The Foo Fighters' 'The Pretender' is one of the most successful alternative rock songs in the history of American music. It enjoyed the longest ever spell at number one in the Modern Rock Tracks chart (18 weeks), it was the most played alternative rock song on US radio in 2007, and as this article went to press it had been nominated for three Grammy awards, for Record of the Year, Best Rock Song and Best Hard Rock Performance. The song didn't quite break the hegemony of the all-pervasive hip-hop/R&B genre, only reaching number 37 in the regular Billboard charts and number eight in the UK. However, it did help propel the album on which it features, Echoes, Silence, Patience & Grace (also boasting two Grammy nominations, for Album of the Year and Best Rock Album), to the top of hit parades around the planet.
Clearly, the Foo Fighters hit a home run with 'The Pretender'. Whether this is due to the in-your-face arrangements and performance, the exuberant chorus, the political subtext or the confrontational video is a moot point, though one imagines it was a bit of each. The success of the song is impressive by any standard, but neither mixer Rich Costey, nor the band, nor producer Gil Norton, saw it coming. Costey, who has become a highly in-demand producer and mixer in recent years, specialising in artists known for their musical integrity — such as Muse, Mars Volta, Fiona Apple, Franz Ferdinand and Interpol — mixed the whole of Echoes, Silence, Patience & Grace.
"While we were mixing all the album's songs through," recounts Costey, "we didn't really have a conversation as to what would be a single. I mixed 'The Pretender' later on, and we knew it was a strong song, but for us it was just another mix."
Costey was speaking from Avatar Studios in New York, where he had been on mixing duties for most of the winter. The mixer and producer is unusual in that he tends to mix entire albums, rather than just individual songs. "I think I was hired to mix the Foo Fighters' album because they wanted a second, or perhaps rather a third or fourth, pair of ears on the project. There are two schools of thought in mixing, one being that you give the mixer something raw that he cooks the hell out of, and adds loads of herbs and spices to make a dish out of it. I have done projects on which this was the case.
"My work with The Foo Fighters fell into the second category. They had already done a lot of prep cooking and things were in very good condition. So I had to preserve what they had done to a fairly large degree, and this meant that my work mainly consisted of balancing and rides to get the dynamics to happen. Obviously, a certain amount of processing was still necessary, but that was more to enhance what the band, Gil and engineer Adrian Bushby had already done. It's sort of like the mixer's Hippocratic oath — the first rule is Do Not Kill The Patient. I think there are all sorts of people out there who are killing the patient. I try very hard not to add to that problem.
"There were two main challenges in mixing 'The Pretender'. The first was a general one concerning the Foo Fighters, namely their endless walls of guitar overdubs, almost like a swarm of bees. It is something the band goes back and forth about, because Taylor [Hawkins] always wants the drums to be louder, and Dave [Grohl] prefers the power of the bees. The guitars on 'The Pretender' are quite full on, with countermelodies and so forth, and they all tend to be in the same range, so it gets quite dense. The challenge of this type of mix is to retain the power of the track, yet define a space for everything. Handling the guitar balance was a slight chore, and in comparison the drums and vocals were quite easy.
"The other difficulty with 'The Pretender' was in getting all the dynamics to work properly. The song starts out with a very stripped-down intro — a small string section, guitar and vocals — then ramps up into the chorus, after which there's another breakdown, then it ramps up again and there is a series of breakdowns in the middle eight, including a repeat of the intro, and finally the last chorus with octave guitars coming in, that you really want to be able to hear on top of the wall of everything else. The only way to do this type of mix is to simply do loads and loads of fader rides. I don't think any amount of EQ can substitute for doing rides.
"That's not to say I'm shy on the EQ. I'm not sure where I developed this habit, but I don't cut many frequencies when I mix. I used to cut a lot when I was younger, but eventually I almost entirely stopped doing it. It wasn't a conscious decision; I guess I just don't like the sound. I'll cut a little bit on the drums sometimes, and with vocals I occasionally need to control problem areas, and I also may be quite brutal with synthesizers. But that's about it. I know there are people who cut frequencies to try to layer a track, but look at it this way: if you are conducting an orchestra, can you filter the woodwinds so the oboe doesn't stick out so much? What you do is have the oboist play a little bit more softly, so you can hear the rest of the reeds."
