The 'process' of composition

[Tom P]

Hullo all.
I was a little hesitant about posting this, as I wasn’t sure if it would be considered the right place. But then it is rather a ‘general’ question so I guess it’ll be fine here. Anyway...
Music has jumped back into the foreground in my life recently, and I’m currently making a start on my A Level music composition task. I’m writing for a string orchestra with a little bit of percussion (timpani and cymbals), and I feel I’ve made a decent start. But I can’t help but feel rather inexperienced with this whole composing thing – as indeed I am. I’ve read various materials on music writing, including some of the lessons in this forum’s interactive ‘Principles of Orchestration’ (notably the sections for melody and harmony with strings). I’ve also studied four part harmony at school, and I have a good idea of chords and inversions for major and minor keys and suchlike.
But for all that, I find I’ve never been given a great deal of guidance about the process of composing. After all, it’s not something they really teach at school. What I’d really like to know is how people here work; in particular, how do you get from the idea in your head to a finished piece?
I’m aware that a lot of composers write piano sketches before orchestrating their work. For those here that do this, I was wondering how detailed you make these sketches before orchestrating. Do you just write down the melody and a chord structure underneath, or is virtually every note that would be in the final piece written in the piano sketch? If the latter applies, how has each voice been developed from a simple framework of chords?
Some of these questions may be a little awkward to answer without giving me an entire lecture, but all I really want to know is how composers here would describe the process – what happens between the mental idea and the finished piece?
Thanks for your time, and a very happy New Year to you all!
Tom

[qccowboy]

Since finishing my studies with Professor Belkin, my "method" has pretty much solidified and I basically write utilizing the same "process" for most of my pieces.

From a melodic or harmonic element, I will decide a form. Sometimes, I will sketch purely the melodic contour on paper. From that contour, I might highlight certain elements which will be repeated and serve as development later (that process is particularly marked in my "Symphony in C").

Often, I will through-compose large sections of a piece, putting in as much information as I can... to which I generally return once the movement is completed. Upon this completion, I start the trimming process. Here, I cut things that I may have added in the "full fervor" of my compositional passion :)

I also adjust harmony, melody, rhythm... there are always bits and pieces that are not quite to my liking when I first write a piece.

Most times, during the "first draft" phase of my compositional process, I might make the counterpoint leaner, and come back later to either augment and thicken it, or further cut it down in texture.

Composition is a process of construction. The form is its architecture. The melody and harmony are its roofing and siding. The instrumentation might be the additional "decorative" elements of the construction, like the windows, sunroofs, decks, sliding doors, the in-ground pool. And the dynamics and articulations are the furniture and decor.

[reberclark]

In essence, what cue-see said.

[Ken-P]

http://www.wisemanproject.com/education-orchestrationtip-by-KentaroSato.html

[PaulR]

Hullo all.
What I’d really like to know is how people here work; in particular, how do you get from the idea in your head to a finished piece?
[FONT=Calibri][SIZE=3]I’m aware that a lot of composers write piano sketches before orchestrating their work. For those here that do this, I was wondering how detailed you make these sketches before orchestrating.

Well Tom - since I'm a supporter of Kent County Cricket, I would say it depends on how good a keyboard player one is. I don't work on too detailed orchestrations because most of the time I don't write it down. Computers make me very lazy, plus I don't write much anyway. Orchestration is an additive/subtractive process to me. It also depends on the type of music - I would imagine melodic writing and orchestration is a long way off something more atonal or minimalist for instance.

If you're interested in the more melodic style, then I would suggest orchestration may be dependent on what it stands for - i.e. is it happy or sad etc etc. You start with your melodic structures and then you can orchestrate that in thousands of different ways. All I would say is try and be original in orchestration and experiment. Try to be contrapuntal and ornament main melodic statements, Q & A style just as an example.

[Tom P]

Thanks for the ideas, all. Those are some interesting techniques, qccowboy. Sketching melodic shapes on paper is something I've thought about but never really practised. It's certainly worth a try. I've tried some through-composing before now on one or two occasions, though I found it hard to establish a sense of balance and structure (though, to some extent, that may be the nature of the technique). Then again, if it can give me one or two melodic gems which I can lift and use in piece in ternery, rondo, sonata and all other forms, I'll try it out if I haven't got any clear ideas in my head.



Composition is a process of construction. The form is its architecture. The melody and harmony are its roofing and siding. The instrumentation might be the additional "decorative" elements of the construction, like the windows, sunroofs, decks, sliding doors, the in-ground pool. And the dynamics and articulations are the furniture and decor.


I shall remember this forever. :)

Thanks very much for the link, Ken, it looks very informative indeed. It's getting late now, so I'll look at that in greater detail tomorrow.



All I would say is try and be original in orchestration and experiment.


