Bob Clearmountain has been the world's premier mix engineer for three decades — but Martin Scorsese still managed to challenge him with his ideas about how the Rolling Stones in concert should be presented.
As well as classic movies like Taxi Driver and Goodfellas, Martin Scorsese made his name as a rock film maker with the Band's concert film The Last Waltz. Last year saw him return to the genre with perhaps the biggest live band of them all: the Rolling Stones.
The band had already released a four‑DVD set documenting the Biggest Bang tour, officially the highest‑grossing tour of all time, so Scorsese's Stones movie risked appearing superfluous. However, Shine A Light was widely hailed as a triumph, and the associated double‑CD soundtrack sold well on both sides of the Atlantic.
Shine A Light is primarily a theatrical release rather than a DVD, and a rockumentary rather than a straight concert registration, with Scorsese's camerawork putting much other concert footage in the shade. The track listing is rather different from the Biggest Bang DVDs, containing many rarely-performed gems from the Stones' back catalogue. And in terms of audio production, the mix is tied to the film edit in a unique and revolutionary fashion.
Shine A Light was recorded and mixed by Bob Clearmountain, a legend in his own right, whose association with the Stones goes as far back as 1979. Having mixed a number of earlier live albums for the band, he assumed that he was on home territory and estimated that his part in the project would take no more than a few months. Instead, the project took up one and a half years of his life, and he found himself stretched into new areas by Martin Scorsese's insistence on an entirely untried and untested roving microphone‑on‑the‑camera perspective. Creating separate mixes for theatrical film release, Imax, DVD surround and DVD stereo, plus an entirely separate CD stereo version, Clearmountain ended up mixing the material up to 15 times, in half a dozen studios in Los Angeles, New York and London.
Shine A Light was recorded at the grand old Beacon Theatre in New York on October 29 and November 1, 2006. The second night supplied all concert material for the movie and the DVD, apart from Bill Clinton's speech, which was taken from the first night. The CD is also mostly taken from the second night, with three tracks added from the first.
"It was a pretty straight‑ahead recording,” explains Clearmountain. "I worked with David Hewitt and Phil Gitomer of Remote Recording Services. Phil was dealing with the DAW system on which everything was recorded 24‑bit, 48kHz, with backups made to DASH 48‑track half‑inch digital tape. I simply used the microphones that the live PA guys used, which were very standard: Shure SM57 on the snare drum, Beyer M88 on the bass drum, Mick's vocals with a wireless SM58, guitars with SM57, and so on. There were 10 audience microphones, I don't remember which ones, probably mostly Neumanns: a couple on the stage, and the rest hanging in various places inside the theatre. I had brought some Apogee Mini‑MP mic preamps for Mick's vocal, snare, bass drum and the overhead mics. The Mini‑MPs are the best mic pres on the market, but nobody knows about them, and Apogee has stopped making them.
"The mics went into an on‑stage splitter box that fed the FOH, the stage monitors and the truck. In addition to my Mini‑MPs, Remote Recording Services had a rack of mic pres on stage and sent line level back to the truck, which makes for less noise and a clear top end. Unfortunately it's also awkward, because you have to communicate with someone on the stage about setting the levels. We ended up with about 72 to 80 channels altogether, including audience. I recorded everything flat, no effects. Once a song started, I tried not to touch anything. You don't want to change settings too drastically, otherwise it becomes a bitch to mix later on. I was fairly conservative with the levels, and mainly spent my time doing a monitor mix, with EQ and effects post‑recording. My monitor mix went to the film team, who were recording everything on a bunch of video machines. Everything was filmed, but these movie cameras have little video cameras in them, which are recorded to video tape that's used to view the rushes. These were recorded with my rough mix as a reference.”
On returning to his Mix This! studio in Los Angeles, Clearmountain transferred all recorded material to his Pro Tools system, and during December and January he did what he called "preliminary mixes” of both the October 29 and the November 1 concerts in 5.1 and stereo, sending everything through his 72‑input SSL G+ series desk and pulling down the sample rate to 47.976k while mixing, because the movie editing was being done on an Avid system. "These were very good basic mixes that were meant as a guide so the film team had something to work with while editing. The live mixes I'd done were OK, but there's only so much you can do in a live situation.”
