The success of Disney's Frozen underlines the importance of music in film, whether it be as songs or soundtracks.
Two of the hottest releases of 2013 were the Disney animated musical Frozen, and its soundtrack. Loosely based on the Hans Christian Andersen fairytale The Snow Queen, the movie tells the story of princess Anna and her sister Elsa, who has the power to turn everything into ice and snow. Despite being released towards the end of November, the movie became the second highest-grossing animated film of the year after Despicable Me 2, and looks set to become Disney's highest-grossing animated movie ever. One of the movie's prime attractions is its Broadway-style songs, which explains the immense success of the soundtrack album. Frozen (Original Motion Picture Soundtrack) spent several weeks at the top of the Billboard albums chart — where it remained at the time of writing — and reached number three in the UK album chart, with international sales exceeding one million.
The standard release of Frozen (Original Motion Picture Soundtrack) contains 32 tracks. The first 10 are songs written by wife and husband Kristen Anderson-Lopez and Robert Lopez. The other 22 tracks consist of one wordless choralpiece called 'Vuelie' composed by Norwegian-Sami musician Frode Fjellhiem and used as the film's title music, and 21 orchestral pieces written and arranged by movie composer Christophe Beck (one incorporating a reprise of 'Vuelie'). Nine of the Anderson-Lopez songs were arranged by Dave Metzger, and incorporate Robert Lopez's original piano ideas, rhythm section and an orchestra. One of these nine songs, 'Let It Go', sung by Broadway actress and singer Idina Menzel, who voiced Elsa in the movie, has also been released in an alternate version, sung by Demi Lovato, with a beefier arrangement and production by Emanuel Kiriakou and Andrew Goldstein that is more in line with today's pop music. Both versions of 'Let It Go' were released as singles, but the Broadway-style original proved more popular, reaching number 22 and 21 in the UK and the US charts while the Lovato version managed 56 and 38.
Leaving aside the Levato version of 'Let It Go' and Fjellhiem's 'Vuelie', the Frozen soundtrack consists of two very distinct sections, and each was recorded and mixed by a different engineer during different sessions. David Boucher took care of the nine songs, while Casey Stone captured Beck's orchestral score. Boucher is best known for his work as engineer and mixer with legendary producer Mitchell Froom. Growing up in Atlanta, Georgia, Boucher studied music engineering at the University of Miami and went on to cut his engineering teeth at various studios in New York before moving to Los Angeles in 1998, where he worked for three and a half years as Bob Clearmountain's assistant. He began his collaboration with Froom in 2001, and has since recorded and/or mixed every album Froom has produced. "Mitchell really believes in the engineer-producer partnership,” remarks Boucher, who in 2001 had faced the considerable challenge of taking over from Froom's previous long-time sidekick, the inimitable Tchad Blake. While mostly based in Froom's analogue studio, Boucher has also expanded into working independently, for example as engineer, mixer and producer for indie rock-violinist Andrew Bird.
Boucher's involvement in the Hollywood film world is the result of his and Froom's association with Randy Newman, who wrote the score for the 2009 Disney movie The Princess And The Frog, which Froom produced and Boucher engineered and mixed. The quality of Boucher's engineering and mixing work, and also, apparently, his capacity for breaking the ice and getting singers to relax, resulted in return calls from Disney. One such call came in October 2012, asking whether Boucher was up for recording the songs the Anderson-Lopez duo were writing for a forthcoming Disney movie. It was the beginning of an 11-month off-and-on process, which culminated in full band and orchestra recording sessions and mixdown at the Warner Brothers Eastwood stage in Burbank. These 11 months began with Boucher recording vocals at Sunset Sound, and later tweaking and mixing the recordings in the box at his own home facility.
"The first thing I received were the song demos,” Boucher recalls, "with Bobby [Lopez] playing piano and singing the male vocals, and Kristen [Anderson-Lopez] singing the female vocal lines. Bobby works in Logic, and his demos came in as a stereo bounce of his piano and a stereo bounce of their vocals with effects. There was no click, because he doesn't think that music should be tied to a click. In some cases Bobby also included some sketches of the rhythm section. We gave a stereo bounce of the vocals and piano to the singers, so they could practise them, and then recorded their vocals at Sunset Sound to Bobby's piano recordings. These were the final vocals that you hear in the movie and on the soundtrack album. We used either Studio 2, which has a Fred Hill-modified Neve 8088 desk, or Studio 3, with the studio's custom API-De Mideo console. The singers would have Bobby's piano in their headphones, and I recorded Idina with a vintage Telefunken ELAM 251 and Kristen Bell [who voiced and sang Anna's parts] with a Neumann U48, while I used a [Neumann] U67 or a U47 for the male vocals. The female singers' mics went into my Little Labs Lmnopres mic pres and vintage Urei LA3A compressors, and for the guys I used Sunset Sound's API mic pres and the LA2A or LA3A. Whenever possible the vocals went into Pro Tools via the Apogee Symphony I/O, and if that wasn't there, the Avid I/Os clocked to an Apogee Big Ben.
