Writing (Video Game) Music: The Structure of a Song

- Writing Music: The Structure of a Song -

A flood of use(less) thoughts in the early morning
by Francesco D'Andrea, composer & sound designer 

We all know that having a strong hook, a singable melody or a solid rhythmical pattern is essential to forge a memorable song or soundtrack piece, but what about structure?

As we're in the process of making a track, we should set our mind free and unleash ideas fresh as they come out of our minds. This means experiment, write down ideas without having to worry to much about the piece structure, develop the theme across the song in several ways to enforce it, but subsequently this may cause a scenario in which, at the end of the work process, we'll have a bunch of good and bad "parts", or maybe only good ones, which independently from their single value may stick together and follow each other in two ways: in a good one, or badly. And that depends on structure.

After writing all the stuff, all the melodies, riffs, counterpoint, choruses and verses, and even once we're satisfied with the final result - and we think we've got an astonishing piece of music - we should not rush towards the final steps of mix and mastering! 

Before to get through these conclusive steps, why don't we let the piece "rest" for a day or a couple? It could be extremely beneficial to its final quality. 
We’re proud of the work we’ve done, we went mad about panning, orchestration, arrangements, EQ, and worked and re-worked it for days...so we’re probably impatient to “close” this track forever in its definitive version, upload it on soundcloud or bandcamp and post it to our friends, right? 
But what if, one or two days later, after “forgetting” that song and listening to other music (or not listening to anything at all for our ears’ sake!) we’ll discover that what we did is freaking good, but with a (more or less) simple change of structure could be 200% better?? 

It happens to me all times, well....not all the times, but often: at a certain point, listening again for the 100th time to a rock ballad or an orchestral piece I did and I was proud of, there’s something annoying in that toooo looong bridge passage or that chorus that repeats for the 20th time, or that interlude which is good in itself, but completely out from the rest of the composition’s timber, mood or tempo?I 

So, what are factors that could help us to consider choises to make when doing the final “cut and paste” work? 
In music, we’d say, there are no rules, so what follows is just some considerations deriving from my sweat, modest experience:

1 - What listeners expect - VS - What you want
Let’s say that we’re listening to a Madonna pop song. Or to a renaissance classical piece. 
We, as listeners rather than as composers, would expect a structure which is the most simple and linear as possible. But why is that? What does it mean? 
Our modern cognition of a linear, simple and effective structure is the result of at least two components: logic, and habits.
Logic: listening to music can be compared to listening to a talk. If President of United States of America speaks in tv about taxes increase due to crisis, then he skips talking about her mother’s boobs without making a logical connection, wouldn’t you be scared a little bit, or surprised? You would choke with your hamburger sitting in front of TV, or at least you would regret that vote you gave to him, maybe? 
Well, in music it could be translated as: if you move from a sweet part of flute and violins with pizzicato cellos to a punching metal riff played in sixteenths on cymbals and thai percussions, your next listener will probably close your track in that point while gladly filling the air in his room with the words “WHATTA MOTHERFUK!!”. 

Well, you as composer may reply in two ways: 
1 - Sorry man, I was really tired when doing that soundtrack, and I was already near close to my client’s deadline, maybe I had too much coffee, but I’ll change that part, be sure, it will be better and more “linear”.
2 - Shut the fuck up man! I know usually songs don’t do that, but there’s no rules, I’m an artist and this is my piece, and I want it to sound in this damn way! I want to sound different!

There’s no correct or wrong answer! Both may be valuable choises, unless you’re doing a work for a client and have to fulfill his/her guidelines because his/her is fulfilling your pickpocket.
But generally speaking, make sure that after a tenth listening, the structure is good enough at least to your ears.

2 - Accept criticisms
Have some trusted ears to listen your stuff and remember other’s observations for the future: even if they suck compared to what you think now, they could be useful informations to use for next works.
I use to have some people to listen to my works because I trust their judjments, their ears, and because they’re hard to please. I even let them listen to the work in progress as they’re developing, if I think they can focus on the ideas I propose on the run, and not to the actual “non definitive” sound achieved.
Having someone to criticise you can be an important constant in your musical growth. It’s also useful to understand why people like a certain thing or not, and if you have an indomitable self-exteem, they could help you to keep your feet on the ground and remember yourself that while you’re a brilliant musician, you may do better sometimes.
Anyway, just remember that they’re there to propose and discuss, don’t let their advices become your music. You’re the writer, you’ll discern bad observations from good ones!

