Recording Les Misérables - Part 2: Implementing the Plans
by Simon Hayes AMPS
Photos by Laurie Sparham/Universal Pictures
Beginning my assignment on Les Misérables, I had some enviable, even unprecedented, advantages. I had support from the producers at Working Title Pictures and the Director, Tom Hooper, to use every resource available to achieve live recording of all the vocals without any ADR. And I had a crew of seven skilled associates to help achieve this goal, all handpicked from the best technicians I know, all excellent choices for their ability to work together as a team. But it still remained to coordinate with other departments and develop a plan for how this goal might be accomplished.
Meeting Supervising Music Editor Gerard McCann was the next step and a defining moment in the planning stage. Right away we agreed to join forces and merge his four-man department with my seven-man team. Whatever demarcation had existed, we relegated to history and agreed that the teams would share all the tasks of the daily technical grind including rigging, cabling and loading gear.
Music Supervisor Becky Bentham was also part of this first meeting. She is a legend in the UK film industry. Both Gerard and I had worked with her before and had great respect for her abilities.
The three of us discussed the project in detail and worked out a plan of attack. We would have two live pianists on set at all times. Both were part of Cameron Mackintosh’s team and had years of experience with the orchestrations of Les Mis. One pianist would work with the shooting crew and the other would be available at all times for warm-ups and rehearsal. Whichever one was on set that day would work inside a soundproofed plywood box fitted with ventilated Perspex windows so that the mechanical sound of the Korg electric keyboard would be confined. The player would wear headphones with an IFB feed of the vocal mix in one ear. The pianist was also fitted with a radio mike for direct communication with the actors via their “earwig” feed.
The piano would then be routed both to Pro Tools Rig #1 and also to the sound cart for transmission to the actors’ earpieces. Says Gerard McCann: “We had our live piano performing and three Pro Tools systems, operated by Music Editors Rob Houston, John Warhurst and myself. Simon was able to route that live piano feed into earpieces worn by the actors who were then able to sing to live accompaniment. Our Pro Tools systems had three roles: one was dedicated to playback for tracks that required a fixed tempo, like chorus material. For the larger crowd songs, we would record a rehearsal of the ensemble cast on set on the day, and use that as a playback for shooting so that the crowd could follow along singing in the correct tempo, and this live singing recorded by Simon. This was to allow Tom maximum freedom to use as much of this sometimes rough, raw, but very real sounding live chorus as he chose, together with additional layers he might record later in post. A second machine was dedicated to recording the live vocal and piano mixes from Simon, and the third was used to turn around this recorded material almost instantly for playback.”
In working out the production sound methodology, I was keen to stick to a comfortable workflow; this wasn’t the time to be introducing new or untested equipment into the recording chain. I needed to be using equipment that was second nature to me so my attention might be on capturing performance rather than technical issues.
I chose to gang together two Zaxcom Devas, one the Deva 16 and the other a Deva 5. This would give us 26 tracks. I would give the picture editor two mix tracks to use on his Avid timeline: Mix-1 had the vocals and the mono piano; Mix-2 had the vocals only, without the piano. This gave the editor the facility to adjust the blend of voice and accompaniment as needed.
We linked the two recorders together so they would have identical timecode. The Deva 16 had the two mix tracks plus isolated mikes on tracks 3–16. Machine 2’s ten tracks were all assigned to ISOs.
The two linked machines gave us a total of 24 tracks. Since we might need to use radio links for the two mono booms and the stereo boom, we were limited to 20 radio mikes. I already had two fantastic Audio Developments’ mixers with eight channels each. They were modified to supply either analog or digital signal on all the outputs so we were well equipped for 16 tracks. We reasoned that we would not need all available tracks recording the solo performers, only when recording the chorus, so we could connect directly to the Devas and use the front panel faders on those occasions.
I also ran a safety copy of the mix tracks on a 24-bit Nagra V in case of a hard disk failure on the primary machines. That covered us in the event of an equipment failure on a magical “perfect take.”