"Following each vocal recording session, the vocal-piano recordings were sent as Pro Tools files to David Metzger, with the exception of the song 'Do You Want To Build A Snowman?', which was sent to Christophe Beck. David and Christophe then wrote their band and orchestral arrangements, and made MIDI/sample mock-ups of them, which they printed as audio stems. I would then create mixes of the vocal-piano recordings with the stems of the orchestral and band arrangement mock-ups, trying to make everything sound as close as possible to the finished article. If we could get enough emotional impact from the arrangement mock-ups, we knew that when we replaced these mock-ups with real instruments it was going to be great. So it was important that my mixes were as good as possible. I did them at my home and with Andrew Page [Disney's Director of Music Production] at the Disney Animation Studio complex, initially in stereo, so the Disney guys could listen to them as songs, and the animation guys could work with them. Once they were happy with the arrangements, and the animation was close to completion, I mixed the recordings again in LCR, for screening purposes.
"When we started out recording the final vocals, all we had were animation storyboards with perhaps tiny amounts of animation. Animation normally is drawn to the finished vocals, because the animation usually shows the singer's face. Once the music mix and the images were completed enough, there were screenings in the Avid rooms for the directors and producer and music department. These rooms are LCR [which is normal stereo with continuous panning, plus an additional centre speaker], to make sure the voice sticks to the character on the screen. The dialogue comes out of the centre speaker, so you want the singing in the centre speaker as well, otherwise you'd go from a real centre to a phantom centre which would sound really disconnected. I didn't only put the vocal in the middle, but also added some bass and kick and other low frequencies for a bit of power and to support the vocal, and make sure it feels more connected to the LR speakers. The aim of these mixes and screenings was to get the final OK from directors about the music arrangements. You want to have a decent facsimile about what you will be getting before you drop 100 grand on an orchestra date. Once the lead vocals and images are locked, you are stuck with that, and from that point onwards it's a matter of getting the instrumentation to flatter the vocal as much as possible.”
The orchestra and band recordings for the songs took place at the Warner Brothers Eastwood scoring stage on 22nd to 24th July, 2013. Named after Clint Eastwood, the 465-square-metre stage can accommodate a full orchestra, and features a 96-channel AMS Neve 88RS-SP mixing console and four Pro Tools rigs. Two months earlier, on May 20th, Casey Stone had engineered an orchestral session at Eastwood for Frozen's trailer. With Stone scheduled to record the orchestra for Christophe Beck's movie score later in the year, Boucher liaised with him to make sure their orchestral sounds would not be too dissimilar.
"I called Casey to see whether we could align our setups, so the transitions of the orchestra sound from the songs to the score would be smooth and the audience wouldn't be noticing an entirely different approach. He sent me his microphone arrangement for the trailer session, and I picked the things he had done that I normally do anyway, though I made some adjustments for some of the close microphones. The most important thing was that we used exactly the same microphones for our Decca Tree, outriggers and surrounds. Casey had used [Neumann] M150s for the trailer sessions, but I sent him an audition of the M50s that I typically rent for the tree and he was happy with that, so we settled on them. So with our Decca Tree being exactly the same and the song and score sessions being recorded in the same room, there was enough continuity. The Disney people don't like things bumping into each other in an unpleasant way. The other thing that I did was rearrange the strings on the stage, with the first violins on the far left, as usual, the violas inside left, the cellos inside right, and the second violins on the far right, because it's the way Beck likes his orchestra set up. I believe it was the norm at one point in music history and I have used that layout on other records. It's often done when the composer writes contrapuntal parts for the first and second violins, rather than using all the violins as one section.