Given Clearmountain's decades of experience and his track record with the Stones, he felt confident that these basic mixes would form a solid foundation to work from, and would be met with only minimal requests for tweaks and changes. However, Scorsese, editing the movie in New York, came back with a unique suggestion. "I do a lot of DVD concert mixing, and normally it's a rush thing and you're not working to a final picture edit. Instead you're working with a line cut of the images, which was produced during the live recording. You're watching the picture just to get a general vibe and a basic sense of where everybody is standing on stage, which then reflects in the placements in the mix. Shine A Light was different. Marty's direction was: 'I want to hear what I'm seeing on the screen. If the screen shows Ronnie's guitar, I want it to be featured in the mix, I want to feel what he's playing.' That means that you're mixing with a moving listening perspective: every time there's a full shot of Keith, for instance, his guitar goes to the centre.
"So I had to wait until they could send me the final cut, or something close to the final cut, of each song, and I would remix each song according to Marty's request, which was challenging to do! It meant using a lot of automation and I had to be careful that things didn't jump out so much that they started to sound unmusical. Because I was working on an analogue desk, I came up with a system that I called 'pan bus'. Instead of sending each instrument to the stereo bus, I sent each track that needed to be panned to fit the picture — Keith's and Ronnie's guitars, Mick's voice, Mick's harmonica, Bobby Keys' saxophone — to a multi‑channel bus off the fader. So Ronnie's guitar would be on bus 17, Keith's guitar on bus 18, and so on. These buses were patched as aux inputs into Pro Tools and assigned to an L‑C‑R bus output from Pro Tools, which came up on three subgroup channels on the SSL, panned L‑C‑R in both stereo and 5.1. So if Keith had to move to the centre, I used the Pro Tools automation to do this, drawing it in on the automation line. This was really cumbersome, but it worked well, and I could be very precise.
"The big challenge in doing these mic‑on‑a‑camera mixes was to give you the experience that you're standing on stage with the band, but at the same time to make sure it's not sounding ridiculous. I'd have to balance out what Marty was asking me, while also keeping a musical flow. I didn't really want people to notice it, but it was an interesting experience, because it really sounds like you are there, you are on the stage. We were mixing a movie, we were not mixing a record, so we treated it like a movie. I also took a lot of my cues from watching The Last Waltz. I was really impressed with how this was mixed — nothing extreme, but the way they used panning was really effective.”
Just when Clearmountain, still at Mix This!, was getting the hang of this novel way of mixing to picture, he received a panic phone call from New York with some disturbing news: his new surround mixes sounded dreadful. It's not the kind of feedback the mix luminary is used to receiving. So what was going on? "Marty was editing and listening to my mixes on a regular THX Dolby 5.1 system, like a typical movie theatre. He said that they sounded really harsh and that there was no treble and it was all too mid‑rangey. I was like: 'What!? How can that be!?' I know my system well and the mixes sounded really good here. I'm using five Dynaudio BM15 speakers and a KRK S12 subwoofer. So I booked some time at the Sherry Lansing movie theatre at the Paramount lot here in LA, loaded a few of the mixes onto a DA88 tape, because that's what they had, and played them through their system. Sure enough, it sounded dreadful. Nothing but harsh upper mid‑range, no top, and very little bottom. I was wondering 'What the hell is going on?'
"It was a shock to me to find out that movie theatre speakers don't sound anything like home theatre speakers. I asked a lot of people, and many knew of it, but no‑one knew why, or everybody had different reasons. When people mix movies on mixing stages they're used to the speakers being voiced just like a movie theatre, and they adjust their EQ accordingly during the mix. They're not worried about the home theatre thing. But I was mixing on speakers that are voiced just like a home theatre system. In the end I hired a film mixing stage here in LA, and using an EQ plug‑in, I came up with an EQ curve that corrected my mixes so they sounded right on the film stage. After that I processed all my 5.1 film mixes with this EQ curve. Of course I got rid of that EQ again for the DVD.”
