Drums
The layered kicks, snaps and claps are thus fed to their own buses for processing, which involves EQ and various forms of saturation ranging from subtle tape and console emulation to more obvious distortion. "FabFilter Pro‑Q2 is my favourite EQ. I'll use it any time I want to pull something out or add something. I'll only use another EQ if I want the specific sound that EQ adds — Neve, SSL, API and so on. I use Slate's Virtual Mix Rack generally for console emulations, especially the Neve desk one, and sometimes the API one. I think they sound really good and they are really subtle, and they help a lot if you have many instances of the plug-in. That warms up the sounds in a nice way. It's similar with the Slate Virtual Tape Machine, which I think sounds slightly more subtle than the UAD tape emulation. SoundToys Decapitator and Devil‑Loc and FabFilter Saturn are all good for distortion. The Pro‑C2 is an amazing compressor that does exactly what you want it to do. I use the FabFilter stuff a lot, because it is so good and so straightforward to use, and it does not colour the sound."
All Kersting's audio drum tracks end up going through his 'Drums' aux, either directly or via the individual instrument aux tracks. Some of these latter tracks, and the 'Drums' aux itself, have sends to three parallel tracks: 'Parallel Drums' with the SoundToys Decapitator and Kush UBK1 compressor, 'SPL Drums' with the Native Instruments Transient Master and 'Big Parallel', with the Kush Novatron compressor and Pro‑Q2.
Kersting: "This main 'Drums' aux bus has an insert to my Drawmer 1968 hardware compressor, and there's a Decapitator adding some saturation, and yet more compression from the UAD API 2500. The Drums bus also has sends to the 'SPL Drums', adding some attack and sustain to the drums, and to the 'Big Parallel' bus, which has the Novatron set to 'Punish' mode, and a Pro‑Q2 to roll off some high end, because when you do parallel compression you sometimes end up with super-bright stuff. I added the Novatron because I felt that the drums in the chorus needed more impact. I created a blend of these four drums aux tracks to get one cohesive sound."
Konstantin Kersting: "You want to have the artist walk out with a song that they are really proud of, and I feel like you get that way more when the artist is in the room with you, rather than just presenting them with the finished track at the end of the production.
Music
The 808 which Kersting added to the pre‑choruses is part of the bass section, and is the first track of this section. The bass and guitar have a slightly different routing structure, with no parallel processing, and there are tons of plug-ins on the inserts of the Model D and Arturia Moog audio bass tracks. A send from the Model D bass goes to a 'Bass Chorus' aux with the SoundToys Microshift plug-in, and the two main guitar parts each have sends to a 'Guitar Slap' aux with the SoundToys Echoboy and a 'Guitar Reverb' aux with the FabFilter Pro‑R.
"There's a Decapitator on the 808, because you need a bit of distortion on them to make them cut through on small speakers. I check this on my Auratone 5C monitors. I even try to make sure you can hear 808s on a phone speaker. You have to make sure that your mix sounds good on all those devices, as well as on your main studio speakers, which in my case are the Quested VS2108s.
"There are no fixed rules. It is about whatever gets me to make it sounding the best. In this case that meant lots of plug-ins on some of the audio tracks, particularly the Model D bass, which is the main bass you hear in the track. I'm pulling out quite a bit of sub and high mid-range with the Pro‑Q2 and sub with the FabFilter Pro‑MB, because the bass was blowing out my sub. The Decapitator and the Dada Life Sausage Fattener add distortion in the lower mid-range, again to make sure the bass sounds good on small speakers. In addition, there's a Kush Audio Novatron compressor to keep the bass in place. Regarding the keys, many of the plug-ins on them were added during production, because I was layering several sounds to get the right textures and vibe."
Vocals
As already noted above, the vocals are divided into three sections: harmony vocals (which sound like a medieval choir), the group backing vocals, and the lead vocals, which are split up over 10 tracks. Each of these three sections has its own group aux, with tons of plug-ins. Most work clearly went into the lead vocals, as the audio tracks also have a fair amount of plug-ins, mostly Waves Tunes LT, FabFilter Pro‑C2 and Avid D3 De‑Esser. The 'Lead Voc 1' aux has nine inserts and 10 sends, plus there's a supplementary 'Lead Voc 2' aux, with five inserts and five sends, adding yet another layer of effects.
Toni Watson's lead vocal passed through numerous plug-ins: both the EQ and the multiband compressor here are controlling a strong upper-mid resonance in her voice.
"There is actually not that much happening!" insists Kersting. "All these plug-ins do a tiny bit. I split out the lead vocals over several tracks to be able to treat each slightly differently. I felt that the Waves Tune plug-in sounded best on Tones's vocals, so I used that, just to nudge a few notes here and there. Plus there's a compressor, the C2, and a de-esser on each of the audio tracks. Three tracks are marked '-8', and they have the SoundToys Little Alter Boy, which lowers them by an octave, and there's a track with the Tal Vocoder, which is the vocoder for the middle eight.
"All these vocal tracks apart from the vocoder track go to the master vocal bus, 'Lead Voc 1', and on that I have the FabFilter DS de‑esser, and then a hardware insert to a Urei LA2A, which is not doing a lot. It's quite a slow compressor, and I always have it on the limiting setting, just taking off little tiny bits. Then I have the UAD Neve 1073 emulation, set to the drive input, like you would on an actual Neve preamp. So I turn it up a little bit, and then turn the fader volume down, which I find gives a really nice sound. Then I have the Slate VMR, with a Neve console emulation, an 1176 emulation, another harmonic saturation thing, and a slight compressor.