"You can see from the microphone list [available for download, see box] that I use different spot microphones than Casey does, but my orchestra also had to compete with drums and electric instruments and vocals. I also knew that the soundtrack was going to be mixed for the Dolby Atmos system,so I added a couple of M150 ambient microphones specifically just for that, and to make sure it would not drive re-recording mixer David Fluhr crazy when he went from song to score during the final dub mix. I used the Neve desk preamps for many of the orchestra mics, exceptions being the Millenia HV3D mic pres for the Decca Tree mics as well as my Schoeps wide left, bonus centre and wide right mics, the POM Audio Design PomVision mic pres for the two RCA 77DX trumpet mics, and EAB mic pres on the two [Neumann] KM254 piano mics. The EABs are some old German broadcast microphone preamps that I hadn't ever heard of until my tech dropped them in my lap, and they sound really nice. You can drive them a bit when you need to, and they have a very sweet EQ. The orchestra was recorded via Warner's Genex GXA8 and GXD8 A-D and D-A converters. I would have preferred to bring my Apogees, but the studio prefers to go with its own equipment, because when you have a recording date with that many people on the floor they don't want to run the risk of any outside equipment going wrong.
"The drummer, Matt Chamberlain, was in a separate booth. I had an AKG D20 on his kick drum and a Shure Unidyne 545 mic on the snare, and both went through API mic pres and 550 EQs. The hi-hat [AKG C451B], mono drums [RCA KU3A], and tom [Neumann U87] mics went through the Grace Designs M801 remote mic pres, and the two overhead mics [U67] through API mic pres and the Manley dual Pultec EQ. The acoustic [U67] and electric [AKG C414EB] guitars also went through the Grace mic pres. The bass, electric guitar and acoustic guitar were in the room with the orchestra, the latter in some cases in the front booth. I brought some of my own gear, like my Urei LA12 and UA 1176AE for some on limiting the drum set. The rhythm arrangements were all written out, but every time we gave Matt some leeway, he would come up with things that made it better. It's this funny thing in music of the whole being more than the sum of the parts. Once you take the collaboration out of music, you are screwed. And the nice thing about working with the guys from Disney is that they know that about music. You can agree or disagree with their taste, but they know how they want to feel when they hear things, and this makes it easier to work together and get a great end result.”
Boucher: "Before mixing I had been working regularly on the Sunset Sound recordings and orchestral mock-up stems in my studio, comping the vocals, volume-automating them and getting them to sound as good as possible. I prefer riding vocals to using compression, especially with a singer as dynamic as Idina Menzel. When she sings quietly she has this really beautiful, evocative, low voice, and when she belts she can pin a condenser microphone right to its limits! Also, there sometimes were lyric changes as the story developed, which meant that we had to go back and punch in vocals, and there also sometimes were picture changes that meant that I had to edit the music to make sure it fit to the new picture cut. For these reasons it was much easier to keep everything in the box. Going back to adjust outboard gear every time there were song and image changes would have been impossible.
"I recorded at 32-bit/96kHz, and had up to 128 tracks in Pro Tools, but one of the problems was that not all of the AAX DSP versions of the plug-ins [as needed for Pro Tools 11] were available, and every time I used a native plug-in, I'd lose a voice. These sessions are massive, and I had to start bouncing. 'Let It Go' may have been the biggest session, and the final mix session consists of 96 tracks, after having bounced the main orchestra tracks to two 5.1 orchestral stems at the top. The top track is a 5.1 mix of the entire orchestra, and below that is a 5.1 mix of only the low strings, where we split the orchestra at the composer's request. They wanted to get some more bite and aggression out of the low string lines. I also added the Waves Renaissance Bass [plug-in] to that track to try to get it to sound a bit more like the low strings were sent through an electric guitar amplifier. It was the only way to get them to cut through.
"I delivered stems of several different instruments and orchestra sections for the final dub mix, and to have enough separation between each of the stems, I recorded the arrangements in sections. The strings, woodwinds and brass went down together, and all the orchestral percussion was overdubbed in separate passes — broken out to give the greatest flexibility at the dub. With each pass I also recorded the Decca Tree and all the other room mics to make sure I had the same room ambience in each stem. Even if an overdub required only one track, I would also have the full complement of tree, room mics and Atmos mics, resulting in enormous track counts for just the one instrument! There were tons of such overdubs for this song, with marimba, vibraphone, taiko drums and so on. You run out of Pro Tools tracks eventually, especially if you're working at high sampling rates. You then either have to lock up an additional Pro Tools rig, which dramatically increases your costs, or bounce things down.”