3 - Does that melody passage enroll itself enough?
I do that mistake continuously. I’m caught in the process of delivering new ideas to the track, and sometimes they’re so many, if I’m prolific, that those ideas cross their legs one over the other, and what I have in the end is a track that doesn’t breathe enough!
What to do? We all know that less is more sometimes. And yes, this is not paramount, you can disagree with this sometimes, or if you want intricate, complex arrangements. Which is, we have to say, fascinating nonetheless, sometimes. 
But another thing, other than being complex and make crowded arrangements, is to let a music sentence develops itself.
We’re scoring an adventure theme, and after a beautiful, quite long intro....here it comes! The main theme! We want to give it at least 8 bars to let “him” express and say what he should say, right? 4 bars could be not enough, some other times we just need to extend it to 8 bars, and then develop it in other 8 with different orchestration and flavour, right? 
What I mean is: let the track have as many different parts you want, but give them all their needed time to express. Otherwise, you’ll miss both tunes’ strenght, proper structure, and will weaken your song, no matter how many good ideas you throw in it!


4 - Progressive/additive structure or not?
As a music consumer and listener, I notice that the most of times, rock, orchestral, classical, funk, electronic music, whatever....have a progressive and additive structure.
What is it? Listen to “Billie Jean” by Michael Jackson as an example. 
We have a 4/4 drum beat, then we add the memorable bass line, then synth/keys are added, then finally comes, Michael’s voice. That’s a popular, simple and effective example on how adding instruments on a pair count basis works well. Each few pair beats, we add something. And nowadays, in most of cases, we’re used to that, we expect that, and thus we’re pleased to hear that. This approach, also, helps to have a repetitive approach. While other instruments come over, the first one we heard is still there and keeps playing since some bars, and we’re already getting used to it, we are helped to acquaint ourselves to that musical bite, it doesn’t matter how simple, good, bad, weird, innovative or common it is.
This is a good approach because it’s, in most of cases, instinctive. Like music is: when we compose, we start from a single idea in most of cases, and then what? We’ll have another idea that we will choose because it is overlapping the first one in a way that it works. And here we have additions, in a progressive way. 

Do we have to follow this approach? Yes, and no. Of course. 
Everybody knows that there’s not a rule. If we’re running out of musical instinct and genius, it can be an help. Because it’s a linear, organized way to write down stuff.
Otherwise, we don’t have, we’re not forced to start from one item, and then grow with more and more items added. If you believe you’re a classic example of “additive, linear composer”, you may, if you like, experiment with your talent and new possibilities, and try to move away from this approach, for a while, just to see what happens. 
You may find that a more unpredictable strategy to link parts and instruments could be annoying, or refreshing to your ears. Give yourself the chance to experiment new ways just to have the possibility to understand the difference, and then make your choice in complete freedom. 
Billie Jean is a great, insane, memorable track. Hands down. On the other side, we have Holst, Beethoven, Williams, or Hendrix that certainly didn’t embrace the additive approach for all that they wrote. Sometimes you just unleash yourself starting a track with 7 instruments playing in octave, and then want a solo instrument to break that crowded ensamble just few seconds after, and then “grow back” to a super-multi-ensamble section for a soundshake galore!

5 - Watch your Wawe-form!
Charles Bukowski, my favourite writer and poet, drunk character of this century, testosterone-driven artist, often presumptuous and insecure at the same time, once said (I’ll try to quote by heart): “They all believe to be good writers, they crowd my desk will all their poetry collections, but when I just watch them on the paper, they look boring, they don’t have bravery, and they don’t have boldness”
Mr Bukowski talks about “watching at the paper” before even to read the poem.
What if he was a musician? He would have probably said: “Their waveforms are lacking bravery, they don’t have boldness” ?
Waweforms can be boring, and sometimes they’re just like music.
Given that this shouldn’t be the primary way of judgement of a composition, it could be useful nonetheless to make some conclusions. 
I noticed this also thanks to soundcloud, which notoriously shows you waveforms of the songs. 
Looking at your track’s waveform in your favourite DAW may lead you to some conclusions (or not, that depends on your sensibility and personal taste):