"After that, I have the Pro‑Q2 again taking out some mid-range, adding a little bit of top end from 5k onwards, and getting rid of low end. Then it goes into the Slate VMR, which is adding harmonic saturation again, then it goes to the Pro Tools de-esser, because I added some highs with the EQ. And then it goes to the Pro‑MB multiband, which is working pretty much in the same mid-range as where I cut with the EQ. I must have wanted to tame her vocals a little bit more. It's also taking out some of the 's' sounds again, around 10k. Finally, there's a gate, to take care of any noise from the LA2A outboard compressor.
"These are the inserts on the lead vocal bus. After this are the effect sends. The first five go to a slap, a long delay, a reverb, a small plate and the vocoder track. The second round of sends go to the five orange aux tracks, which all have compression. I'm mixing into compression with the levels of the compressors set, Michael Brauer-style, and I then blend these five tracks together to create a cohesive vocal sound. The compressors are the Kush Audio UBK1, which can be super-saturated and quite squishy-sounding, the Slate emulation of an 1176, a Slate VCA compressor [wrongly labelled in the session], an Empirical Labs Arousor, and for the Vox Distortion bus, the Decapitator.
"It looks like a lot, but again, it is just a whole bunch of compressors doing small things together. I find that my vocals sit a lot better in the track since I started doing this. They are easier to mix in, and I need less EQ. I actually create my blend of the compressors first, because generally my vocal sound will be kind of dictated by the blend of those compressors."
Master Bus
Konstantin Kersting still uses hardware processing, especially on the master bus, where his Smart C2 compressor was employed on 'Dance Monkey'.
The vocal tracks go directly to output 4 of Kersting's Apogee DA-16X, while the rest of the session is sent to the unit via his four 'global' aux tracks, respectively keys (A) and backing vocals (D) to input 7-8, drums and bass (B) to 1-2, and guitars (C) to 5-6. This makes the entire session fit his eight-channel X‑Desk.
"The X‑Desk sums to a Crane Song HEDD, and from there into a hardware passive JLM PEQ and a Smart C2 compressor, again doing very little, at the most taking off 3dB, and then back into Pro Tools. In Pro Tools I have whatever sounds good on the day on my mix bus. In this session I had the Slate VMR, with a Neve console emulation and iZotope Ozone 7. That goes to my Slate Digital FG-X mastering processor, which is adding some transients, and from there it goes to my Kush Audio Clariphonic DSP, which I love. You put it on and it is like a blanket has been lifted off your mix! At the end of the chain are the Slate VTM tape emulation, and the FabFilter Pro‑L, which is just there to make sure that there are no unexpected peaks. I never add limiting even when I send out listening copies. My mixes generally come in at -12LUFS, which is already louder than Spotify. The mix that I send for approval is also the mix that I send to mastering!"
Sun, Sea & Studios
Konstantin Kersting is originally from the German capital Berlin. He arrived in Australia in 2007 as a teenage exchange student and, wanting a beach and a non-urban, Australian lifestyle, settled in the coastal town of Caloundra. "I played violin from age four or five," recalls Kersting, "and played in orchestras, but let go of that when I turned 16, because playing the violin seemed hopelessly uncool. I switched to bass guitar, and played in jazz bands for ages, and continued to have music lessons, which gave me a background in music theory. I also play guitar and keyboards, and a little bit of drums.
"A few years after I arrived in Australia, I got a Bachelor of Music degree at Queensland University of Technology in Brisbane. After one year there I also started working as an assistant in the studio for one my lecturers, Yanto Browning, who became my mentor. At Airlock Studios — owned and operated by Ian Haug of Australian band Powderfinger — I worked my way up from assistant to second-in-house engineer, and after a few years I eventually went completely freelance."
It was only four years ago, when he moved to his current studio, that Kersting started getting into "the more electronic side of music production", a move which was kickstarted when he participated in a 2016 Mix With The Masters seminar with star mix engineer Michael Brauer. "I learned things about mixing that have become part of my approach, but mostly it was seeing how the other people at the seminar worked that pushed me towards a different perspective on things. After that seminar I got more into the electronic side of producing, which is a whole different thing. Some people are amazing studio engineers, but they would not know how to create a cool trap beat. I had to relearn a lot of things, and today electronic productions are 90 percent of what work on, mostly with female pop singers."
Vocal Tones
One of the things that has helped drive the success of 'Dance Monkey' is Toni Watson's very distinctive vocal sound. Konstantin Kersting explains that this isn't the result of monkeying about in the studio. "I used my Miktek CV4 large-diaphragm microphone on her, and that went into my SSL XR627 preamp, which sounds really good on vocals. I ran it through the SSL XR425 EQ after that, and then the Urei 1176 and into the computer via my Lynx Aurora 16. In general, I try to minimise the amount of compression on the way in. Tones' voice is so expressive, the 1176 was just about containing the peaks, as opposed to adding a sound.
"We did maybe four or five takes, and that was it. She's a really good natural singer, and the way it sounds is just the way she sang it. It's unique, and unlike anyone else, and I guess it's why people have taken to the track. I barely did any tuning, apart from manual corrections on a few notes, and there's no added distortion, or things like that. I thought of the song as a left-field pop song, and that kind of stuff does not need to be 100-percent perfect. The main thing I had to do was to control the sharp, high mid-range that her voice has, using EQ and multiband compression. I did this during the mix."