David Crosby’s latest album has gained much praise for its sonics. The man who tracked and mixed it offers SOS readers the inside story — with audio examples.
The pureMix.net crew and myself were putting the final nails in the coffin of NAMM 2016 when Michael League’s face lit up my phone. Michael and I met a while back working on a Lucy Woodward project that he produced and I mixed, and then we collaborated on a crazy one-off song with his band Snarky Puppy featuring Stolar that he arranged and played and I produced and mixed. We also spent a lot of time co-producing Banda Magda’s expansive soundscapes over the years, so that day when he offered me to come work with him on the next David Crosby record I knew it’d be a cool project.
I don’t usually ‘only engineer’ records. I tend to produce, record, mix and sometimes master the whole thing myself. In this case, considering the calibre of the personnel involved, I thought it’d be fine to not carry the whole record alone, but I did make Michael aware that it is hard for me to not jump in with production ideas when I feel the record needs it. He made it clear that was partly why he called me, so I signed on. I’m glad I did.
David and Michael had been working on the record for a while already. They wrote the songs together. They cut demos. They already had decided to keep the record drum-and-percussion free. They had a direction. I thought it was wonderful. When I heard the first demo, I immediately heard how I wanted the record to sound in the end. There was so much space to exploit, large areas between the notes that I could see myself really taking advantage of. Plus, Michael gave me carte blanche with a bold “Do your thing!” and never asked a question. Paradise for experimentation and highly reprehensible shenanigans.
My immediate goal, sonically, was to give a nod to David’s legacy, to hint at the tones and blends of his work with Crosby, Stills & Nash and Crosby, Stills, Nash & Young, but to do it with a modern aesthetic using the tools we have today (no tape was harmed in the making of this record). I decided to try to make it so you could feel like David Crosby is sitting right in front of you with his 1969 Martin and singing you a song — a fun experience I’ve been privileged to hear at first hand — with Michael League playing bass and guitar (also tremendous fun). I had already experimented with similar aesthetics on Will Knox and Cyrille Aimée records, and it seemed to me that a David Crosby record would be the perfect occasion to go further in that direction. In the end it came out pretty close to what I envisioned when I first heard the demos.
I usually work at Flux Studios, my spot in NYC, but David wanted to work in California and had already booked his friend Jackson Browne’s place, Groovemasters in Santa Monica, for a whole month. There was little information online about the place: no real pictures, no gear list, nothing. A call to Ed Wong, the lovely studio manager at Groovemasters, provided some info, but not enough for me to project what it would take to track a major record in there. My friend John Paterno, who lives and works in LA, knew the place, had worked there years before, liked it, and even recommended the assistants as being great. So I got some positive vibes, but still no current details — which was kind of scary considering we were planning to track keepers on day one.
I really only discovered the place when I showed up there at 10am the first Monday of the session. In essence, Groovemasters was a time machine with the dial set on 1988-ish. It had all the trappings: a big Neve 8068, a Neve BCM10 sidecar, a big producer desk filled with all the gear that has made us (and still makes us) all salivate when the names are mentioned, more Neve 1073s than you can count without fingers getting involved, drawers full of ELA M251s, U47s, RCA 44s and even a Fairchild compressor or two.
The control room was setup in classic style with lots of room in front of the board, gigantic but completely unusable main speakers embedded in the walls, old Tannoy/Manley Gold passive nearfields on the console bridge, and then movable racks on the side with the newer gear like the standard early 21st Century Apogee converters and assorted computer silencing box. There was so much gear and associated paraphernalia that the bathroom attached to the control room had been turned into a machine room for the Neve automation. It was kept so cold that no-one really went to the bathroom in there more than once over the course of the session.
Since I had little info about the place before I showed up, I had decided to plan ahead and bring some of my own stuff just in case. I brought my Lauten Eden prototype and a couple of Lauten Atlantis mics, a Dangerous compressor, BAX EQ and Source, three pairs of Focal Spirit Pro headphones so we could all listen to the same thing when tracking and a Universal Audio Apollo Twin just in case. We ended up using all of it. Including the Twin.
