Inside Track: Justin Bieber 'Purpose'

Most of Justin Bieber’s vocals on Purpose were recorded in Studio 3 at the Record Plant in LA.

In a transition overseen by long-term collaborator Josh Gudwin, teenage prodigy Justin Bieber has become a grown-up pop star.

Since breaking through to an international audience in 2009 at the tender age of 15, Justin Bieber has at times demonstrated how not to grow up in public. Bieber’s fourth studio album, released last November, was thus a make-or-break proposition. Three years after his previous studio album, Purpose was to be his first fully fledged adult project, and likely career-defining.

Unsurprisingly, Bieber’s manager, Scooter Braun, and record company took no chances, and explored all possible avenues to make sure the new album would be a success. Writing camps were organised, and Bieber spent time working with household names like Rick Rubin and Kanye West. In the end, however, Bieber reduced his team to the bare essentials: just two guys with whom he felt entirely at ease.

On the writing side, this meant a central role for Jason ‘Poo Bear’ Boyd, while Josh Gudwin held everything together on the studio and production side. This involved logistics, taking care of album A&R, engineering all the overdubs to the writers’ backing tracks, including Bieber’s vocals, mixing most of the songs, and in general being responsible for delivering the album. It earned him an unusual ‘album producer’ credit.

Solid Support

Josh Gudwin has been Bieber’s sidekick for half a decade. The engineer and mixer has enjoyed a fairly meteoric career, studying at Florida’s Full Sail University during 2005-2006, interning at Track Records Studio in LA, working with songwriter Esther Dean, and most of all studying for two years with America’s number one vocal producer, Kuk Harrell. In the process Gudwin clocked up an array of big-name credits, including TI, Lionel Richie, T-Pain, Quincy Jones, Rihanna, Jennifer Lopez, Carly Rae Jepsen, Celine Dion and Pharrell Williams. However, his first meeting with Bieber in 2010 proved the most significant of his career to date, and Gudwin has since worked with Bieber on studio, remix and compilation albums like Under The Mistletoe, Never Say Never: The Remixes, Believe, Believe Acoustic, Journals, and Live At Madison Square Garden.

From his home studio in Los Angeles, Gudwin recalls the earliest beginnings of what was to become Bieber’s first adult album. “The sessions with Rick were Scooter’s idea, and while he and Justin were working I was literally just sitting on Rick’s couch and watching. Kanye and Justin hook up from time to time and develop big creative plans, but they both end up doing their own things, so nothing ever comes of them. We also did not use any material that come out of the writing camps. It was just a matter of seeing what was out there. In the end, Justin didn’t want outside people around who he doesn’t know that well, so I was fortunate to be in a position to really support his vision and help him create a solid body of work.

“Poo Bear also was central in the making of the album, and Scooter always gave opinions and suggestions. We had discussions along the lines of helping Justin to find his feet and direction, but a lot of what we did was just writing and recording new songs, and seeing what came out. There were some songs, mostly written by Poo Bear, that rose to the top, and we realised we had to build around them. It kind of naturally progressed. In the end it took Justin about two years to get clear on where he wanted to go. Poo Bear and I knew what Justin liked, and we brought in many other songwriters, and when Skrillex came in it went one step up again. But Justin was directly involved with the entire process of making the album and he has full accountability on this one.”

Producer, engineer and mixer Josh Gudwin has been Justin Bieber’s right-hand man for many years.

Purposeful

Gudwin, Poo Bear and Bieber engaged a proliferation of songwriters and producers in the making of Purpose that is impressive even by the sprawling standards of 21st Century pop/R&B albums. In addition to Skrillex there were dozens of others, amongst them Diplo, Ed Sheeran, Benny Blanco, Soundz, Mike Dean, Michael ‘Blood Pop’ Tucker and Ian Kirkpatric. Normally this would be indicative of a scattered and disjointed album, but in fact, Purpose hangs together pretty well, which probably contributed to the album’s remarkable commercial success. After Adele’s 25, Purpose is probably the most successful album of the Winter of 2015-2016, and at the time of writing it had already spawned four major hit singles, including three UK and US number one hits — ‘What Do You Mean?’, ‘Sorry’ and ‘Love Yourself’ — with ‘Where Are Now’ not far behind. At times these songs were battling for the highest chart positions amongst themselves.

