How to: Comes to Be
Updated: Oct 4, 2020
Today's extended blog post is about how the latest track I've posted, Comes to Be, was assembled from scratch during quarantine. Originally, I had toyed with the idea of speaking to some of the concept and inspiration behind the song. Some of that subject matter is still a little fresh and painful, though, so I'm going forward with a more technical breakdown of how this track came together. This post details how the song was assembled using the talents of multiple people in multiple home studios, with zero facetime, and a LOT of file-sharing.
Back in the late-March stage of early quarantine, I started doing a number of livestreams from my studio that were a combination of actual songs, and free-form improv segments over themes that I commonly pull from when performing instrumental gigs. One of those themes, which you can check out here in a very "early" clip, started out as a simple four-bar motif that I'd riff over using the TC Flashback X4 loop pedal, and by early April had started generating a consistent melody and some recurring lyrical ideas. I have a ridiculous number of these kinds of snippets that haven't yet become an actual musical work, but this one kind of took off and it felt like it needed immediate attention.
There are a number of crossover skillsets that come in handy when putting together a song, or an EP, or an album - one of which I've found incredibly important and have tried to grow over time is project management. A fully produced album is thousands of hours of work shared by multiple people, and that's speaking just about the final audio product. So, in the context of trying to manage an EP I've been working on for nearly a year now, it's not necessarily best practice to write a brand-new song and see it through to completion. In this case, though, perhaps out of a need to distract myself with something different, I found it cathartic to put aside the other five tracks for the space of a few months and shift focus.
As with most of my music, the acoustic track was laid down first. I'll usually build a click-track type beat in Boom where I can dial in a tempo, some drum velocities, and find a swing percentage that establish the groove. Once that's done, my workflow involves tracking a primary acoustic guitar, secondary acoustic guitar (either doubling the primary, changing chord voicings, or establishing a melody/hook), bass, and then whatever else feels appropriate to the tune.
I've had great success with getting very open and expressive acoustic sounds using just the Neumann TLM170R in cardioid and some creative placement of my two ASC TubeTraps, grabbing a pure mono take of the performance. For this, I used the Neve 33115 preamp with a little bit of EQ for flavor, into the Neve 2264ALB compressor, and then into the Pultec EQP500X. The 2264 was taking perhaps 2-3dB off at most, likely on slow attack, to mellow out the more aggressive and longer-lasting transients, and the Pultec likely had a small presence boost dialed in via the high shelf set at 5kHz. Lately, I've just enjoyed running signal through that unit purely for the API op-amp on the input stage and the depth it brings to any signal. Bass would have been the same signal chain, with a Rupert Neve RNDI up front of the 33115 which would have been set to DI with a 5dB boost going into the 2264 for extra presence and drive.
Following this, I believe I did the first of what would be many electric guitar passes and re-tracking sessions. To condense a long story, I fought with the electric part's melody, chord voicings, and even the instrument itself up until the week before I had a final mix done. Part of that was trying to re-integrate what I'd recorded earlier with the updated character of the song once keys, horns and vocals were all in place, and the other part was that the guitar I was using (in this case, my Fender David Gilmour signature Stratocaster) had fallen woefully out of shape and needed a set-up badly. I finally brought it in to Russ Blake at Portland's 12th Fret guitar shop in early August for his wizardy, but listening back to those tracks recorded in April revealed tuning issues that wouldn't fly for the final mix. So, that work was scrapped and re-done.
Re-tracking the chorus electric in August
In May, I reached out to Josh Baruch about drums and sent him a skeleton track with acoustic, electric, bass, some keys, possibly an early MIDI trumpet, and the original Boom-generated groove map. Within a week, he'd sent me a stereo bounce of a fantastic pass for the track, recorded at his home drum studio, which I dropped into the master Pro Tools session and began to edit and track against. Josh and I have found that it is super helpful to work together using a stereo bounce of the drums before we get into swapping gigabyte-sized file loads containing all of the drum tracks, since we'll sometimes go through a few iterations and re-tracks as I adjust elements on my end and he comes up with exciting fills and patterns to integrate into the performance. In this case, I think the original stereo drum take was incredibly close, if not exact to the performance that was used in the final mix.
