My approach to music composition

A couple times a year, I’m asked something about my approach to composing music, such as how I learned or what books or websites I might suggest.

So what follows is a somewhat unorganized info-dump about my approach to composing music. I certainly don’t claim to be a musical expert or composing professional, or even very experienced. I know I have much to learn. But since people have asked, I know there’s at least a few people out there interested…

How I started composing

If you think about it, music, like talking, is something the human brain just starts “doing” naturally. When I was five or six years old, I enjoyed hauling around a tape recorder and improvising really awful annoying-sounding songs. A lot of children will improvise songs naturally. Of course, the structure of these improvised songs and melodies are sloppy and unrefined, but the point is that the seed of musical composition is an automatic subconscious process. It may be that some brains have more of a natural tendency to improvise music and rhythms than others, but I think the ability is there for all, as everyone has the natural ability to perceive music in the first place. The ability to improvise, then, is just the formation of a feedback loop. Anyway, I reckon if you remain in tune with that part of your brain (or perhaps try and develop it), you’ll have a sort of “musical engine” from which to work and gather inspiration from. With this, you can hear a melodic phrase and automatically “hear” in your head a possible way it might continue.

Even with this ability, which is really no more profound than being able to creatively construct a coherent sentence without thinking, getting that music out of the head and onto paper (or a computer, in my case), can still be very difficult, as it’s an entirely different process that usually requires much more thinking. But everyone is naturally musical, whether or not they develop the skill of composing, because music is a natural human phenomenon.

In fifth grade, some very basic music-reading lessons were part of the music curriculum. I remember the music teacher explaining whole notes and half notes and quarter notes, etc., using the example of an apple that continually gets cut in half. I was good at math, so this felt intuitive to me. And then of course there were the rhymes, like: “If you want to know the lines, just remember Every Good Boy Does Fine. If you want to know the space, just remember it spells FACE.” I never actually use these rhymes, though, because when I’m composing I really don’t care about the letter of a note. Anyway, a foundation for being able to read music was thus built. (On a side note, it was around this time that I decided I’d like to learn to play the keyboard proficiently. The idea of being able to simply play whatever music I wanted, to be in control of the music, really appealed to me. But after realizing how much practice it took, I didn’t care that much, and I still can’t do much more than pluck out a melody with a few fingers. Maybe someday… but probably not. I’m too interested in other things.)

When I was eleven or twelve years old, we got a digital keyboard that you could connect to a computer and record MIDI. I was interested in writing music for video games I would never finish programming (another long-standing hobby of mine). I had no experience playing the keyboard, save for those very basic music-reading lessons, but I didn’t care about that. So I improvised several horrible pieces, recording each instrument in turn, having no clue what I was doing. Here’s one of those awful pieces: The Toy. (1997) Wow, so beautiful.

I remember listening to music and thinking that the key to great-sounding music was that more than one note at a time were playing. “Harmony” was still a vague musical notion to me, and I had no clue what chords were, but just by fooling around on a keyboard, I began to tease out the basics of triads. So around the time I was in high school, I would write music by first coming up with a chord progression (though I didn’t know that terminology at the time) through trial and error, and then writing a melody for it. I kept everything to white keys only (everything in C major!) and found three notes that sounded good together (which will naturally result in all C major triads… except the diminished one… because that’s ugly… you will not find it in any of my early pieces). What made a melody sound good was simple: At any given time, it has notes that are also in the chord. Beyond that, you need only to listen to that natural improviser in your head and use trial and error. Where does it feel like the melody should go? Can’t decide? Well, how about this? How about that? Ah, I’m sensing it should go here… Rather than recording my keyboard performances, I began clicking notes into a MIDI sequencer, Microsoft’s DirectMusic Producer (which was for creating interactive music for video games and such; it was pretty cool, though I only used it for its MIDI sequencing capabilities). My earliest piece from those days: The Workshop. (2002)

So that was my method: find chord progressions that sounds good, write melodies for them, and vary the orchestration in different ways, all through trial and error.

