All it took to make a star of Carly Rae Jepsen was one memorable song — and, in Dave Ogilvie, a mix engineer who understood how to make it stand out.
Producer Josh Ramsay called mixer David 'Rave' Ogilvie in March 2011, excited about a new song he'd written and recorded with the relatively unknown Canadian singer Carly Rae Jepsen. Ogilvie recalls, "I enjoy everything Josh works on and like mixing his stuff, so I was eager to hear what he'd done. I went over to his studio, The Umbrella Factory, and when he played me the song I thought it had one of the biggest hooks I'd heard in years. I couldn't wait to mix it, and did so a couple of months later. I knew that the Canadian radio would love the song, and when it took off in Canada I felt vindicated in my initial opinion. But I had no inkling at all of its worldwide potential.”
Very few people had. 'Call Me Maybe' was released in Canada in September 2011, and was in the top 10 by the end of the year. Then Justin Bieber heard it on Canadian radio and tweeted that it was "possibly the catchiest song I've ever heard” — whereupon 'Call Me Maybe' went on to become the big Summer hit of 2012. It reached number one in 20-something countries, including Canada, the UK and the US, went multi-platinum in Canada, Australia, New Zealand and the US (where it sold a whopping four million copies), and turned Jepsen from a former Canadian Idol second runner-up into a global star.
Back at home in Vancouver, Ogilvie is well known as a top mixer and producer, although he is more strongly associated with electronic and industrial music than breezy, catchy pop. He has worked with Skinny Puppy, Nine Inch Nails, the Birthday Massacre, Marilyn Manson, Einstürzende Neubauten and many more, and in recent years his activities have increasingly crossed over to the pop market, in part because of his regular work with Ramsay, who is also a member of the pop band Marianas Trench.
Ogilvie: "One of my biggest influences was Neil Young, so I can relate to the folky, acoustic guitar stuff, but also to more experimental things. For example, the first record that I ever bought was by Kraftwerk. So as a producer I've always been completely comfortable working with synthesizers and computers and rock and folk guitars. When I finally got round to working with Nine Inch Nails , one of the most striking things for me was Trent [Reznor]'s love of melody. It's also where I came from: when I was a kid, I listened to a lot of pop music on the radio. To be able to enjoy melody in industrial music was a big turning point for me: finally, there was somebody who understood that you can have cacophony and all these crazy, aggressive sounds, and yet you can put melody in it as well. There is music in everything, and these days there's a big convergence of genres. You can now use anything from anywhere, and people get excited about it. The pop music that has fallen into my lap during the last two years has been easy to deal with, because usually it's well recorded with lots of slick sounds. I may then throw some really extreme drum samples into the mix, and people will say: 'Hey, that's amazing, where did that come from?' Often I don't have the heart to tell them, because they may get nervous about it!”
'Call Me Maybe' thus conceals some darker elements beneath its radio-friendly surface. At first listen, it's a rather fluffy, lightweight, bubblegum pop song, but closer inspection quickly reveals another dimension, which is most apparent in a hypnotic four-to-the-floor bass drum that becomes monstrous in the choruses and a muscular string hook. The song also comes across as a genuine representation of Carly Rae Jepsen's character.
"Yeah, Carly has this quirky aspect to her personality,” comments Ogilvie. "Josh and I both know her well and it was very important for us that you could hear and feel who she is in the song, that you could feel her attitude. It's something that so often gets lost in the production or the mix, when the lead vocal gets homogenised and you can't really identify with the singer. For me as a mixer, this is one of the biggest issues, and I'm always very careful to make sure that the singer's identity comes across, which in this case was a bright and sassy young woman with lots of confidence. Mixing this song was a matter of really making sure that the reality of the song and Carly's personality were matched by the technology. It is a teenage song and most people would not have taken it any further than that, but Josh and I did whatever it took to make it something special. We didn't want to take the easy route of using presets and adding loads of bells and whistles. Instead we were looking for a more creative and different approach. Compared to a lot of the other pop stuff I mix, this session was quite minimal. The arrangement doesn't try to fit every hole, and mixing the song was more about making what's there jump out at you.