The first thing I do in a new studio after tasting the coffee is listen to the control room at the tracking/mixing position using my reference tracks. I quickly realised that I really did not like what I was hearing. Or, at least, that I could not understand what I was hearing. I rarely sit in front of a huge console like that, and I am not used to the low-mid reflection you get from bridge speakers, I feel it makes things really muddy when they are not, and forces me to track or mix bottom-light. I also could not deal with being that far away from the speakers, and the Tannoys felt incredibly bright and scooped to my ears.
Replacing the Tannoys with a pair of ATCs that were lying there did not help much for familiarity, and I quickly arranged for Troy Manning, the local Focal rep, to drop a pair of Focal Solo6s. I tried putting them in many places on the Neve bridge, but in the end we scoured a couple of mic stands and I set a small side-station up to the left of the console, where the computer cart already was, and built a little studio within the studio. Now I could understand what I was hearing. And the coffee was great, by the way.
Of course the first thing that David, whom I had never met or spoken to before, said to me when entering the room — even before he said hello, but right after he recommended that I should really do something about my hair — was “Jackson’s studio ain’t good enough for ya, gotta build your own?” I knew right then and there we were going to have a lot of fun.
Michael and I work pretty fast, so I took time to set up separate ‘tracking stations’: two in the live room for guitars, and one in the vocal booth for David and Michael’s vocals. For the guitars, I set up two identical stations, facing each other, in case they chose to track a song together (they did). On both stations I set up one Lauten Atlantis and a bunch of other mics I chose from the resident RCA and AEA ribbon collection. All mics came up on discrete channels of the 8068 in the control room so I could blend the different mics to my liking and record the blend to one track in Pro Tools. This system allowed me rapidly to craft the sound of the guitars depending on the instrument, the key, the vibe and planned instrumentation of each song. In some cases, I recorded both the Atlantis and whatever ribbon I liked, to create more of a stereo spread or to play with the blend in post. In all cases I also recorded a DI (when we did not forget to plug it back after impromptu control-room arranging sessions).
Michael had plans to add a few colours to the record to provide variety and keep everything interesting. We were graced by visits from Corey Henry and Bill Laurence, from Michael’s band Snarky Puppy, who respectively laid B3 organ and piano on two tracks each. I used house mics on those, so as to not disrupt any of the existing setups and to keep the sound of the record as unified as possible, since those guys came in at random times in the middle of the tracking process.
Electric and upright basses were tracked quickly, one-take style, by Michael. It went so fast that I am not 100 percent sure what gear we used. I’m pretty sure the electric bass went through the Apollo Twin rig into an Ampeg amp that was lying there. For the upright, I can tell from pictures that I used my usual two-microphone system, with a large-diaphragm cardioid condenser in front of the beast and a small-diaphragm omni shoved in the bridge. In this case I used one of their ELA M251s and an AKG C451, but I vaguely remember they only had a cardioid head. I also remember that, for fun, I added a Coles ribbon next to the 251. I ended up using a blend of all three during the mix.
In the end, we managed to track all nine songs from scratch in 10 days, editing and comping as we went. So, by the end of the second week we were 100 percent ready to mix. I had seriously lobbied for us to mix our record in my room at Flux Studios in NYC instead because, after tracking kind of blind, I really needed to be in my environment with my monitoring system. David and Michael were gracious enough to agree to it, and David decided to use the rest of his month of studio time at Groovemasters to make another album with his son James Raymond.
We reconvened in NYC for the mix a couple of weeks later. My only premeditated decision about that session was that I wanted to use my outboard analogue reverbs for a little vintage taste. I used my AKG BX10 and BX20 springs, as well as my EMT 140 stereo plate. Half way through the first mix, I felt the need for a chamber tone for some songs so I added my Bricasti to the mix. (Not analogue, but very unique and fun to use). All those were set up as inserts in Pro Tools so in case they died (they do die) I could quickly replace them with plug-ins without having to re-route anything or lose time.