- Vocals: McDSP G Dynamics, Cranesong Phoenix, Waves VEQ4 Neve EQ & C4, Massey De-esser & P2, Avid ReVibe & Lo-fi
"When I mix a song for a movie I always start with bringing out the tone of the vocal and make sure it's pleasant to listen to. I try to get it to sound so that if you were to listen to it a cappella, it would be compelling. After that I take the vocal level down again and add the instruments, and do more vocal rides to make sure it all sticks together. Sometimes I make choices that I would not normally make when mixing rock records with regards to the vocal and the instrumentation levels, because intelligibility is of paramount importance when mixing for on-screen elements like a singer in a musical. The viewer has to stay with the character on the screen all the way through. So, by rock & roll standards, the vocal will sit a bit more on top of the music. This is vocal music with powerful singers; it's not rock & roll, even though it may have rock & roll elements.
"As I mentioned, at the top of the 'Let It Go' session are my full orchestra 5.1 and the low string overdub 5.1 stems, and immediately underneath is the lead vocal track comp, on which I did extensive volume automation, and the lead vocal aux track. The Disney guys preferred me to have an aux track so that if the vocal comp changed, it didn't change the effects and automation. I had five plug-ins on the insert of the aux track, the first being the McDSP G Dynamics, which is a compressor with a nice side-chain built in. Idina has a really big voice and I didn't want the low end to pump the compressor, so I tailored it with a side-chain EQ so it wasn't quite as aggressive-sounding. I then had the Cranesong Phoenix, which is a tape emulation plug-in, the Waves VEQ4 Neve EQ, the Waves C4 dynamic EQ to make sure her voice didn't sound too brash when she was belting, and the Massey De-esser, which I love, and which works great when you automate it. There's a send to another vocal effects track which has the Avid ReVibe on it, which acts like an outboard reverb. It sounds really good to my ears and I use it a lot. The ReVibe track also has a Lo-fi, which I like to use on most in-the-box reverbs just to degrade the sound slightly, so it doesn't sound quite as digital, and the McDSP P2, a two-band parametric EQ to tailor the reverb so it sat around Idina's voice.”
- Instruments: Waves Renaissance EQ, SSL Compressor, Renaissance Bass, CLA 1176 & CLA2A, Avid ReVibe & Lo-fi, Bomb Factory Sansamp PSA1, ReFuse Lowender, TC System 6000, Lexicon 960, desk EQ & compression
"Underneath the vocal and vocal aux track is Bobby Lopez's original piano track, which has a Renaissance EQ and is then sent to an aux track with another instance of the ReVibe, on a different reverb setting than for the vocals. Between these is a track with a percussion loop that David Metzger made to reinforce the bridge of the song and on which I have the Lo-fi. Following this are four intro tracks with glass harmonica, bowed strings, synths and Wurlitzer, and below that one are several instances of the Decca Tree and room and Atmos mics. The Decca Tree tracks have the SSL Compressor just to add a bit more punch.
"Below these are all the band tracks, starting with the drums. Tracks 23-24 are room mics for the drum kit. For this track I had the rhythm section play to a previous pass of the orchestra, and I opened the door of the drum booth and used a couple of percussion mics out in the room to get a bigger sound on the drum kit. Track 25 is a drum reverb track, with the ReVibe, which is sent to the 5.1 drums stem. Other plug-ins on the drums are the RBass on the kick, the lo-fi and Waves CLA 1176 on the snare, the Renaissance EQ on the mono drum mic, as well as the overheads ride and overheads hi-hat. Next is the bass DI, on which I had the Sansamp PSA1 and Waves CLA LA2A. Below the band tracks are several percussion passes, including the ambient mic complement, with another ReVibe on a tambourine overdub. The tambourine was probably in the booth when it was played, so I needed to add a little bit of space to it. Track 72 is another 5.1 live orchestra stem, and track 81 a percussion submix on which I had the ReFuse Lowender plug-in, which is a really good emulation of the Dbx 120XP [Subharmonic Synthesizer] for the low-frequency effects channel.
"I did almost all the above treatments at home, so there would be no surprises by the time we went to the Eastwood scoring stage for the final mix and mix printing. I laid everything out over the 96-input desk at Eastwood, and also used the small faders for reverb returns for the outboard reverb I used, which came from the TC Electronic System 6000 and Lexicon 960. I also used the analogue EQs and compressors on the desk. I prefer to mix on an analogue desk, and I do this all the time on the Amek Media 51 desk at Mitchell's studio, or the Trident Series 65 that I have at home. But that doesn't mean that I can't get good results in the box. It just means that the knobs get turned differently. The reasons to go to Warners for the final mixdown was to take advantage of the monitoring, and of the analogue summing in the Neve desk, which makes things sound a bit warmer than if you sum them in Pro Tools. Also, while being at Eastwood the Disney guys could make their comments, and I would tweak things accordingly.