“We began work on the album in 2013 and for the last year we worked full-time on it,” says Gudwin. “I kept an overview of the entire project, and managed all the files, and the contacts and exchanges with the writers and producers. We wrote and recorded in various studios, in LA, Atlanta, New York, Toronto and even as far away as Greece, but mostly we recorded at the Record Plant in LA. The SSL 3 room is one of the best sounding rooms in LA. I love the monitoring system. I bring my own main monitors, the ATC SCM25As, which are also my favourite for mixing, as well as my NS10s and little Bose Freestyles.

“Justin loves that studio, and regards it as home away from home. Sessions can take place anywhere, but when you remain in once place it’s easier to maintain consistency and make sure things always sound the same. I did almost all Justin’s vocal recordings at the Record Plant, apart from for a couple of songs, when I was not available. It’s a pretty easy process, because Justin has been doing this for so long that he has a very good knowledge of how to sing in the studio and the techniques that are used. Sometimes we will do a few takes and I will comp that later, sometimes I record him line by line. Basically my job is to capture him in the right way, and then I design his vocals and place them in the track.

“I use various microphones to record him, like the Sony C800G, Telefunken ELAM 251, Neumann U47 or U67. It depends on the sound we want. It’s pretty intuitive. We simply pick a mic, he starts singing and if it isn’t right, we try another one. We like to keep it moving. The mic pre usually is a Neve, and I like to use the Tube-Tech CL1B compressor on the way in, because it’s smooth and gives me the control I’m looking for. After that everything is in the box, apart from some hardware effects I like to use right at the end of the mix, for which I go through a few specific channels on the desk. For the rest my sessions are entirely inside my laptop, so I can travel and set up anywhere. I have an expansion chassis with an HDX card and two UAD cards, and can easily plug in to any studio.”

Speedy Sound

The current direction in R&B/pop sees artists and production teams using a lot of sound design to chisel out a distinctive sonic identity. Gudwin explains how he went about creating a distinct sound world for Bieber’s album, including the singer’s vocals. “Often a major part of the sound design already comes from the producer. Skrillex is really good at creating his own sounds, and I then create vocal sound designs to go with that. But with other tracks I often go in heavy to create additional sounds and effects. Once I get the session from the producers, it’s pretty much just me.

“Every song is different, and therefore every vocal requires different ’verbs and delays. Justin gives me full creative freedom when it comes to his vocal sound, but if he doesn’t agree, he’ll let me know. I edit the vocals as I go, and I groove them, in real time. When Justin walks out of the vocal booth, I like things to sound as good as possible as quickly as possible, also because he wants to hear it back like that almost immediately. I then continue adding sounds and effects, and get into doing a rough mix, dialling in the finer details.

“I learned about the importance of speed from Kuk. When I worked with him he’d have the vocals recorded and edited and loaded into the track, and I would come in and it used to take me two hours to do a rough mix. Kuk said, ‘Just take 30 minutes, bro.’ So I learned to do a rough mix very quickly, and at the end of the day, that often is the sound of the record. The rough mix is often more important than the final mix, because it sets the direction for the song, and if the artist likes that, you don’t want to lose that by redoing sounds and so on. So my final mixes almost always are refinements of the rough.”

Getting The Rough

‘Sorry’, the second single from Purpose, a US and UK number one, is a good example of the way the album came into being and of Gudwin’s mix approach. “Originally Skrillex was going to do his own mixes,” says Gudwin, “with me just doing the vocal mixes, sending him the a cappellas. But he was on tour in India and did not have time to do anything. So he sent me all his sessions in Ableton, and I loaded them into Pro Tools and I started to mix off that. I first did a rough mix, which in this case involved me digging in for 30 minutes to an hour. I’m not paying a lot of attention to the technical aspects at this stage, I’m just putting on plug-ins quickly to get them to do what I want them to do. I do a lot of EQ’ing, but carving out frequencies, and only rarely boosting.