Running 20 radio microphones without any inter-channel modulation or interference is not easy. Luckily, the UK was in the middle of switching the legal film industry channels from one band to another to make way for digital television, and we took full advantage of the temporary window available to us to use both channel 38 and channel 69. As Gerard worked out the need for five different Comtek feeds—that’s right, five mixes—our special good fortune became more apparent. Our plan called for Mix-1 to be piano and vocals while Mix-2 would be piano only for members of the music department who needed to concentrate on that element. (Quite a few members of the music team kept two receivers on their belts so they could swap between these two mixes as they wished.)
Mix-3 would be vocal only for use by dialog coaches working on accents. The pianists also used this mix while listening to a direct feed from the electric piano in the other ear.
Mix-4 was a special mix that Tom Hooper and Danny Cohen required for themselves and the camera crew. The music was such a large part of the tempo and timing that the camera crew needed to hear the piano and voices to motivate their action. We added a talkback mike—a Shure SM58 with a transmitter—to permit Tom to communicate with camera operators and grips even during the takes.
Mix-5 was the boom operators’ headphone feed, much the same as Mix-4 but with my voice alongside the singing and piano instead of Tom’s. I was, of course, using the onboard talkback mike on my mixer rather than a handheld SM58. This permitted me to talk to the three boom ops throughout takes about lens sizes, shadows, etc. With the 20 radio mikes, five wireless headphone feeds and Tom’s SM58 transmitter, we would be using up to 26 separate frequencies at any time. The responsibility for wrangling all these frequencies fell to 1st Assistant Sound Robin Johnson. Without his skill and experience, I doubt we would have been able to run that many channels.
All of this equipment would live on two sound carts that could be moved around on location. We were becoming technically ready. The next step was to consider the “in-the-ear” monitors for the actors.
We considered several in-ear monitors and made a decision early on to use a traditional induction loop system over the newer radio systems. To fit within the ear, all of these systems are limited to a very small driver that severely limits sound quality. None of the present designs sound very good. Since the units with a built-in radio receiver offered no audio advantage, we couldn’t justify their extra expense particularly considering the number of units we would need. We concentrated our efforts into finding the best induction loop amplifiers and in optimizing the performance of the traditional design.
We confronted two problems with the available earwigs: their small driver size severely limited bandwidth and they were not very loud. An orchestra with a broad mixture of bass and high frequencies would confuse the tiny driver and the output became muddled. We found the problem was less acute using the Korg electric keyboard as its output is simpler and tends toward the midrange. The pianists were a great help with this by adjusting their play accordingly. We also adjusted the EQ settings on the keyboard to suit the earwigs.
The loudness issue was not so easily resolved. These earpieces were originally designed to assist people with hearing difficulties, not to be used as a reference while singing “Who Am I?” or “I Dreamed a Dream” at the top of one’s voice. We contacted the manufacturer and they were very helpful and supplied us with louder units. We also had them come out and make ear casts of each principal actor to supply them with custom-fitted earpieces both left and right. This helped in several ways. The custom earwigs fit deeper in the ear canal and were less visible to camera. Also, a precise fit ensured that the earpiece was optimally positioned, and its tiny outlet hole unblocked, so it could deliver its maximum output. Having both left- and right-fitted earpieces also gave the option for using both if an actor were struggling to hear. This was really a last resort because it would interfere with the actors hearing their own vocals.
We decided early on not to feed vocals into the earwigs both because of the frequency response issues and also because we would forever be discussing individual preferences on the balance between vocals and piano. This would present an impossible situation because we could only provide one earwig mix on the induction loop. But there are always exceptions—on “I Dreamed a Dream” Anne came to me after the first take and asked to wear both earwigs with the piano as loud as possible and a tiny amount of her own vocal added. Since she was singing a solo, and we didn’t need to provide earwigs to others, we were able to accommodate her.
For a couple of monumentally challenging sequences, Tom staged two actors at locations hundreds of yards apart, harmonizing together in real time but shot with separate cameras. In those instances, we fed their vocals to their earwigs so they could keep pace with one another. This created much hilarity on set as Hugh and Russell realized they could communicate with each other and began comparing progress on the setup and which camera crew might be ready first. There were other exceptions to our no-vocals-in-the-earwigs rule but we generally tried to keep the playback practice as simple as possible.