To better be able to tackle the challenges of a novel approach to mixing, which required a lot of dialogue between Scorsese and Clearmountain, and the sonic incompatibility between traditional music studio and film theatre monitors, Clearmountain decamped to New York in the summer of 2007. He worked there for another six weeks on the 5.1 mixes at Battery Studios. During this and the previous stage he also did a few fixes to the material. "There are no overdubs, except for about four seconds of a solo that Ronnie wanted to fix. I also did a few minor edits, for example fixing a chord Keith missed just by cutting and pasting, things like that. But it's very minimal. I don't know whether people realise that most artists today spend weeks in the studio redoing things on 'live' albums. By contrast, virtually everything on Shine A Light really is live.
"Battery's desk was eight channels smaller than mine, and my own SSL had been maxed out, so to be able to work in 5.1 there I had to compress things down and rent in an SSL AWS900 for the surround returns. Once Marty was satisfied with these mixes, my assistant, Brandon Duncan, ran stems of all the mixes and we continued at Soundtrack, a film mixing stage in New York in September, for the final film mix, which I did with Tom Fleischman. We did these mixes on a large Euphonix System 5 digital console, fine‑tuning levels with Marty's direction. Each instrument had its own five‑channel stem, which included the panning. But there were some reverse shots in the movie, for which we used the Euphonix to switch the perspective of the drums and backing vocals 180 degrees, stuff which is kind of hard to do on an analogue desk.”
In February 2008, Clearmountain went to Pacific Ocean Post in Santa Monica with mixer Ted Hall to prepare the DVD soundtrack, which involved re‑EQ'ing the movie 5.1 soundtrack and doing a fold‑down to stereo. The stereo mix allows home listeners without surround equipment to have a similar mic‑on‑a‑cam experience, but to achieve this, Clearmountain again had to negotiate some AV format idiosyncrasies. "Usually, folding down to stereo is done with a process called 'LT‑RT', which involves analogue Dolby Pro Logic encoding that was developed in the '80s and '90s, ie before digital. But, using that process, the audience recording ended up in mono and the whole stereo mix ended up sounding weird and mono‑ish. So I spent quite a bit of time doing a straight reduction from 5.1 to stereo. A little later Previously I had also spent a week in Toronto at an Imax stage, folding down the 5.1 mix to the Imax format, which only uses five channels and no LFE channel. So you have to mix the subwoofer signal in with the other five channels.”
Clearmountain and Jagger spent a week in January 2008 at The Townhouse in London, doing a stereo mix for the Shine A Light double‑CD soundtrack. "We had to do a new mix for the CD version, because all the sonic close‑ups from the DVD stereo mix, with things moving around and getting louder for no apparent reason, make no sense without the picture. So I did a desk recall of the stereo mixes that I had originally done in December 2006 and January 2007; the fixes that I had since done in the DAW of course remained in there. Mick and I were remixing, really. He was pretty involved, not pushing faders up himself, but having very specific requests, like wanting more bass drum or less guitar. We talked in detail about things. I had to re‑EQ my original mixes because everything sounded so different, and this was a bit tricky, because the Townhouse had G‑series EQs, whereas my desk has E‑series EQs. I don't really like the G‑series EQ that much. In the end the CD mixes sound totally different from the DVD mixes, plus there are some extra songs on the CD which are not on the DVD.”
A listen to the DVD and CD mixes certainly bears out the differences in sound, and, unsurprisingly, there are further idiosyncrasies. For example, on the European DVD release of Shine A Light, his stereo fold‑down mix isn't available, something that shocked Clearmountain. "That's horrendous, that's a terrible thing, because I was really adamant about that.”
Scorsese's mic‑on‑a‑cam effect may well have contributed to the many positive reviews, but has remained largely unnoticed by critics. "I'm glad people haven't noticed,” says Clearmountain, "because it should be part of the experience of the film and I didn't want to make it too obvious. But if you really pay attention it does become very obvious.”
"There are several differences between mixing a live concert and a studio track,” explains Clearmountain. "One, obviously, is the presence of the audience and ambience. If you want ambience on a studio track you usually have to add it, whereas with a live recording you're having a constant battle to make sure you can hear the audience reaction and that the music doesn't get too washed out by the ambience, especially if the PA was cranked up really loud, which it was with the Stones. You have more control with a studio recording, and things tend to be tidy and separated, and there are not as many mistakes. You also often are creating the arrangement, deciding that certain tracks don't work or are not necessary or adding an additional instrument or vocal.
