"I do a lot of riding anyway, so I'm not relying on the compressor solely for level control. It's usually more for the sound, and the SSL compressors are pretty aggressive-sounding. One exception to that is that I used to like to use the old LA-2As for vocals. When I was recording, I would almost always use them."
"I listen to the song and try to get a feel for it. If it's a rock song that has a section that's really rockin', I'll sometimes go to that first and just work on that section, getting the band to really kick ass. I'll do that just to kind of make sure that that's my level. Then I'll work it down and build up from there, but that's the place that's really got to be rockin'. And that's usually where I find out how hard I want to hit the quad compressor. I mix through the compressor right from the beginning while I'm still working on that loud section." Note: Consistently set at 4:1 on automatic release running between -4 and -6 db.
Begin by defining the impact. And I usually do it pretty fast. Usually, when I'm working on a mix after I've gone through and done my little road map and decided, “I'm going to work on this part of the song,” within about 15 or 20 minutes, I've got the basic energy slamming the way I want it to sound or pretty close. Then I'll just keep going over and trying to find out how loud I can have this guitar and have it still feel right or how quiet.
On Cymbal Rides:
I ride them a lot of times because sometimes a cymbal won't be as loud as another cymbal or something. So there's that. But also, in a greater sense — and I think that this is what you're referring to — every cymbal crash will be ridden up maybe 5 or more dB.
I use the samples more to drive reverbs. If you killed the reverb, you'd still hear the sample. And the thing I like is that I can EQ them so that I can really tune the ambience and where it sits in the whole frequency response.
Again, more so than I can with the overheads because I usually EQ those so that the cymbals sound the way I want them to sound. Not always, but often, when the cymbals are sitting where I want them to sit, I'll hear more ambience from that. I'd rather keep that down and be able to shade with a little more control using my ambient sample.
Just ride the cymbals for the actual hits and then after I end up getting the guitars happening, the vocals in and everything, I'll find that the cymbals to be effective need to be a little louder. But, as I said, I don't like to flood the drum sound with too much ambience. It depends on the nature of the song, but especially if it's a dense song with a lot going on in it as opposed to an old Led Zeppelin thing with one guitar and bass and drums for miles. You can get away with a lot more interesting ambience with that kind of a thing. But there are not that many records out there that are that sparse.
"I will go through and listen to individual tracks to get a sense of what's going on. I may adjust some group balances or flip the phase on some guitar mics. In general I'll familiarise myself with the audio bits that are there. I'll usually be throwing things up really fast, almost like doing a quick monitor mix, and balance everything very quickly, just to see how all the elements are supposed to sound together, and I have a basic feeling of the entire track."
"After that I'll focus on the things that really need attention. I'll look whether samples are provided in the session, usually kick and snare, and I'll listen to them and how they work together, and will try them in and out of phase with the kick and snare mics and so on. But it's not a matter of initially working on just the drums until they are perfect, more of listening to everything and then seeing what catches my ear. I also like to cut channels when there's nothing going on. Once I have things mapped out with all sections of the song feeling pretty much the way I think they should sound, I'll do a first pass of the automation. As I carry on refining the mix after that, I'll try to preserve the basic feeling that I have for the song and not to disturb it. The whole process from when I first sit at the board and start listening to what is in the session, until the time when I switch on the automation, usually takes about three hours."
"Riding the faders is definitely a feel thing. If I feel that a certain balance is not quite correct, I'll back up and ride everything again, until it feels right. In the early stages, while I am still roughing out the entire mix, I will document all the levels in the different sections with small pencil marks next to the faders. I will spend a lot of time listening to certain sections loudly, to make sure things have the right impact, and also very quietly, to make sure the balances are really coming out well. At some point, I will spend more attention on the drums as a unit, and will get them working together with the bass and the guitars to get that basic band sound. I'll usually focus on a larger section of the song, like the verse or the chorus or a solo section, and I'll often have that section loop to be able to go over and over it until it is really moving me and rocking and kicking ass, and so on. It's the reason why I don't work in automation yet, because I don't want it to be writing mix settings all the time. It is just me adjusting levels manually and making pencil marks.
"After about three hours of work I'll be at a point where I'm ready to commit the mix to automation. I'll be clear on the general outline of the mix, and the detailed rides are becoming too complex to handle manually. Once I am working in automation, I'll go into more detail. For example, when there are many guitars in the track that have to work together with the bass, certain notes in the bass track will often stand out more than others. Compression just does not level that out for me. When I compress the bass, it tends to be for the sound rather than for the levels. Instead, I'll spend a lot of time riding certain notes up that seem to be getting lost, and notes down that are jumping out too much. It also makes more sense to get into this kind of note-by-note detail after I have most of the other elements in the mix balanced, because you can hear the details in the context of the mix as a whole. But even at this stage I will sometimes still drop out of automation, if I want to try something new, and I don't want to keep cancelling mixes. But overall, this stage is mostly a matter of fine-tuning the automation, riding this bit up a bit, swoop something else, maybe have a part 3-4 dB louder when the chorus hits and then bring it down gradually. I'm playing around with things that I feel add to the drama and architecture of the mix.”
"I tend to use effects in relatively broad strokes. It's not unusual for me to really make a compressor slam, to hit it pretty hard. I'll often put it on an extreme setting, to get a sense of what the ballpark is, and then back it off to what sounds right to me."
