[ back to Tom 7 Radar ]

s
o
n
g

o
f

t
h
e

m
o
n
t
h
Song of the month, July 2007 = Theme from The Goog (16 Apr 2008 at 17:32)
Whoa hooaaa. It's been a long-ass time since I've done a Song of the Month. I am going to do something moderately stupid to jump-start this dead battery: I'm going to write about one of my own songs. I think that's okay because Song of the Month isn't some kind of award that I bestow so there's no like conflict of interest, and even if I did give awards nobody cares about that and I certainly don't care about awards that I give to myself. (Except for: Most Ambivalent About Self-Congratulation. Rock on, dude.) The real point of the Song of the Month series is to give me an excuse to write, especially about inconsequential things and things that are difficult to describe. What do they say, that writing about music is like dancing to Architecture in Helsinki? Is that right?

Anyway just to be clear that this is an out-of-process Song of the Month, I am doing something patently absurd. This entry is backdated to a time before this song even existed. In fact I wrote the song Tuesday night. This here post is basically about the process of songwriting, rather than the song itself, since that is fresh in my mind.

Here is the scourge of songwriting, for me: When I'm writing a song I just wanna write the songs that already exist and that I love. My cold rational mind knows that many of the songs that exist and that I love were written while explicitly avoiding writing songs that already exist, and so I ought not fear mining the outskirts, but fuck if I don't want to just be like Am F C G all day. It is a constant problem. Nels will testify that I am particularly implacable on this point, in that I'll frequently claim when we're writing a song that it's already a song that already exists even though I can't say which one. (Actually, this kind of thinking pervades the way I approach creative processes like research and programming, too.) This wouldn't be a problem—after all, it naturally leads to more original material—except when you get stuck at a point where the next move (note, chord, lyric, etc.) seems oh so right and so natural but you feel you can't do it because it's not original. When I say you I mean me. Maybe you if you write music either just go for it and write that same song, or maybe you don't find yourself in this scenario because you're better at laying out an original plan from the get-go. When I'm nouveau-paralyzed in this kind of situation I often end up doing something pretty awkward sounding, by necessity, and maybe sometimes later it grows on me but usually it's just a failure.

Not that I don't write songs employing formulas, though. For example I can't tell you how many times I have a chord progression like X/Y/Z/W and then the "change" is to something like Y/X/Z. This is my go-to technique when I'm feeling pretty good about the song so far and I want the change to sound natural. It doesn't make me feel like I'm rewriting a song that already exists because if the first part is original seeming then some permutation of it probably is too, except in relation to itself but songs are allowed to be similar to themselves. There are plenty of other formulas. Sometimes I learn a new formula, often by studying someone else's song carefully but sometimes just by luck. This is favorable. When it happens, I happily use it in combination with other formulas until it and they mutually dry out. A good example would be a song called Setlist (I never uploaded it, sorry) in which I first discovered for myself the partial capo. It's true that probably a third of my songs since then are using the partial capo in some way, but the remarkable thing is how many of them relate in some way to the specific structure of Setlist and not just the way the guitar is configured from an objects-attached-to-it standpoint. I also remember having this realization when working on AAD #13, specifically on both Advanced and Codeine Tornado, which was there's nothing forcing me to make the chorus fit neatly into a specific even size filled with repeating chords. I could do like XYZW-XYZW-XYZWVU and sing an extra-long melody over the WVU part that resolved some kind of repetitive uneasiness of the XYZW part that would have sounded really annoying on its own. It's not that I really thought I was forced to do it that first way, but something about the way I usually wrote songs in those days (figure out a guitar part almost blind to its melody potential and then, having recorded it, come up with something to sing atop) led me to avoid aperiodic chord progressions. It is funny to think back to that time because now this seems totally obvious and I do it all the time.

So what are we getting at here? Some kind of new formula or what that Tom is so proud of? Kind of. Let me go back to the scourge of songwriting. I think if you are me, or let's just say I'm me, there's one sort of natural thing that I do to try to be original, which is to start with the thing that my heart desires but then to fuck with it until it no longer sounds obvious. One easy way (especially with the video game music where I don't need to worry about actually performing the song) is to give it a funny rhythm. Many of my songs are made this way, I think, subconsciously. The problem with it is that my heart's desire is sitting there just within my reach and I'm deliberately thwacking my stupid greedy hand away from it, and this feels bad. I don't think it leads to more originality either because the reinforcement is mostly negative. When I put it this way another obvious direction is to start with something that's really fucked up, but then brush its hair until it is a pop song. I've actually done that plenty of times. It's different though when I can be deliberate about it, and being deliberate about it is the formula that is the subject of this post.