“I’ll start at any place in the track where I feel inspired to start. Basically I’ll go in and start scrubbing, like washing a car. I go all the way down the session, and I’ll dig into the vocal design, and will add EQ and then effects, and I’ll do drops and filters and whatever is needed to create space and movement. Once I can listen to the session without anything bothering me, that is the rough mix. I’ll put it in a folder and I’ll send it to Justin, and he’ll listen to it a thousand times, and sometimes the rough becomes the final mix. If we’re going with the rough, I’ll just do some final detail at the very end. For example, with ‘What Do You Mean?’ we literally went with my rough mix. The beat and song are very straightforward, there is nothing crazy, and the rough had the feel, the vibe, and that was it.”

The mix of ‘Sorry’ was partly done at Henson Studio D, as were various other sessions for the album. This photo from one of the Henson sessions shows the room’s SSL G+ console with, left to right, engineer Peter Mack, Josh Gudwin, producer Maejor Ali and Justin Bieber.

By contrast, ‘Sorry’ was a complicated session and the final mix took a lot of time; so much so, in fact, that Gudwin got in some extra help. “I did the final mixes for the album mainly at the Record Plant,” recalls Gudwin,” but also at Henson Studio D, and even did some mixes at Jungle City in New York. The workload got so heavy towards the end that I asked my buddy Andrew Wuepper to help me out for the last two weeks. Even together we were doing 16- to 20-hour days! We’d go to and fro, with him mixing a song for a few hours and then I would go in for a bit, and then he’d mix some more. Doing these mixes together was a treat, because normally our job is so self-obsessed. We started the final stage of the mix for ‘Sorry’ at Henson and worked on it through the night, and then took a 6am flight to New York and went straight to Jungle City and finished the mix that same day/night. Julia Michaels came to the studio to listen and she added some quiet background vocals in the hook. Once I got her parts in, the mix was done. The song was mastered the next day.”

Andrew Wuepper adds: “I have known Josh for many years, so it was natural for him to call me one day, saying he was totally overwhelmed with the Bieber project, and would I come in and help him out on a couple of mixes. He was having all the responsibilities for delivering the album, and wasn’t only mixing, but also still finishing production on some songs, with parts still being recorded and replaced. He had to deliver 17 mixes in two weeks! In the end he and I finished the mixes of seven songs during those two weeks. I also went with him to New York, because we still had to start the mix of the song ‘Children’ — we were waiting for some files from Skrillex — and we needed to finalise the mix of ‘Sorry’. They also wanted Josh to be there for mastering, and on top we had literally gone for several days without sleep by that stage, so he needed a second pair of ears to fall back on.”

Sorry State

‘Sorry’ is a 100bpm EDM pop track, with Jamaican influences in the rhythm, and featuring a characteristic five-note horn hook. The music and vocal sounds are characteristic of the album: light, open, smooth, airy, breathy. The Pro Tools mix session consists of about 125 tracks: not unusual by today’s standards, but still considerable.

This composite Pro Tools Edit Window screen capture shows the entire mix session for ‘Sorry’.At the top of the session is a master and a mix print, and below that are three group tracks in red for all drums, all bass, and all keyboards, followed by nine drums and percussion tracks, five bass tracks, a mish-mash of 16 tracks that include horn tracks, some aux tracks and outboard effects, more percussion, a piano track and a string track. Clear order is restored again with Bieber’s lead vocal tracks (blue) and associated delay tracks (green), live room and outboard prints, tons of backing vocal tracks (11 by Bieber, 30 by singer Trevon Trapper, all in light blue, and four by co-writer Julia Michaels), and finally, in green, a group of aux effect tracks at the bottom.

“This is an actual working mix session, and yes, it is big,” agreed Gudwin. “I’m not so focused on organisation because when you’re on deadline, you don’t always have time to label the session from a visual standpoint. The stem sessions that I create from my mix sessions are super-clean, though. I do usually have the vocal delays as separate tracks, on which I pull the audio I want to treat, and at the bottom of the session I normally have a template of aux effect tracks, in green, with a hall reverb, a plate reverb, and a long reverb; three delays — half-note, quarter-note and eighth-note — and two ping-pong delays and a UAD Roland Dimension D. They are my utility delays and utility reverbs, and they’ll be modified for each session.”