"Regarding gating, I often engage the SSL gates when working on the kick and snare. I'll start by listening to the phase relationships between the miked and sampled kicks and I'll play around with the compression as well, to see how it affects the attack and body of the kick sound. I use the gate to clean up ambient noise on the kick mic, or sometimes to trim the body, so I get a punch rather than a longer sound. Depending on what the sample sounds like, I may shorten the end of the decay with the gate. Sometimes there's a click or a bump at the end of a sample that's interfering rhythmically, and I use the gate to clean that up, but occasionally I'll use a super-fast attack on the gate to create a small pop at the beginning of the sound. If I don't have a sample that fits, and the kick drum has a kind of mushy attack, I will adjust the threshold and the attack to sharpen the transient. This is tricky to do, because it varies a lot on the level of the kick drum, and that's why I do this less often.
"I may also throw in a kick sample to generate reverb, and I will then gate that, to adjust the reverb. I don't add an ambient kick sample to replace the kick, but instead I gate the miked kick so it's shorter, and the kick sample will give me a controlled ambience. I send the sample to various reverbs, sometimes drastically EQ'ing it, sometimes not much at all, to create an ambience that will fit the sonic landscape.”
"Occasionally the kick needs to be bone-dry, if it's for a techno-electronic track for example, but for rock tracks I like to add a room sound. Again, this is to do with the architecture and landscape of the mix. If it's a rock track I like the sense that everyone is playing together in the same room. I always imagine the listener being in the same space as the performers. If things are very dry and clean, it can sound unnatural and lightweight.
"I love live performance. This doesn't mean that I try to mix all records as if they were recorded in an auditorium, but I do whatever makes sense to me to place the listener in a space with the band, whatever that space is. Sometimes it will be a big auditorium or stadium sound, sometimes that of a small club, sometimes it will be the type of recording where you have many different ambiences going on that are interesting dynamically and sonically, without placing the band in a specific place. Sometimes I'll put the band in a room just for a certain breakdown, purely for effect.
"When I came up in the 1970s, all studios had very dry rooms, because in those days the idea was to just pad everything. There was something cool about that, and I still like that dry sound. One thing is that it gives you a lot of clarity; it's an easy way of getting definition. It puts on more pressure to make sure the dry sounds are right. When I start putting a mix together I'll usually go for that dry sound, because I'm trying to get it to sound like it's hitting me in the chest and punching me. I may use gates to make things like the drums sound really dry. Later on, during the stage when I'm working with the automation, I will fill out all the ambiences and reverbs and echoes and whatever, until the mix feels right to me.
"That first stage, when the mix is dry and I'm going for punch and clarity, is done while monitoring at higher volumes. I listen loud to get a sense of the general impact feeling, and to work with the low end. I spend probably the majority of my time listening at low volumes, and sometimes at extremely low volumes, to the point where if there's any noise in the room, it'll distract me. I'll almost welcome ambient sounds in the studio because it's like listening in the real world.
"I have often described my philosophy as me wanting to make sure that the mix still sounds good if you are listening to it on a ghetto-blaster on the beach that has been kicked over in the sand. I want to make sure that you can, even in that scenario, still hear all the things you are supposed to hear. In my experience, listening at low volume is the best way to fine-tune my balances, making sure that I can hear that one small guitar part that plays an accent in one place, and that all the details of the reverbs work well.”
On FX:
Distressor on the vocals, but also the SSL compressor.
Reverbs Lexicon 480 for a longer reverb
PCM70 for smaller rooms
PCM42 for delay
Eventide H3500 for doubler
I usually have three or four 42s set up to do various things. One of them will work in conjunction with the longer reverb to give it a bit of a pulse kind of thing. All these effects would be on sends on the board, apart from the Distressor, which I tend to have on the inserts. In general I used the reverbs subtly on the vocals and lead guitars, and occasionally on one or more of the small synth parts. But sometimes these elements are bone-dry. I use delays on the rhythm guitars when I want to create a double in certain sections and pan the original track and double hard left and right. Normally I like to create these things from the performances that are already there by taking the double from another moment in the song where the part is played, so I actually really get two different performances. But in cases where the part is only played once, I'll use the delay.
"I use an Alan Smart C2 outboard compressor on the stereo mix. I found out a long time ago that when a mix is broadcast and really squashed, it changes the dynamics of the low end in particular, and if I can do some of that compression ahead of time, I can minimize the damage done by extreme compression. I like the SSL G-series compressor, and the C2 also gives me that sound. I ride the master fader during mixdown, and put the compressor before the master fader, so the compression isn't changed while I'm working. If I push the master fader up by 1.5dB I want the mix to get 1.5dB louder! I am very concerned with preserving the dynamics of each mix, as it is part of a greater picture that I try to paint. Riding the master fader during the mix is part of my emphasis on the sonic architecture of the mix and of maximizing the feel that I get from it.”
On Mixing Nevermind
"I mixed the album at Scream Studios in Los Angeles, using methods that are fairly similar to the ones I'm still using today, with an SSL G console and not a lot of outboard gear. While mixing I did not try to polish it at all. I was trying to keep the sound in your face and raw, while at the same time providing the dynamics and architecture that I like to hear. And I'm still very happy with the result.”
