Kpopalypse explains good vs bad production in k-pop

Hi everyone, it's that time again - time for one of those technical posts that you all love.  Recently people have been asking me variations of the following question:

Yes I am!  If you've ever wondered what constitutes a decent production, Kpopalypse is here to help!  Read on for all the trufax about good production vs bad production!

Let's start off by defining a few concepts - what is sound production anyway?  In the most general sense, sound production is the conceptual process of getting the "sound idea" from the brain of the composer onto some kind of recorded format.  The musical process is "performance" and the physical process is "audio engineering" - but production is the realm of conceptually translating sound ideas.  Of course a producer may be an audio engineer as well - in k-pop this is common.  They may also in some cases be the songwriter and/or the performer, as k-pop tends to use the words "producer" and "songwriter" interchangeably, but I'm talking about a producer in the strict sense for the purposes of this post.  If this paragraph makes no fucking sense at all I wrote another explanation here which may help.  My posts on vocal and instrumental production may also help if you're confused by any technical terms here.

So how do we know if the sound production of the producer is good or bad?  The answer is through how successfully the idea from the composer is conveyed from the original concept onto the recording.  For example, if the original sound idea is someone taking a shit for five minutes, then if the producer was able to bring the right ideas to the table to turn that concept into the smelliest, most visceral piece of recorded shit possible, then this is good production.  If on the other hand the shit didn't seem that shitty, then the producer didn't do as good a job.  This is of course assuming that the original goal is the sound of shit.

Of course, k-pop doesn't aim for shit, what they're aiming for instead is what any pop producer aims for - a final product which has the following characteristics:

• All the important defining elements of the song elements can be clearly heard at all times, especially any vocal as in pop music the vocal carries the all-important melody plus the sound of the singer's voice that fangirls will jizz over
• The beat, where it exists, will also be clearly audible, because making people want to dance is important - dancing makes people happy and happy people earn more money to spend on products endorsed by k-pop idols
• Will kick ass on a big speaker system AND will also still sound reasonably clear on a tiny phone or radio speaker
• Will sound just as loud and clear as someone else's song if they are played back-to-back on the radio

Finding examples of k-pop songs which meet these requirements is dead easy, and that's the problem with doing this kind of post about k-pop - I need to give you some examples of the bad stuff, and k-pop production quality is almost universally excellent.  Just about any k-pop video that you can pull up randomly for your favourite groups these days is going to sound great from a pure production perspective.  Sure, you might not like the sounds in the final song, or the song itself, but that's a different issue to whether the producer did a good job.  It's not just the big three that nail it either, even nugu groups these days usually have great production and you really have to look very hard, or go back through k-pop history fairly deep to really find anything that sucks so you can contrast it to the good stuff to give a picture of good vs bad production... but Kpopalypse will do his best!


H.O.T's "Outside Castle" is typical of the absolute rubbish that SM pumped out and fangirls sheepishly lapped up about 15 years ago and demonstrates one of the most challenging aspects of producing a pop song correctly (if you don't know what the fuck you're doing, that is) - volume normalisation.   Anyone bred on a diet of today's k-pop will instantly notice that when the first singer's voice goes into the lower register at 2:02 he becomes briefly really hard to hear, and when the rapping starts, that's also really hard to hear at certain points.  Yes, they're rapping in a whisper, but that's no excuse, because there's tools in an audio engineer's arsenal that can bring out the sound of a whisper so you can actually still hear the words properly.  Think about how clear JYP's signature "JYP" whisper always sounds, that's because the whisper's volume levels have been compressed, something which H.O.T's vocals have not been (or at least not enough).  H.O.T also haven't been pitch-corrected - at 1:55 the singer is actually hitting the note flat - this would never be allowed on a k-pop recording in 2015, it would be "fixed in the mix".  On top of this, the mix overall just sounds dull... I'm not talking about it being a dull song, but there's a real lack of brightness across everything and I'm not just talking about the crappy video transfer, the actual original audio is this way too.  Play this song and something from even a nugu modern boy group these days back-to-back at the same volume and notice the difference in clarity.  Modern pop mixes are heavily equalised, post-processed and mastered meaning that frequencies which allow the songs to cut through on radio broadcast or perform better on smaller speakers are enhanced.  There's definitely not much of this happening on H.O.T's recording, and while the final result was probably an acceptable production standard for k-pop in 2000 when the standards for a finished pop recording were lower, an in-house SM producer would be sacked for creating something with this level of audio quality today.


M.O.A's "Run For Your Dream" is on the other hand a song that I really like, but the production of it is still pretty poor, and it's a testament to how good this song is that a botched production job can't fuck it up.  So what's wrong with the production?  If you listen closely to the acoustic guitars, the strumming is actually not quite in time with everything else, it's especially noticeable in the second half of the song where the beat speeds up.  Also once again the vocals aren't normalised correctly, the best example in this song being that the word "run" is always a little bit lost in the mix every time the chorus is sung, and losing the first note of the hook in a chorus line is a big deal.  The raps are pretty lumpy in volume levels too, to the point where when the chorus kicks in it actually sounds a bit quieter than the peaks of the rap verses - not what you want in a song like this.  These factors all combine to generate that uncanny valley feeling that "something is wrong here", and of course self-appointed k-pop music industry experts with their incredibly asinine "vocal talent" obsession but no real knowledge go straight for the explanation "oh they can't sing":

This is wrong.  Actually the vocalists are perfectly on-key.  It's the volume levels that are the problem, but people are not accustomed to hearing incorrect volume level compression in k-pop, so when they do hear a producer fuck this up, they immediately assume something is wrong with the singer, however it's not the case here.  The vocalists were ready to debut - the producer wasn't.