I wrote a song called Sensations sometime last year. If I were giving awards to myself I would give it to that song instead, because I'm really happy with the way it turned out, to the point that I put it on repeat and try to study it to figure out why it sounds good to me and how I can extract principles and formulas from it. My conclusions are thus: The song is based around a (rhythmically) fucked up bass line in 21/16 time. Its rhythm is grouped as 5-5-6-5 for the whole song. Layered on top are two melodic lines whose rhythms have the same period (that is, 21) but different phases and groupings. (For example the square wave in the "verse" starts two beats before the measure boundary, and the sawtooth one beat after.) The thing about this is that it sort of smooths over the underlying rhythmic strangeness of the bass, because at any given time there is at least one fairly simple melody part descending or ascending or holding out some chord, and doing so in a way that makes rhythmic sense locally. But my attention keeps drifting from one to another, meaning that there's no particular point I can say where there is an "extra" beat or a "missing" beat. (That can sound good but it is more challenging to try to write something that sounds like it is "naturally" 21/16 time.) The third observation comes from the way I came around to this. I started with something fucked and put some wandering melodies over it until it started to make some sense to me, then I started reinforcing them with other lines, even tweaking the melodies that I had already put down so that they would be less weird, and finally it was done. Very clearly the "brush its hair" approach. You can sort of hear the progression in the introduction to the song itself, which I like. Those are the principles and formulas as best as I can articulate them.

So finally, to the song of the month. I can't believe how long this post is. It is a Tom 7 Entertainment System tune called "Theme from The Goog" sort of by request (but that is a different story) and when I wrote it, I set out to employ the principles learned from Sensations. It begins with a bass line based on the sequence 1-2-3-4-0:
Theme from The Goog bassline
It's just repeating the same notes to introduce the rhythm. It makes no sense, like when they tell someone in chef school as a hazing ritual to make a dessert out of lobster and meringue. Next some other parts are added to "explain" the bass, which then changes in notes (but not rhythm) to be something cheerful and sensible. Even though I am using the "out-of-phase" trick a little bit, I also "reset" each measure or two with a blank spot, which means I have to establish any kind of overlayed smoothing rhythm very quickly rather than rely on a long term drift lasting several measures. Eventually, having demonstrated that this song has not been written before, I can start letting loose with some of those wailing catchy riffs that my heart desires. The only trouble is to match these back up into the original verse so that the song can properly loop (this is critical for video game music, of course, since video game music must be incessant). I got a little lucky here, but helped things along with another formula/principle: I repeat the dying robot motif (series of four descending half-steps from the verse and second part) in a blank slot preceding the loop back, reminding the listener like "Oh yeah, I'm listening to that song still." In the end I am happy with the result, which spans 13/8, 9/8, 4/4, and 11/4 time. It's a bit less thrifty than Sensations and it has some transitions that don't flow perfectly, but it is sprawling and it gets stuck in my head and that makes it a Song of the Month in my book.
c
o
m
m
e
n
t
jcreed (pool-72-95-158-84.pitbpa.east.verizon.net) – 04.18.08 09:49:43
Song from the Goog is excellent stuff.

And boy do I ever know what you mean about Am F C G. Also "Say it Ain't So"/"737 prechorus"/"III-instead-of-iii variant of measures 3,4,5,6 of pachelbel's canon" Am E F C.
c
o
m
m
e
n
t
Cortney (pool-138-88-116-72.res.east.verizon.net) – 04.18.08 20:45:14
Hooray for Song of the Month being back!
c
o
m
m
e
n
t
matus (cse-dhcp-10-158.ucsd.edu) – 05.16.08 21:56:39
theme for goog is incredible, holy wow. having the melody line up with gaps in the bass at first, and then break out, was brilliant.

I think I have listened to it a million times since I downloaded it yesterday (went through the effort of bending space-time JUST to have those 2.6 million minutes instead of 1440).
c
o
m
m
e
n
t
Tom 7 (pokemon.wv.cc.cmu.edu) – 05.17.08 20:03:51
Thanks Matus! I'm glad you liked it. More to come.
c
o
m
m
e
n
t
Tom 7 (h-68-166-207-58.phlapafg.dynamic.covad.net) – 05.18.08 09:48:52
If you can share your time-bending secret, please, it is my greatest wish...
p
o
s
t

a

c
o
m
m
e
n
t
Please enter the code, unless you are spammer!!
[ Tom 7 Radar  •  Tom 7 on Google+  •  on Twitter  •  on Facebook ]