Distortion Ain’t a Dirty Word
How Rich Costey and Chvrches learnt to stop worrying about the meters and trust their ears.
The Sound City movie really pissed Iain Cook and Martin Doherty off. “They just blamed ProTools for everything!” said a clearly outraged Martin. “For the death of creativity and spontaneity,” chimed in Iain. The pair are two thirds of Glaswegian synth pop band Chvrches who just dropped their debut album The Bones Of What You Believe. The tour cough had caught up with singer Lauren Mayberry, but a producer herself, it’s likely she would have been weighing in too.
“It was like they don’t get it, but in an equal and opposite way,” commented Martin. “Sure, it’s not just about digital, but anyone who says it’s not useful in some way, and that you can’t strike a balance marrying beautiful analogue signal paths to the almost endless possibilities of digital is short-sighted and a fool.”
The pair aren’t proselytising on behalf of Avid. They don’t even use ProTools. Nuendo and Cubase in the studio, and Ableton Live on the road, are their tools of choice. But that’s exactly their point. There are so many different music-making tools in the digital environment, so resorting to name calling is to miss the creative forest for the trees.
“You ask any huge rap producer and he’ll probably tell you he uses Fruity Loops,” continues Martin. “Really man? Fruity Loops? C’mon! But that’s the thing, it’s your process, your personality, the way you work. The software you use is important, but more importantly, it’s how you use it. And the guys using FL Studio are doing incredible things that are really difficult to do in a classic, linear multi-tracking situation.”
“We always talk about how gear and software inform the writing process,” said Iain. “Because a lot of our ideas come from a drum sound or sample or a synth sound, and that technology inspires our songwriting. Maybe we should try FL Studio.”
MANY DAWS TO ENTRY
The pair have been serial DAW hoppers. Before Nuendo and Cubase, Martin used Logic, back when Emagic still owned it. He eventually switched because he didn’t want to pony-up for a new Mac. These days it’s Apple laptops all the way, but they haven’t managed to gel with Logic again. Though when Chvrches crossed paths with Sigur Ros at Rich Costey’s studio in California, he was mixing their record in Logic because that’s what they recorded it in. Which got the wheels turning again. “If those guys can do their whole album in Logic,” said Iain. “There must be something weird about it.”
It’s actually a working philosophy of Costey’s: to mix a record in the same DAW it was created in. “I’m a big believer that whatever program somebody uses to record their music, is the one it should stay in,” he said. “Most people who mix records use ProTools and I do as well.
“I got rid of the TDM system and switched to the native card, which enables you to boot up whatever program you want and still access loads of analogue outputs. Chvrches recorded that album in Cubase, so I mixed it from Cubase.”
One reason Costey likes to keep it native is because more and more he’s finding himself mixing albums while bands are on the road or on the other side of the world. So it pays to have the same session as the band in case they’ve forgotten to finish a line, or didn’t backup a part properly in the haste to get to the airport.
The Mother We Share from Chvrches’
The Bones Of What You Believe
Case in point, Chvrches. The band had only just finished tracking in Iain’s basement studio in Glasgow and already their management team was pushing them out on the road. “When a band breaks on the internet, and not in an individual territory or one territory at a time,” said Martin. “You’re suddenly in a really fortunate position where people want you to be everywhere at once.”
There were even “a few things we had to do on the hoof, because we had to finish the album on tour,” explained Iain.
Iain and Martin are not only savvy musicians — having had successful indie band careers prior to Chvrches — but engineers too. “In both our own band contexts, we were always the guys who were put in charge of audio technology,” said Martin. “We both learned how to use computer audio technology from an early age. I think the first computer I had was an Intel Celeron 800. I couldn’t afford the P4. I was broke.”
“P4?” said Iain. “Try the P1 mate. That was my first one.”
Iain had recorded a couple of previous band’s albums in his studio, which is set up primarily for his TV and film post-production work. And the band mixed two of the tracks on the album, Tether and By The Throat. But you can’t fight fame if it’s knocking at the door. So as the band stepped out on an exhaustive tour schedule, they had to find a mix engineer. And not just any mix engineer. “Handing it off was a very difficult thing to do, but we just ran out of time,” said Martin. “Not to be conceited, but we felt we had enough knowledge in the locker to see it through to the end. And if we were going to get someone, we wanted someone world class. Specifically someone who would take it and look at it more as a rock recording than a dance or electronica recording. So then you’re really talking about one of three or four guys in the world — Rich [Costey], Flood or Alan Moulder. We were lucky Rich was free.”