That's some advice I'll try my best to remember. I think some of my problems come from worrying excessively about the rules. Don't get me wrong, I respect them. In fact, I think breaking the rules is a bit like so called 'free-verse' in poetry. It can be done, and it can be done well, but it can also very easily be done badly. Still, I guess my primary goal should be to, as you say, experiment, and if it sounds good and original, then I'll worry about the rules. I don't think they're as restrictive as some people make them out to be, anyway.

Thanks again, everyone. I'll try some of this out tomorrow. :)

[garymosse]

Hi Tom,
In most cases, it's best to have a reason to write any piece.
An elegy, a heroic piece, a piece of sadness, etc.
Start with an idea of what you want to express. Dig into your mind to find each note & the following note. The instrument color should match the mood.
Dark colors-cello, euphonium, low register clarinet, English horn, french horn. They convey emotional, pensive lines.
Bright colors are trumpets, high pitch woodwinds, etc.
These sounds are bright & cheerful/active.
The instrument choice has lines which move up or down, sustain to identify the instruments best register and give the melodic line interest.
After this is about done, come up with an equally strong bass lie which compliments the melody.
Don't add any more parts than absolutely necessary.
There is no one way to do this, but it's how I do it.
Gary

[boulifb]

Hello,

On my side, when I have a melody in my head, I use my keyboard to check the tonality to get the key I'll work my piece. Once I have the tonality, I switch the keyboard off: it won't be needed anymore. Then I start to write the bass line (contrabass) and I try to apply as much as possible the rules of harmony on the bass line (progression chords etc.). Then I write the melody I had in my head above the basses. The melody should be automatically corrected according to the bass line. I use the 1st violins to write my melody. Once both are made (basses & melody) I orchestrate according to the rules of orchestration and to what I have in mind to create the effects I want to do.

I hope this will help you.

Fred.

[Fabio]

The possible processes are of course a lot, all working well, according your skills, your attitude and your target. By the way I think that some common starting point can help a beginner to find his own way:

- you may write without any target (e.g. as if you were improvisating at piano, and simply you take note of the outcome, then you refine it with a little recursive process of re-play, refine, re-play, up to the end).

- you may write with a specific target. actually it helps a lot, and I consider it the best way of starting to understand the process.

The "target" is a kind of pre-designed structure: e.g. the text and the duration of a song, or the exact duration of a video to soundtrack. This is a common actual practice so it's probably easy to understand for every modern composer. But a classical way exist also and is following a structure (a sonata or a fugue were kind of "empty boxes" to fill in...).

If you create your structure before, you may for instance take a sheet of paper and write a kind of little project. how long it is, and a possible amount of bars; the key and the meter, tempo etc.; how many sections you want to use for structuring it; the proportion of the sections.

This first structure can create the "empty boxes" you will fill with music. Then you can start writing or playing keyboard to find ideas, but still in your project sheet you should before go a step forward: you may assign to sectionsa kind of "colour" and understand how hard or soft, slow or fast, flowing or beating, melodic or rythmic you expect the section sounding. It will help a lot to drive your search for ideas, adn will anticipate orchestration ideas too (then you can find during improvisation melodies and rythm already fitting the instrument you imagine you will use in the finished work).

The finalization is creating paper or notation file with all the decided structural elements and put in the musical ideas.
The usual process of "theme development" (usually working as exposition of main idea, followed by kind of answer/development/digression/contrast etc., and finally conclusion with-or-without re-exposition) would be used to fill the number of bars you need for every section.

An important note is that this process is based on recursivity and flexibility, so never be rigid, the opposite frequently happens that writing you will find some idea good enough to push you changing the initial plan, sometime even in a totally unexpected direction, and it's a creative improvement. (the only very rigid "forced" structure is the video duration if you make music for video, but all the rest is flexible enough for dramatic changes work-in-progress:p)

For orchestration of ideas, I think that 2 possible approches are possible:
- you have still a very basic chord+melody perception of music
- you have already a polifonic perception of music and you can write countermelodies and counterpoint.

In the first case don't spend time in refining the "keyboard scretch", move urgently to the orchestral score and fill it like a painter (using sounds as colours): put notes that make chords to the right instruments, and the melody to your favorite solos according the plan, then listen and refine.

In the second case and if you are a fine keyboard player you may look for nice countermelodies during composition, or use counterpoint skills for writing voices and polifony directly during orchestration: this is usually a final stage, not really a beginner starting point, but as soon as you do it, as soon you will improve your "orchestral thinking" (feeling the instruments in the "role", with pedale effects, countermelodies, rythmic effects etc. etc.)

I hope it give some point for experimenting: "structuring" it was the main lesson I learned in classical music composition classes in Conservatorio.
No way of writing large forms without it.

2 unwritten additional rulesa are:
1) use the eraser more than the pencil!
2) and always enjoy!
the bad results are always coming by the lazy attitude of keeping first version or to finish soon becaus you get bored...