"The main thing that Josh wanted me to do during my mix of 'Call Me Maybe' was to make the kick drum really prominent and punchy. He wanted me to do everything I could to make it super-special. I think I spent a good couple of hours on the kick drum alone, and kept pushing myself to create the ultimate kick drum, and have it super-loud and driving the track. When I finished the mix, I thought it really was very, very loud, and I was a little apprehensive about the power of the bottom end, wondering whether I'd gone too far with it for a pop song. But Josh and I looked at each other and said: 'This is what we are shooting for.' So we printed it like that and when I later heard it on the radio I was like, 'Wow, this is awesome.' It seems to work because of the space in the track, which I was careful to leave, which means that you can hear the punch the kick drum is intended to have on computer speakers and in cars and in stores, in fact pretty much everywhere.
"My background in industrial music definitely helped in achieving this. I was trying to get the same feel in 'Call Me Maybe' as in a Nine Inch Nails song, making sure it had a pop sensibility, but with people not even noticing how aggressive the kick drum is. It was the same thing with the strings, which are canned, taken from a Miroslav sample bank, an Orkester patch and an EXS24 patch. I find that it's so easy for strings to get lost in a track. You often mix them to a level where you think they are good, and later you hear them on the radio and you think, 'Man, I should have put them up a dB or more.' Josh and I both knew that the strings provided a big hook for the song, and so he had doubled them with electric and acoustic guitars to make them sound beefier and more aggressive. I mixed these guitars in just behind the string samples, so you don't really consciously notice them, but they definitely gave the strings more kick. I normally wouldn't think that strings and guitars would combine well like that, but in this case it gave the strings such an aggression and a striking sonic image. We had to get over our fear of mixing the strings and kick too loud and for them not to drown out the vocals, which were the third crucial element. The vocals also had to be clear as day and aggressive-sounding.”
David Ogilvie mixed 'Call Me Maybe' on a 72-channel G-series SSL with E-series EQ in The Warehouse, one of the big studios in Vancouver, owned by Bryan Adams. He explains his requirements in choosing a studio. "Where I work depends on the project. Budgets have changed, so I can't always work on a massive SSL with a huge Pro Tools rig. I don't like working in the box. I'll do it if absolutely necessary, but I'll generally try to work out a deal with the studio so they'll give me a break and I don't have to work in the box. When working on a desk, I'll always mix on an SSL, and one of the main issues is that it has to be in good working condition, because there's nothing worse than mixing on an old board that's not working properly any more. For monitors, I still like the NS10s with subwoofer powered by a Studer amp, and I also like working with Genelecs, Dynaudios and KRKs. I generally work at one of three studios: the in-house studio of 604 Records [Jepsen's label, which is closely related to Simkin Management, which has Jepsen, Ramsay and Ogilvie on its roster], Hipposonic Studios, or The Warehouse Studio A, which is normally Randy Staub's room.” (See January 2012 issue: /sos/jan12/articles/it-0112.htm)
"The first thing I do when I start a mix is push up all the faders and listen to everything. I walk around the room and tweak little things, but I'm otherwise not doing very much for the first hour or so. If there are people in the room, they can get quite confused, because it looks like I'm not doing anything, but I'm busy in my mind trying to assemble a picture of the entire Session, and formulating a vision for where I want to go with it. Of course, I will also already be listening for what sounds work and don't work and whether I have to add any samples. But you can't tell what a sound is going to do unless you hear it in context with the other sounds. You could spend three hours working on a snare sound and make it sound fantastic on its own, but when you put it in with the guitars, all of a sudden it may sound terrible. So I really like to assess all the sounds at the same time, in context, sometimes hearing them in groups, as opposed to listening to them individually.
"Phase two will, 90 percent of the time, be me going to the drums and working with them, and working on the bottom end. I try to get the bottom end of the drums to the point where I can bring in the bass, and get the entire low end tight. From there, I will start looking at the width of the drums, and after that I'll establish the middle. So I'll first work with the rhythm and then I'll figure out where the space is for the vocals. At that point, it's my priority to establish the middle, the centre, especially for radio. So that's kick, snare, bass and vocals, and that centre has to be really strong before I start thinking about the edges, like the strings and guitars and so on. Once my centre is strong enough, I'll turn off the vocals and I'll bring in all the instruments. With 'Call Me Maybe', it was definitely a matter of first bringing in the strings, to make sure that they had enough power and wouldn't disappear the moment I brought the vocals back in.”
Drums: SSL EQ & compression, Metric Halo Channel Strip, GML EQ, Dbx 120XP, AMS RMX.