A little later, I had a friend in the high school chess club who, in addition to always beating me in chess, was a genius prodigy piano player and a composer in his own right. Amazingly, he just happened to know what triads were. I showed him some of my work on the way to a chess tournament. He pointed out that in my piece Flight of the Dragon (2003) I used the chord progression from Pachelbel’s Canon. (At about the 2-minute mark. Gah, listen to that stupid stereo effect. Why did I find that interesting?) So just through trial and error, I had teased out Pachelbel’s popular chords! Although at the time, my response was something like, “Chords? What is this ‘chords’ you speak of? What arcane mystic secret is this?” And for the rest of the ride to the tournament, I got a personal lesson on the basics of triads: the difference between major and minor triads, writing them as roman numerals, the basic flow chart of what chords tend to lead to what chords, what inversions were, etc. While it took a bit of time to digest the new names for things (that diminished chord in particular baffled me), it all made sense, as it gave context to everything I had already been doing through trial and error! Woohoo! And thus my journey into exploring music theory began.

So that’s my long-winded little history of my composing origins. I’m not sure that little lesson on the basics of triads would have meant so much to me if I did not already have actual experience with them. So my biggest piece of advise for anyone who wants to try composing is this:

There is no substitute for practice and experimentation. You will learn the most by doing, by trying.

At least, that’s my opinion from my experience, for what it’s worth. And I think it applies to any creative act. I once asked a musician friend in college whether he was interested in composing his own works. He said something like: “I’m waiting until I learn how.” To which I thought: Then you will never learn. (Of course, I think the answer was really just ‘no’, but he didn’t want to say that.)

I’ve heard a similar anecdote about Mozart. It’s just a story, but it goes like this: A young composer asked Mozart, “Herr Mozart, I’d like to write a symphony. How shall I go about it?” Mozart smiled and said, “A symphony is a pretty big undertaking. Might I suggest starting with something small, like a sonatina?” “But,” the young composer replied, “you were writing symphonies when you were a child!” To which Mozart replied, “Yes, but I didn’t ask anyone how.”

Anyway, my approach to composing hasn’t actually changed much since those early days; I still think in terms of chord progressions and melodies, and my style is very song-like. My musical “vocabulary” has expanded a bit, and my overall structures are less repetitive (I still haven’t tackled anything as large in scope as a symphony), but it’s still just chords and melodies. This can work fine, as all music can ultimately be understood in terms of chords and melodies, presented and orchestrated in different ways. (Digression: Some might argue that atonal music does not employ the use of chords and melodies. But of course atonal music is not music at all. It’s just crap. Yeah, that’s right, I said it. If someone plays atonal music in the forest and nobody hears it, does it sound good?)

Random composing exercises to try

About a year ago, someone on YouTube asked for advice about how to expand his composing skills, and among other things I suggested these exercises which have helped me, based on my basic chord-and-melody approach:

1) Compose a chord progression first, then compose several different melodies on top of it.

2) Take the chord progression from a popular song and compose new melodies for it.

3) (In case you haven’t already) Study music theory, learn the “lingo” and study how classical composers did things (this is of course an ongoing process… I’m only now beginning to understand how classical composers used augmented 6th chords, for example; just a few years ago it seemed like magic) (more on this in a bit)

4) Write multiple possible chord progressions for a single melody and try to figure why one sounds better (or at least different) than another.

5) Take a classical score (or MIDI) and just try changing things around. Change a melody here, shift some chords there, mute some instruments here, and just listen to the affect your changes create.

6) With a classical MIDI piece, listen to various instruments muted, and then on their own, and see how they contribute to the whole. And then try doing something similar in a piece of your own.