Rich mixed Recover as a test track, and there was no question he was the right mixer for the job. In all, Chvrches were only able to attend a single day-long mix session. With “their tour bus parked out front,” said Costey, they recalled as many songs as they could get through. They spent the rest of the time watching vicariously through a live stream of the console, and communicating via Skype.
IN THE ORIGINAL BOX
“The way the Chvrches album was mixed was not unusual,” said Costey. “I’m often doing mixes for an artist who is on the road somewhere. Occasionally I’ll be mixing and there’ll be a keyboard or vocal part missing from the session. By using the same session they can easily send the part right from their laptop, and I open up the session and it’s fine. They don’t have to bounce anything out.
“We also keep all the plug-ins running live. Whatever ones they’ve got, I’ve got. So if we’re doing filter sweeps, or echo feedbacks, we can interact. They can do a pass of delay feedback and send it to me from wherever they are.
“The other reason I use the same DAW is I think each program has a peculiar sound to it and if an artist has built up their character and identity around using this particular program, not using it goes down the road of diluting what they’ve built up.
“For example, Logic has a very particular sound to it. You can pick it out a mile away. I worked on the first Foster The People album and for Mark [Foster], Logic is like an extension of his body, he’s so comfortable and quick with it. That album has a particular sound because almost all of it was done in Logic.
“I started this approach when I was mixing a record by The Shins for Greg Kursten, who produces a lot of pop records. I found bouncing out Greg’s tracks to ProTools wasn’t really working because he had so many different aux channels running into one another. With his drum sounds, when you bounce out the kick or the snare, you’re not hearing how all the different compressors react with the whole kit. So to me it didn’t sound right unless I just ran his sessions completely.”
He gets a bit of masking tape out and just starts taping up the meters and draws a reading where it’s nicely out of the red
BUILDING THE CHVRCH
Chvrches is built around the perfect pop marriage of Lauren’s pure, almost nymph-like vocal delivery sitting easily atop an aggressive, but spacious rhythm-driven synth landscape. It didn’t start out that way though.
Martin Doherty: It happened organically. Initially, I was singing on most of the demos, but they were just the sketches coming out of our heads. Iain played me some of Lauren’s stuff and we talked about a totally different timbre for the recordings. She came in, tried some stuff out, and we had a proper moment in the studio where we knew it was cool.
When we realised how well it could possibly work we started writing songs together with Lauren and she became an intrinsic part of the creative process. She allows us to push things harder because we have something to bring balance to an audio recording — a voice that’s sweet and a pure tone. As long as her voice holds that top line, the song can still be pleasing to the ear. That was something that clicked almost instantly when we heard her sing.
There are some ‘brick wall’ elements on our album but the classier you get the more you start to understand and realise that it’s about where the space is rather than filling up the space.
Iain Cook: It’s a really hard thing to learn. Sometimes you go down the line of filling up an arrangement until there’s no room left, and you’ve lost the original energy it had. In an alt-rock context you fill everything up with cymbals and loud guitars. For this project, it was about finding ways for synths to sound big like that.
AT: How have you done that?
IC: By layering up wide sounds with three oscillators at a really wide pitch base. Distorting things, using a lot of arpeggiators, making four or five passes with different voicings of chords and then chopping them up and stacking them into a sampler so you’ve got this huge sound that’s triggered with individual hits. We’re giving away all our secrets!
MD: When you don’t have a drummer accenting all the divisions between the kick and the snare on a hi-hat, and hitting the toms all the time, all this high end suddenly appears which you can either fill up with shite or let breathe. Or you can find new ways to accent those rhythms — a piece of percussion, some lines of deliberately chopped vocal.
We’re all about crude sampling, and place those samples in the triplet or eighth-note divisions. We put a lot of work into how the rhythm tracks are composed and it usually starts with a big dirty kick drum and a driven snare, which is usually three or four sounds morphed or amalgamated together.
IC: We’re always figuring out when not to fill up the space. Take Depeche Mode. A lot of the time they only push limited frequencies in their arrangements, but there’s so much room for the song.
MD: Or When Doves Cry by Prince, that’s more or less drums and vocals.
IC: No bass line.