"When mixing a live recording, you basically try to recreate the live experience, what it was like sitting in the audience, only better, because often things get lost in the PA. On Shine A Light I had a couple of slap delays over much of the mix, 100ms and 125ms using my Roland SDE3000 and panned left and right, something like that, and most things go through that in different amounts. It adds overall excitement and makes it sound less as if it was done in a TV studio and more like it was done in a big place. The songs were mixed one by one, because someone will change a guitar or an amp or there'll be a different keyboard. Too many things changed with each song to mix it all in one go. I don't even know whether that would be possible. Although each song was mixed separately, it was not like I started from scratch with each song. The drums are basically the same on each track, and the general sound also remained the same.”
- Vocals: Urei 1178, 1176 and LA3A, Roland SDE3000, Audio Ease Altiverb, Digidesign Pitch, SSL EQ & compression
"Part of the challenge in mixing the Stones is that many of their instruments are concentrated in the upper mid‑range. Luckily, Mick Jagger's voice easily cuts through the guitars. You don't have to turn him up much to hear him. Mick kind of compresses himself, but I did use an 1178 on him, for just a little bit of compression. I also used the SDE3000 slap delays on Mick, and a little bit of Altiverb on a big Hall setting. The only reason for the latter was that when the song ends you push the audience mics up for the audience response, and this adds hall ambience to the voice. The voice is pretty loud in the PA and is picked up by the audience mics. Adding a little bit of reverb to the vocal and some slap delays during the songs obscured this difference. That was all with Mick. He's just dead easy to mix. I treated Keith's vocals more or less the same as Mick's, though the EQ would be different. His voice is a little bit harder to deal with because it doesn't cut through quite as well. Luckily the band was very quiet during his big song, 'You Got The Silver'.
"I also used a bit of 1178 compression on Christina Aguilera, and some desk EQ, adding treble I think, and a little bit of Digidesign Pitch to tune her vocals in a few places, very minimal. Her vocal was difficult, because the song she sang in, 'Live With Me', was very loud, and she had a wireless mic that sucked everything in that was on the stage, especially if you added a bit of compression, which made it 10 times worse! She has a bunch of mics that can be colour‑coordinated to what she's wearing, so she had to have that mic. No‑one used in‑ear monitors, so they had wedges, and this meant that spill was a problem on all vocal mics. When a singer wasn't singing I'd duck the mic signal right down on the SSL with the fader, using the automation, though sometimes my assistant would cut things in Pro Tools. Buddy Guy's voice was very easy to mix, just some compression and desk EQ, nothing special. I had an 1176 on Jack [White]'s vocals, and LA3As on the backing vocals. But in none of these cases did I use a lot of compression. I tend to ride the vocals quite strongly by hand with the desk automation.
"I don't use de‑essers. Instead, I have a patch on the SSL. I mult the vocal to the channel next to it, on which I'll roll out all mid‑range and bottom and boost around 7k as far as it will go. I'll then send the output of the fader into its own bus, and select a compressor on that channel to pick up on that bus, so basically it becomes a sort of loop in the channel. I put the compressor on fast attack and fast release, and link it to the original vocal channel, on which I turn on the dynamics. It works like a side‑chain on the compressor and it has the effect of a fast high‑frequency limiter, its threshold being controlled by the fader on the parallel channel. It's really effective, very fast, and it de‑esses very smoothly and much better than any other de‑esser I have ever used. Plus you can automate the threshold. It uses a lot of buttons and knobs, but once you see how it works, it's really simple and very effective. I used it on Keith's and Mick's vocals.”
"Mick played acoustic and electric, and Blondie Chaplin, one of the backing singers, occasionally played acoustic. Most of the acoustics had two DI outs which went to separate tracks. Getting all these guitars in there was a matter of panning and EQ. There was nothing else very specific. When you mix, you just do what's appropriate, and I often can't remember what it was. You just turn the knobs until it sounds right. You hear a guitar sound, and you think: 'OK, this frequency is going to conflict with the voice,' and you adjust it. That's all — no big secret!