Good luck and enjoy!

[wrayer]

Just put a bunch of notes in a bag, shake it up, pull them out and record the notes as they come out of the bag...

ummm, sit for 4' 33" and see what the audience does...

uhmm, put a bunch of trash cans, lids and drumsticks in the audience and tell them to humor themselves...

You see, music is everywhere...

The other answers were much more to the point, I just got attacked by the "John Cage" syndrome.

Bill

[reberclark]

When you need a title consider this GREAT ADVICE on how to title a piece from another member of this forum:

"How to title a piece of music:
Decide on a central unifying concept
Come up with a title reflecting the concept.
Plan out the song to effectively articulate the concept.
Execute the plan with an eye toward realizing the concept.
Realize [you've] screwed up big time and failed grievously at all the above.
Give it a random title involving planets and faeries." - Chris P. Lorán

I really love that post!

[dudefromthebronx]

It always seemed to me that composers instinctively begin with the melody, then they weave the rhythm, harmony and tempo around it. Nothing wrong with that, but I'd be just as inclined to start with rhythm, harmony or tempo and worry about the melody later. Sometimes we need to set the mood first.

Another common process is that most composers will pluck, plunk, blow or bow a tune on an instrument they know how to play. Thanks to soundsets like GPO, we can just put any instrument on solo and come up with inspirations we'd otherwise never would've gotten.

It also helps to look for ideas in the real world. My inspiration for the ostinato pattern in my composition "Afternoon In Central Park" came from squeaky windshield wipers on a taxi I rode home in one rainy evening.*()

[Tom P]

I'd just like to thank everyone once more before I get some much needed sleep. This is all really helpful stuff! The information I've picked up from everyone here has made the process seem much less overwhelming. I've spent the day trying a few things out and have got off to a promising start with my A level piece. I feel privileged to be given advice by experienced composers such as yourselves. You've been a great help - thank you so much. :)

Tom

[dominick]

Some great posts here - and hope this thread has a long life!

[Pingu]

Well I have a slightly odd way of working. Generally I'll start with a very small fragment - maybe a single motif that I find quite powerful - and literally just add what seems to need to happen next. I never know what I'll be doing in two bars' time, never have a clue what the overall structure is likely to be.

I probably ought to learn how to compose properly/more methodically, but I sometimes enjoy my 'method.' Just occasionally it produces something truly powerful, as though my subconscious has been working things out by itself, and all my 'random' decisions seem somehow unified.

On the other hand, producing orchestral material is slow beyond belief. I don't even sketch things out then orchestrate later. I quite literally produce one bar at a time, adding the parts that seem to be missing. Then I play back what I have, and work out where each idea needs to go for the next bar.

[danveach]

I'm usually inspired by melody. My ideas mostly come from playing around with the keyboard late at night, using a string voice for the overall orchestral line and individual instrument voices for solos. (I also sometimes jump up in the morning with a melody in my head that I must write down immediately!)

In a symphony, for instance, once I have a first and second theme and some variations that I like, I write out the complete melody line for the exposition, and then harmonize it (again using string voices).

Now I set up a "sketch pad" in Finale with just string ensemble, cellos for the bass line, and solo winds and brasses. I transfer the whole melody to the string ensemble part and then work out a nice "singing" bass line for the cellos. Then divvy up the melody between the solo parts and strings, and fill in the rest of the harmony.

For the development, which is more harmony-driven, I fragment the melody lines into interesting motifs and take them to far-off harmonic locales, all building up to the triumphant return home of the recapitualtion.

Then all you have to do is pop everything into the full score and write in the second parts, percussion, etc. While you're putting off doing this, you can get started on your second symphony! :rolleyes:

[Tom P]

For the development, which is more harmony-driven, I fragment the melody lines into interesting motifs and take them to far-off harmonic locales, all building up to the triumphant return home of the recapitualtion.

I love those moments, they sound delicious! :)

I'd just like to say thanks once more to everybody here. I think I've got the makings of my very own 'method'. The way I'm planning to do this A Level coursework is this: think of a nice phrase or melody line (not for the entire piece, just one section), write it out on one staff and make a simple piano accompaniment to establish the chord structure (really simple: just chord, chord, chord...). Refine the melody and add ornamentation and suchlike (with instruments in mind: if I know I want horns to play a phrase, I won't cover it with trills). Then enrich the chord structure where appropriate, for example, adding chords with sevenths where the melody implies them. After that, I'll write the outer parts of the harmony followed by the inner parts. Then orchestrate, et violà, a section of my piece.

As this is for A Level coursework, it would be considered pretty dumb of me to post it to the Listening Room, but when it's out the way with, I'll get to work on some of my own stuff. And how refreshing that will be, what with those horrible 'briefs' the exam boards give you. :(

Thanks again, everyone!

Tom