"If you look at the 'Pretender' Edit window, you see that it's really neatly organised and ready to be mixed for someone who has never seen the Session before. That's Adrian Bushby's doing. I'm a bit of a slob with my tracks; if you see a Session that I produce it will look like a mess, because I know where everything is. I have assistants that help me organise stuff, and they will also do this when a Session comes in for mixing. I'll then join them when it comes to laying things out on the console. Often I'll also have to sort through the Pro Tools Session anyway, before I get far into mixing, because frequently there are more music tracks than console inputs. I'll spend a while familiarising myself with what's there and developing track-sharing schemes.
"I think on most of the album there's a little bit of guitar track-sharing, as it was mixed at The Pass in Los Angeles on a Neve 8078 console that had 40 full in-line channels with EQ and 32 monitor channels. 'The Pretender' had 70-odd tracks, so if I wanted the possibility of EQ on each track, some sharing of similar tracks was inevitable. I don't mind doing that, but I rarely submix in Pro Tools. You might later find that a viola in the string section is obscured by a baritone guitar and you want to be able to adjust for that.
"After getting a sense of what works and what doesn't work in the rough mix, and what the artist wants, I'll start a mix by pushing all the faders up for the vocals and the rhythm section and so on, to see how different parts are interacting. The lyrics and vocal delivery greatly influence how the track will sound. The dynamics of the vocal absolutely shape everything. The vocal should be driving the band wherever possible. So while mixing it's pretty critical to have the vocals in there early on, otherwise you might find when you push the vocal up later that you have to start all over again.
"I tend to start with getting the tones for the vocals and different instruments right, using various combinations of compression, EQ and distortion, and after that it's just rides, rides, and panning. My strategy usually is to mix in sections. Once the sonic relationships between the instruments are more or less stable, I will start getting the intro to sound great. I try a few different approaches, because with any song, if the intro isn't great it doesn't matter what happens to the rest of the song. That's something I learned a long time ago: if the song is going to work, the intro has to grab you immediately. The importance of this cannot be overstated.
"Once the intro makes sense, you also have a template for the way the rest of the song works. I'll work on the verse, and then the chorus, and so on, and gradually build up the track. Then I start listening to everything from the beginning in context and make sure all the transitions are OK, and then I try to push it up to the next level. That's all just rides. While I'm working in sections, I listen to everything, not just to individual tracks."
Artist: The Foo Fighters
Writer: David Grohl
Producers: Gil Norton and the Foo Fighters
Drums: Quad 8 310, Pultec EQP1, API 550a, Neve 33609, API 2500, Lexicon 960, Urei 1176, Smart C2, SPL Transient Designer, room mics
"I had just gotten some Quad 8 310 EQs before I began the Foo Fighters project, and I found that cranking the shit out of the mid-range with the Quads helped the drums quite a lot. On most of the album the bass and snare drum went through the Quad 8s, with severe amounts of mid-range added. Quad 8 grew out of Electrodyne and was very popular in the 1970s, particularly with film companies. I obtained some channels that came out of the Motown Sunrise console and they're basically three bands of EQ and a mic pre and output level, and the Q gets tighter as you push it up. The Quad 8s are a little bit rounder and warmer-sounding than similar API modules. I also used the Pultec EQP1 and API 550a on the bass drum.
"I had a bunch of different side-chain compressors on the drums that would change from song to song. On some songs it would be the Smart C2, medium ratio and fast recovery, on others the Neve 33609, the Urei 1176, the API 2500 or the Empirical Labs Distressor, and the SPL Transient Designer on toms. You can use the SPL to lengthen the sound of the toms. I don't compress all the drums at the same time, I'll compress individual parts and mix the compressed sound in with the natural sound of the drums. There was relatively little compression implemented on the drums in this song, because the band didn't care for it. They wanted the drums to sound more raw.
"Throughout the album mix I might have used a bit of Lexicon 960 on the drums for reverb, but the room mics — amongst them a heavily compressed foldback microphone — were so good that I tended to use those. I tend to like room mics that are on the darker side. If they're too bright, you can't turn them up very loud because you then also get all kinds of messy cymbal noises. One other thing to note is the drums that come in after the intro of the song. When the whole band is slamming accents like that, you want to make the drums sound very aggressive, and this meant pushing room mics, pushing compression mics and so forth. As I said, there were a lot of rides."
