Philosophy Of Mixing: Views From The Pros

When it comes to mixing, there are no rules, and all engineers and producers have their own preferred approach. We kick off this 'Mixing It' special feature with a look at how the professionals mix it.

<h3>EUGENE ELLIS</h3>

On mixing Soul II Soul's 'Joy'

"When people talk about going for a mix, this usually means balancing what's on tape and perhaps adding some effects. In our case, however, it's a totally different concept. It's almost like rearranging the track; putting new sounds on, changing the drums, and this carries on right throughout the production process. It really amounts to a constant state of mixing. Then, when the last thing has been laid down, the finishing process is very quick.

"I didn't really go to town on outboard gear for 'Joy'. I put the usual compression on the vocal, used the Lexicon 480 for some general purpose reverb, and put the Summit compressor on bass just to get that extra kind of rumble which I like. Overall, however, there's not much in the way of effects on either the album or the 12" version. It's just straight down the line; a listen‑to‑the‑vocal kind of song.

"When I approach a mix, the philosophy is to decide on what is going to be the most important element and then create the track around that. This task was made much easier on 'Joy' by having a great vocal to start off with, and everything else took shape after that." Richard Buskin

Stephen Street

On his approach to mixing

"From time to time everyone needs to fix something in the mix. But if it's not sounding right after four or five hours, there's something seriously wrong. I use automation because I like to be able to do a move and then forget about it. I work quickly on my mixes and I don't spend days on one track — I like a mix to be there in 10 hours maximum. It should be quick unless you are mixing something you have never heard before. If it's your own track and you have always had a clear image in your head of what you want to do with it, I cannot see why it should take much longer than eight hours to mix. Sometimes people do rough mixes that are great and they try and re‑create it later but can't, because they have too many effects plugged in or too many different reverbs going. There's only so much you can have going on, because there's only so much the human ear can take in.

"When I'm mixing I work to a definite pattern. I work for six hours, go home, sleep on it, and then come in the next day and finish it off in another couple of hours before starting on the next track. I can make more decisions in one hour the following morning than I can in four hours late the previous night. You need to be fresh to hear what was wrong — the night before you always tend to try and work around it.

"It is possible to fix some things in the mix, but there is only so much you can fix — you can't fix poor playing, so it's worth spending a bit more time in the studio getting it right — so long as you don't lose the spontaneity.

"I use EQ at source but people are so used to hearing bright records these days that sometimes I have to put some EQ on later in the mix. But my favourite mixes are the ones where I've spent as little time as possible and I've used as little EQ as possible, because if you have a nice, natural balance you shouldn't have to use a lot. Mixing isn't frightening — it's just balancing something so that it sounds right to your own ears. It is the one time you have to trust your own judgement and let your own ears be the benchmark." Sue Sillitoe

Tim Palmer

On Tears For Fears' 'Laid So Low'

"I always record a lot of bits and bobs as I go along — subtle delays and things like that — but on this track a lot of things, such as the vocal delay in the verses, were done at the mixing stage. At the point where the guitar comes to a stop in the intro, for example, I put a big single‑beat delay on that in time with the track, sending off a single channel to an AMS.

"I would normally split vocals and guitar solos over a couple of channels, and just use one of the channels to send to effects units so that I could control the level of a particular effect all the way through the mix. Sometimes, at the end of a guitar part for instance, you might want a lot of reverb and then have another part very dry; so rather than just put reverb on it, I'd sooner control it on another fader.

"While much of the percussion is essentially very dry — as if in a very tight room — I used the AMS reverb 'non‑linear' sound on a lot of the marimba parts and the tight snares, just to open them out a little bit. On the chorus there's a bigger reverb on the snare, achieved with an AMS using the 'ambience' setting. The main keyboard chords were fed into the Quadraverb, which has a nice, floaty delayed reverb sound that Alan [Griffiths] set up. The rhythm guitars had just a bit of plate reverb on them, and the lead guitars, for the most part, were treated with a very subtle delay, with a much more noticeable delay in a couple of special places. On the guitar part which links into verse 2, I faded up the chord and then cut it dead in order to give a sort of backwards sound, to create a little more impact and interest.