I saw the movie Amadeus for the first time as a freshman in high school which introduced me to Mozart. I was already interested in orchestras because of film scores (John Williams, woooo!), but Amadeus introduced me to classical music (and how similar to film scores it is). Anyway, if you haven’t seen it, there’s a scene near the end of Amadeus when Mozart dictates a composition from his death bed, and as he does we can hear each instrument individually, and then altogether. That scene had a profound influence on me because it broke the music down into parts, and while I had no idea what stuff like “tonic and dominant” were at the time, I remember thinking, “oh, I could do that!” (Not right away obviously, but I mean, I realized I could learn to do it by breaking things down into understandable pieces.)

Studying music theory

(This is all just my point of view from my experience.) In terms of the art of composing, I have found music theory books to be of no practical direct help at all. That is, they are not instruction manuals on how to compose, or even how to get started. They provide no framework for the actual craft. That’s not to say that they’re useless; they can certainly help. But they only help indirectly. They can help you to analyse and understand music in ways that you can then apply to your own compositions, but they’re not going to guide you through that process; you kind of have to figure that out yourself, for better or worse.

In this sense, there are no “rules.” You might say there are “guidelines”, but I’m not sure I’d even say that. Rather, there’s an analysis of common elements we tend to find again and again in the wide landscape of music, and music theory books survey the landscape, attempting to organize and give names to these elements and attempting to make sense of how they relate to one another and how they affect a composition. (For example, it’s not that parallel fifths are necessarily bad, it’s that notes that are a perfect fifth apart will tend to blend together, so if they move in parallel you’ll lose sense of their independence. Whether or not that’s bad depends on whether or not you care about keeping the voices independent.) Music theorists will sometimes disagree with each other. (Though probably not about the basics.) Really, I think music theory is an art in and of itself. It’s sort of an open subject, and you’re free to have your own opinions about why something works or doesn’t, or whether or not a particular way of analyzing something makes sense. It’s not like math, in which statements can be proven or unproven. After all, at the end of the day, all that matters is whether or not you’re pleased with the sound of your music. “If it sounds good, it is good.”

(For instance, a lot of music theory books will differentiate between a major key and its relative minor. I prefer to just keep my analysis in major mode with an understanding that the vi chord can act as a tonic.)

So, while I certainly recommend studying music theory, I’d also recommend not thinking of it as a substitute to actually getting your hands dirty and doing some exercises and experiments on your own. Without some actual experience with composing, music theory will lack any useful context anyway. You can become a brilliant theorist, but a lousy composer, and vice-versa. Music theory and music composition are certainly intertwined, but they’re ultimately two different disciplines.

What music theory should you study? Well, whatever you’re interested in, really.

(By the way, as you’ll see from my history, aside from plucking out notes on a keyboard, my musical foundation includes no instrument playing at all. I use the computer for my musical needs. So I don’t know what sort of music theory comes standard in an instrumentalist’s lessons beyond scales and reading sheet music (and chords for instruments that can play them), as that’s simply not my experience.)

You’d probably want to start with the basics, as reading a chapter about diminished seventh chords is going to confuse you if you don’t know what diminished and seventh chords are. So an intro to music theory will do; you just have to find the resources that explain things in a way you’re comfortable with. An “Idiot’s Guide” or a “For Dummies” book might work. (I don’t mean for that to sound insulting; I think those are actually quite good series.) Nowadays a simple Google search will serve you plenty of results, and there are plenty of YouTube lessons out there. My aforementioned friend in high school taught me a lot of the basics, but I also remember browsing websites and The Complete Idiot’s Guide to Music Theory from the local library. I would say the basics would include:

• How to read sheet music
• Musical scales and keys (what notes are in, say, F major?)
• Chord basics
  • Constructing chords (major, minor, diminished, augmented, sevenths, etc.)
  • Notating chords
  • Inverting chords
• Tonic and dominant (and scale degrees in general for that matter, but tonic and dominant are special)
• Cadences (authentic, plagal, deceptive, etc.)
• The film Amadeus… just because it’s the best movie ever

Really, Googling the subject or reading a book on the subject will present you with the basics; they’re well established.