"As you can see on the screenshots, I far prefer the desk and outboard for treating the drums. The only inboard [plug-in] I used was the Metric Halo Channel Strip on the 'Verse Kick' (5) and on the 'Ghost Snare' (15). All the drum tracks would have had some treatment from the GML EQ. I would have used all five bands on the GML, all set pretty aggressively, with +5 or +6 or -5 on the frequencies. I'd also have parallel compression on the drums, using the SSL board compressor. My main focus in working on the drums was making the chorus kick drums as massive as possible. There's only one bass drum in the verses, which had a Pultec on it for some more punch, and five bass drum tracks in the choruses, and I spent almost two hours adding three bass-drum samples in the choruses to the two that Josh already had in there, and making them sit right with each other and in the track. I have amassed an enormous sample collection over the years, from recording different artists, being in different spaces, in general recording things anywhere where I could get good sounds. It just kept building and building and it's now so massive that I feel sorry for the assistants I work with, because when I say to them 'Find a kick sample,' they are so overwhelmed. But I can go to a sample bank and I'll immediately find what I am looking for just from looking at them.
"In this case, I went for more dance-orientated kick samples. There have been many pop songs in the US in recent years that use that four-to-the-floor Euro-style dance kick — Katy Perry's 'Fireworks' certainly was an influence — and that's funny, because for me, it's a matter of things coming full circle. A lot of what I was doing in the 1980s used the big four-on-the-floor dance thing. It kind of disappeared in the 1990s, but all of a sudden it's made a comeback. For me, it's stepping into a comfort zone, as opposed to trying to be really subtle with things. Instead it's a matter of how to drive that four-on-the-floor thing. It's easy to program it, but much more of a challenge to turn it into something special. One thing I did in this track was to have some Dbx 120XP [subharmonic synthesizer] on the kicks, and a little bit on the snare, adding some subharmonics. That's a trick I picked up from Randy [Staub]. Another thing I did was feed a little bit of the bass guitar into the drum bus, so it will go through the drum EQ and compression and it'll also get some subsonic effects from the DBX, and then all of a sudden the kick, the snare, and the bass blend together as a solid unit in the middle. With the snare, we were taking the approach of not having a traditional snare, with it being made up 60 percent of handclaps, 30 percent from the 808 snare, and the real snare quite low in the mix. It was quite fun not to have to worry about a traditional snare sound. The clap snare sound worked much better in the track. The Ghost Snare (15) is subtly thinned out with the Metric Halo Channel Strip. I also had some AMS RMX reverb on the snare, and on the drum loops, as well as some compression from the SSL.”
Guitar & bass: SSL EQ & compression, Pultec EQ, Metric Halo Channel Strip, Sansamp PSA1, Urei 1176.
"The guitars were all treated with Pultec EQ and SSL desk EQ and compression, no inboard. The acoustic guitar was just there for rhythmic support, and would have had some SSL board compressor. I didn't do much to treat the two synth- bass tracks, just some Channel Strip and some parallel compression on the board, while the bass guitar in the chorus also has the Channel Strip to tighten it up, plus a Sansamp PSA1 for some grit. In addition, I added some parallel compression on the bass guitar from an Urei 'blackface' 1176, which is a standard thing for me. At the start of every mix session, I'll set up an SSL compressor going into an SSL EQ for the drums, an 1176 for the bass, and an [Empirical Labs] Distressor for the vocals.”
Vocals: Empirical Labs Distressor, Waves SSL Channel, Renaissance Vox, Renaissance De-esser, L1, Dbx de-esser, AMS DMX, Avid Revibe.
"Josh had recorded the verse vocals with a Neumann U87 going into a Neve preamp and then into a Distressor for a cleaner tone, and used a Neumann U87 going into an API mic pre into an [Empirical Labs] Fatso Jr for a more aggressive chorus vocal sound. I used a lot of plug-ins on the vocals in general, because I will painstakingly go through each vocal track and will generally put a compressor, limiter, EQ and de-esser on each track. There's normally not enough outboard to do this on individual channels on the desk, so this is where I get quite deeply involved in using plug-ins, and I'll then do a little more stuff on the desk and I'll have parallel compression from a Distressor.
