This article is the first in a short series that discusses the methods that I use to write a song analysis. The methods can be transferred to just about any song, irrespective of genre/style.
The Bar/Timecode/Structural (‘BTS’) diagram:
What is a BTS diagram?
When I discuss the general track information of a song in any one of my articles, I like to supplement the bullet points with a detailed graphic (e.g. see fig A below) that indicates:
- The number of bars present
- The exact duration of the song
- How that time is partitioned within the structure of the composition.
‘BTS Diagram’ is not an academic term. I’m not aware of a formal term that is used for describing this diagram within academic circles. Therefore, to avoid confusion, I have labeled the diagram in my own work as a ‘Bar/Timecode/Structural’ diagram (which conveniently abbreviates to ‘BTS’
– side note: I’m not in any way affiliated with BTS, I’m just trying to exercise a very poor sense of humour!)
(fig A – BTS Diagram used during my analysis of ‘Oppa’ by Brother Su)
Why are BTS diagrams useful?
These diagrams can be used as a quick and precise referential device when discussing a specific aspect or section of a song. They are easy to read and understand and they are accessible for most readers (regardless of their level of musical knowledge/ability).
The diagrams also assist in identifying symmetrical/asymmetrical trends in a song’s composition or indeed between multiple different songs in a genre. Think of it as a non-musical ‘map’ of a song.
What you will need to create a BTS diagram
- A Pen/Pencil
- A song to analyse
- (optional) freeware audio software such as ‘Audacity’ to pinpoint the exact timecodes for section changes.
- (optional) A Metronome App for approximating BPM (I recommend ‘Pro Metronome’ which is free on iOS and Android)
- (optional) a musical instrument
How to create a BTS diagram
I generally draft BTS diagrams by hand (as can be seen in fig B below) as a first step when listening to a new song. The process is quick – it normally only takes a couple of plays of a track to map it out entirely. I would recommend this process as a great exercise for anyone trying to improve their listening skills as you are effectively forced to ‘map out’ what you hear:
Once you’ve chosen a song for analysis, listen to it from beginning to end. Do not write any detailed notes at this early stage, but instead make a conscious effort to establish how the beats of the song are ‘grouped’ (so, if you tap your foot or pencil along to the underlying ‘pulse’ of the song, do you find yourself naturally tapping along to groups of 2,3,4 or more beats? Does that grouping ever change during the song?).
Providing you are confidently able to detect the underlying rhythmic pulse of a song and figure out the groupings of those beats, you can thereby determine that the song/extract of a song that you are listening to has a time signature of ‘X’ number of ‘beats’ in a bar (the ‘bar’ is, in other words, the ‘group’ of beats which we previously discussed). From that, you should be able to figure out how many bars are in a particular section of a song simply by listening to a passage and counting the beats in your head whilst you listen.
In theory, providing there are no unexpected complications, you should by this stage be able to roughly map the different sections of the song (I would recommend labeling these as ‘intro’, ‘verse’, ‘bridge’, ‘chorus’, ‘instrumental’ etc) and noting down the corresponding number of bars for each section. Now all that remains is to add the time code for the start points of each section (this is by far the easiest part of process).
(fig B – handwritten draft BTS diagram used during my analysis of
Areas of difficulty:
There are 2 significant obstacles to be mindful of – how does one confidently know if the beats picked up from the underlying rhythmic pulse are grouped into bars of 2, 3 or 4 beats or otherwise? And what happens if a song physically changes the way the beats are grouped, once or perhaps several times throughout the track?
The first challenge is hard to overcome as you are reliant on your ears and instinct to determine which recurring beat in a group is ‘strongest’ and from that determine the gap between this strong beat and the next strong beat – I don’t propose to go into ‘methods of detecting a pulse’ in greater detail here as there is already a plethora of content available online which deals with this matter – a simple Google/YouTube search should set you on the right path.
The second issue isn’t so much a difficulty but rather a factor which listeners should be cautious of. Although my previous articles collectively suggest that there is a trend in k-pop compositions to remain in a consistent time signature throughout the whole song – this is not always the case. Should you come across a change in time signature, you will need to pin-point exactly where it occurred and how many consequent bars it affects. There’s no quick-fix solution for detecting this, it takes practice to pick up and the difficulty of doing so varies depending on what song you’re listening to!
I should also point out before going any further that I am explaining perception of time signatures in a rather elementary fashion and have not taken into account the liability of your perception of a simple time signature being unfairly manipulated by syncopated rhythms (e.g. mistakenly recognising a 2/4 time signature when you are actually hearing 2x dotted crotchet (quarter note) beats underlying what is really a 6/8 time signature) – again, experience and practice are key to picking up on this.
Creating a BTS diagram – visual presentation
After completing the BTS diagram, I then neaten everything out using Paint X on mac (although Microsoft Paint is just as good) and fetch some additional information specific to the track:
Working out the…
I tend to use ‘Pro Metronome’, an app on iOS and Android to determine the metronome mark/bpm. The app is free to install and features a ‘tap’ button on the bottom-right of the screen, which (as long as the app has been set up correctly) allows you to closely approximate the bpm by simply tapping along with the song whilst listening to it. Of course, the app does not provide 100% accuracy due to the fractional delay in hearing the beat from the song’s audio to physically tapping it into an app in real-time. There is a human element to this process and naturally a margin for error!
– That said, I have always found the programmed bpm to be extremely close to the original track and I take measures to satisfy this when playing back my score transcriptions on the notation software that I use (more about ‘notation software’ another time).
I was extremely fortunate to be born with perfect pitch. Consequently, working out key signatures and transcribing audio has always been relatively instinctive. However I appreciate that this is not a universal approach. The only method I can suggest is to try matching pitches by physically playing along to the song on your preferred instrument and using your instrument as a tool to ‘figure out’ the scale and key signature the song relates to. The approach seems rather long-winded though and again, it may be worth another search online for tips.
Why is it useful to know what the key signature is at this stage anyway? This knowledge serves as a foundation for transcribing the song if you want to make an arrangement of the song, or if you want to analyse a particular type of musical progression that you enjoyed so you can implement something similar in your own work in the future.
Knowledge is power… on the provision that you obtain it with an understanding of its function. Otherwise, it is seemingly an unharnessed embellishment. [memus989: 2017]
Although this article does not teach how to detect a rhythmic pulse or coordinate perceived beats into groups of bars, I anticipate that it details an appropriate use for such knowledge once you have figured it out(!) and (if you’ve made it this far) I hope applying it to something like the BTS diagram is of interest or relevance to you!