The best engineers thrive on pressure. Which is handy when they’re recording the farewell tour of one of the world’s biggest rock bands, and timecode trouble is brewing...
On shows 17 and 18 of their “last major tour”, the most prolific band in prog rock came home to Toronto to blow out the Air Canada Centre. Celebrating 40 years of epic musicianship, with over 40 million albums sold, Rush graced the cover of Rolling Stone magazine. Coming only a couple of years after their induction into the Rock & Roll Hall of Fame, it is in some ways the culmination of their entire career.
Fittingly, the R40 show was designed to take fans on a trip through the band’s history. Throughout the performance, the stage shifts and changes, pulling the audience through that lifetime, and on occasion projecting cameo appearances from Hollywood’s elite. All the iconic Rush backdrops make an appearance, from the washing machines to the giant amplifier stacks, until we end with humble combo amps on wooden gymnasium chairs, recalling the band’s earliest days.
With the tour marking such a significant milestone, Rush chose to record their Toronto homecoming shows for the R40 Live DVD and album, and had no hesitation when it came to choosing a producer for the job. “After having worked with Rush on a few previous projects, we’d built a great working relationship,” David Bottrill tells me from his mixing studio, where he also mixed the R40 show. David and I have been working together for years now, so I was fortunate enough to have been brought on as Assistant Audio Producer.
My own audio career started out in post–production for TV, working on smaller shows that had found their way onto bigger networks. I’d then spend my nights recording music, and ended up building a career juggling studio work, mixing music for broadcast, and recording classical music. I met David through a mutual friend and colleague, producer/mixer Brian Moncarz. David and I chose to start working together as colleagues in production when the timing worked out, starting with smaller independent projects and evolving into larger projects, like Rush. These days, our work is closely interwoven within his company Mainstation Music.
David’s credentials mark him out as one of the exceptional producers and mixers of the past 20 years. Having started his career working under Daniel Lanois at Grant Avenue Studios near his hometown of Dundas, Canada, David was swept over to the UK during the making of Peter Gabriel’s career–defining album So. Following that, David became the head engineer at Gabriel’s Real World Studios, where he co–produced the former Genesis frontman’s score for Martin Scorsese’s Last Temptation Of Christ, then engineered Gabriel’s Us.
David eventually left Real World to become a freelance producer, and has produced or mixed artists including Muse, Placebo, Smashing Pumpkins, Kid Rock, Silverchair and Stone Sour. In particular, his work with Tool defined them as one of the most original and important metal bands of all time.
“Being a teen in the ’70 in Canada, I was absolutely a Rush fan,” David told me during the tracking the R40 shows; and prior to R40, he had already been responsible for a remix of the band’s Vapor Trails album.
Preparation is paramount in live recording, so the weeks leading up to the Toronto shows were mostly spent organising the Pro Tools session and liaising with Doug McClement at LiveWire Remote Recorders, who provided the recording truck. The LiveWire truck houses a 96–input Solid State Logic C200 digital mixer, which only fits into the truck because one side of the trailer expands to reveal a 16x12–foot control room. Designed by Pilchner–Schoustal, the studio is a remarkably well laid–out workspace given the size constraints.
Recording takes place to a Pro Tools HDX system, but there is also an unexpected amount of vintage outboard equipment. “Most of the outboard has been accumulated over the 40–plus years that I’ve been doing live recordings,” says McClement. “We can insert into any track if needed, though these days we tend to rely pretty heavily on the SSL’s compressors and EQs.”
The most consuming part of pre–production was arranging the signal flow so that on the day there was no confusion between the front of house and the recording truck. The truck’s signal path was completely independent from the front of house with the exception of the mics themselves. Given that there would only be a short, shared soundcheck between the two camps, we knew that organisation would be vital. There would be no time to deal with issues while Rush were soundchecking (which they still choose to do, despite having full virtual soundcheck capabilities).
Brad Madix, Rush’s front–of–house engineer, runs an Avid Venue system, and provided us with a virtual soundcheck Pro Tools session well in advance so that we could configure our Pro Tools session long before the shows. Besides having a session that was easy to navigate in the heat of a live recording, the priority at that stage was to minimise the amount of management that would have to go into the session when starting to mix. Often in mixing projects you’ll lose hours trying to fit people’s files into your workflow. Given the immense size of this project, we knew that we had to mitigate those setbacks as much as we could. David and I have spent the last couple of years refining our system of mixing, and the prep work culminated in a 191–track Pro Tools session — quite busy for a three–piece band! No fewer than 48 of these were drum tracks. David had the virtual soundcheck tracks laid out in a rough mix within the Pro Tools session so that we could easily test everything out in the recording truck.