"The mix was a balancing operation more than anything else. In all, it took about a day and a half. I took a DAT copy of the mix home with me, then made a few adjustments and put it down. We lived with that version for a few days, and then there were a couple of things which Roland [Orzabal] wanted to try and a couple of things which I wanted to try, so we recalled the mix on a couple of occasions and experimented. Each time this took between five to six hours, and then we chose the version which we thought was best. Even if sometimes you don't agree, it's really important to try ideas out, because talking about something and actually hearing it are completely different." Richard Buskin

Hugh Padgham

On Sting's Ten Summoner's Tales

"When I'm mixing I really rely on Sting to be absent a lot of the time, because I need him to come in towards the end of the mix and say, 'Oh, that sounds great,' 'That sounds terrible,' or 'Let's try this'. You've got to have somebody with an objective viewpoint. He's not all that technically minded and doesn't want to hang around in the studio all day. That was great, because I could get on with it by myself without being hassled, and then he would come in at the end of the day and be able to provide that objective viewpoint.

"None of the songs were heavily produced, so we pretty well knew what we wanted to hear when it came to the mix. The main thing was to capture the right feel. Sometimes, when you rattle off rough mixes, there's a vibe about them which technically isn't good but is great emotionally. Sting is really good at coming in and saying, 'Do you remember that rough mix you gave me three weeks ago?' or 'Listen to this. I really like this rough mix'. And I'll say, 'Well yes, it is great, but it stinks because of whatever,' and so then it will be a case of trying to get the same kind of mix but technically more correct, and that's good. It's very challenging working like that." Richard Buskin

Robbie Adams

On mixing U2's Zooropa album

"We used the SSL room downstairs in Windmill Lane Studios for mixing. Daniel [Lanois] would often sit at the back shouting encouragements like: 'yeah, come on, louder!', just to get your excitement going and put you on edge. Band members would also sit in on mixes, and do a kind of cheerleader thing. It all induces a nervous energy in you and creates a lot of pressure, and gives the whole thing a performance feel. The reason we work like that is that mixing, like music, is a very personal thing. It's the people that are important, and the machines and the quality of them comes very secondary. It's like recording Bono's vocals with an SM58 and loud monitor wedges. The important thing is the way he sings, not whether there's spill or not.

"It's the same with mix automation. Computers are supposed to be like a clone of what you did, but I feel that they don't always do that 100%. Sometimes a tiny move on a fader can make a big difference, but it won't always come back the same way. I might sometimes save my very last mix in the computer, just for reference, but generally I get a much greater buzz out of mixing manually, just like the band." Paul Tingen

Phil Kelsey

On remixing Erasure's cover of the Abba song 'Take A Chance On Me'

"Normally I get the tape, put it on the machine, and see what there is. I've never once kept anything from an original mix except for vocals. I always strip the track right back down to the vocals. I've always seen a remix as exactly that, it should be different to the original. If you start keeping things from the original then it's not really that different, is it?"

I find that the more I go on, the less I resort to effects and tricks in the mix...

"With the Abba track I thought I'd try and do a disco version of it, then sacked that idea fairly quickly. I then made this massive decision not to use any of the song apart from a sample of two or three of Andy Bell's words, which I took from the middle eight/bridge section where it goes into minor chords. I took a few of the vocal lines from there, chopped them up, put one word from one sentence with another word from another sentence and made up a few new phrases. I came out with a ragga/toasting kind of thing. As soon as I did it I knew that was the one. Then it went all trancey and ended up 14 minutes long!" Wilf Smarties

Malcolm Toft

On how mixing used to be done back in the Sixties

"When I joined Trident, we were the first 8‑track studio in Europe, and we had a Sound Techniques mixing console that had eight groups with quadrant faders, 20 inputs, very simple EQ and a few aux sends. The sound of a recording was very much down to mic techniques and skills, with a lot of emphasis on how the musicians themselves played. You did a lot more in the studio than you did in the control room. You were always recording live instruments, doing sessions that ranged from two guitars, bass and drums to a 30‑piece orchestra. You had to take that down to eight tracks and get as much separation as you could. There was no timecode, overdubbing was relatively limited, and you didn't have the ability to click‑track drums and sync them in. You knew that if you didn't get it right going to tape, there was nothing you could do to salvage it in the mix. If you were recording strings and brass together, you were far more concerned about keeping the brass off the string tracks than you were about anything else. It was all down to mic technique and the use of screens to get separation, so that when the producer said to you that he wanted to drop something out of the mix, you could do it without still having spill in all the other mics. You only had three hours to get all this down and so there was a lot of pressure. There were no second chances, and the only things that would get overdubbed were usually the vocals.

