Used on "Purpose", as stated by Purpose mix engineers Josh Gudwin and Andrew Wuepper in this May 2016 Sound on Sound interview about the mixing of the title track and others.

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.

“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.

(...) 'Love Yourself'

(...) Next are Justin’s vocals, with I overdubbed in New York. Almost all the regular vocal tracks have the UAD SSL Channel, LA2 and the Waves C6 multiband compressor. These three plug-ins work great on his vocals, so I tend to stick with them. Each of the vocal tracks has slightly different settings from them. They all go to Justin’s lead vocal bus above them, which has the FabFilter Pro-DS de-esser and the Pro-2 EQ. Justin’s lead vocals also go through the Bricasti, and the effect is printed. My own four backing vocals went through a vocal bus on which I had the SSL Channel, and a compressor, but they didn’t do much, nor did the sends. These vocals are very much in the background, I just wanted to add some texture really quickly. Finally, at the top of the session everything went through a Master track, on which I had the UAD SSL channel, mainly for compression set to mid-attack and auto-release — without drums you don’t need the slow attack and quick release — a Massey EQ, boosting 100Hz and 16kHz, a FabFilter Pro-L for level, and the Sonnox Oxford Inflator to add some sheen.”

'What Do You Mean?'

(...) “There were a few more complicated vocal effects, like the ‘JG FX’ track, on which I used the Vitalizer doing some spatial expanding, a UAD Fatso to thicken it up, and I then cut some low mids with an EQ, and used a de-esser. This effect track gave more depth to the vocal and made him sound more like an angel! Underneath that are two ‘Radio’ tracks [which sound like they could have come straight from Peter Gabriel’s ’80s Fairlight experiments]. I used Auto-Tune to pitch the vocals up, and then put on a SoundToys AlterBoy, which changed the formant, and then I’m filtering 350Hz and below.

“Below the ‘radio’ tracks are all my vocal delay tracks, in green as usual, with the bus for all delay throws at the top. One delay track, ‘Splaater’, is a flutter delay with Auto-Tune and the Cooper Time Cube, and the other two have the Waves H-Delay and SoundToys Echo Boy. They’re both quarter-note delays, but with different feels. Below the delay tracks are some ad lib tracks, and all the actual lead vocals, pulled out over several tracks. Most of them have my regular trio of UAD SSL Channel, LA2 and Waves C6 plug-ins, and all vocals are sent to the ‘JB Buss’, on which I have the FabFilter DS de-esser, a Manley EQ doing light cuts at 330Hz, 560Hz and 3kHz, and then a whole bunch of sends to my regular aux tracks at the bottom: a hall reverb, a plate reverb, a light quarter-inch delay, a light ping-pong delay and a Dimension D.

“Further down are four Justin backing vocal tracks, which go to a bus above them, ‘JBG1’, on which I have the SSL Channel, boosting highs and cutting lows and doing some heavy compression, a UAD 33609 compressor, and then the Waves Enigma [phaser/flanger] on a Mutron setting, adding some sweeping sounds. The sends are once again hall and plate reverbs, a quarter-note and a ping-pong delay and a UAD Roland Dimension D. Once Justin had decided to go with my rough mix, I spent another half hour on it, doing some EQ adjustments, and that was it.”