Once the R40 tour kicked off, David made a point of catching an earlier show in Buffalo, New York, to ensure that our recordings would translate accurately to the show experience. “Knowing the atmosphere that you’re trying to capture is vital to a project like this,” David explains. “A big part of that is the audience, but it’s also hearing how Brad represents the music. From that I’m able to evaluate the direction that I’d like to take.”
On the morning of the first Toronto show, LiveWire’s incredible crew (Gary Tompkins, David Johnston and Alex Halayko) ran a 500–foot fibre–optic snake through the backstage folds of the Air Canada Centre to the SSL console’s external, digitally controlled mic preamps. These lived under the stage, next to the mic splitters that also sent the mic signals to front of house and monitors. Back in the truck, we used a comms system to talk to David Johnston, who had the gruelling job of being beneath the stage during soundcheck, making sure that everything was patched correctly. With two drum kits and an incredible 14 audience mics, the soundcheck was barely enough time to check all 96 lines. David and I are more used to the pace of studio life, so in the gap between the soundcheck and the first set we couldn’t help but acknowledge the looming question: “That sounded good, right?” We had certainly taken for granted the luxuries of being in the studio. But the instant Rush kicked off, we were pummelled by their force and knew we were in good shape. Or so we thought...
There were dozens of cameras stationed around the Air Canada Centre, which is a 20,000–seat arena. In order to ensure sync, all of those cameras and the recording truck had to be slaved to timecode, which was to be generated by the camera department. Given the distance between all the cameras, a battery–powered timecode generator was used for portability. We had reservations about that, but had no choice but to trust the camera department. Once all of the cameras had been sync’ed, the recording truck was left with the timecode generator.
The first half of the show on day one went flawlessly. Knowing that we had an intermission, we stopped the recording. The room fell very silent moments after when we realised that the timecode had stopped because Pro Tools could no longer read it from the generator. The intermission countdown had started, so we had 20 minutes to solve this problem. A bedlam of camera assistants clambered to remedy the issue, running laps around the arena to get the timecode generator sync’ed again and then get back to the mobile truck in time for the second set. I felt my heart pounding all the way up to my throat as I watched the last scene of the intermission’s video play. In the final seconds before Rush stepped back on stage, with no sign of a camera assistant, I was about to hit Record and just deal with the consequences of a wild take later, when I realised that I had Insertion Follows Playback switched off in Pro Tools. This, of course, meant that the cursor was back at the beginning of the first set, and would have recorded over it if I hadn’t quickly thrown the transport forward in time to a random spot where I knew we’d be OK to record the second set.
The show had initially started around 20:20:00:00. We knew that the total running time of the show, including intermissions, was just under three hours. As planned, that would have had us ending the show around timecode 23:20:00:00. However, in my frantic attempt at heroism, with a Pro Tools session zoomed out to see the entire first set, the random spot that I’d thrown the transport to was 22:49:02:19. Doug McClement, who’s done thousands of live recordings, relayed a story about having seen the Pro Tools transport stop cold when it crosses 24:00:00:00. None of us knew whether that bug had been addressed or not — but, looking at Rush’s set list, we realised that the transport would pass midnight sometime during the encore. We were living an audio nerd’s rendition of a Brothers Grimm story.
We deliberated on the possible solutions, and settled on stopping the recording in a between–song gap, throwing the transport back to 00:00:00:00 and then re–sync’ing it later. I pushed for this solution because I knew that I could fix it with the Tascam X48 backup recording that we were running, which was gain–matched to our master and could easily be digitally transferred into Pro Tools. It was an hour of editing work that I was perfectly happy to do if it meant saving the first day’s recording. I just couldn’t live with the idea of potentially losing the recording because of an obscure transport issue. David and Doug were on board, and with a little bit of editing finesse, everything turned out just fine. As for the timecode generator, it turned out that it was a faulty BNC cable, as it always is. This was one of the most elaborate concert productions of the decade — and it was all teetering on a single copper wire. Needless to say, the next morning we ordered our own AC–powered timecode generator and extra BNC cables, and they held our sync perfectly for the entire second show.
The weeks that followed the recordings were spent doing general work on the session while trying to get approvals on the set and the songs to be used. Knowing that there was a hard deadline for the release, David managed his time by starting to shape the sounds of the second day’s show. It was a really strong performance, and we all felt confident that it would make up the bulk of the release.