"For most pop stuff, I'll have a centre vocal and a left and a right channel vocal. So I'll have three vocals, and I'll treat them almost as one, with the main one in the middle a little louder. The other two are each in a speaker, so the lead vocal jumps out at you in every way. You can see that there are quite a few treatments on the middle vocal in the chorus (36), which has the SSL Channel, the Renaissance Vox and the Renaissance De-esser. Before those, there's a Trim, so I can determine how hard I hit the SSL Channel, which will mainly apply some EQ. The RVox smoothes the vocal out a bit, and after the Renaissance De-esser it goes to the board, where I'll have a Dbx de-esser on the insert on the channel as well. Different de-essers work differently.
"On the vocal doubles in the chorus (37, 38), I'll have the same plug-ins, but the L1 [limiter] instead of the RVox, because I want to square those doubles as loud as I can for size and power. I put the L1 on anything that I want really loud and strong, like vocal doubles and triples. It is a fantastic plug-in that I use extensively on vocals. If I want more forgiving compression on a lead vocal, I'll use the RVox or the C4 [multi-band compressor], and also the Distressor. The latter is always a big part of my lead vocal sound, and I slam it very, very hard, probably 8:1. On its own, that's kind of unlistenable, but I blend it right in underneath the normal vocal sound, and I'll then group these two channels together and I'll move them as one for the rest of the mix. Of course, the balance between the Distressor vocal sound and the non-Distressor vocal sound is critical. I also used the AMS DMX on the vocals for very tight stereo harmonising, and inboard I used the ReVibe for some reverb. Incidentally, all vocals in [the doubled] tracks are actual vocal overdubs sung by Carly, I didn't copy the tracks across. Oh, and track 43 is a last-minute vocal overdub that I asked Josh to do live in the studio while I was mixing. I felt that there was something missing there, and it takes five minutes to set up a microphone when you have an idea. So that vocal was thrown in at the last moment and definitely made the song stronger.
"If there was any Auto-Tune on the vocals, it would have been done very minimally by hand before I began the mix. Carly is such a strong singer that tuning her vocals isn't really necessary. I really have a problem with people who run Auto-Tune in automatic mode, because it means that the vocal will sound very processed. I listen thoroughly to the tuning of the vocals, and if something bothers me, I'll fix it, but if it doesn't, I let it go. I will not look at the graph. The average listener now expects perfection in tuning and timing, but I will always go on what my ears tell me. Yes, there's a greater demand for perfection than in the past, but you need imperfection, otherwise there's nothing special about what you're doing. Everything will sound the same. So I'll always do my best to get away with as much imperfection as I can. The other thing that has changed is that musicians and singers have become much better because of advances in technology. They understand timing and tuning on a much deeper level and have incorporated that into their performances. I really like that. And because singers are better, their breaths have become bigger and louder. As opposed to editing these out I prefer to use volume automation and my ears to ride these breaths, and also the esses, to make it sound as natural and effective as possible. You can see these volume rides on track 33, for example.”
Strings & keyboards: SSL EQ, Metric Halo Channel Strip, Avid Revibe, Focusrite Red 3.
"With synthesizers, I try to make sure that the sounds that are there are the sounds I want, rather than treating them with inboard or outboard after programming them. Josh and I are very in tune with each other on this, and he creates great synth sounds. He works with many synths that I also have and that I showed him how to use, like Vanguard, Massive, Nexus and Sylenth, and he's taken them to the next stage, where I go, 'wow, how did you do that?' So very little was done to treat the few synth tracks in this session. The string tracks that you see in the session are all pre-blended, pre-EQ'ed, and pre-compressed and bounced down to stereo tracks, probably using a lot of Metric Halo Channel Strip, which is my go-to inboard EQ. It's a plug-in that you can use multiple instances of in a session without it collapsing in on itself. I find that with some plug-ins, if you put them on one channel, they sound OK, but if you put them on 16 channels and then mix them together, something really strange happens, perhaps through phasing or something. With the strings on this session, you can see that there's no inboard, apart from the Revibe aux. The rest is outboard, mostly the SSL channel EQs and, most importantly, the Focusrite Red, which is a fantastic fattening limiter. I used the Red to slam all the strings together and give them more size. The Red colours the strings to make them sound warmer and bigger and bring them to life. I used the Red as parallel compression that I mixed in with the original signal.”
Stereo mix: Sontec MEC 432C.