With recorders, track assignments, piano accompaniment and earpiece distribution worked out, Gerard McCann and I had a good plan for recording the vocals. But we needed to meet with Orchestrator and Music Producer Anne Dudley and her team to confirm that our efforts would meet her needs. We met her and Music Supervisor Becky Bentham at the famous Abbey Road Studios in London. They told us that their engineers would like to hear the mikes we intended to use so we set up some test sessions. The Neumann U87 is the standard condenser microphone in a music studio. Its accuracy is unexcelled and its large diaphragm produces a smooth response to rapid transient changes. The music studio also offers acoustic excellence and the ability to place the microphone in optimum position. No location recording plan we might devise would ever be able to equal that performance. But the live recording offers the advantage of immediacy and an emotional link to the acting so the operative question was whether the fidelity of our system would meet listening expectations.
I chose the Schoeps Super CMITs for our boom operators. These new microphones use DSP noise-canceling technology to reject off-axis background sound. This capability is a great advantage but demands a high level of skill from the boom operator. When the Schoeps were used in testing it became clear that, if they were in an optimum position, the kind possible while shooting a close-up, they could compete on a level playing field with the music studio mikes.
We also tested the DPA lavaliers and Lectrosonics radio mikes. In my opinion, the DPA matches the Schoeps Super CMIT more closely than any lavalier I’ve heard. During the demo at Abbey Road, the engineers, despite initial skepticism, were suitably impressed. They felt they were getting approximately 60 percent of the quality of a Neumann U87 when I believe they were expecting much less. When you consider that the studio mike is placed on a stand in the best possible position while the DPA is rigged on the actor’s chest, that is an excellent result.
Paco Delgado, the Costume Designer, was extremely helpful and collaborative in this process. To hide the lavalier mikes, he and his team supplied us with the necessary cuts of fabric from each costume and also allowed us to make the holes needed to hide cables. He encouraged us to take the lavalier rigging to a level that enabled us to record absolutely clean singing with no clothing rustle. As we started shooting, it became clear that the process of mic’ing the cast was far more time-consuming than on a “normal” film not just because of the need to match fabrics but also because there were so many radio mikes used.
It’s always my aim to deliver as natural a dynamic range as possible so I was in full agreement with the engineers’ request that we not use any compression or limiters in the recording chain. To make the full 24-bit dynamic range available, this meant not only refraining from using or tripping limiters in the equipment but also not riding gain during the take. We used the Lectrosonics transmitters at a very low-gain setting to ensure that limiters would never be engaged. Historically, the higher gain setting needed with radio mikes to stay above artifacts meant that limiters were needed to prevent overloads with louder signals. The ability of the current generation of Lectrosonics’ gear to capture clean signal at lower settings, even with whispered delivery, was impressive and a key reason we were able to take on the project. By agreement between the Music Department and the Sound Department, we used no limiters or EQ anywhere while recording Les Misérables.
With everyone in agreement on the methodology, we turned our attention to the challenges of recording live singing on a movie set. We had to consider the scale of the Paris street scenes and how to manage them. Tom asked me if I would prefer to shoot the exteriors on a soundstage or on location. I knew that Tom wanted to shoot the scenes, some as long as 14 minutes, from start to finish without a cut. I didn’t see how this would be possible outdoors in a modern, aircraft-infested environment but the only stage large enough for the planned scenes, the 007 stage at Pinewood, is not really a soundstage and has poor acoustics. Just a few weeks into preproduction, Tom contacted me to tell me about a new stage being built in Pinewood—the Richard Attenborough Stage—that would be the biggest in the UK. (After our good fortune with the transitional availability of radio frequencies, we began to think someone upstairs was smiling on our project.)
Eve Stewart, the Production Designer, asked me about ways that set design could help with Tom’s vision of a live musical. I commented that for live sound we wanted reality. If they are in shot, the cobbles should be real cobbles, the oak door frames should be real oak, so that any sounds we picked up would be as authentic as possible. She took my suggestion and filled every inch of the 30,000-square-foot stage with sets built with the characteristics of permanent structures.