David and I mix entirely ‘in the box’, with mirrored setups based around Universal Audio Apollo interface/DSP units. We both own outboard, but have found that the quality of plug–ins and converters these days, coupled with the universal expectation of recall, means that the ITB approach just makes more sense. For this project, we also brought in an SSL Nucleus: David has tried a lot of control surfaces, but having spent the majority of his life working on SSL desks and having owned an AWS900, it seemed like a natural progression for him. “Using a controller is all about tactility for me, because that’s how I grew up working. It’s about being able to connect with my mix, not my computer.”
Once everything was in place and the song approvals had come in, David dug into the mix. The system of mixing that we use relies heavily on aux channels and fairly complex routing, which means that you really have to be mindful of headroom, especially in larger sessions like this one. We’ll use the Pro Tools Clip Gain feature to rough–in how hard we’re hitting inserts; then the output of the last insert in the chain is used to optimise where the fader sits, because, if possible, you really want to avoid it being at the bottom of its travel. The routing of the session is such that each set of instruments ends up in a subgroup. These subgroups are basically David’s ‘master layout’, which he balances to create the mix as a whole. They feed into the master bus, which has some processing on it: David is really fond of the UAD Shadow Hills Mastering Compressor but only tends to use the discrete compressor, often running through the iron transformers and never pulling more than a dB or two. Following that is a Brainworx BX_digital v2 equaliser to add some air and roll off unnecessary sub–bass frequencies.
Neil Peart used two drum kits in the show, one for each set. The kicks were recorded with either an Audio–Technica AE2500 or a Shure Beta 91. The snares were miked with either an Audio–Technica 23HE or an AE3500, and there were Audio–Technica AE3000s on toms, ATM450s on hats, AT4060s as overheads, and Neumann KM184s on the ride cymbals. There was also a medley of Shures, AKGs and Audio–Technicas filling in the gaps on the percussion.
The two kits each sounded quite different, which raised the issue of having to make both work within what is essentially the same mix. We wanted to retain the character of each kit, without having one sound better than the other or leaving holes in the frequencies of the mix. This was a bit of a balancing act, and in the end, each kit was processed separately but treated in much the same way. We didn’t compress anything on the way into Pro Tools, but at the mix, the kick tracks were usually sent through a UAD 1176 compressor with a medium attack and a fast release, which was then fed into a UAD Neve 1081 EQ to add some bottom, some ‘point’, and some air.The snare mics each fed into their own compressor: usually, either a UAD 1176 or Fairchild 660, but David occasionally reached for a UAD Empirical Labs FATSO. Regardless of what compressor he uses, David has a light hand with compression.
The snare EQ was usually a UAD Neve emulation of some sort, but when it comes to more surgical EQ duties, or just as a general go–to plug–in, David usually reaches for McDSP’s FilterBank, which was on all of the cymbals and percussion. The different elements of percussion also had McDSP’s CompressorBank on them. Both plug–ins are incredibly versatile, while remaining musical, even if you push them. The audio community seems to be really obsessed with plug–in emulations of old gear, as though there’s some sort of vintage elixir in them, but we don’t think that the GUI is necessarily indicative of how musical a plug–in is.
For drums, David often set up a plate and a non–linear reverb, but with Rush, David also added some surround hall, which is used on other instruments as well to add some glue and give the recording a more suitable environment. You can really get a sense of how the spaces meld together during the drum solo.
The drums all ended up being bused to a Fairchild 670 for a bit of excitement, and were then lightly equalised with an SSL X–EQ from the Duende Native bundle to add some air and a bit of low end. Finally, there was a multi–band compressor at the end of the drum chain, which was just used to tame the odd rogue transient.
Geddy Lee’s bass tone was impressive without any treatment to it, thanks to his playing technique. In particular, every time he used his Zemaitis bass, which has custom–wound pickups by Tom Brantley, we couldn’t help but be completely amazed by the tone. It was one of those sounds that makes you turn your head, which I remember doing when the band played ‘Losing It’. The bass sound is a mix of amps and DIs that David usually split into one of two buses. Broadly speaking, one bus was stereo and was used for any widening, while the other was mono. They got blended together in various degrees depending on the song.
The main compressors were UAD LA2As, and the EQs were a medley of Neves and APIs, chosen to suit the application. “For me, Neves are used when the low end needs to be filled out more,” David clarifies, “whereas APIs, 560s in particular, allow me to carve the sound better and add some grit.” The low end of the EQ usually focused on frequencies below about 80Hz and also around 200Hz to 300Hz, leaving a bit of ‘pocket’ for the kick. Cleaner bass tracks sometimes needed a bit of help from Waves’ Renaissance Bass. A distorted bass sound from Geddy’s SansAmp, was used at a fairly low level to help with definition.