"The entire Session was in 24-bit/48kHz and I mixed back into the Session, going via the Sontec MES 432C mastering EQ. I always set that up very early in the mix, probably when I'm working on the rhythm section and the centre, and I'll put the Sontec in the chain and create an EQ that'll be 95 percent correct, and I'll continue listening to what I'm doing via the Sontec. 'Call Me Maybe' was mastered by Gene Grimaldi at Oasis Mastering in Los Angeles, and he's one of the three or four mastering engineers that I trust. I want the mastered mixes to come back to me sounding better than when I sent them off, and also loud enough to combat other music out there, but not so loud that there's no room any more, otherwise it hurts my ears. When stuff is slammed too hard, it drives me nuts. I really don't like the extra distortion it adds. If I want distortion on a mix, I'll add it myself! A mix is my baby, but I've learned that as long as it's one of those guys that I trust, I mix and produce, and they master. Gene knows what he's doing and he did a great job on this track.”
The 'Call Me Maybe' Pro Tools mix Session consists of 56 mono and stereo tracks, a dramatic reduction from the number originally recorded. Dave Ogilvie explains: "Josh and I did a lot of comping and cleaned the session up before I began mixing. Carly is a strong singer, with whom tuning and timing are not an issue, so her vocals were also a matter of simply finding the good takes and putting them together for a final comp. Strings we pre-mixed down quite dramatically. For example, stereo tracks 52 and 53, the verse strings, would originally have consisted of 12 to 14 tracks.
"At the top of the Session are a rough mix ['Faders Up'] and the master mix. I had the piano swell and tambourine on the very left side of the board, and that's why they're the two tracks underneath that. Normally, the drums would start immediately under the master mix. There are 19 tracks of drums, which include just one track of verse kick drum (track 5), five tracks of chorus kicks (10-14, two from Josh and three that I added); 15 is a snare, 16 an 808 snare, 17-18 are handclap samples, 20 is a crash sound, and 19 a hi-hat, though most of the hi-hat sounds came from the drum loops (21-24). Tracks 25-29 are guitars, then there are two bass synth tracks and a bass guitar track: after that, 11 vocal tracks (33-44), and finally, in purple, the keyboards and strings. There's an Aux track for some Revibe reverb on the strings (50), and immediately after that a string run going into the verse strings (52 and 53), which are beefed up by the acoustic guitar (29). I put the bridge guitar with the strings (54), because it is doubling the bridge strings (55). It is adding to the string track rather than being a real guitar track. The chorus strings are further complemented by Mellotron string samples (56). I'm very lucky to have a friend who painstakingly sampled all his Mellotron and Novatron units, and they are wonderful to add underneath strings.”
'Call Me Maybe' took two days to mix, but the process was accelerated by some prep work done by Josh Ramsay and Dave Ogilvie at the former's studio. "We prepare the Sessions beforehand in Pro Tools, so as not to waste time in the studio,” says Ogilvie. "We'll submix things like strings and backing vocals to make the Session manageable on a desk, and edit and tune things, and in other ways make sure the Session is the way I want it. If there's a big vocal bounce to be done, with many layers of vocals, generally speaking I'll do that beforehand with Josh, which saves hours of time. Josh was raised in a studio — his father was the owner of Little Mountain Sound — so he learned to have his Sessions in really good shape. His sessions are 95 percent the way I like them, so I can get started on them immediately by the time I get into the mix room. It's the same with Ryan Stewart, the other producer who worked on Carly's forthcoming album, his sessions are also in really good shape.” [Jepsen's Curiosity EP, featuring 'Call Me Maybe' was only released in Canada, and an album by the singer, mixed by Ogilvie, was in the works at the time of writing.]
"Before I start work on a mix, I make sure that the Pro Tools Session itself is laid out exactly the way I want it, which is usually quite standard, with — from the top — drums, bass, guitars, keyboards, vocals, and then effects at the bottom. Colour-coding makes it far easier to move around the Session and find things faster. I then lay things out on the board. When moving from side to side on the board, the difference in sound can be brutal, so I lay things that I'll be listening to a lot out in the middle of the board, usually the vocals.
"I'll always ask for the rough before I begin a mix. If you have a starting point of how the artist, producer, label and management hear the balances in a song, and you then work on improving the sonic qualities, it can prevent lots of miscommunication and save hours of work. You don't want to have spent six hours on a mix and for them to then say, 'Oh, no, we wanted it completely different.' What's nice in Josh's case is that he won't give me a mix that's supposed to be a final one; he simply gives me the Session the way they last listened to it, with the faders where he left them. So I can just throw up the faders, and even though it's still messy, it gives me a rough basic balance and I'll take it from there.”