Again, this doesn’t mean you have to be proficient in even the basics to begin composing; I certainly wasn’t at all. I’ll say it again: I think it’s a lot easier to learn the basics when you have some experience, however simple it may be, to apply it to. Compose a bit first, get your hands dirty, then explore theory.

Beyond the basics, it’s really up to you what to study, if you even want to continue studying music theory at all. Personally, I think I have most of the basics down. (Though I’ll admit I don’t have everything necessarily memorized… quick, what key has four flats? I’d have to go look it up, or visualize a keyboard for a moment. An instrumentalist might have that memorized, but composing-wise, it really doesn’t matter. The answer is, ah… A-flat major.) I usually study harmony nowadays; I’d like to get better at voice-leading and counterpoint, and using chromatic chords, and using more inverted chords to spice up my bass lines, which are a bit boring at the moment.

Some books I’ve found useful

I have read none of these books cover to cover, or even studied any of them religiously; rather, I review a chapter now and then, pondering what it has to offer, and perhaps trying a few new things out in a composition or sketch. As I said before, I don’t claim to be an expert in any of this. Heck, that’s why I open these books in the first place! So here are some books I’ve found useful and/or am continuing to study.

Tchaikovsky actually wrote a nice succinct book on harmony called Guide to the Practical Study of Harmony. It won’t help you compose like Tchaikovsky, though, so don’t expect that. I was hoping he might have some insights into voice-leading, but he actually writes: “Theory can supply but a very general insight into the nature of the voices. It is for the students–provided he is sufficiently talented–gradually to learn and appreciate by diligent application all the finer points and peculiarities of free-voice leading — matters which cannot be laid down in formula.” Gah! Thanks a lot, Tchaikovsky! Still, I find it a surprisingly well-written guide, and I still have more to learn from its pages.

Tonal Harmony is a standard textbook on the subject. I bought an older edition used for far cheaper than the list price. I still refer to it now and then.

I don’t mind admitting that I think Arnold Schoenberg was a lousy composer with silly composing goals, but I have found his music theory books to be very interesting. He can get wordy though, with lengthy paragraphs that sometimes seem to digress into philosophy and such. Which can be interesting, but not necessarily directly practical. He’s got his Theory of Harmony, a beast of book, with long paragraphs aplenty. His Fundamentals of Musical Composition provides more examples and much less writing, and may be the closest book to a practical guide to composing that I’ve come across, as he actually talks about constructing phrases and themes and putting them together for a structured composition. Finally, his Structural Functions of Harmony provides… actually, I’m still slowly studying this book, but I think it provides some insights into the structural functions of harmony or something like that. The first few chapters alone I find to be interesting. For example, he differentiates between a succession and a progression of chords. I think he was onto something that a lot of books on harmony ignore or take for granted: chords do not exist in a vacuum; they indeed serve structural functions that it helps to be aware of.

Audacious Euphony: Chromatic Harmony and the Triad’s Second Nature is a pretty fascinating book. I’ll admit some of it is a bit advanced for me, but basically it’s about chord relations, and how it may make more sense to analyze some progressions not from a “tonal function” point of view, but rather from a… chromatic voice-leading point of view? I’m not even sure how to describe it. (See Neo-Riemannian theory, eg.) I actually did read this one cover to cover, though like I said, some of it was too advanced for me. But it did help me understand and think about chords and their relationships in new ways I hadn’t considered, and I love its various geometric diagrams that map out chordal relations (which you can see examples of on the cover). Quite a fascinating book.

The Study of Counterpoint is considered a classic, but beyond “avoid parallel fifths” I haven’t found much practical application for species counterpoint in general. I’m just not sure how to apply it to the melody-driven style of music I write. I’d love to know exactly how Haydn, Mozart, and Beethoven, who supposedly studied the book, used what they learned from it.