MD: Realising that was a mind-blowing moment. And so many electronic recordings are clean and soul-less. It’s part of the reason we don’t use a step-sequenced or scene-based approach, because we want to retain some of the roughness of an alternative rock group recording, which is our background. The idea that everything is played, and some things are a wee bit behind. You can still use dance production and styles — like the whole song just being different notches on a filter or one keyboard line — but there has to be that organic and played human element for it to excite us.
The same thing goes for tonality. There was a really inspiring interview with Adrian Utley from Portishead we watched when we started on this project. He said he had no interest in sounds that are sonically perfect or beautifully in tune, and that appealed to us.
IC: Soundtoys’ Decapitator and Devil-Loc had a big sonic stamp on the record.
In fact, distortion was liberally stamped on all aspects of the record, which is somewhat of a specialty for Rich.
Rich Costey: “The main directive was to try and dirty it up and keep it from getting too clean. They wanted it to be powerful in a way that most people don’t think of synthesiser albums sounding. Most clients I get want their albums to work on many levels, but shiny-ness is usually not one of them.
AT:: What are the fundamental reasons behind dirtying it up?
RC: In the case of Chvrches, they wanted the music to have a slightly more punkish approach and more power than a traditional synth record. When you add distortion to the synths and drums, and put everything back together, they start to live in a state they wouldn’t otherwise have. Things are slightly less discrete sounding, on purpose. I’m sure it’s partly due to the band being constantly on the road too. They probably wanted their record to sound a little more familiar to them, like what they’re hearing coming off the stage every night — to have some of that energy.
AT: What are some of your dirty tools?
RC: A go-to plug-in on that record was Decimort from D16 Group. It’s like a bit-crusher, but it’s got some filters and a built-in distortion. It’s supposed to mimic the sound of an E-mu SP-1200 or something. I don’t know if it sounds anything like an SP-1200, but it sounds good. I would use that and Soundtoys’ Decapitator a lot. I’ve also got an ARP2600 I would occasionally run stuff through because the preamp sounds really nasty.
Occasionally we’d clip a couple of Neve channels, but that’s mostly it. I used a lot of plug-in distortions, because we upload a mix and then we might not hear from them for a couple of days, so we wanted to keep things easy to recall.
AT: Have you been through a lot of saturation devices over the years?
RC: In any given case the type of distortion has a huge effect on the presentation of whatever it is you’re distorting. But types of distortion to me are more like fads. I get excited about using one on everything for a while, and then I move on to something else and use that on everything.
I had an EMS Synthi A for a long time and went through a period where I ran almost everything through that. It’s the world’s most expensive distortion box. The Synthi A doesn’t come with a keyboard. It’s a nightmare even trying to get that thing to play an octave in tune, so it’s basically just for external processing. I still occasionally use a Symbolic Sound Kyma system for distortion.
AT: How did you come to use the Kyma — it’s not something you often come across in pop record mixing?
RC: I’ve had the Kyma maybe eight or nine years now. I remember reading about it and it seemed like a really difficult piece to get your head around, but if you spent time with it you could probably get it to do almost anything. That sounded really appealing to me, so I went and got one.
It’s similar to Reaktor in that it has modules and sub-modules that you can route into one another to create samplers, keyboards, processors, whatever you want. It’s incredibly deep. It has its own sequencer but the sequencer isn’t for recording notes, it’s for recording changes in the processor over time. I have to admit I have been a bit too busy to ever really master it, but I can fumble my way around the thing enough to make pretty reasonable use of it in a session.
Having never worked with Rich before, Chvrches had some big-name preconceptions of the talented mixer that probably wouldn’t be much different from those anyone else would hold — vintage gear piled to the ceilings, and perfectly honed processes. Their in-studio day was a bit of an eye opener.
IC: We had this idea that Rich used the most expensive outboard vintage gear. He does have a lot of that stuff, but much of the time we’d ask him how he got a particular sound, and it would be some freeware distortion plug-in he just downloaded. What drives that guy is just what sounds good.
His whole approach to this album was quite surprising to me. It’s amazing the way he throws anything at it and doesn’t give a f**k what the meters are saying.
MD: In the studio with him and looking at some of the meters, we were like, ‘Is that cool?’
IC: Have you actually taken 16dB off that whole mix there?
MD: Then he gets a bit of masking tape out and starts taping up the meters and draws a reading where it’s nicely out of the red.
