How do you write large-scale music?

So the question for everyone here is, how do you personally approach writing large-scale music. This can be music that is "large" in size, in orchestration, in length, etc, etc. I thought it could be helpful to have a few sources, and since I'm going to be tackling my most ambitious project so far (symphony for small orchestra) I've started looking really analytically at "how" I wrote and how I was most productive. I found it was helpful, and it allowed me to break some habits of laziness when it comes to writing.

So, I'll get started.

I've started by giving myself a sketch, in words, of what I'd like to see from my entire piece. I broke down the movements, and described what I'd like to hear within each movement. Though I doubt the finished product will match completely (or at all) with what I've written, it seems to help give me a sense of direction or "where to start".

After that, I started writing a sketch of what I have the most developed in my head. I'm doing the sketch on manuscript allowing 9 voices, but I've found I'm only using 4-6 for the most part. I've started by writing out the long lines, of what I know I want to hear. Then, I start filling in the blanks. As I'm doing this, I find myself making little notes here and there of possible instrumentations and doublings.

So, what does everyone else do?

Oooh, I just read something relevant to that a couple of hours ago in the book "Musical Composition" (subtitled: 'A small treatise for students') by Stanford (a very good introduction to many aspects of composition, and not incredibly technical either):

"[...] "The greater the number of staves in the score, the fewer the number of ideas," was an astute comment on the situation by a great conductor. The best antidote to this plethora of material is a solid grounding in writing for a small orchestra, such as sufficed for Mozart, and, in the main, for Beethoven. A compose who can produce such colour as Wagner did in his Siegfried Idyll, with the same limited means, will find it far easier to dispose larger forces when he is called upon to command them, and his experience in controlling smaller bodies will lead gradually up to a greater ability in handling army corps. He will know his units and their capabilities, and will not strain them beyond their powers. He will, for example, treat his horn-players as human beings, not as padding to fill up chords, and he will never lose sight of the two most valuable assets of an orchestrator, contrast of colour and economy of material."

This text is public domain (the author died in 1924 - I think the actual hard-copy of the book I am holding right now has a signed dedication dated 1915 :O ). But in any case, here's the source: Stanford, Charles Villiers "Musical Composition: A Short Treatise for Students" (MacMillan and Co, 1911, London), p.173

Also, a good knowledge of the instruments, how they are played and their capabilities/special techniques is very important, because there is nothing players hate most than composers who don't know how to write for their instrument (if you write, for example, an unplayable double-stop on the violin, or if you write a note that does not exist in the range of the piccolo). An orchestration book (I'd suggest either the Adler, the Blatter, or the Piston one - you can also find Rimsky-Korsakov's book on orchestration online) helps largely on that, but it would take a lot of time and effort to try and apply all of these details for each and every instrument in the orchestra right away in an orchestral piece, which is why Stanford suggests you start with smaller ensembles, so you can apply what you know about a family of instruments first (or a few instruments from each family), so you get to know the instruments gradually and in a very natural and comfortable way.

You don't have to play every orchestral instrument, but it helps in assimilating the characteristics of the instruments if you've at least tried to play them. Berlioz only played the guitar, while Hindemith played virtually all the instruments of the orchestra quite decently, and they both wrote greatly for the orchestra :)

But what do I know, I've never written for orchestra :P (well, I have tried once, but it was nothing more than an experiment to get acquainted better with the instruments of the orchestra - which I found is more effective if you're writing for small ensembles, as small as trios and duos, even solo pieces)

Also, for "large" in terms of length, you'd probably need a good knowledge of developing your material (if that's the way you work), which means you could study Strauss's works (he is a master of variation and development, especially in his Don Quixote), or Berg's piano sonata (an incredible piece, which almost completely derives out of three leit-motifs presented in the first page), or read a book, such as Schoenberg's Fundamentals of Musical Composition, which is really a very good book for theme development and deriving form out of smaller phrases (and the same is true for his music as well, whether you take the early works such as his Verklarte Nacht, or the later, dodecaphonic works).

"The greater the number of staves in the score, the fewer the number of ideas,"

I think an even greater instruction would be:

"Don't let your music notation program prevent you from writing what you really want, or mean, to write."

Flute 1 and Flute 2 on a single staff? Boring parts for both. Flute 1 and Flute 2 on their own individual staves? Interesting parts. Why? The limitations of putting two separate lines on a single staff within my notation program was limiting the way I thought about writing their parts.

As far as advice for the OP, plan everything out before you start writing. That way you are not suddenly stuck in a dead end wondering where to go next. There's a thread I started a while back ("How do you compose?") that shows how I plan. I would suggest you check out that thread for additional ideas from other members that could apply.

Sorry, I wasn't so much asking for advice. Thank you both for your comments; they were incredibly helpful. I guess what I was asking for is for other people who currently write in this medium (orchestral/concert band) to explain their approach.

As I've said before, I tend to sketch everything on a few staves. In this sketch, there are notes for instrumentation, but until I actually start writing the piece, I don't want to worry as much about this. I work on manuscript, not because I dislike people who compose with notation software, but because I find my own work to be less interesting or expressive when I just start plunking notes out.

Thanks, Flint, for letting me know about that thread, though. I'll definitely have to check it out.

Well, when I write my pieces, I first think of the number of movements and their styles.

I will use my current symphony as an example. I decided to write a symphony in four movements I - allegro II-andante III-scherzo IV -Andante/Presto. This is even before I come up with a theme. Then, I take my random sketches/ideas of the past and construct theme into various themes. Then, (for the first movement) I formulate an introduction and the various parts of the sonata form. Then, I begin writing. For me, the orchestration/distribution of parts is what I hear in my head, which is directly influenced by other classical music (mozart, beethoven). Traditionally, sonata form is used for the first and second movements of symphonies, tertiary form for third movements, and rondo form for fourth movements.