Our interest in solid oak and stone applied only to areas seen in the shots; outside what the cameras saw we tried to make the set and crew sonically disappear. Our efforts extended even to fitting rubber shoes on all the horses’ hooves.
For Eponine’s number, “A Little Fall of Rain,” we faced the additional challenge of recording the entire number in the rain. We worked with the Special Effects Department to get the best possible rain that would show on camera without drowning the mikes or making too much noise. We covered every part of the set not seen by the camera, every rooftop and every piece of floor, with rubberized horsehair to deaden the raindrops. We had an entire truckload of horsehair delivered to Pinewood. We also had a horsehair cover to provide quiet protection for the camera and asked the camera technicians to wear black “Bolton” cloth (Duvateen) ponchos over their Gore-Tex to soak up the sound of the water hitting. We even had a second boom operator shadow the primary boom with a horsehair roof on the end of his boom pole to shield the primary mike. That was the attention to detail that we exercised and it was possible because of an outstanding seven-man team. With a truck full of rubber-backed carpet, this team padded every dolly track and every walk-and-talk to keep the set as quiet as possible and recovered the carpets as soon as the shot was completed so they never held up the shooting. These efforts paid off not just by reducing noise from footfalls but they helped to deaden sound reflections throughout the set and augmented the many sound blankets we hung for that purpose.
Wind to flutter hair and costumes is a necessary element to create the illusion that players are outside and not on a set. Traditionally, large fans or wind machines provide this but they are quite noisy and compel ADR whenever they’re used. We coordinated with the FX Department to place the wind machines outside the stage and pipe-in the wind through flexible air-conditioning hose. The mikes didn’t pick up the sound of the electric motors at all, just the sound of moving air that mimicked the sound of actual wind. And, since its frequency fell outside of normal voices, it could be effectively removed in post.
After all the technical planning, we were ready to put our methodology to the test. The film had engaged the actors for an eightweek rehearsal period directly prior to shooting. Such a lengthy rehearsal period isn’t the norm but Les Mis was a complex project. I felt it important for the whole sound crew to be involved from the beginning but there was a move to exclude us. I can certainly understand the budget implications of adding a large sound crew for an extra eight weeks. And, the performers can be self-conscious as they develop their performances. Working with playback or with a piano accompaniment will mask errors in pitch or delivery but singing a cappella leaves every performance mercilessly exposed. I could understand the reluctance but I felt it important that everyone become committed to the live recording protocols from the beginning. I worried that, after eight weeks of rehearsal with the blanket of protection afforded by an amplified piano, the cast might balk at the introduction of the earwigs on the first day of shooting. If they felt they couldn’t work without the live piano, the whole plan of live recording would founder. We needed the collaboration between Cast and Sound to begin on the first day of rehearsals.
I also felt that the long rehearsal period was important to more than just the cast. I wanted to use earwigs and radio mikes on every rehearsal so that the Pianists, Roger Davison and Jennifer Whyte, could become comfortable with the process of working within a sound booth and following the pace of the singers from their own headphones. And, I wanted the practice time for the Sound Department so that we might become familiar with the songs, the staging, the head turns, the extremes in dynamics, and work out solutions to the challenges in advance. Sometimes a single performer would need two mikes, one on each side or one close to the mouth and one lower, to handle these variables.
Even more important than the technical issues was the opportunity to become acquainted with the cast and earn their trust that we would deliver quality recordings of their live performances. I pressed these points with the producers and with Tom and eventually we were invited to participate.
By the end of the rehearsal period, the cast was completely unfazed by using the earwigs and having direct communication with the pianists through their lavaliers. They would arrive at our sound carts upon entering the rehearsal stage to ask for their mikes and earwigs before proceeding to the set and enjoyed being able to communicate directly with the pianists without raising their voices to draw the pianists’ attention.