Another thing we couldn’t fathom before we got there were his vocal sounds — it sounded like a 100 grand. Then we got there and saw his vocal chain… and it’s worth 100 grand!
RC: I don’t know if it’s quite that much, but you’ve got to get the vocal right. If you can’t get the vocal right, people are going to stop hiring you pretty quickly.
AT: So what did that vocal chain look like for Lauren?
RC: The vocal will first go through a Lang EQ, and then the Avalon 2055, then split into three channels on my console. When you set the frequency to the uppermost limit on the Lang, which I think is 20kHz, and turn up the top end, it seems to add harmonics on top of the vocal without affecting the content of the vocal. In other words, you can sort of brunt it up without it getting more percussive on the Ss or the Ps.
Then each split has a different piece of outboard gear. One channel has an UREI 1176, the other channel will sometimes have a Fairchild on it, sometimes just an API EQ, sometimes a Teletronix LA2A, and sometimes a Universal Audio 610 module from our vintage UA 610 desk. It depends on the song.
The third channel is often a Standard Audio Level-Or 500 series module. I wouldn’t always use it, but it’s there if I need it.
Then I mix them together. I run a modified Dolby channel on a parallel which also brings out the top end of the vocal a little bit without it getting too ess-y.
I also have another compressor, a Neve 33609, that all the vocals are paralleled into. It glues them together a bit.
AT: Are you compressing them much, or is it more about the combination of sounds?
RC: It’s more about the flavour of the devices because, honestly, most things come in reasonably compressed these days, no matter who the artist is. They don’t usually need a whole lot of level control.
AT: Was keeping Lauren’s vocal pure, compared to the dirty synths, a focus of yours?
Rich: If her vocal was distorted on top of the background, you’d start to lose the character of what she’s doing and it would start to sound a bit of a mess. You can run a dirty synth and a dirty vocal, it just depends on the style of music. These guys are writing pop songs with an old-school punk energy, so I tried to get the tracks to sound really clobbering and have her vocal in there, but not too loud. They were always mindful of making sure she never got too up in front of the track.
Notably, nearly all the vocal delays are plug-ins, except for a little bit of TC Electronic D-two. I’m really into Fabfilter’s delay plug-in, it’s awesome. And there are a few passages on the album that have really effective filter delays using Soundtoys plug-ins. We did use all hardware reverbs — AMS, and Lexicon 480 and 224 reverbs. There are a couple of plug-in reverbs I think sound okay, but hardware reverbs are much better whenever you can get them. Even if they’re cheap and shitty, they have more character than a plug-in.
While the vocal chain might not have topped the 100 grand mark, Costey’s mix bus surely does.
RC: The mix bus is terrifying! It’s something you don’t enter into lightly [laughs]. All the gear is on a super high-end power conditioner because you just wouldn’t try this kind of setup without it. If you get even the slightest mismatches between gear you’re going to have loading problems. This whole system has taken a long time to get to this point and everything has been A/B’d to the billionth degree to make sure it’s working.
It starts out with a couple of Fairchild 660s, one on each side. Then it goes into a Shadow Hills mastering compressor. Sometimes with Chvrches there would be a pair of Esoteric Audio Research 660s, and then into a Maselec EQ into Millennia EQ, which would sometimes be swapped out for the GML EQ. And then for most of the Chrvches record it would go into a pair of 610 console channels to boost the bottom end.
The Fairchilds are not compressing at all, it’s just about the sound. It’s hard to explain. The Fairchild does the same kind of thing the 610s do. I’m not going to them for a particular compressor or EQ, but it does this weird thing where it makes the music sound more important. It’s the only way I can describe it. And then we’re compressing a bit on the Shadow Hills depending on the song. Some of the songs I would hit at a harder ratio and some would be a ratio of 2:1.
AT: Sounding ‘more important’ is a little vague. Is there any factor you could put your finger on?
RC: With digital recording, there’s always a bit of a battle. In many ways it sounds empirically better — if you’re running really good converters and a good clock — than your average analogue recording. But when you have 80 or 90 tracks of digital source material running, then everything starts to sound kinda small and ill-defined to me.
The intent of the artist is better preserved in analogue recordings than in digital ones. In digital recordings you’re always trying to enhance the intent. If you listen to Back in Black and it sounds tough as f**k, then you are correctly receiving the intent of the artist. So the goal is to make sure that intent is coming across. Certain pieces of outboard gear help that intent come across.