To get a feel for how better production can make a difference, try listening to the first few seconds of these recordings of S.E.S and Red Velvet performing "Be Natural", back to back (make sure you crank the volume of the S.E.S version so they're equal as Red Velvet's version has a much better transfer to YouTube).  Both versions use exactly the same backing track so the differences are subtle, but in Red Velvet's version vocal volume is more consistent and it's a little easier to hear the details in the mix.  The S.E.S version isn't that bad but the Red Velvet version is better in terms of how pop production quality is assessed.

Yes I'm aware that there's a video which places one version of "Be Natural" in one speaker and the other version in the other, but using this isn't practical for comparing production values, as it's impossible to fairly judge production quality of a stereo recording from just listening to one speaker of it.  The manipulation of the stereo field to create room in the mix for each element comprises part of what creates a clearer sound on a complex multi-track pop music recording.  So what is "room in the mix"?  To answer this, it helps to imagine sound as a physical space.  I've used these diagrams before when talking about the deluded bullshit circus that is assessing MR Removed videos, but here they are again:

I'm too lazy to redraw everything, so time to shamelessly plagiarise my own article from two years ago.  In the above picture, the horizontal axis represents stereo, and the vertical axis represents pitch.  Divide it up and we get this.

And here's a typical result of elements that you might hear in a pop mix:

This is an oversimplification of course, but it introduces the general concept of everything needing its own space (and now you know why I couldn't use those mono recordings - remove the stereo aspect and we have three "boxes" to work with instead of nine).  If you have two things sitting in one general area, they will compete, and if one element is louder or present more often than something else, but occupies the same space, it will mask the other element.  This phenomenon is called "audio masking" and it's why snipers sometimes time their shots to happen at the same time as other loud noises.

In the H.O.T song, the sung vocals come out quite clearly and cut through when they're high, but the lower vocals and whispered raps compete heavily with the keyboard/orchestral noises that are happening at around the same time and frequency, so the vocals get "swallowed up" to some degree.  This could have been fixed by:

• Adding more high-end frequencies to the vocal so it can cut above the strings
• Removing the frequencies in the strings part that directly compete with the vocal
• Panning the strings out so they sit more in the L and R boxes while leaving the vocals central
• Compressing the vocal so the volume is more consistent and quieter words don't get buried
• Hiring the aforementioned sniper to remove the problematic producer

The above deals with the issue of basic pop music production standards, and this is a very objective standard, because without certain audio treatments, songs will sound too quiet or too "muddy" or simply amateurish when played next to other songs.  In a highly competitive field it's important to create something that can stand up favourably next to everything else around it with equal clarity and professionalism.

Of course, k-pop fans are notorious for not bothering with (or even knowing about) objective standards, so they don't tend to use the word "production" in this objective sense.  For a k-pop fan, comments about "production" are usually more about "creative production choices" than "production craft", so when a k-pop fan (including me, at times) says "I like a song because of the production" what they're usually actually saying is "I like the way the backing track sounds".  Here's some examples.


I single this song out occasionally for oh-my-god-why-did-they-make-it-sound-like-that slings and arrows a bit but there's nothing wrong with T-ara's "Yayaya" just from a technical production point of view.  The issue that I have with it is that I don't feel that E-Tribe's wall-of-noise style on this track really suits pop music, generally speaking.  The fact that the song fulfills modern pop production requirements is an objective truth, but the fact that I think it sounds like rubbing a cheese grater across my ears is a subjective point of view.


I'm also not a big fan of the marching drum beats in the verses of Oh My Girl's "Cupid".  Marching drums can actually sound great in other k-pop songs but their implementation here doesn't work for me at all.  However they're not inserted badly or anything like that, production-wise this is fine, I just don't think this type of instrumentation suits the material, therefore "I don't like the production" rather than "this is produced badly".


This type of yoloshit production style is everything that's wrong with popular music right now.  Because of this, there's a possibility that CL's American debut will actually do reasonably well, I think that given the right marketing she'll fit right in with ratchet-lite like V-Nasty, Riff Raff, Lil' Debbie, etc no problem.  "The Baddest Female" isn't badly made, either - YG's production here is as usual excellent.  Sure, they forgot to actually write a song, but in the yoloshit genre I guess songwriting is considered like sour cream on your baked potato, an optional extra that not everyone wants.

I hope that's enough examples for you to get the idea, because I really couldn't be fucked listing any more shitty songs that suck but happen to also be produced okay, I could be here all day doing that as it would account for about 90% of all k-pop made since about 2008.  Anyway hopefully this post has been sufficiently confusing, annoying and contradictory to prevent and completely discourage any sort of productionfagging from obsessive-compulsive fangirl crazies who want to prove "my bias is best", and hopefully also educational and interesting enough so the rest of you didn't fall asleep reading!  Yay!