"In those days, a typical session would end up with the drums on two tracks — the snare on one track and the rest of the kit on the other — one track for the bass, one for the guitar, one for the piano, one for the vocals, one for backing vocals, and then you'd maybe bounce the backing vocals down with something else if it started to get busy. You'd fly by the seat of your pants, and if you got really stuck, you'd bounce the two drum tracks down to one.

"In the control room, we had about four limiters and, apart from the EMT plates and tape echo, that was it for outboard gear. When it came to mixing, we'd usually set up a basic balance with a bit of reverb on the drums and vocals, and then the producer would tell us what he wanted. It would usually be the engineer's job to sort out the stereo image; there was a standard philosophy then that the drums and bass should be positioned around the middle, with the vocal in the centre, and then you might pan one guitar three‑quarters left and one right. Of course if you'd bounced two or more tracks together, you were limited as to what you could do with them in the stereo mix. Usually you'd end up with the drums, bass and vocals in the middle, and it would be the fills such as guitar and piano that came out of the sides. In those days that would be enough — if you could hear piano out of the left and guitar out of the right, then it was stereo!

"We were very fortunate in that our desk had panning. Some of the very early desks had bus switches so that you could route a signal either left, right, or to both buses to get it in the centre.The only way you could get in‑between pan positions on those desks was to send the same signal to two channels, route one left and the other right, and then alter the relative levels. But it used up too many channels, and most of the time, people didn't bother. Things were very crude, and you were also limited to what you could do with the reverb during the mix. Usually we'd bring that back through two channels panned hard left and right with a slightly longer delay on one side than the other. Everything on the track would share the same one or two reverbs. One was set to around two seconds, and the other to maybe one‑and‑a‑half to two seconds with a slapback echo on it." Paul White</string>

<key>footer></body</key><h3>EUGENE ELLIS</h3>

<string>On mixing Soul II Soul's 'Joy'

"When people talk about going for a mix, this usually means balancing what's on tape and perhaps adding some effects. In our case, however, it's a totally different concept. It's almost like rearranging the track; putting new sounds on, changing the drums, and this carries on right throughout the production process. It really amounts to a constant state of mixing. Then, when the last thing has been laid down, the finishing process is very quick.

"I didn't really go to town on outboard gear for 'Joy'. I put the usual compression on the vocal, used the Lexicon 480 for some general purpose reverb, and put the Summit compressor on bass just to get that extra kind of rumble which I like. Overall, however, there's not much in the way of effects on either the album or the 12" version. It's just straight down the line; a listen‑to‑the‑vocal kind of song.

"When I approach a mix, the philosophy is to decide on what is going to be the most important element and then create the track around that. This task was made much easier on 'Joy' by having a great vocal to start off with, and everything else took shape after that." Richard Buskin

Stephen Street

On his approach to mixing

"From time to time everyone needs to fix something in the mix. But if it's not sounding right after four or five hours, there's something seriously wrong. I use automation because I like to be able to do a move and then forget about it. I work quickly on my mixes and I don't spend days on one track — I like a mix to be there in 10 hours maximum. It should be quick unless you are mixing something you have never heard before. If it's your own track and you have always had a clear image in your head of what you want to do with it, I cannot see why it should take much longer than eight hours to mix. Sometimes people do rough mixes that are great and they try and re‑create it later but can't, because they have too many effects plugged in or too many different reverbs going. There's only so much you can have going on, because there's only so much the human ear can take in.

"When I'm mixing I work to a definite pattern. I work for six hours, go home, sleep on it, and then come in the next day and finish it off in another couple of hours before starting on the next track. I can make more decisions in one hour the following morning than I can in four hours late the previous night. You need to be fresh to hear what was wrong — the night before you always tend to try and work around it.

