Although many people seem to write off mainstream K-Pop hits like this song as musical fluff, there’s actually a lot of quality craftsmanship going on under the surface. Yes, the first minute of the song suggests we might be in for another ‘here we go round IV-I-V-vi’ snoozefest, for instance, but thankfully the chorus brings with it not only a little variety in the chord pattern (changing to IV-V-I-V-iv), but also a welcome change from the chord-per-bar harmonic rhythm, by virtue of the G-D/F_# chord progression in the third bar of each four-bar chunk.

[For the sake of easier discussion from this point let me quickly label how I’m hearing the song sections, so it’s clear what I’m talking about from now on!

• Intro (0:00-0:14)
• Verse A (0:15-0:30)
• Verse B (0:31-0:46)
• Prechorus A (0:47-1:01)
• Chorus A (1:02-1:17)
• Chorus B (1:18-1:34)
• Mid-section (1:35-1:50)
• Prechorus B (1:51-2:07)
• Chorus A (2:08-2:22)
• Chorus B (2:23-2:38)
• Chorus C (2:39-2:54)
• Outro (2:55-end)

I realise that what I’ve called the Prechorus here operates a lot like a drop chorus, but bear with me, because I think it’s worth its own label…]

From a rhythmic standpoint, it’s nice to see a dance-pop track using a 12/8 time-signature, which is one of the things that I think really helps this song stand out from the market. But the producers haven’t rested on their laurels here, because they’ve taken good advantage of the same kinds of syncopations and cross-rhythms that are so often used to leaven simple time signatures. The hemiolas on “I couldn’t find my own place” (0:35) and “now that’s how I’m getting paid” (0:43) provide a 12/8 substitute for the dotted quarter or eighth notes frequently used to add polyrhythmic texture in 4/4, for example, and I love the stark contrast between the heavy use of syncopation in Verse B and the resolutely straight Prechorus that follows. I also think there’s a subtle momentum generated by the vocal rhythm’s gradual development from the disordered, fragmentary Verse A (0:15-0:30), through the syncopation of Verse B, and then onwards into the straighter, stomping rhythms of the Prechorus and Chorus sections (give or take a few hemiola fills at the ends of phrases).

But what really caught my attention, from a production standpoint, is the vocal melody writing. First of all, the sheer range covered here is enormous (two octaves and a fourth) – only a couple of semitones short of A-Ha’s ‘Take On Me’, but without sounding like quite such a freak-show act! Beyond this pure bravura, though, the way the vocal register is paced through the song is very smart. First we get a gradual increase in both pitch range and pitch register, with each section reaching an ever higher top note: D3 in the first half of Verse A, then F_#3 in the second half; B3 in Verse B; E4 in the Prechorus; G4 in Chorus A; and A4 in Chorus B – as shown here:

However, this tactic has now effectively been all but maxed out – there’s not much further they could go with this idea without the singers moving into high coloratura or whistle register! So the writers at this point take the opportunity to kind of ‘reset’ this parameter by settling back into a low register for a while (remaining below A3 until the second half of the second Prechorus) so that it can then begin ratcheting things back up again during the remainder of the song, moving to E4 later in Prechorus B, and then to G4 and A4 in the Chorus sections as before:

The melodic writing also relies heavily on the idea of establishing a melodic phrase in the first half of an eight-bar section, and then varying that phrase slightly for the second half, thereby balancing those perennially competing pop imperatives of (i) lodging the melody clearly in the listener’s mind by virtue of repetition and (ii) maintaining a constant steam of attention-grabbing novelty. In the Verses and Prechoruses, this variation is achieved by altering the final two bars, whereas in Chorus A the variation occurs in bar 5, where the line rises to G4 compared with the E4 in bar 1. Clearly the demands for memorability won out in Chorus B, however, where both halves are effectively the same melodically, but it’s worth noting that Chorus C then introduces a whole new riff (at 2:39-2:46) to refresh our attention before the final iteration of the hook melody at 2:47.

And, speaking of the main Chorus B hook, one of its most characteristic features is its series of three soaring melodic arches, the second and third of which both showcase an unusual compound-ninth construction made up of two rising leaps: a third followed by a seventh. But it’s important to note that this contour is deliberately weakened during both Prechoruses. Partly this is a ramification of harmonic differences between the Prechoruses and Choruses – in the second and sixth bars of Prechorus A, for instance, the second note in each case becomes G for this reason, rather than the F_# that appears in Chorus B at that same point. But it’s also clearly a deliberate choice to save the third phrase’s high notes for the forthcoming Chorus sections in each case – only the first half of Prechorus B (1:52-1:59) is allowed to reproduce the full hook-line’s pitch range, presumably because (a) it’s all being sung down the octave and (b) the third bar’s expected compound-ninth contour has been weakened to a compound seventh by shifting the first two notes up a third to B and D respectively at 1:55.

And one final nice touch is that the new melodic idea introduced in the first half of Chorus B is an A pedal note, and a slightly unusual one too. In traditional harmonic practice, pedal notes normally start and end on chords in which they constitute one of the consonant notes – the idea being that the pedal note starts as a consonance, then becomes a dissonance to set up some musical tension, before releasing that tension when it finishes as a consonance again. Here, however, the pedal note is dissonant both when it arrives (as an added sixth over the C major chord at 2:39) and when it finishes (as a suspended fourth over the E minor chord at 2:45). But if you consider its musical role here, I think it makes perfect sense for it to do that – its purpose isn’t to create the traditional ’tension-release’ dynamic, but rather to provide just the tension, from which the return of the hook line is the release.