 3 kHz range in all recordings. But I didn’t get it on this. It doesn’t hurt. Listening to your previous records, on the loud parts I can hear them hitting the ceiling a bit; but on this one the headroom is higher. Even when it gets the loudest, it doesn’t feel like it’s smashing up against anything.

Oh, that’s interesting. One thing I noticed, looking back, is that I did a ton of [mix] automation. Tons. Individual drums. Fader rides on the kick out mic. On the overheads, on the bass. A bunch on the drum bus. Some on the mix bus. Maybe that helps fit everything into the space.

Actual written automation? Not Pro Tools clip gain?

No. Lots of written automation.

Oh, wow. Do you do that much on other peoples’ records?

Maybe? I’d like to do more than I do already. I think it’s really important.

Is it accentuating loud passages? Bringing up the room on quiet parts, and that type of automation?

It’s both. If Dan [Sneddon] starts leaning into the cymbals, I’ll pull down the overheads a little bit. If the song lifts, maybe bump the whole mix bus up half a dB. For instance, on this record there are lots of little guitar details. Jon jumps back and forth between being the rhythm guitar and the stunt guitar.

Is he the right side?

Yes. The good guitar side. So, when he plays something that’s important, a quick 1 dB lift makes a big difference.

You don’t have a control surface. If you’re doing multiple fader rides, are you doing them one at a time?

Yeah, I do one thing at a time, all with the mouse. But I’m in a zone to do that. I’ll run through and do a pass of only tom or overhead automation, and, once I’ve done that, it’s logged in my head to listen for other little tweaks as I go.

Most bands don’t have a budget to put that level of care and deliberation into details like that.

I don’t think it takes that much time. Maybe it’s more a matter of letting the other decision processes chill out enough to get to that point. If every set of mix notes is huge changes, as well as trying to figure out if I can even make the band happy, at what point does it make sense to commit to automation moves?

Right. When I’m working on somebody else’s music, I default to, "Well, this was their dynamic. This is the thing they worked on at band practice." I’m not trying to mess with it too much. I can accentuate it, but I don’t want to fuck it up, or put a bunch of time in and have to backtrack when they’re like, "Why did this jump so much on this part?"

Well, hopefully the automation is mostly invisible. If I notice that the hi-hats are overwhelming on every verse, I’ll go through the song and ride those down. No one should notice that, like, "Hey, dude. I’m trying to smash the fuck out of my hi-hats there." You’re not really an automation person, right?

I’ll clip gain to even out a vocal or guitar part if something’s jumping out. But I don’t often ride faders. A lot of times, with the music I’m dealing with, everything’s loud all the time. I can grab a whole section and turn it all down, rather than having to ride it, because it’s so consistent and compressed.

A lot of the moves that I’m talking about are that. They’re coarse, across a whole section of a song.

I've seen screenshots from some people's mixes – I'll look at the automation window, and it’s like a constant crawl across every track. Everything’s automated as a performance.

So, they’re doing automation passes with a control surface?

Yeah, or on a console. They’re riding everything the whole time, doing one or two things at a time. That always blows my mind.

I know the traditional automation method is playing back [the mix] with your hands on the faders. That doesn’t work as well for me. I find myself chasing what just happened too much. With a mouse I can be pretty quick and decisive, listen back, and move on. It sucks for my wrists, but works pretty well otherwise. My impression with you, as we’ve spent a lot of time listening to recordings together, is that you compress a lot and you crush a lot. That does a lot of the automation for you.

It evens it out, for sure. I was looking at the Deafheaven song that I’m prepping for NailTheMix [online mix school]. It’s an 11-minute song, and I said, "Hey, if I don’t build this from scratch on the show, it’s going to be the shortest episode ever." It’s like, "Let’s push these faders up, and... we’re done!"

I’ve nailed the mix! [laughs]

A big part of my thought process is "let it lie." The band played. We evened out the overs enough. I crush some things. I like to think that the band is doing a lot of the work, especially because they’re all recording live. Did you guys record this live?

Yep. Since the beginning of this band, we’ve recorded everything together. We’ve mic’d everything up, stood in the same room, counted off, and played the song.

Amps in booths?

Amps isolated, yeah. Then, once we have the takes, we do a day or two of, "Now what do we need?" It’s like making a movie, when they do multiple shots to make something translate better. "What can we add or redo to bring out sonic details?" For instance, there are a handful of quiet sections that we did play live, but we’d try redoing the guitars so we could open the iso booth doors and use room mics.

Are there click tracks?

Oh, god; no.

How do you redo a section where it’s wide open?

With hi-hat counts. Dan’s counting the whole time.

And you edit those out?

We generally leave it. He does a lot of quiet pedal hat counts, and, to me, they’re part of the song. I miss them when they’re not there.

Whenever bands want to take that out, I’m always like, "Come on. It sounds like people playing!"

Yeah, exactly. I leave crap at the beginning and end of mixes all the time. Guitar hum on breaks. Always on purpose. But, inevitably, I get asked to take it all out. This is the nice part about doing your own band. No aesthetic debates!

We can take way more liberties when it’s our own music, right? We have this idea of how it all should go.