Books on orchestration don’t help me much as I don’t compose for real orchestras; if I was ever hired to work on a film and they planned on recording a real orchestra, we’d have to hire an orchestrator. That said, my favorite book on orchestration is Rimsky-Korsakov’s classic Principles of Orchestration. His practical examples from his own compositions work just fine for digital orchestras as well. I spent time listening to each and every example he offered. (Partly because I was involved in a project in which I animated each and every example with a red line going across the score. What fun!)

A book on a theory of voice-leading would be nice to have, but alas, I have not found one.

Quick digression: I have not studied any books on melody for two reasons: Firstly, melody has always come pretty naturally to me (as it probably does to a lot of composers), so I rarely desired one. Secondly, when I was interested in the subject, I came up with my own theory, which I hope to write a book on myself someday, as I’ve mentioned elsewhere on this blog before. For now, I’d like to expand it into a theory of harmony so that I can write a computer program that will generate Mozartean symphonies. It can be done and it will be done, whether I do it myself or someone else beats me to it. Unless I die, slip into a coma, or remain as broke as I am now (which is a possibility (give me money)), we’ll have it in the next five years, ten at the most.

Another resource: I recorded a few episodes of a podcast on composing music called The Compose Pile. I wanted to make it a regular thing, but it was too much work for no real reward, but might be interesting for anyone who wants to hear me blather on and on about my composing process.

When it comes to actually studying musical scores, I would recommend diving in with a preconceived plan to study something in particular (after familiarizing yourself with some theory) rather than just reviewing the notes another composer chose. Maybe you want to study structure, or harmony, or voice-leading and counterpoint, or orchestration, or melody, etc. Just reviewing the notes without any plans or context might otherwise be a bit useless. (Maybe not; maybe something interesting will pop out at you, but it doesn’t happen often for me.) Unfortunately scores are really designed for easy reading by performers on instruments, not really for composers looking for patterns, so it can take practice and experience. It always takes me way too long to analyze harmonies, and I particularly loathe transposing instruments. Still, it can be good practice. There are plenty of orchestral scores of symphonies and such to buy, which you can find by Googling. The International Music Score Library Project has many public domain scores available for free, and the Digital Mozart Edition offers the scores to most of Mozart’s work. (Because Mozart is awesome.)

Another short digression, but some good books related to the meta-subject of teaching oneself something include: The Genius in All Of Us, Secrets of a Buccaneer Scholar, The Little Book of Talent, and Mastery. Awesome books, especially if you ever have any self-doubt about learning any art. It’s all a process and it all takes practice. There’s really no such thing as “genius.” (Or, at the very least, it’s a label we use to praise others rather than some life-long self-condemnation of forever lacking a skill when what you really lack is motivation. I’m always bit annoyed when people claim to lack a natural talent for something, as if that’s all it ever takes. No, you’re not un-gifted, you’re lazy. Get to work. It’s like that stupid line from Good Will Hunting when Matt Damon’s cursed with awesome character claims that Beethoven and Mozart looked at a piano and “could just play.” It’s a stupid childish romantic view of intelligence. It’s complete nonsense and does a horrible disservice to anyone who wants to learn something new. (Sorry, I really enjoy ranting about this subject.))

Music theory classes?

Should you study music theory in school? Like I said, it’s really a different discipline than composing. Anyway, it’s up to you. I never did and I’d never want to. I don’t learn well in that sort of environment. I loathe tests and quizzes and assignments. Some people like that sort of structure, but personally it just makes me stressed, and turns the subject to be studied into a chore. I much prefer to guide myself. (That said, if I could afford a personal tutor, I wouldn’t mind that, but I would probably fail miserably in a classroom setting.) At the very least, I would say it’s definitely not something you need. You can certainly learn to compose without ever studying music “formally.” After all, in the end, composing is a skill you develop, not a set of formulas from a textbook, and that development is going to happen outside of the classroom and over many years anyway. Even if you take a “formal” class, there’s really is no such thing as studying any art “formally”, because you will have to develop the skill on your own, through your own efforts.