Maybe the track lacks of dynamics? What could be described as lack of courage, in a track, it it means something to you, if you care about it? 
One thing we should avoid, in my opinion, is to have a beautiful track which is all beauty and no heart. This translates in a pleasant melody, but with an overall flat sound. No change in “punchness” and “played volume”, or tempo, or timbers and orchestration.
Maybe you encountered this same problem in the last song you wrote. It is good, but there’s not enough crescendo parts, or dynamics, or for example, that part of the song would be better if played without a compressor, because it is moving continuously from a calm to a more angry king of playing? It is also good, expecially in longer compositions, to have raging sections, and moments of rest. 
Contrariwise: you may notice that after so much calm, it’s a bit annoying to have that sudden strings and brass sections rips to break that atmosphere without something to introduce them. 
Well, waveform could be just an additional help to notice these big and little things. Bukowski would thank us for being so “interesting at first glance”!

6 - The Blues Theorema
Bluesmen they do this all the time: they find a riff, something good, that works: and they stick with it. And I mean, they are right.
You may do the same. Regardless of your musical genre, if you found something memorable, keep going with it until you’re sure that the identified, precious pearl inside your composition stands out against the crowd of notes. If you have many hooks, make a choice and give more space to the best one. This will help listeners to endear to something, which may be better than leaving him/her with the doubt of what to care for, in a song. Okay, this is something similar to what I said before, but repetition, that’s a key of success! 

7 - Lenght
One minute may be too much, or not. 
Three minutes and 46 seconds could be not enough. One hour and 41 minutes could...
Okay, you got it: Take care of the track lenght. 
Sometimes I write something, and I like it so much, and I’m so proud about it that I’ll have a smile on my face for all that blessed day. It does happen often, because I am a music kid and I like my toys (read: instruments, real ones or VST). Even if I do really good things, it happens just once in a while. But when I’m feeling that I’m doing a badass track, I tend to be afraid to finish it too early. Or sometimes I am afraid to continue it, because I don’t want to screw what is, in my opinion, so good so far. 
What should be the right moment of conclusion? The perfect lenght that leaves you satisfied as a listener, and well fed up as a composer? 
There’s no answer to this! Gne gne gne! 
Seriously: I just discover it when it’s too late, sometimes. So, I would find myself to open an old Cubase project which I thought I finished 2 years before, just to rework it, because I realized that what I thought to be a song, it was actually only the first part! 
Shame on me, maybe that’s just because I wanted to keep producing and working on a track that was actually already complete, but hey, that’s what I felt!
But also in this case. Before to say “it’s finished”, or “it’s too long”, or “too short”, take some pause. Take a walk, hang out with friends, make sex or go to the catholic camp meeting (?!) and when you really start to forget how your track is precisely, get back to listen to it. You may find out that it’s too short, or incredibly long and boring as a musical rendition of Fedor Dostoevskij’s “Crime and Punishment”.

8 - Don’t listen to what I said
If you had chance to listen some of my works (http://francescodandrea.bandcamp.com/ ) and you liked them a lot, or even find them a source of some inspiration, well, forget a good part of what I said, because all I did here was to write a sum up of all mistakes that I did, and I still do sometimes. I’m trying to grow as a musician and composer, like you’re probably doing, and some of this issues mentioned are still bothering me! I became aware of certain limits or weaknesses in my songwriting, but I’m still trying to overcome them and translate my musical ideas in a better way. 

If you’re reading this line after all the previous, thanks a lot! Woke up this morning with a need to write something, so I hope you found this useful in some way. 
This “article” or whatever it is, could be even more useful if you share your ideas, if you’re a songwriter, a composer, or even just a listener. We do need also their advices, as they’re the ones we’re supposed to delight with what we do!

- Francesco

--
Francesco D’Andrea is an italian soundtrack composer, song-writer and sound designer. He’s currently working as freelancer in France from his private recording facility, and scores music for cinema, videogames, cartoons and advertising. You can listen his music at: http://francescodandrea.bandcamp.com/