Mix Analysis: Chandelier by Sia (Audio)

Hey folks, Matthew Weiss here — weiss-sound.com and advancedmixing.com.

This is going to be an in-the-dark analysis, which means you’re going to get no visual component here at all, we’re just using our ears because that’s what we do when we are producing music.

So this is going to be an analysis of Chandelier by Sia, mixed by Manny Marroquin. It is one of my favorite references to use for a really big, bombastic Pop song. It’s got a huge chorus, it’s got these really broken down verses, and it’s just a cool song in general.

So I’m going to play a little bit of it, and we’ll talk about a couple of things.


So it’s a very interesting mix, sonically speaking, because it has just an ass-ton of compression going on, but it works. Like — and it’s technically — it’s over compression in the sense where you hear it and you go, “I know technically that is incorrect. Transients should not square off like that, that’s not how it’s supposed to work,” and yet, it’s compelling. It still keeps me there, and so the analytical mind can go, “Well, that’s bad,” but the feel of it is pretty undeniable.

So there’s a couple of things going on. This is one of those kinds of records where when I listen to it, I feel like this probably had a lot of compression going on from Manny specifically, not from the mastering engineer. I think that the mastering engineer probably got this and either had a version with a limiter or several limiters going on, or just very fast compression that was printed by Manny, and maybe one that didn’t have that, or just the version with the limiters on there.

So I want to go through a couple of ideas in terms of the way the record is arranged sonically, and how some of these tones are coming to life, and how all of this dense, dense stuff stands under such high compression.

So the first thing that I want you to listen to is the tonal consistency in the verse, particularly Sia’s voice.


So Sia, part of her style is specifically to almost mumble. It’s very much like, [emulating mumbling vocals], like, it’s purposefully slurred. That’s part of her artistic delivery.

So to have that much clarity in the vocal shows that there’s some pretty heavy handed EQ going on. Manny in particular, his style is very much based on compartmentalized EQ and dynamics, so what I mean by that is that he will very frequently separate sources into frequency bands, so a voice might be treated as a low band, a mid-range band, and a top end band, and in this particular case, I either am hearing a lot of multi-band compression going on, or literally crossovers set and the vocal divided into several different frequency bands, and treated independently with different compression to keep things very much in line, and the one to really pick out is the upper mid-range.

If you listen to where her voice pinches, which is at maybe about 4kHz, 3kHz to 4kHz, somewhere in there, maybe even a little higher. It’s hard to tell exactly, but you’ll hear this tone. If you just try and ignore all of the changes of her voice, and everything else, you’ll literally hear a tone show up, and I’ll see if I can isolate it on an EQ.


So it’s like, 2kHz. I was over exaggerating it, but maybe 1.5-2kHz, somewhere in there, where it’s just super consistent.

[mix, filtered]

Like, that specific range stays locked in the entire time, and the reason why this is important is because once everything really kicks in and we get into this extremely dense section…


Our arrangement is super full. We don’t have anywhere else to put the vocals, so we need to make sure that it’s going to be that consistent if we’re going to have this style of mix.

There’s a number of ways to create this dynamic manipulation through separating frequencies. One of the traditional ways is to literally use high-pass filters and low-pass filters at similar corners, so you would take, like, let’s say you were doing it with the vocal, you would take the vocal and you would put maybe a high pass filter on — maybe copy the vocal three times, and you’d put on the first version a low-pass filter at like, 500Hz, and then you’d put on the second one, a high-pass filter at 500Hz and a low-pass filter at like, 2kHz or something like that, and then on the third one, you would put a high-pass — just high-pass at 2kHz.

So each one independently kind of sounds like a band-limited version of the vocal, but summed together, they all add up to a complete frequency spectrum.

I don’t do this with EQs, usually when I’m doing it, I’ll use one of the iZotope programs to do it, because they have — you know, they have built in crossovers right away, and then the other thing is you can kind of kill two birds with one stone, like you can use the Ozone 7 multi-band compressor, and you already have compressors right there on your crossovers. So it saves you a step of having to put on the different compressor. You know, unless you wanted the tone of a specific compressor for the low band or something like that.

Sometimes I’ll do it with the Ozone Exciter, because I might want to work some excitement into each band, or sometimes if it’s like a source like a synth of some sort, or a bass that has a lot of frequency content, I might do it with the iZotope Ozone Stereo Imager so that I can take the top band and I can make the top band stereo and wide, where all the buzzy bass stuff is happening, and I’ll talk about that in a second.