Alex Lifeson puts a lot of effort into his guitar sounds, which are constantly changing throughout the set. Primarily, the sound comes from either a Lerxst Amp or a Boogie, mostly running through Palmer speaker simulators, unless it’s a DI’d acoustic like in ‘Closer To The Heart’. We also had separate stereo feeds for his effects, which are driven by two Fractal Audio Axe–FX units. You always want to honour the effort that a musician has put into his sound, but at certain parts through the set, David felt that the guitars needed a bit more weight. These days we rely heavily on a Kemper Profiler Amp for most re–amping duties. We’ve had one for a couple years now and have profiled all of our own amps through our own recording chains. The results were indistinguishable from our live amps, so we often just use our own Kemper profiles on mix projects, then record the re–amped signal back through Chandler Little Devil pres, or the UAD Neve 1073 Unison preamp emulation. As with drums samples, though, the re–amped tracks are really just tucked under the original sounds to help fill them out and add some harmonic complexity.
The UAD SSL E–Series channel strip is usually David’s go–to for guitars. The compression is always light but it’s the smoothness and clarity of the SSL EQ that really shines on guitars, which can get quite muddy. I think that David is also just accustomed to how they react in a mix. Usually, one of the EQ bands is pushing some weight into the part, so somewhere below 200Hz, and then another is adding some definition and presence. That frequency is much more dependent on what the other instruments sound like but it tends to be in the upper mid–range. Sometimes David will also add a high shelf to brighten up a darker guitar sound, if a song calls for it. David tends to use additive EQ more often than not, rather than subtractive. If a guitar doesn’t have an SSL on it, it’ll likely be an API 550, which always seems to work well on electric guitars if they need some bite.
A handful of different delays are used on guitars to either help give a sense of space or to add character to a part. The delays range from tape slapbacks to longer, analog–style delays with really characterful repeats. Which delay David uses depends entirely on what part is being played. Waves H–Delay and SoundToys EchoBoy are two of our favourites because they have so much personality, but it’s not uncommon that David will just use the generic Avid Mod Delay. If you’re looking for clarity in the repeats, a standard digital delay is often the way to go. Guitars may also have a bit of reverb added, which is usually either the UAD EMT140 or a spring reverb of some sort. Often the different reverbs are used to help differentiate guitar parts. For the most part, spatial effects were subtle in the R40 mixes, especially since Alex’s own effects were the priority. There were a couple of great moments though, like in ‘Anthem’, where David really used delays in the surround channels.
All the band’s vocals were recorded using Audio–Technica AE6100s. David always started his mix chain with a McDSP 575 noise filter, which is perfect for notching out problematic frequencies. In this case it was just being used to roll off the bottom end around 160Hz so as not to hit the 1176 on Geddy’s vocals too hard. You have to be really mindful of those low, often unnecessary frequencies because they can really push your compressor, squashing the mid–range. The 1176 is the Anniversary Edition, which has a 2:1 ratio. It sounds really natural on vocals, even when pushed, but it can still have that 1176 aggressiveness if you want it to. The main EQ was a Pultec, which is used for that smooth top–end air that Pultecs are known for, while some bite was usually added in the mids from a Trident, Helios or SSL, depending on the song and on which band member was singing. The vocals got sent to numerous buses, each with a different effect on it. The UAD EMT 140 plate is a favourite on vocals, as is the Waves H–Delay. If a song calls for a different kind of reverb, David will turn to Valhalla Room or the AMS RMX16, both of which have their own character.
The band were still on tour for the initial mixing stage, so it was often difficult to get them in a listening environment where they felt comfortable giving mix approvals. But given the release timeline, David had to keep moving forward with mixing, the hope being that the band wouldn’t require any fundamental changes to the mix that would unravel everything. Thankfully, with a few minor tweaks the band were happy. Rush are so seasoned at releasing live material that once they’ve entrusted someone to mix, the process seems to run smoothly. Their main concern was the stereo mix, so David constantly had to make sure that the stereo fold–down was producing accurate results, especially in the vocals, which can easily collapse when the centre channel is removed. Ideally, you’re aiming to get both the surround mix and the stereo mix to be reminiscent of each other, and everyone agreed that David had achieved that.
It wasn’t lost on David or me how fortunate we were to have been asked to do a project that was so important for Rush and their fans. It’s so rare to go in knowing that the outcome will hold significance for people, and the fans who came out to support Rush for the R40 shows were a testament to that. Ultimately though, these Toronto shows were more than just moments of hometown pride. They were a kind of liturgy where fans could show their devotion to a band that has been with them for most of their lives. In return, adding to their immense catalogue, Rush has given fans a new chapter in their legacy. It was a beautiful send-off. A Farewell To Kings, one might say.