“When I first came in,” recalls Wuepper, “and we were working on the first couple of songs, I showed Josh the template of how I like to work. He isn’t really picky about how the sessions are organised, but I am, so I asked him if I could move certain things. He agreed and so we got used to working in that format. I always have sub buses at the top of my session, in this case ‘All Drums’, ‘All Bass’ and ‘All Keys’, and where appropriate I will also often have ‘All Guitars’, ‘All Lead Vocals’, ‘All Backing Vocals’ and ‘All FX’. Part of the logic behind that is that master faders have a deeper bit resolution than auxiliary tracks, so if you are attaching a master fader to control the input of those aux tracks, in theory you can push a little more gain into those aux tracks.

“I tend to put instruments together that musically sound together, so in this session the horn sample, to which I added a sub-bass, is next to the 808 track, which is the track called ‘bs2-1’ at the bottom of the bass tracks. In my mind, I treated the horn and the 808 as one thing. I am thinking of it in feelings. Whatever group of instruments collectively adds a certain feeling to a song, I’ll treat as one entity, rather than treat every piece as a different thing. We wanted the low end to be big throughout this song, and the verses had the kick drum, and in the B section the 808 comes in, but it didn’t have enough warmth or low end, so we got that from the horns.

“When Josh and I started work on this song it was already pretty close. The production was stellar, Justin’s vocal sounded astounding, and the rough mix was really good. Many people were already in love with it and the record company had picked it as the next single, so we were not going to change the direction. Basically our approach was to get it ready for commercial release, which involved making sure the bass is tight, the vocal pops out, that the general feel is there, and that the ideas that are already there are enhanced as much as possible. Most of my contribution consisted of detailing the reverbs, adding processing so that the bass sounded bigger, and making sure the compression treatments on the vocals and drums were smooth. It really was a matter of bringing the track even more to life with a last sheen of polish.”

Vocal Focus

This trio of plug-ins makes up Josh Gudwin’s default vocal processing chain: the UA SSL E-Channel and LA-2A compressor, and Waves C6 multiband compressor.“Most effects and treatments in this session are on Justin’s vocals,” explains Gudwin. “The other tracks didn’t need that much. Skrillex’s and Blood’s drums and percussion tracks sounded pretty good when they came in. I use a lot of UAD SSL E-channel plugins, and I have that on the two kick tracks at the top. I can get a great overall sound using just the compressor and EQ of that plug-in. The first snare also has a FabFilter Pro-Q 2, which is a great EQ, and the second the Xfer Records LFO Tool and the SoundToys Filter Freak, both to get the snare pumping.

“There are only three keyboard bass tracks active, and they’re all the same part, which I pulled onto different tracks so I could affect each differently. I prefer that to automating one track. The first bass track has the Brainworx EQ, doing some M-S processing, and the Little Labs Voice Of God, adding some sub. Another track has the Waves Renaissance Compressor, not doing much, and the iZotope Ozone, which is widening and EQ’ing, and the third bass track has the UAD LA2, keeping things in place.”

Wuepper: “Josh asked me if I could make the bass sound bigger and fatter, with some more sub as well. So I inserted UAD’s Voice Of God, which really is a subharmonic synth, and I used an outboard Fairchild 670 to control the sub, to make sure that lowest octave would not blow up your speakers! The 670 was one of the few pieces of outboard that we used, the other one being a Bricasti M7, which supplied reverb to every mix we did for the album. At the beginning of ‘Sorry’ you can hear these big percussion hits with a big reverb tail, which came from a ‘London Plate’ reverb on the Bricasti. I cranked up the reverb as much as possible, going too far with it, and then backed it off a bit. Another piece of outboard that we used on the entire album, particularly on big synth sounds, was the Dolby SR unit. Originally designed for noise reduction on tape, it acts like a high-harmonics generator, a like the Aural Exciter, and we found that it added some breathiness and air that sounded cool.”