"Once the mixes were approved, I would stem them out and send them to David Fluhr for the dub mix, which took place at Disney Digital Studios Stage A. I delivered many elements of the mix as 5.1 stems, but the guitars might have been 4.0, meaning 5.1 without a centre speaker and a subwoofer, and the bass and the drums would each have been 3.1, meaning LCR and a subwoofer, the piano probably 5.0, and the orchestra and percussion was all in 5.1. So there was a hodgepodge of different stem layouts, plus the Atmos tracks. When David Fluhr put all the faders of all the stems on his desk to zero, he would have my mix. At the same time he could endlessly tweak it if he chose to. I'm not a big fan of giving people that many options, but as [sound engineer and re-recording mixer] Michael Semanick explained to me once, 'You would rather give me the option to duck an instrument than risk me trying to EQ it out and wreck the whole mix.' If a director says, 'Get it out of there,' it goes.
"After having printed all the stems, I then mixed them to stereo, again going via the Neve. I did not use any compression or EQ on the stereo bus, and sent the stereo mixes to Bob Ludwig, who mastered the soundtrack album.”
Scoring mixer Casey Stone has amassed an impressive credit list, having recorded and mixed the scores for close to 200 movies, including Boogie Nights (1997), American Wedding (2003), The Muppets (2011), and Frozen. Many of his scoring credits originate from his ongoing collaboration with Canadian film composer Christophe Beck. Stone is originally from Iowa in the US, and studied at the University of Southern California, where he obtained a BSc in Music Recording. Following this he worked for a while as a freelance music engineer, before moving into recording and mixing film scores. He has been based in the UK since 2008.
Stone recorded the orchestral score for Frozen at the Warner Brothers Scoring Stage in LA, to Pro Tools HDX1, running at 32-bit/96kHz. The orchestral score recordings lasted six days (September 3-6 and 9-10), and mixdown seven (September 11-13 and 16-19). Stone mixed the orchestral recordings at Christophe Beck's studio in Santa Monica, on a 300-channel Euphonix System 5 digital desk, using two Pro Tools rigs plus Stone's own dedicated Altiverb computer and favourite Klein & Hummel O300 monitors.
Stone: "One Pro Tools rig was for playback, and the other to print the mix to. Because I was splitting out the mix in several different 5.1 stems — strings, brass, woodwinds, percussion, and so on — I was printing up to 64 channels. Hence the need for a second Pro Tools rig. I did an offline sample-rate conversion of the original 59 orchestral tracks in the first Pro Tools rig from 96k to 48k, because the dub mix is done in 48k, and then sent these 59 channels to the Euphonix, on which I balanced them to 5.1, using desk EQ and my Altiverb. I do very little mixing in the box, though there was percussion for several action scenes that I premixed in the first Pro Tools rig (see 'Summit Seige' box). I was monitoring in 5.1, using a combination of my K&H O300 monitors for LCR, and the studio's own Genelec monitors for the other three channels. I did not make any special allowances for the fact that the soundtrack would also be mixed in Atmos.
"I usually learn the styles and musical themes in the score as we go through the recording process, and when I'm mixing it's important that I understand the context of the music within the film. I can usually tell from the music alone how the cue is supposed to function for each scene, but it really helps to get the sound and balances right if I mix the music while also monitoring the dialogue and effects — I probably do that nearly half the time. Although the composer has crafted the score to interact with the dialogue, it is also part of a scoring mixer's job to ensure that the score works well with the dialogue and the sound effects. Our primary focus is the film, we are not mixing some nice music for an album!
"I prefer to mix the cues in the order in which they appear in the film. This allows me to see and hear how the score develops, in the same way the viewer will. Having a digital desk is crucial here. When I have my first loud, driving orchestral cue, I'll save it, so I can use that setting as a starting point for the next cue that is similar in nature. The same when there's a cute comedy scene, with clarinet and pizzicato strings. I make a note of it and can reload it for the next comedy section. On an analogue console I'd have to do it the old way, by mixing all the cues of a certain type together before moving on to different styles.