"It is possible to fix some things in the mix, but there is only so much you can fix — you can't fix poor playing, so it's worth spending a bit more time in the studio getting it right — so long as you don't lose the spontaneity.

"I use EQ at source but people are so used to hearing bright records these days that sometimes I have to put some EQ on later in the mix. But my favourite mixes are the ones where I've spent as little time as possible and I've used as little EQ as possible, because if you have a nice, natural balance you shouldn't have to use a lot. Mixing isn't frightening — it's just balancing something so that it sounds right to your own ears. It is the one time you have to trust your own judgement and let your own ears be the benchmark." Sue Sillitoe

Tim Palmer

On Tears For Fears' 'Laid So Low'

"I always record a lot of bits and bobs as I go along — subtle delays and things like that — but on this track a lot of things, such as the vocal delay in the verses, were done at the mixing stage. At the point where the guitar comes to a stop in the intro, for example, I put a big single‑beat delay on that in time with the track, sending off a single channel to an AMS.

"I would normally split vocals and guitar solos over a couple of channels, and just use one of the channels to send to effects units so that I could control the level of a particular effect all the way through the mix. Sometimes, at the end of a guitar part for instance, you might want a lot of reverb and then have another part very dry; so rather than just put reverb on it, I'd sooner control it on another fader.

"While much of the percussion is essentially very dry — as if in a very tight room — I used the AMS reverb 'non‑linear' sound on a lot of the marimba parts and the tight snares, just to open them out a little bit. On the chorus there's a bigger reverb on the snare, achieved with an AMS using the 'ambience' setting. The main keyboard chords were fed into the Quadraverb, which has a nice, floaty delayed reverb sound that Alan [Griffiths] set up. The rhythm guitars had just a bit of plate reverb on them, and the lead guitars, for the most part, were treated with a very subtle delay, with a much more noticeable delay in a couple of special places. On the guitar part which links into verse 2, I faded up the chord and then cut it dead in order to give a sort of backwards sound, to create a little more impact and interest.

"The mix was a balancing operation more than anything else. In all, it took about a day and a half. I took a DAT copy of the mix home with me, then made a few adjustments and put it down. We lived with that version for a few days, and then there were a couple of things which Roland [Orzabal] wanted to try and a couple of things which I wanted to try, so we recalled the mix on a couple of occasions and experimented. Each time this took between five to six hours, and then we chose the version which we thought was best. Even if sometimes you don't agree, it's really important to try ideas out, because talking about something and actually hearing it are completely different." Richard Buskin

Hugh Padgham

On Sting's Ten Summoner's Tales

"When I'm mixing I really rely on Sting to be absent a lot of the time, because I need him to come in towards the end of the mix and say, 'Oh, that sounds great,' 'That sounds terrible,' or 'Let's try this'. You've got to have somebody with an objective viewpoint. He's not all that technically minded and doesn't want to hang around in the studio all day. That was great, because I could get on with it by myself without being hassled, and then he would come in at the end of the day and be able to provide that objective viewpoint.

"None of the songs were heavily produced, so we pretty well knew what we wanted to hear when it came to the mix. The main thing was to capture the right feel. Sometimes, when you rattle off rough mixes, there's a vibe about them which technically isn't good but is great emotionally. Sting is really good at coming in and saying, 'Do you remember that rough mix you gave me three weeks ago?' or 'Listen to this. I really like this rough mix'. And I'll say, 'Well yes, it is great, but it stinks because of whatever,' and so then it will be a case of trying to get the same kind of mix but technically more correct, and that's good. It's very challenging working like that." Richard Buskin

Robbie Adams

On mixing U2's Zooropa album

"We used the SSL room downstairs in Windmill Lane Studios for mixing. Daniel [Lanois] would often sit at the back shouting encouragements like: 'yeah, come on, louder!', just to get your excitement going and put you on edge. Band members would also sit in on mixes, and do a kind of cheerleader thing. It all induces a nervous energy in you and creates a lot of pressure, and gives the whole thing a performance feel. The reason we work like that is that mixing, like music, is a very personal thing. It's the people that are important, and the machines and the quality of them comes very secondary. It's like recording Bono's vocals with an SM58 and loud monitor wedges. The important thing is the way he sings, not whether there's spill or not.