Collecting melodies

Sometimes melodies flow from my subconscious more easily than other times. Usually after I’ve been composing for a couple of hours, I’ll be able to hear that subconscious improviser more easily than when I first begin. So to help me get started writing new pieces, I usually begin with a melody that I already have written. (Not always; with a bit fooling around, I can always compose a new melody from scratch, thank you very much, but a melody already stashed away can save time at the start.) So when I’m not composing, I stash away melodies that either pop into my head randomly or that I pluck out on my keyboard with a bit of fooling around. I have a keyboard beside my computer desk (yeah, I’m not professional enough to have a keyboard actually on my desk, as I click in my notes), which helps establish the habit of fooling around with melodies now and then. To save melodies, I usually use my smartphone, a super handy device for this sort of thing. I either record a video of my hand playing the melody, or I use an Android app called Musical Pro to record a little MIDI file. (Unfortunately I don’t see the app available anymore… gah, who does that, releases an Android app and then removes it? *cough*)

So if you have a quick and easy way to record melodies or phrases that pop into your head at random times, and/or make it habit to sit down and play around on an instrument for ten minutes a day, I think it’s a great habit to get into and can give you something to work with when you sit down to compose.

(By the way, though I mention melodies “flowing from my subconscious”, I don’t necessarily mean to suggest that they flow out fully-formed and perfect. Sometimes they actually do, but many times I will continually tweak a melody over and over again to get it just right, to find something that not only sounds pleasing and natural, but something that really elicits an emotion from me. Sometimes I’ll have to take a break and walk away from it for an hour or so and come back to it with fresh ears. Sometimes I’ll even scrap the whole melody and try something completely different. It’s very easy to write a dull melody, one that technically works but just doesn’t pull me in. (Though what makes a melody sound ‘dull’ is of course subjective.) I try never to settle for dull. I want something that will really evoke a feeling from my gut, even if no one else will really hear it but me, something that I will want to listen to again and again, even if I don’t have the ability to breathe the life into it using my computer that a live performer would. So this can mean a lot of tweaking and trial and error, over and over. There’s no substitute for it. It’s part of the process.)

Digital orchestration

Software that I use:

I still use Overture 4 for composing and rendering my pieces. Looks like Overture 5 is out now, but I haven’t bought it yet. Can’t afford it at the moment, but it looks nice. Anyway, I’ve found Overture 4 to be fantastic at combining the ability to compose using notation, yet retaining the ability to edit lots of MIDI data. I love it.

I don’t use that many sample libraries; they’re just way too expensive for me, and I’m too pure of heart for pirating. (By the way, if you hope to make lots of money, composing digital orchestra music is not a great source of income. At all. (Give me money.)) For sounds, I’ve been using Garritan Personal Orchestra since 2004. Again, it looks like a version 5 is out, while I’m still using version 4… anyway, for it’s price, it’s most certainly the best orchestral sample library out there, in my opinion. I use some of Garritan’s other libraries for other instruments. I also use Anthology: Celtic Wind from time to time for its beautiful whistles.

I tried EastWest’s ComposerCloud for a month and really liked it, but can’t afford it at the moment.

Reaper may also be worth looking into. Looks very nice for its price.

Finally, I think RapidComposer is pretty neat. It’s a bit too expensive, and was a bit too buggy for me when I tried it, but I like the concepts behind it.

I think that’s all… for now…

That’s enough blathering from me for today, isn’t it? I hope this is helpful or interesting to some people out there! I guess for me it mainly comes down to three things: Chords, melodies, experiment. Lots of other little things to consider, but those are the main elements of my approach to the craft. Find some chord progressions, write some melodies, and continually practice and experiment with orchestrating them.

Good luck!

Let us end with a prayer… “Lord, make me a great composer. Let me celebrate Your glory through music, and be celebrated myself. Make me famous through the world, dear God. Make me immortal. After I die, let people speak my name forever with love for what I wrote…” Wait… I don’t think things went so well for that last person who prayed like that… let us not be so vain.

Mediocrities everywhere, I absolve you!