The rest was going to be 90 percent plug-ins in Pro Tools, using my usual Dangerous Music analogue summing chain: two Dangerous Convert 8s into a Dangerous 2Bus+, feeding a Dangerous Liaison and Universal Audio 2192 converter. Some songs had the Crane Song HEDD A-D instead because the 2192 had too much colour.
My recall notes say that the only processing on the stereo mix was a couple clicks of transformer emulation and a couple clicks of para-limiter on the Dangerous 2Bus+, and then my Neumann PEa stereo EQ, engaged but flat (so applying more transformer smoothing but no EQ) and my Manley Pultec with +1.5 clicks at 20Hz and +2.5 clicks at 12kHz. The Neumann and the Manley are hardwired to my Dangerous Liaison inserts 2 and 6. The stereo bus chain was the same for every song on the record, like a fixed console setup would provide.
On ‘The Things We Do For Love’ the challenge was to create a viable stereo image with what was one of the truly one-guitar basics. Once the tones came together, David’s arpeggio guitar in the middle felt puny, for lack of a better word, against the super-lush 12-strings, enormous background vocals, trippy electrics and deep upright. I found myself stuck into the situation where it was either too loud or too soft. It did not fit. So I set out to give it width and girth using a combo of reprehensible tricks like a stereo Haas delay, gated sends in long reverbs, parallel compression and other things like that. It would take a whole article just to describe the process of this, but I go into it in the companion video to this article if you are into guitar stereo-ising geekery: www.puremix.net/fab-mixing-david-crosby.
Once I got the centre of the mix feeling record-like, the rest came together quickly. The 12-strings were double-tracked and got panned full stereo. Easy. The electric sat nicely a bit to the left, with slightly too much reverb, for sauce. The background vocals were just submixed to logically musical groups and all high-passed quite high, with not much life left below 400Hz. Some of them needed extra sheen when blended with the rest and I used the 16kHz band of the UAD Pultec Pro for that. No compression, no tuning, no phase tricks, just the singers and a bit of the same reverbs as the rest of the track.
As always when mixing, things that sounded perfect two hours ago feel like they need love right now. Everything is relative in mixing, like in love. So once all instrumental tracks got cleaned up and the centre beefed up, the lead vocal felt like it needed to be opened up. It felt dull. My go-to plug-in for this is the UAD Pultec Pro, using the 5kHz and the 10 or 12 kHz bands. Of course, once you start pushing the shine, you get into balance problems in the rest of the spectrum. Then you need to de-ess, and then stuff sticks out, so you need to compress a little... and then you’re screwed.
My aim was to stay true to the tone that David fell in love with during the tracking sessions, so I kept the processing to a minimum, constantly checking the bypassed sound to make sure the rest of the track was not skewing the vocal in the wrong direction. After I got the top I wanted with the Pultec, I beefed up the very low to rebalance the tone, and then I high-passed the whole track with a Sonnox Oxford EQ (thus achieving kind of a modern, more controlled, version of cutting and boosting the same frequency on a Pultec) and I set up an Oxford Compressor for the couple of peaks that resulted from boosting the high end. I also added a short Haas-style delay to push the sound slightly backward and lose a tiny bit of the presence and crudeness of the fairly close miked vocal (this one was tracked in the booth, not in the live room).I ended up dulling the guitars a bit to allow for the vocal to stay closer to the original. In the end, nothing sounds absolutely perfect in solo but it all sounds pretty glorious ‘in the sauce’, and the world only gets to hear the sauce. But not you! You’re special. I took the time to print both raw and processed stems so you can be ahead of the rest of the world and hear pre- and post-mix tones on every track of ‘The Things We Do For Love’. They are available here for free: http://puremix.net/crosbystems.