"While mixing, most of what I do is riding faders for levels, and I pan everything in the same way as the orchestra was seated in the room. I don't really use any compression on the orchestra, but will apply EQ on the desk, and add Altiverb. I have my own dedicated computer that runs Altiverb, under a host program called Plogue Bidule. My computer has a MADI card, so it connects via that with the Euphonix desk, which uses MADI as its native digital I/O. MADI is a 64-channel format on a single BNC cable. I really like the way Altiverb sounds. My go-to reverb for big, orchestral halls is 'Mechanics Hall'. Since I am mixing in multiple surround stems, I need one instance of Altiverb for each stem, like strings, brass, winds, percussion, vocals, and so on, so I may need as many as 45 channels of reverb returns!
"There is no separate process for printing the stems while I am mixing, they simply get printed all together when I press record to print my mix. After printing the 5.1 mix and having sent the stems to David Fluhr for the final dub mix, I spent the last day folding these 5.1 stems down to stereo for the soundtrack album. I did this in the box. I don't remix, it's just a matter of finessing things. I listen in stereo as things are folded, and rebalance where necessary. The composer may also want certain parts of the cues mixed differently for the soundtrack album. In this case it was really nice for us, as the score team, that the score shared an album with the songs, because I think we are on our way to a platinum album now!”
A comparison of Casey Stone's list of orchestral microphones for the Frozen sessions with Boucher's shows up many similarities. Both used Neumann KM130s as surround mics, M50s for their Decca Tree, and Schoeps MK2S for wide left, right and centre mics. They also use similar close-mic setups, although many of the close mics themselves are different — Stone favours Neumann KM140 mics on the violins and violas, Sennheiser MKH40s on the cellos and MKH800s on the basss, MKH40s and MKH406s on the woodwinds, while Boucher uses a far larger variety of mics, for example Neumann U67, M149, M49, RCA KU3A, AKG C414EB and C24s just on the strings. The main close mic overlap is in the percussion section, where both seem to like the Neumann KM184. Stone used 59 orchestral mics and inputs, plus eight for the choir when required, while Boucher used 55, plus 14 for the band and two Atmos mics.
Stone: "When you look at the microphone setups of a wide variety of scoring mixers, they would look pretty similar. Some people use more microphones than others to give themselves more options in the mix, but overall I don't think it's that different. Many of the differences occur in the mix in terms of how the amount of room microphones is balanced against the amount of close microphones, and with that the amount of artificial reverb that you use. Many of today's scoring stages are on the dry side, so artificial reverb can have a big effect on the final sound.
"Of course I get input from directors, producers, the movie studio, and the composer as to what orchestra sound they want. Christophe definitely likes things on the drier side. He doesn't mind reverb, but he doesn't want to hear it. I also prefer to hear the instruments clearly, within a big and wide soundscape.”
Although Casey Stone mixed the entire Frozen score on the Euphonix desk at the composer's studio, he executed some nifty in-the-box moves for a percussion part of an action sequence in the track 'Summit Siege'. At the time of mixing the working title was still 'Hans Captures Elsa', as can be seen in the accompanying screenshot.
Stone: "We recorded this percussion part, which was for an action scene, during a separate session. The live recording consists of 10 tracks, marked '44852' in the screen shot, incorporating two surround mics, the Decca Tree, a piano recorded in stereo playing some low crunch chords, and three percussion mics. To the right of them are two stereo percussion synth tracks, named 'LowPercSweeteners' and 'TomSweeteners' in the comments window, and they were used to beef this section up, which is all about big, layered hits. To the right of the 'TomSweeteners' track is a bus called 'Big Percussion Bus'. The two synth tracks and the bus all go to the Euphonix desk via a 5.1 output called 'Big Perc'.
"You can see the surround panners on all the live tracks. I placed them in 5.1 and then sent them, via outputs marked 'BP51', to the input of the Big Perc Bus, marked 'BgPrc51'. The bus had a multiple-mono instance of the SoundToys Decapitator [distortion plug-in] operating on the six channels, to crunch things up a little, plus the Waves L360 5.1 limiter to remove some peaks, and there's also an Avid EQ III, shown here doing ridiculously little. I think it was automated, and it may have done more in other sections. Because the two synth tracks were stereo, I had to spread them across 5.1, and I used the Waves UM226 Stereo Surround plug-in for this. The two 5.1 synth tracks and the 5.1 Big Percussion Bus came up on six channels on the Euphonix, where I'd have one fader to control them.”