The song as it existed in mid-April with a number of unedited rough passes and Boom
At the same time I sent Josh the song, I'd also fired it off to Jeff Simpson for tracking horn parts at his Additional Biscuit studio, and had also split my Pro Tools session into three separate sessions; the master session, in which I was starting to try out mix ideas involving some processing-heavy plugins in a mix sense that made overdubbing without latency rather difficult (even on a TDM platform!), a vocal session in which I could work on the dense vocal stacks and effects programming, and an overdub session for trying out ideas and where I would sculpt the synth-heavy intro and BT-inspired bass synth elements. A fourth session would eventually be added when I got Josh's full drum file, as I typically use a standalone session to micro-adjust some elements if desired, play with sub-millisecond alignment and phase coherency, and sometimes fly tracks to outboard gear for processing before bringing the drums into the master session. In this case, I did end up sending the drums through a Handsome Audio ZULU, which is an incredible passive tape machine emulator I am borrowing from Colorfield Studios.
For vocals, 100% of the tracking was done using the TLM170 into the Neve 33115 -> 2264 -> EQP500X chain. I've since added a Klark-Teknik KT-2A unit that I am so far extremely pleased with, being a longtime user of the UA-reissue LA-2A hardware unit, but these vocals were done prior to that addition. Various settings were used for different sections and layers of the song, which I think is audible in the final product - the close-up, almost breaking-up sound of the verse was achieved by running the 33115 with considerable preamp boost into the 2264, giving me instant harmonic distortion if I leaned into the microphone at all. The harmonies, choruses, and effects layers were all done with less preamp gain, higher compression ratios, and some adjustments in the gain staging in order to achieve a strong and clear pop-chorus vocal tone.
Vocal editing, comping and effects were all done within the standalone vocal session. I'll be honest with you - being vocally out-of-practice, and not necessarily a star-level vocalist to begin with, means I did tune up the performances at points. I am not a fan of using a single auto-tune product slapped across the entire track; what I typically do is, once a performance is fully comped, I will create tuning lanes each with Auto-Tune, Melodyne, and native Pro Tools tuned vocals. Each of these lanes has been manually adjusted and meticulously dialed in - and from there, I will play the song back repeatedly and if anything in my original vocal stands out in terms of poor tuning, I will find the edited vocal that sounds most natural and drop it in. This is a fantastically tedious and time-consuming process, but the end result is fairly natural (to my ears). There are some sections of this song, specifically the first words of each stanza in the verse, where I leaned heavily into tuning the primary and background vocals for effect, but otherwise the track follows this manually-tuned approach.
Vocal effects were done using two Altiverb reverb impulses (one Cello Studio echo chamber set to a gated half-note, and one Lexicon 224 long plate verb with a little shimmer), and four Echoboy delay instances that created a super-delay engine. This is a trick I learned from sitting in on Devin Townsend's mix sessions - each delay is set with either different character, panning, or timing, on its own bus, and at 100% wet. I'll then open up four bus sends on my primary vocal tracks, harmony tracks, background vocal bus, and automate those bus sends throughout the song to either lean into, say, a 30ips tape delay during an entire verse segment, or throw a single part of a word into a quarter-note left/right stereo delay for emphasis. This latter automation is heavily influenced by a great conversation I had with singer Christian Burns (BBMak, All Hail The Silence), whose productions are absolutely beautiful, and well worth checking out.
Vocals were finished around mid-July, at which point I took about a month off from working on the track - this was due to a personal tragedy that unfortunately is very close to the subject of the song, and is a little too near to the heart still to dig into here. Perhaps in another blog post, we'll go there.