This song was the second one we mixed during the week-long session at Flux. It came together quickly. I was done in maybe three hours. Everyone signed off on it but I did recall it a couple times after more of the record came together because I wanted to unify the overall vocal placement and amount of reverb (and bottom) from song to song. Greg Calbi mastered the record at Sterling Sound a few weeks later. I attended the mastering session and I was pleased to hear about the same stuff there that I heard in my room at Flux — always a good omen. He still managed to coax a bit more openness and width out of the mixes, and also did a great job at bringing all the songs together into what felt more like a connected piece of work.
You never know what song is going to be picked as a single when you mix an album. On this one I was pretty convinced that ‘The City’ would be picked but I was hoping for ‘The Us Below’ or ‘The Things We Do For Love’ instead, because I felt that they showcased the purest and rawest quality of tone, sound stage and performance, as well as a good balance of classic and modern. So I was very pleased to hear the team’s choice in that matter. It was wonderful to be part of a project at the level of a David Crosby record where everything mattered equally: songs, arrangements, tones and sound stage, and where both David and Michael were willing to be a little more patient than usual with setup and geeky trickery to allow for a more immersive-sounding end result. All in all, the whole album came together in 15 days, start to finish, which still has us all baffled to this day. Listening back to it months later, while there are little details I’d love to be able to address (there always are), there is nothing I feel I would do differently. And that’s a good feeling.
Fab Dupont is an independent producer and content creator for pureMix.net.
Knowing David Crosby’s propensity for stacking vocals, I decided to anticipate and be ready for it. There is nothing I hate more than getting a track to mix with heavily stacked vocals that have been recorded by the same guy, the same day, on the same mic from the same spot, with the same preamp. It’s hell. Every problem gets magnified by being layered so many times. So I needed a solution for that problem before it happened, and I needed a signature sound for David’s solo vocal too: very present, intimate, super-clean and smooth.
Although Groovemasters had a dizzying array of choices for microphones, including David’s reportedly favourite U47 in the whole world and more 251s than I had seen in one place myself, I chose to use my Eden. I knew from working on the design of it with Brian at Lauten Audio that it would be great for David. Of course it would have to be David’s choice, so I also put his favorite U47 up. I put the two mics next to each other, and I set them up so that I could run them through identical preamps and easily insert my Dangerous Compressor and BAX EQ on either of them instantly. Consoles are good for this stuff. I had the Dangerous Compressor running in dual-mono mode, with the left side set as a limiter feeding the right side set as a compressor. And then I used the BAX and its 18kHz band for shine and the filters for a clean top and bottom. It worked like a charm.
Halfway through the third or fourth day, on the louder songs, I started to develop a distaste for the tone of the vocal booth that the vocal rig was in. I anticipated some grief at the mixing stage trying to remove the low-mid bump and some of the reflections that came out in higher keys. I had made the mistake of gauging the vocal sound on the quiet, super-intimate songs and had not checked a louder one before I settled for it (won’t do that again). So, to the team’s great dismay and much teasing, I relocated the whole rig to David’s side of the live room. I liked the new location so much better that I wanted to recut whatever vocals we had already done. I was instantly declared certifiably insane by everyone and we just cut the subsequent vocals in the new location. So, if you pay attention you’ll hear two different lead vocal colours throughout the record. See if you can figure out which is which. In the end, David was kind enough to declare that he thought the vocal sound was the best he ever had in his career. Joy and happiness.
For background vocals, I organised a system where each singer would sing each layer into the Eden from up close, then the 47 from up close, then the Eden from far away and last 47 from far away. It was a bit tedious to keep track of and I got plenty of rough handling from David about it. Until he heard it. It sounded just like the old CSN but super-present and without muddy bunching up: no peaks, a super-smooth blend and and clean ambient vibe for days. Of course the sessions quickly started to look like a very tall house of cards, but with just two singers it was the fastest and most beautiful way I found to do it.