Now, another thing that I want to point out is the relationship between the bass, the piano, and the kick.


So that bass is a composite of a number of things. There’s a low bass that’s really living in the sub-range that pushes out that 50Hz, maybe even lower, maybe like, 40Hz kind of thing, and then there’s also a piano that’s being played on the lower end, which I believe has some really complex filtering going on to make it so that the fundamental tone of the piano is there, and strong, but nothing in that sort of like, woofy zone is there at all, like that 200 has been carved very carefully to my ear, and also everything below the lowest fundamental tone of the piano is like, way stripped away, so it’s like, just the key fundamentals in the low range of the piano are there. The overtones of the piano are still there, everything that’s 300Hz and up, a lot of that stuff is still there, but not that, and the reason is because there’s also a bass pad. There’s some kind of a — maybe it’s like a low string pad, or maybe it’s just a synth bass pad that’s got some sort of weird ambient thing going on, and that’s sort of living right in that same 100-200Hz as well.

So we have a kick, we have a bass with a lot of sub, we have some kind of a pad thing, and we have this low piano, and they’re all living in the same space. So you know, if I was mixing this record, it would sound like a crazy, jumbly space, because I kind of do that on purpose a little bit, but Manny did this obviously, and so everything is just very crisp and in its own spot.


Now, there is some kind of a quick ducking thing going on with the kick, so that the low basses all kind of like — very quickly get out of the way when the kick hits, and then they come back, and it’s really just over the attack of the kick, it’s not really the sustain of the kick is running into the bass, but the hit of the kick is definitely going into it, and the kick itself is like, smashed. It’s got this like, very aggressive kind of square-ish shape to it that is — it’s like an EDM kind of thing, but taken to the nth degree, and I suspect that the sample started that way in some way, shape, or form to begin with, and Manny just ran with it and said, “You know what, if this is going to be a squared off kick, fine, let’s let it be a squared off kick and rock n’ roll.”

Which I think works, actually, I think it gives it this sort of raw and polished at the same time, and so while it’s technically totally wrong, it’s still functionally, musically, totally working.


The other things to note are that this — you know, Pop gets this sort of rep for being sterile, and that’s not true at all. Pop mixes are very heavy handed, generally speaking, there’s a lot of the mixers print right in the record, and the mixers usually act in a very musical way, so the last thing that I want to point out here is there’s this buzz going on, and I suspect that it’s a product of the low bass synth, like, maybe it’s a synth from Massive where it’s got this very deep low end, but it also has this buzzy, dubstepy kind of top end, and since there’s obviously a Reggae influence in this record, it’s probably a dubstep style bass.

So what I think happened was I think that again, going to the split frequency thing, I think that Manny put the bass onto two tracks: The low end of the bass, and the high end of the bass, and he’s ducking out the low end of the bass, but the high end of the bass, that buzzy thing, is not only staying very consistent, but it’s also way, way, way turned up in the mix.


So that’s what like, 10kHz on a bass sounds like, when you listen to that buzzing thing that’s just streaming throughout the chorus, and I love that. I think that it is brilliant, I think that it helps to define modern pop, and what we can do with digital sounds and digital synths.

It’s fearless, and it’s really important to have that going on, because that energy is irreplaceable, it adds texture, it adds character, it’s the intensity, it’s the drama of it, because it’s got this fiery tone to it, and it matches the emotional energy of the chorus.

So when you’re doing your mixes, you know, when you find stuff like that, don’t shy away from it, don’t say, “Oh, man this bass is buzzy as hell, let’s take the buzz out so that it’s cleaner,” I mean, sometimes you do need to do that, but a lot of times, it’s that stuff that you wouldn’t normally like that ends up becoming one of the best features of the sound of a record.

So alright guys, I hope that you learned something, and until next time.

Matthew Weiss

Matthew Weiss

Matthew Weiss is a Grammy nominated and Spellemann Award winning audio engineer from Philadelphia. Matthew has mixed songs for Snoop, Sonny Digital, Gorilla Zoe, Uri Caine, Dizzee Rascal, Arrested Development, 9th Wonder, !llmind & more. Get in touch: Weiss-Sound.com.
Smiley face