It was going well but we were developing a new process and everyone, Tom Hooper, the Producers, the Music Department and our own Sound Department, wanted a test to confirm that it would all work through editing and mixing to a final product. The “Red and Black” number performed by the students in the café was a good selection for our test. With multiple solo lines from the cast and an ensemble of about 20 students, it provided a taste of most of the circumstances we would encounter throughout the film. From the beginning, I had requested that rehearsals take place in the proper acoustic environment so that we might make test recordings and check the results later through studio monitors. Consequently, our rehearsal space was a proper soundstage at Pinewood that was suitable for a film test. Tom decided to shoot the test with a full camera crew and three 35mm cameras.
The test shoot proved challenging, exciting and interesting. Although Tom had discussed the visual style he had worked out with DP Danny Cohen, nothing quite prepared me for his singleminded enthusiasm for shooting every take all the way through from beginning to end. For the sake of performance and energy, Tom would shoot numbers in their entirety so I needed to be ready at all times. For me this meant multi-tracking and mixing 20 mikes on every take from 8 a.m. to 8 p.m. It was mentally demanding; I had to find a zone and stay focused. My own mixing improved with the constant practice but that was a small benefit as we always intended to remix from the ISO tracks in post. More importantly, the boom operators thoroughly learned the intricacies of every move by both cameras and cast members and became adept at following the singers exactly. Since the long takes forced a camera reload for nearly every take, my crew had an opportunity to act on every little problem revealed by the previous take. Carpet placement could be optimized, a cast member standing on a squeaky floorboard could be shifted slightly and chorus or extras that were whispering when they should have been miming, could be advised. (Many members of the chorus ensemble came from a theater background where ad-libs would enhance the performance. It took awhile before they became comfortable with the understanding that film editing needed consistent, i.e. silent, backgrounds.)
Silencing the ad-libs and background action was a huge undertaking that continued throughout the movie and was a constant negotiation with Tom. He liked the way the ad-libs tended to increase energy in the performances and used them to motivate the soloists to project their singing to rise above the clutter. But I maintained that working this way would force ADR when the adlibs and chatter didn’t match in the cuts. Tom understood; while he encouraged active participation in the rehearsals, he recorded the takes with mimed background action.
We finished the test shoot and I was mentally and physically wiped out. It had been the most challenging day I had ever recorded and it dawned on me that we had 70 days of this in front of us, many without the comfort and acoustic security of a soundstage. Every single day would require immense focus and energy from all of us. We got word very quickly as the test was edited and orchestrated that the vocal recordings were a complete success. Everyone was incredibly euphoric that our workflow had been proved not just possible but hugely successful. There were lots of extremely happy producers after the test.
I’m glad I experienced the test before we started shooting because it gave me a chance to prepare myself for an incredibly demanding shoot. The first part of the shoot was a reduced unit in the French Alps shooting Valjean (Hugh Jackman) traveling on foot from the port to the Bishop’s chapel. We arrived and the 1st Assistant Director told me Tom had chosen a location on the highest mountain peak and it was impossible to access it in vehicles. He asked me to go ‘handheld’ because carrying the kit up the mountain would be impossible. I told him that I wasn’t prepared to compromise sound quality in any way and we set about carrying my 180-pound sound cart up the mountain. It took four men nearly an hour to make a 20-minute trip across the boulder-strewn pass. It was a Herculean effort but we arrived at the summit with all the equipment—the proper D/A converters, the big mixing panel, the high-gain antennas—we needed to do a first-class job. It was just this kind of single-minded purpose and resistance to compromise that got us great production tracks.
Quickly, we learned from Tom and the 1st AD exactly where Hugh would be walking and singing and we set about running a battery-powered induction loop under the rocks. With Tom and Hugh’s permission, we had prerecorded the piano track in rehearsals. If Hugh was comfortable setting a pace in rehearsal, we could run playback from a Mac laptop using Audacity rather than take a piano and Pro Tools up the mountain. Valjean was to walk across the summit covered by a single handheld camera. Arthur, my Key 1st Assistant Sound, asked if he could work with a radio boom to help with the uneven surface at the summit. I asked that he remain on a cable for every shot apart from a 360-degree pan so we might minimize radio electronics in the boom signal chain and maximize sound quality. We fitted Hugh with two radio mikes, one tight and one slightly wider. Tom asked us to shoot the rehearsal so I had no idea of the volume to expect. As Tom, Arthur and the camera crew tracked with Hugh and he began to sing, it became clear we were capturing something magical. I quickly listened to the ISO tracks and decided that Arthur’s boom with the Super CMIT was the best sounding track. Due to the tight headroom Tom was maintaining, it was in a perfect position 10 inches above Hugh’s head. I concentrated my attention on Arthur’s boom track in subsequent takes. There was no background noise apart from Hugh’s wooden clogs and his walking stick tapping the granite. They were not compromising the vocal performance and I decided not to bother Hugh about them so he might get on with his acting.