In August, Jeff sent over an incredible horn section built using his trumpet, flugelhorn, and a number of bespoke trickeries that you'll have to talk to him about for the details. I quite literally dropped his multitracks into the master session, brought the faders up to a decent level, and that was that. There was some later automation and effects that I added for blending and balance, and I think I slapped an LA-2A plugin on his solo for a bit of pop, but that was it.
The latter two weeks of August were spent in final mix mode - at this point, the master session had grown to somewhere in the area of 60 to 70 tracks, many of them stereo stems and bounced submixes of elements. My mix session for this consisted of those tracks (each with various plugins still in place, if they hadn't been fixed by bouncing down already), three additional effects auxes containing a reverb and two delays which I use in very small doses to affect the overall feel and environment of a mix, four stereo output busses feeding my SSL for analog summing, and a stereo return to print the result from the SSL. My current output bus scheme, with a little fudging depending on how creative I'm getting during the mix, are currently set up as this: Bus 1 is drums/bass/LFE synth, Bus 2 is guitars/keys/synths, Bus 3 is vocals and other "airy" elements that I want to be super present, and Bus 4 is effects and "etc". I find Bus 4 to be an incredibly useful tool during mixdown as I can fly elements over to that bus that I want to be somewhat external to the "glue" of the other established busses and have them behave exactly as I need to within the overall mix. A great example of this in Comes to Be are the two 12-string guitars (tracked on my Taylor 754CE-L30, again with the TLM170R) that hit during the choruses - I wanted those to sit outside of the core of the chorus and to ring independent of the compression I was using on the electric and acoustic tracks, so Bus 4 was utilized to send those into the air above everything else.
As a side note, the SSL SiX really does a fantastic job of analog summing, easily on the level of the XDesk I used at 12 Media Music, and even when I'm output limited to four busses. If I stumble across another output card for my HD192, the SSL will take another two stereo busses - in which case, I suspect I'll break out the guitar/synth buss into two groupings, and make the effects bus its own dedicated stereo group.
For this final mix, I really only used four outboard processing tools - Bus 2 (guitars/synths) went through the Neve 2264ALBs in stereo linked mode, slow attack, 1.5:1 ratio, 400ms release. At most, I'm taking 3dB off here, with the intent of creating a pretty strong glue for those elements within that bus. On the insert of the SSL main bus, I had the Pultec EQP500X doing duty with some very light presence boosting, and the Handsome Audio ZULU followed that on its highest-fidelity settings to emulate an Ampex deck for mixdown. I then had the SiX's G-series compressor sitting on top of the result, taking 2dB off at the very most, to add a little final glue to the overall package and tame a few transient spots during the very last chorus.
As is probably the case with any detail-obsessed artist mixing their own music, I went through a dozen iterations using this mix process - I quite literally left the HD rig up and running, trying to keep my cat from stepping on the SSL and moving faders, for a week while going through tiny variations in staging and mix adjustment. At the same time, I created the shell of a mastering session in Pro Tools and, using the SSL's quick-switch EXT routing, was auditioning what the mixes would sound like in a mastered sense. That master session ended up being what I used for the "proper" in-the-box master - two Waves linear phase EQs to do some very surgical, no-character adjustment to the mix, the Softube Passive EQ in mid-side for additional shaping, the Waves API 2500 using the Lavyne modern master preset as a starting point, Soundtoys Decapitator (which I ended up removing) for some tiny harmonic excitement, and the Waves L3 maximizer to finish. I have almost exclusively been using the Massey L2007 limiter for final gain during masters, but found the L3's multiband versatility to be necessary for this track. Oh, and of course - the final master was flown out to iZotope RX7 for analysis, any cleanup, and making sure I'm at least hitting around -14dB LUFS averages for streaming delivery.
There you have it - as promised, a technical breakdown of hundreds of hours of work over the course of five months to produce a track that will be on an EP that was already coming together before this track existed. If you haven't heard it, you'll find the YouTube below, and if you like it, please consider grabbing the 24-bit master file from my Bandcamp page.