JUST TAPE OVER IT
AT: Do you really tape over the meters to placate your clients?
RC: Yes, that’s true. I think those were actually the meters on my mix bus, the level coming out of the desk.
AT: Do you drive things into that territory as a matter of course?
RC: It depends on the music. That sound, where it’s not about headroom and it’s just about drive, works for certain pieces of music and not for others. The first time I did it I was mixing a Foo Fighters record. Dave [Grohl] had made a few records at that point and he noticed the pegging meters, so I started putting tape over everything.
AT: To make him feel more at ease?
RC: Yeah. Then I started drawing in meters where they probably should be, so he’d stop worrying about it.
AT: Is that just more a function of your gear and how it performs. Or would you recommend or not recommend people trying it out?
RC: I don’t think I could recommend anybody doing anything, but for a long time my meters would be pinning on a console, and I kept thinking it was because I was terrible and didn’t know what I was doing. Then I would pull the faders down and start again, and every single time I’d end up right back where I’d started. And after a while I assumed I must just like the way it sounds.
But it doesn’t work for everything. Some songs need something that has a bit more detail to it. But if the goal is to sound more arse-kicking, then often that helps.
AT: You mix a lot of rock records, but you obviously have a way with electronic music, specifically the interaction between rhythmic elements and low end. Is there something special you aim to achieve down there?
RC: Compared to a traditional rock album, that’s the whole key to the song: how the bottom end is moving you. And you usually spend a good deal of time getting that right. Having mix bus compression has a big impact on how that feels. How hard you’re hitting it and where the recovery time is set can affect the track if it’s pulsing or reacting to the bottom end elements.
Sometimes if you have a lot of action in the kick drum but you’ve got a lot of sub synth in at the same time, then you usually have to put in a high-pass filter on the detector circuit of your compressor so that it’s not getting whacked every time a sub synth drops in.
AT: So it’s more to do with how it pulses rather than simply how loud it is in the mix?
RC: It’s how it mashes together. I can say I don’t usually compress the keyboards. If it’s a bass synth I wouldn’t really compress that too much, if ever. It will get compressed through a mix bus compressor and as part of a group, but if you start compressing your low end that’s when you begin to lose it.
It’s about balance and feel. The drums most certainly get some sort of parallel compression to make them sound bigger but if you start running a lot of drum compression it will change the length of your drums. With this kind of music, everything is designed to pulse in just the right kind of way, and if you start making the kick drum longer through compression, you’re going to mess all that up.
The goal of a compressor is to make it sound like someone is playing harder — I think that’s a noble goal. But quite often I feel like people just use compressors because it makes shit sound louder and in doing that you’re actually flattening out the music. I’m using a lot less compression on individual instruments than I ever used to. I can’t tell if that’s a factor of my ears changing or what people are sending me is more compressed than it has been.
AT: So your approach with synths is not to compress them, but to widen them?
RC: If it’s a major synth riff I wouldn’t want to use a widener plug-in because it starts throwing the signal out of phase and you’ll lose impact. But if it’s more of a textural element or a dreamy element, wideners help things stand out and generate their own space. I use Nugen Audio’s Stereoizer plug-in quite a lot, and occasionally a Brainworks plug-in, on synths.
PICK OF THE CROP
MD: Why do you think Chvrches picked you to mix their debut? Having never worked with you, it couldn’t have been working methods or a relational bond.
RC: I might have some kind of sonic imprint, but it’s not on purpose. In other words, there are things I like, but I don’t think other things I didn’t do are wrong.
Everything is driven by the music that’s coming in. I’m pretty tough on myself and always trying to improve on whatever it is I’m doing, and will stop at nothing to make an improvement. I’ve wanted to sit behind a pair of speakers and listen to music all day since I was five. That’s all I ever wanted to do.
I was Philip Glass’s engineer for a few years and learned a lot of technical things. Rick Rubin liked the sound of a couple of records I’d done, so for a while I mixed records for him. That was interesting because he does not listen intellectually at all, only from his gut and instinct. Previous to that I listened almost entirely intellectually. He’s an incredibly intellectual person, but when he’d listen to the music he’d literally just listen to it like a teenager.
The goal is to make sure it’s hitting you in that way, because when other people hear the mix or the song for the first time, they’re going to listen to it from instinct. The goal is to get all the technical shit out of the way so you can only work from instinct.