I should add that before shooting, I spoke with Tom, the DP and the camera crew and told them, “Guys, I know you aren’t going to like this and I know we are in freezing temperatures up a mountain but, if this is going to work, I need you all to take off your Gore-Tex trousers and, if you are tracking with the action, just wear your jeans. Otherwise, all I am going to record is the swooshing of Gore-Tex.” This was one of those moments where all the talking about the importance of sound quality and performance was truly put to the test and it was time to see if the crew really understood what that meant. One by one they duly removed their Gore-Tex trousers.
When we arrived home from France and started setting up to shoot in Pinewood Studios, I went to watch dailies at Editorial. I viewed on an Avid machine through near field studio monitors. It was just the raw mix track which in this case was the boom only. As I saw Valjean walk wearily across the mountain range and into a close-up, I could hear his breathlessness due to the altitude and see the fog from his breath on screen. As he started to sing with such fragility from the effect of the altitude, it sounded so real. I was completely spellbound and I knew in that moment that we were creating something special. Never before had I experienced such a connection while watching a musical. As we shot, it became clear to us that we needed to be flexible and use the best method available to record each scene. Scenes like the factory women singing “At the End of the Day” were staged with multiple solos and hard light that made swinging booms to each player difficult. Those scenes were best recorded on radio mikes with the booms playing a secondary role and the stereo boom serving to add dimension to the radio mikes used on the chorus. That was also the technique used for “Lovely Ladies” but for Hugh Jackman’s “Who Am I?” and Anne Hathaway’s “I Dreamed a Dream” and Eddie Redmayne’s “Empty Tables and Empty Chairs,” the boom was the primary recording device.
On “I Dreamed a Dream,” we were shooting with three cameras and, from the first take it became clear that Anne was going to clutch her chest during the emotive parts of the performance. Of course, that was right where the lavalier was placed. To ask her not to do this action, a part of her instinctive body language during the scene, would have been to stifle the truth and honesty in the performance. After the first take, I told Tom that we couldn’t rely on the radio mike any longer and had to get the boom closer. The “A” camera was shooting a close-up, “B” camera was shooting a wider close-up with the same top-line but “C” camera was shooting a classic wide with three feet of headroom. I reasoned that it was unlikely the wide shot would be used for long and the boom could be painted out if necessary, so I asked Tom if he would permit us to bring the boom into the wide shot. The VFX supervisor was present and instantly said, “The shot is static. If you just keep the boom out while the clapper board is going on before the performance starts, we will get a clear piece of the background needed to matte the boom out.” It was this kind of instant answer and collaborative teamwork that enabled Tom to make quick decisions and keep shooting.
The boom was also invaluable on all the sewer scenes where the radios would have become waterlogged. One of my favorite songs in the movie, “Empty Tables and Empty Chairs,” sounds beautiful on the boom and that was possible because all three cameras were shooting close-ups from different angles so the headroom was the same.
Recording singing differs from recording dialog in that the acoustics on singing need to be the same throughout. It would be wrong to have the wide shots sounding “wide” and the close-ups sounding “close” because when the orchestral music is added, the balance of music and vocal would change shot by shot. This would draw attention to the shifts in camera angle. Yet while recording dialog, it is generally accepted that an acoustic change matching shots of differing sizes actually helps a scene to sound real to the audience. For singing, whatever mikes are used must be in a close and uniform position throughout the song.