Samples & Snippets

Gudwin picks out a few other interesting tracks, beginning with a vocal track named ‘vcls’. “That track contains a vocal sample that Blood made of Justin’s vocal. I’m just touching that with the LFO Tool for a bit of pumping, I take out some high frequencies with the Pro-Q 2, and then use the SoundToys Microshift for a slight pitch-shift/chorusing effect. ‘Sry1V’ below that is the vinyl sound in the track. ‘$JBU’ is the main lead vocal bus, on which I have the UAD 1176, Waves De-Esser, Manley Massive Passive, Metric Halo Channel Strip. The sends are to the generic aux tracks at the bottom of the sesion: verb, ping-pong delay, Dimension D. Below the lead vocal bus is the print track of a Bricasti outboard.”

Wuepper: “There are two tracks called ‘Bric’ around the main vocal bus; the one above is the print from the Bricasti effect that was used on the percussion, and the one below is the print of the Bricasti reverb on the vocal. Next are Josh’s vocal delays. Putting delays on audio snippets [ie. copying short vocal clips to new tracks and applying delays as inserts, rather than automating a send from the main vocal track] is a pretty interesting way of doing things. I’ve not seen anyone else do this, but it makes it easier for Josh to manipulate these delays and to go deeper into the effect. He can really fine tune the delay times and decay lengths and so on. Sometimes the feedback you get with plug-ins can act a little weird, and this approach allows him to have more control.”

Gudwin: “The five green tracks are all vocal throws and delays. I don’t like to automate delay throws via aux tracks. The top green track is the master track for the delays, ‘JB Throw All’, and it has a compressor and an SPL Vitalizer. The ‘1147’ delay track has the UAD Cooper Time Cube, with a quick ping-pong-y flutter delay that I use to widen, and the track called ‘1167’ has a basic eighth-note delay from the Echo Boy. The ‘A’ insert is Auto-Tune, but it’s not working on these tracks. When needed, our vocal tuning is normally done by Chris ‘Tek’ O’Ryan in Melodyne. I sometimes do it myself, if I have the time do it, in the stand-alone version.

“The blue vocal tracks below the vocal delay tracks are the main lead vocal comp tracks, and on many of them I have the UAD SSL Channel Strip, UAD LA-2A, and the Waves C6 multiband compressor, and sometimes also the Pro-Q 2 EQ. The ‘DLYP’ track has a delay pan effect, with the SoundToys Primal Tap delay and Panman auto-panner, SSL Channel Strip and the P&M Vinylizer. ‘White’ and ‘Master’ are printed reverbs recorded in two rooms at Henson. They are my main plug-in vocal reverbs, and the green tracks below are pitched with the Elastic Audio X-Form [in Pro Tools] and effected with the Waves H-Compressor for a pumping effect. I pitched the reverbs up an octave or two, and I mixed them in very low. The ‘PCM’ and ‘PC1’ tracks are prints from ping-pong delays from the Lexicon PCM42 outboard.

Formant shifting in SoundToys’ Little AltarBoy helped create an experimental ‘radio’ effect on some of Justin Bieber’s backing vocals.“Justin’s backing vocal tracks all go to the group track called ‘JBG1’, on which I have a Waves De-esser, an SSL Channel and the C6 multiband compressor, plus there are a number of delays and reverbs via the sends. Trevon’s backing vocals all go to ‘JBTR’, which has similar effects. I wanted to fill the song up a bit more, and sometimes it’s not the most enjoyable process for an artist to sing all these background parts. Plus a different vocalist will add a different texture to the song, as long as it complements the lead vocal and the record. As I mentioned, Julia added her vocals during the final mix in New York, and her group track also has the De-esser, SSl Channel and C6. Right at the bottom are some effects tracks, with the Dimension D and group delay throws, and so on.

“Back at the top of the session, the drums, bass, and keys group tracks each have the Waves L2 on them, but barely touching the compression on that, and they go to the Master track at the top, which has the UAD SSL E-Channel, Manley Massive Passive, Oxford Inflator for a touch of processing, and the FabFilter Pro-L compressor. Tom Coyne mastered the album with Randy Merrill at Sterling Sound. I gave them the option to use the mix without and without my mastering bus treatments, and sometimes they did, and sometimes they didn’t. But especially with the Skrillex songs, they were already hitting in a certain way, and many of the mixes were done through the mastering chain that we already had, which was hard to match.”