The R40 concerts needed to be mixed in both stereo and surround. David Bottrill’s approach to surround mixing has changed over the last few years. “One of the first surround mixes that I did was the Smashing Pumpkins’ Oceania: Live In NYC. I began the mixes monitoring primarily in stereo, but I found that when I went to the surround mixes, they didn’t translate as well, so I approached the mixes again starting with the surround and folding down to the stereo mixes. That always seems to translate better.”
Surround mixing for music still feels like uncharted territory. We’ve spoken to many other mixers and mastering engineers, Bob Ludwig and João Carvalho both having been particularly helpful, and anything that remotely resembles a standard is a loose one at best. When David started mixing surround he just went in trusting his ears and mixed to what he felt sounded good to him. Now, after having mixed numerous surround projects, he’s developed something of philosophy.
We’ve heard a few music surround mixes that put the listener on stage, which always seems jarring to us. David wants the listener to feel as though they are sitting in the best seat in the house, so he always mixes to the audience perspective, which means that the Ls and Rs surround channels are used primarily for the ambience and audience mics. At times, Dave will also place synth pads in the surround speakers, but the concern is always making sure that your attention isn’t drawn there because of it. If there’s ever a heavy–handed delay that gets sent to the surrounds, from a guitar for example, it’s for effect. The rest of the time the listener should be focused forward.
For the R40 shows, the Air Canada Centre had temporarily removed the giant scoreboard in the centre of the arena, so the sound was a bit more open. Fourteen audience mics had been rigged at various distances and heights, and David rode the faders heavily to replicate the excitement of being at the show.
The second consideration was the use of the centre channel. David feels quite strongly about its place in a music mix: “The centre channel is great for adding some extra fill to the phantom centre, but you really have to be careful not to push it so far that it starts to sound disassociated from the music.” David also wants to ensure that all listeners get a great experience, even if their speakers aren’t calibrated or set up properly, which means that the phantom centre has to do the job even if there’s no centre speaker. For this reason, Geddy Lee’s lead vocal is only going 30 percent to the centre speaker, and the kick and snare are the only other instruments being sent to the centre, at an even lower level.
David uses the Waves 360º Surround Tools bundle, relying heavily on the M360 Surround Manager & Mixdown plug–in to handle the bass management. The LFE channel was used mainly to showcase Geddy’s bass because, with Rush being a three–piece, it plays an even more integral role in filling out the mix. There’s a bit of the low end from the drums being sent to the LFE as well, just to make sure you really feel the kick and the larger toms.
David Bottrill has the sharpest ear of anyone that I’ve ever met. His pitch perception is unwavering, and I’m pretty sure his years of recording odd time signatures and polyrhythms have given him a metronomic sense of rhythm, too. I actually think that this meticulousness is the foundation of his approach: there’s a precision to his mixing which I think stems from his perception of pitch, time, phase and so on, and how they affect the overall performance. But David walks a fine line in that respect, because he doesn’t believe in gridding sessions or tuning all the vocals. “I try very hard to retain the dynamics and idiosyncrasies of a player,” David says. “So I make sure that samples become a part of their sound, not a separate entity from it.” This seemingly small distinction has a huge effect on the final mix.
When we start a mixing project we generally get our assistant, Mike Monson, to add in drum samples, which is all done manually. However, David is always very clear to point out that he doesn’t ‘sound replace’: he ‘sound enhances’. There are about six to eight different samples that get placed for both the kick and the snare, and just one for each tom. However, not all the samples are used in every song. Each of the samples has a slightly different character, and depending on the qualities of the original recording, David will add complementary samples. Even then, the samples are always very low in the mix — generally about 20dB underneath the live drums — so they’re never a crutch for poorly recorded sounds.
In the case of live shows like this project, where the bottoms of the toms and snares weren’t miked, the samples help add a bit of body and tone, leaving the live drums to give the transient attack and liveliness. The samples are also used to help key gates on the live drums. It keeps everything locked in together and cleans up the ambient noise.
While Monson did his work, David and I cleaned up the rest of the session. A few small edits were needed, which I suppose is no surprise, but I’d actually gone in assuming that there would be more. Rush are so incredibly consistent that the edits were all very minor, and usually involved pulling in moments from the other night’s recording to fix a buzz, noise or tuning issue. If someone pushes their guitar out of tune, is too plosive or knocks their mic stand, it has to be cleaned up — that’s just the expectation of a professional concert release. From a technical standpoint, however, Rush are one of the most accomplished bands in the world. They really don’t need to be ‘enhanced’, and their 40 years together have proven that. David’s job as a producer was finding the balance between honouring the live show and making it as audibly exciting as possible, but the priority was always to capture who Rush are as a live band.