It is possible to use slightly different widths of mike placement as long as there isn’t a noticeable acoustic shift. We often used two radio mikes on an actor if their performance required extreme dynamic range and I would rig one lavalier close to the mouth to get a very closely mic’d performance on the whispers, but another lavalier five or six inches further away to pick up the louder pieces while sounding a little more open. Of course, the mikes were recorded on separate tracks so the dialog editor had a choice depending on what sounded better in the final context of the scene, once orchestration had been added.
Another break in filmmaking tradition was bringing a dialog editor, the extremely skilled Tim Hands, aboard just a few weeks into shooting. He was based at Pinewood while we were shooting and I was in constant contact with him daily explaining how we covered scenes, which tracks I thought were best and pointing out any issues I thought he needs to know. It was his job to clean and edit the vocals on Pro Tools. He was extremely subtle in his work and mindful of Tom’s admonition to not remove anything that would diminish the audience connection to the actor. He concentrated on removing background noises that had nothing to do with the on-screen performance. When a scene was starting to take shape in the Avid, the Picture Editing Department would give Tim the EDL and he would give them a bounce back of the edited audio from my ISO tracks. This meant that Tim was often working on a scene many times as the picture editor and Tom made changes but it also had the valuable ‘knock on’ effect of immersing Tim in the material so that he became completely familiar with all of it. When Alastair Sirkett joined him in the post-production process, this intimate familiarity helped him get the best from the recordings and the pair of them delivered an outstanding finished product.
After the film techniques clean up, the tracks pass to John Warhurst, the Music and Sound Editor. He went through them using music industry technique to make them sound their best going into the final mix. This process exemplifies the special collaborative workflow for this movie. Supervising Music Editor Gerard McCann pointed out at the beginning of our planning that the skills and objectives of a film dialog editor and those of a music vocal editor are very different. For instance, a music editor would be working out of his usual skill set if presented with generator noise or lighting hum while a dialog editor would not be at home adding reverb to enhance vocals. An oversimplification but because the vocals on Les Mis were essentially a crossover of both mediums, we needed to make sure they benefited fully from each methodology.
Although Re-recording Mixer Andy Nelson’s main contribution comes at the very end of the process, his involvement began at the conception. He has extensive experience in musicals including work on Evita and Phantom of the Opera. Tom Hooper was familiar with his work on Alan Parker’s The Commitments, a project that featured some live recording to a prerecorded backing track, so he sought out Andy when he was first considering live recording for Les Mis. Andy confirmed the success of the live recording on The Commitments and encouraged Tom to take on the larger challenge of Les Misérables.
Tom encouraged me to contact Andy Nelson when I was first hired. Gerard McCann and I had a long conference call with him to discuss workflow and methodology, check that he agreed with our plans and receive any advice he might offer. We kept in contact thereafter and he regularly listened to and commented on material as we worked.
Andy was particularly keen on not using EQ or compressors and limiters in the recording chain. He also asked that processing done by the dialog and music editors be “virtual” so that changes could be reversed and the material returned to a raw state at the touch of a button. He wanted to have complete control at the final mix where all the elements of score, sound effects, Foley and vocals could be evaluated together and judged as a whole.
For instance, he wanted us to avoid using plug-ins to clean up camera noise because they often have a slight effect on the vocal tone and he thought that the orchestration might effectively hide the camera noise.
Jonathan Allen, a Re-recording Mixer from Abbey Road Studios, was also generous with help and advice throughout the project. He worked on the orchestrations in Post but also joined me on days with big chorus ensembles and assisted both with advice and mike placement.
The whole project was a collaborative project from the outset. It set out to bring to the audience the in-the-moment emotions and the live singing of the cast. The success of that endeavor demonstrates what can be accomplished with everyone working together.
Cameron Mackintosh offered daily support and input for the project. He commented that “Music, if used correctly, should pull the heartstrings.” I believe that the filming of Les Misérables, as envisioned by Tom Hooper and with the support of Producers Eric Fellner, Tim Bevan, Debra Hayward and Sir Cameron Mackintosh, and each and every crew and cast member, really does “pull the heart strings.” It was a fantastic piece of work.