"It's the same with mix automation. Computers are supposed to be like a clone of what you did, but I feel that they don't always do that 100%. Sometimes a tiny move on a fader can make a big difference, but it won't always come back the same way. I might sometimes save my very last mix in the computer, just for reference, but generally I get a much greater buzz out of mixing manually, just like the band." Paul Tingen

Phil Kelsey

On remixing Erasure's cover of the Abba song 'Take A Chance On Me'

"Normally I get the tape, put it on the machine, and see what there is. I've never once kept anything from an original mix except for vocals. I always strip the track right back down to the vocals. I've always seen a remix as exactly that, it should be different to the original. If you start keeping things from the original then it's not really that different, is it?"

I find that the more I go on, the less I resort to effects and tricks in the mix...

"With the Abba track I thought I'd try and do a disco version of it, then sacked that idea fairly quickly. I then made this massive decision not to use any of the song apart from a sample of two or three of Andy Bell's words, which I took from the middle eight/bridge section where it goes into minor chords. I took a few of the vocal lines from there, chopped them up, put one word from one sentence with another word from another sentence and made up a few new phrases. I came out with a ragga/toasting kind of thing. As soon as I did it I knew that was the one. Then it went all trancey and ended up 14 minutes long!" Wilf Smarties

Malcolm Toft

On how mixing used to be done back in the Sixties

"When I joined Trident, we were the first 8‑track studio in Europe, and we had a Sound Techniques mixing console that had eight groups with quadrant faders, 20 inputs, very simple EQ and a few aux sends. The sound of a recording was very much down to mic techniques and skills, with a lot of emphasis on how the musicians themselves played. You did a lot more in the studio than you did in the control room. You were always recording live instruments, doing sessions that ranged from two guitars, bass and drums to a 30‑piece orchestra. You had to take that down to eight tracks and get as much separation as you could. There was no timecode, overdubbing was relatively limited, and you didn't have the ability to click‑track drums and sync them in. You knew that if you didn't get it right going to tape, there was nothing you could do to salvage it in the mix. If you were recording strings and brass together, you were far more concerned about keeping the brass off the string tracks than you were about anything else. It was all down to mic technique and the use of screens to get separation, so that when the producer said to you that he wanted to drop something out of the mix, you could do it without still having spill in all the other mics. You only had three hours to get all this down and so there was a lot of pressure. There were no second chances, and the only things that would get overdubbed were usually the vocals.

"In those days, a typical session would end up with the drums on two tracks — the snare on one track and the rest of the kit on the other — one track for the bass, one for the guitar, one for the piano, one for the vocals, one for backing vocals, and then you'd maybe bounce the backing vocals down with something else if it started to get busy. You'd fly by the seat of your pants, and if you got really stuck, you'd bounce the two drum tracks down to one.

"In the control room, we had about four limiters and, apart from the EMT plates and tape echo, that was it for outboard gear. When it came to mixing, we'd usually set up a basic balance with a bit of reverb on the drums and vocals, and then the producer would tell us what he wanted. It would usually be the engineer's job to sort out the stereo image; there was a standard philosophy then that the drums and bass should be positioned around the middle, with the vocal in the centre, and then you might pan one guitar three‑quarters left and one right. Of course if you'd bounced two or more tracks together, you were limited as to what you could do with them in the stereo mix. Usually you'd end up with the drums, bass and vocals in the middle, and it would be the fills such as guitar and piano that came out of the sides. In those days that would be enough — if you could hear piano out of the left and guitar out of the right, then it was stereo!

"We were very fortunate in that our desk had panning. Some of the very early desks had bus switches so that you could route a signal either left, right, or to both buses to get it in the centre.The only way you could get in‑between pan positions on those desks was to send the same signal to two channels, route one left and the other right, and then alter the relative levels. But it used up too many channels, and most of the time, people didn't bother. Things were very crude, and you were also limited to what you could do with the reverb during the mix. Usually we'd bring that back through two channels panned hard left and right with a slightly longer delay on one side than the other. Everything on the track would share the same one or two reverbs. One was set to around two seconds, and the other to maybe one‑and‑a‑half to two seconds with a slapback echo on it." Paul White