Bass: Neve 1073, Pultec EQP, Chandler TG1
"The bass consisted of three tracks: one recorded through an Ampeg SVT amp, one through a 4x12 Marshall cabinet, and one put through a Sansamp. I usually push all three tracks up and adjust their balance if I want different colours. I'll also mix in some side-chain compression, which in this case was a 1073, Pultec EQP, and Chandler TG1. The side-chain will usually be bright as hell and really compressed. One of the interesting things regarding mixing bass is that you may think that the bass sound on its own is fairly bright, but when you add the rest of the instruments you find that the top mid-range of the bass is usually missing. It's sometimes quite shocking to realise how much top end you need to add to a bass to make sure it cuts through a track. If you then hear the bass sound in isolation it may sound pretty uncomfortable, but in the midst of a swirling din of a dense track, that amount of top end usually works just fine."
Guitars: Neve EQ, Urei 1176, Neve 33609, Gates Sta-Level, API 550a
"Dave's guitar tone is a little bit more distorted, with more bottom end in comparison with Chris's. So there is method to their guitar madness, where they try to get everything to fit in with each other. I EQ'ed the guitar parts a little differently on the board to make them fit, and did rides when there were melodies that needed to be featured. Another way of creating separation was through panning. I tend to pan the rhythm guitars hard and the melody guitars slightly inside them. Overall the guitar parts sounded great, though, so they really didn't need a lot of processing. I tend not to compress distorted guitars very much when I'm mixing, because they've usually already been compressed. It's strange, but to my ears, when you compress distorted electric guitars while you're mixing, the whole mix starts to sound overcooked. Of course, it depends on the song; I don't have any rules.
"In the case of 'The Pretender' I had a couple of Urei 1176 compressors on Dave's guitar, to pump them up a bit and make them sound more aggressive whilst barely compressing. I had the Gates Sta-Level compressor on the clean guitar in the intro, with quite a lot of compression and a very slow recovery, plus a 33609 with a very short recovery time on Chris' rhythm guitars, to beef them up a little bit. I worked hard to make the octave guitars in the last chorus stick out. I ended up EQ'ing them with the 550a, in addition to the console, to make them cut through this formidable army of rhythm guitars. The rest was just rides. The stereo phase and flange guitars used in the track already had those effects on them."
Strings: room reverb, Neve BCM10
"There were five tracks of strings and two stereo room tracks for the strings. I know I said that I don't do usually do submixes, but I actually did submix the strings through my Neve BCM10 sidecar, and brought them up on a couple of faders on the main desk."
Vocals: Mercury EQH, UA 175, Roger Mayer RM58, Waves DeEsser, Digidesign Reel Tape Saturation, Sound Toys Echoboy, Pultec EQP1A, SPL De-esser, Neve Portico 5042
"If I recall correctly, the lead vocal is doubled throughout the song. I used a Waves DeEsser plug-in on it, which I'm quite a fan of, as I've not heard a hardware de-esser that is competitive with it. The main signal chain after that was a Mercury EQH tube EQ and then a couple of vintage UA 175s. I hit the background vocals with the Roger Mayer RM58. I haven't mentioned any plug-ins yet, apart from the Waves DeEsser, because I hardly used them on 'The Pretender.' The only other ones I used were the Reel Tape Saturation and the Pultec EQP1A on the intro vocals, and the Sound Toys Echoboy on one of the end vocal overdubs.
"I used board EQ on all the vocals. The Pultec EQP1A plug-in on the intro vocals was to help them cut through. The Reel Tape Saturation was a plug-in that I had just bought. I wanted to warm the vocals up a little bit with it and make them a little bit crisper. Nothing too distorted. The Echoboy plug-in was used on a vocal overdub at the end for a tight delay. I don't tend to use plug-ins that much; they're really not that interesting, in my opinion. Finally, I also used an SPL De-esser on some vocals, and the Neve Designs Portico 5042 for a bit of crispness. It has its own sound, and I used it a lot on the whole Foo Fighters album.
"To complete the picture, I used an EMT plate and the room mics for general reverb, and my two EAR 660 limiters were my bus compressors, together with the Manley Massive Passive. We printed the mix to the ATR102, on half-inch analogue tape, from which it was mastered."