Totally. It is a relief not to have to communicate all your decisions for one record. There’s a spot in one of these songs where during a very quiet section you can hear Dan turn on his snares before a pickup. Is that staying? No question.

Are there overdubs or layers?

We overdub the little leads. And there are some guitar layers. They’re subtle. We write a lot of parts with ringy, open chord voicings, and they’re real pretty; but the detail can get lost in the pushed guitar sounds. We’ll often go back and add more “underdubs.” Sparkly, dirty-clean sounds.

It’s pretty hidden though, yeah?

It is, but you’d notice if I muted it. It adds depth. A lot of bands I work with want to add more and more. More scenery. My reptile brain does not want more scenery. It wants to make the scenery more detailed.

There’s a sense of depth, for sure.

That’s fucking great. Depth is hard with music like this. It is just four or five elements all going for it. That Sharkbite Studios live room helps. I tried a third room mic on the drums.

There’s a center mic?

Yeah, a [Shure] SM7 pointed at the far wall and crushed up a bit. It sounds great. The main room mics do their usual thing, and I can make them real loud, but then push a little of this in, darkened up.

Did you "distort everything" on Piecework?

I knew you were going to ask this. [laughs] I think the answer is yes. There’s a mono crush bus on the drums. Kick, snare, toms, going through one of those [Louder Than Liftoff] Colour modules and distorting quite a bit. There’s a parallel drum bus that’s a pair of [Universal Audio] 1176s with two buttons in. It's not quite distortion, but it’s pretty gritty. There’s PSP VintageWarmer 2 [plug-in] on the kick. There’s an SPL TwinTube plug-in on the guitar bus. The vocals are splattered through the Ampex 351 [tape recorder electronics] that I borrowed from you.

The vocals are saturated, but they’re not obliterated.

Yeah, I sang into one of those Extinct ribbon mics. That took distortion in a cool, tasteful way. I was riding the 351 input for different parts. If a section really went for it, I hit it harder. But that was at mix time. Since I’m tracking myself, I run through a plug-in for grit and it’s close enough.

I feel like tracking your own vocals is a gift. You’re at the helm, and no one’s around. You can try weird shit if you want, and then erase it.

One thing about tracking my own vocals, as a recording engineer, is that I remember everything about the experience so I'm dialed in to other singers when I record them. I sang parts over and over, then comped and comped, then re-sang again. Then I listened in the car and did it again. It was worth it. That’s the thing to remember. If someone can’t play a guitar part and we’re doing it over and over, that’s misery. But if someone is finding their way, that’s different. I went to England to record this band Wren, and their singer Owen [Jones] did the vocals in a day. He’s a great vocalist.

For the whole album [Groundswells]?

Yeah. Knocked them out. They were great. But the next morning, he said, “You know what? I can do better.” I was like, "Man, we’ve got other things to do, but all right." Then he absolutely beat it. I would have kept the first takes, for sure. Then, after I went home, he re-recorded a few songs, and beat his takes again.

Damn.

He just knew. But it’s a slippery slope. One of the things I tell bands is, "It doesn’t have to hurt." Play a song well, then move on. If they’re a good band, they don’t need to do 100 takes. But for these vocals, I did need to do lots of takes. I hadn’t sung a lot of this before, and I was working it out as I went.

Do you do vocals at practice?

I do, but we recorded these songs with no vocals written.

Whoa. That’s interesting. I would think that there would be some interplay where one thing would inform the other, and all that.

You would think. I wouldn’t recommend this. It was bad. But we’ve done this long enough, and we do try to write actual songs, so I knew the song structures and where vocals would go. It worked out. Still, it sucked.

Does the band come up with all these songs on the spot, together in a room? Or does somebody bring the seeds?

Jon and I wrote the majority of the record, sitting in my studio together with our guitars plugged into Pro Tools with that “Eleven” amp simulator. We both came to like our fake guitar sounds. [laughs] If need be, I program basic demo drums.

Why that, versus getting everybody in a room and jamming?

We found that it’s grueling to write this material, and it ends up being better doing it when it’s like, "Okay, we’re both down for pain."

How do you feel now that it’s all done? Are you happy with the record?

I am. I feel good about it.

That’s rare, right?

You know me. I’m rough on myself.

Would you call the mastering on this pretty involved? Usually, “Don’t master my record!” is your version of mastering instructions.

I should give a shout out to Carl Saff [Saff Mastering] for his mastering work. We did ten fucking mastering passes. In case Carl is reading this, I want to say, "I’m sorry." He is a patient, patient man! I will say, if I played you the first master, the record got demonstrably better for it. Carl is not a heavy-handed person, no matter what. But I try to get him to juice things up a little. He knows that I like him to push the low end. He’s also a typical mastering engineer, where he’s like, “The mix sounds good! If you’re happy, I’m happy!” He’s trying to make progress rather than open cans of worms. But I’m like, "Carl, I need more from you. Open that can." We’d go back and forth, and he was always very polite and Midwestern. Eventually I’d say, “Okay, I talked myself into it. I’m trying one more move.” It was like talking to a therapist. Me working through things out loud, with him mostly nodding. It made a big difference. For someone whose work one might consider mild-mannered, he improved this record. If nothing else, he accomplished that by being patient. I’m grateful.