"In the stereo spectrum I had Keith on the right and Ronnie on the left, the way they appear on stage to the audience. Depending on the song, I may have used some of the same slap delays and reverb on the guitars, but there was little or no compression on the electrics. I like to keep the dynamics on the electric guitars. I did probably use some Distressor compression on all the acoustics, and I had some Focusrite Red 3 compression on the bass. I also added some low frequencies, around 70 cycles, to the bass. On Buddy Guy's guitar I had a Distressor and EQ. He has a very bright, piercing guitar sound, and for Marty it couldn't be piercing enough. I probably used my Pultec EQP 1A3 for this. But when Mick heard it, he asked for a little less top end.”
- Drums: SSL EQ & compression, Digidesign Sound Replacer
"The drums were pretty straightforward: a little compression on the snare drum and toms, and adding overall treble. You use the EQ to make it sound like drums, to make them sound powerful. Also, Charlie has a very dynamic bass drum, he varies its intensity quite a bit. He's like a jazz drummer, really. For rock, this is tricky to mix, because it may sound good in the verse, and then it jumps out or falls away somewhere else. So I took a sample of him because he has a great‑sounding bass drum, and mixed that in with his original bass drum using a sample plug‑in called Sound Replacer, which follows the dynamics of an instrument. You can specify that you want it to follow maybe 50 percent of the dynamics, so it works kind of like compression. That really helped a lot. It made sure the bass drum was always there.”
The world première of Shine A Light took place at the Berlin Film Festival on February 7, 2008, but even then, there was no time for Clearmountain to rest on his laurels. At Scorsese's request he spent a considerable amount of 2008 flying around the world to check the equipment in each film theatre where the film premièred.
Finally, a whopping two years after the original concerts, Clearmountain is able to look back on a job well done. "Shine A Light really captures the Stones exactly as they play. And they did a great show. I must have seen that movie 150 times now, and I still enjoy it. It's a great experience.”
Bob Clearmountain is one of many top‑flight mixers who still rely on an analogue desk. "I've never done a mix 'in the box', so it's hard for me to comment on what that would sound like. I don't like sitting in front of a computer while I'm mixing, I like having a bunch of knobs in front of me. It's very uncomfortable and cumbersome for me to be opening drop‑down menus and plug‑ins. I'd rather be a plumber than do that! Having said all that, I do like digital. I always had problems with analogue, before I even heard digital. I'd be recording a great live band like Chic or Bryan Adams, and it would sound great during the take, and then the playback from analogue would be so frustrating because it sounded distant, sort of removed from what it was. I could never get the same excitement from tape as when the musicians were actually playing. Finally digital came along, and used with Apogee converters it sounded more live: what came off digital tape suddenly sounded the same as when the band had been playing. Especially now with the higher sampling rates, digital is really good.
"I simply use Pro Tools like a tape machine. I still start a mix by pushing all the faders up, and doing a rough mix, just to get an idea of what the song is about. Only after that will I sometimes solo individual parts, to hear what their contribution is. But I'll never start a mix by just working on the bass drum, or the guitars, or a song section, things like that. For me that's meaningless. I get lost. I can't tell what's going on. I have to hear things in context.
"I mix songs from beginning to end, and not in sections. I'll develop the mix until it is as good as I can get it without automation. After that I'll switch the SSL computer on, and usually the first pass into the automation will be doing the vocal, making sure it sits right in the mix. On subsequent passes I will ride things around the vocal, but I don't bring these in one by one, like some mixers do. My approach is much more random. You might call it stream of consciousness. People often ask me how I do it, but it's like trying to describe how I walk. I simply adjust what hits me at the time and try to make it sound as good as I can.
"One of the few things for which I do use the DAW automation is for vocal tuning, with Digidesign Pitch. It's a straight‑ahead pitch‑shifting plug‑in, and I apply it totally by ear and automate the changes. I find the more popular tuning plug‑ins too cumbersome, and they sound strange to me. And for the Stones project, I used the Pro Tools automation for the dynamic panning that was necessary, because there was not really a good way of doing that on an analogue desk. The other thing about using DAWs is that it's so much easier now to fix things than it was when we were still using tape. But I use few plug‑ins. I like Altiverb, but I treat it like outboard, hooking it up to an aux on my SSL and bringing it back up on my desk. I also use Waves X-Noise sometimes, if I there's a problem. And if I have to change the EQ mid‑song I may use an EQ plug‑in that I can automate.”