I usually sketch out all of the movements before I start the first movement, meaning, I come up with themes and various ideas on how to develop them. I don't ever write a piano sketch, because it is unnecessary to me.

It is important to ALWAYS have coherency. An incoherent large-scale work becomes boring fast and will turn out to be a piece of garbage.

That's my two cents.

How do I write large scale music? I don't. Orchestral stuff don't interest me much. ... Unless we're talking about Webern or something like Ligeti/Penderecki, then I'm all for it.

Though then, really, it's a lot different. It's only large scale because of the number of people involved, but the pieces can be really short.

All the suggestions here are very good. The things you said you were doing are very similar to what I did before I started the large work I am working on. And yeah, like you said, I changed my mind while I was writing as I came up with "better" options but having that basic framework allowed me to get started with something in mind, at least a general idea of the direction I wanted to go in with the piece.

I think you will find that as you write the piece, you will get better and better at it (the technical aspects) and will end up going back and "fixing" the first part of it to match your skill as you complete it.

Listen to some large-scale works by master composers that you admire. It will perhaps inspire you with ideas or how to go about framing your work. That always helps me!

How do I write large scale music? I don't. Orchestral stuff don't interest me much. ... Unless we're talking about Webern or something like Ligeti/Penderecki, then I'm all for it.

Though then, really, it's a lot different. It's only large scale because of the number of people involved, but the pieces can be really short.

Balderdash, I say

And the Webern symphony is no good. Yeah I said it. His Passacaglia Op. 1 is quite phenomenal though. If we're talking orchestra, and that style, Berg was much more successful. But don't mind me, I just wanted to jump on your case because you made such a brash statement.

... I like the Webern symphony...

Me too. And you can't really compare Berg and Webern, in talking about "that style" because they were completely different composers :(

Large-scale works? They tend to flow out of one general idea for me. I usually sit down and write a beginning, and whatever I write becomes the basis for most everything else. I don't really plan all that much, aside from maybe I'll get a quick idea for another section, knowing where I'll put it when I get there. I try to make my music rather goal-oriented, so there's always direction - I've had some problems with that!

This could apply too: If I'm writing a piece for orchestra, I'll write all the music out on a 3-staff piano score first, and orchestrate after into Sibelius. This helps me compose a lot faster - it's also a hell of a lot easier to write down!

This could apply too: If I'm writing a piece for orchestra, I'll write all the music out on a 3-staff piano score first, and orchestrate after into Sibelius. This helps me compose a lot faster - it's also a hell of a lot easier to write down!

I am DEFINITELY trying this! Thanks for the idea.

I think a good method that I kinda use occasionally is to work backwards. This doesn't necessarily have to be literal, but if you start by writing a big contrapuntal climax, you can use the different snippets and parts as shorter motifs or a sparse soloisitic line while "foreshadowing". Once you get to the big part, everything sounds familiar, but fresh, and everything seamlessly flows together more than if you wrote it chronologically.

I tend to plan a general structure and some of the features of each movement (i.e. tempo, mood, individual sections of movements, climaxes). After that I come up with a short motive - it doesn't have to be long. The one I'm using for my 2nd symphony is only 5 notes. This motive then becomes the basis for everything. Every melody (and often a lot of the harmonies) will come from this motive. Then I just write the music. I never write down my plans because I know I'll never stick to them.

Do not know if anyone has mention these:

• Sketch out your music, it will give you a better idea of what you actually are doing and where you want to go - you will have more energy to focus on your orchestration

• Write your piece first for piano - I do not do this, but it is supposed to be pretty good aid.

I'm not so sure if writing just for piano is a good idea. Obviously writing a grand-staff full sketch of your piece that could be played on keyboard is never a bad thing, but pianisitc writing != orchestral writing. If you focus too much on getting the piano sketch to sound good, the final piece will sound like it was written on piano.

I'm not so sure if writing just for piano is a good idea. Obviously writing a grand-staff full sketch of your piece that could be played on keyboard is never a bad thing, but pianisitc writing != orchestral writing. If you focus too much on getting the piano sketch to sound good, the final piece will sound like it was written on piano.

This is why I only compose in full score, never from a reduction.

I'm not so sure if writing just for piano is a good idea. Obviously writing a grand-staff full sketch of your piece that could be played on keyboard is never a bad thing, but pianisitc writing != orchestral writing. If you focus too much on getting the piano sketch to sound good, the final piece will sound like it was written on piano.

Hence the three staffs. Making it impossible to be played on a piano and adding way more voices (as orchestral writing usually indicates)

Very good advice. If you focus on writing something within the span of your hands, the final product is going to be disappointing, especially for a large ensemble. It will be muddy and blockish.

This is why I only compose in full score, never from a reduction.

So many composers write in a short score on the first phase of the process, and do just fine from an orchestrational standpoint, that I don't think anyone can say there is one write way to do it. Just because you can't write something good in the concept of a short score doesn't mean many others can't.

I think this whole mentality is coming from using Finale and Sibelius as a crutch because their abilities on the piano lack or something along those lines. I personally hate writing in Sibelius, I ONLY use for making a polished final product. If it works for other people, fine, but from what I've seen the composers who compose solely using computer programs like Sibelius and Finale write repetitive, dull and unimaginative music. Even John Mackey, a composer who doesn't play an instrument and composes mainly through Finale writes a short score as the first phase of any major work he does. But what does he know, he only went to Juilliard right?

Different strokes for different folks, I guess. *shrug* I won't be arrogant and say that my particular way of doing things is the only way with any merit. It's what I have found in my experience to be true - for myself, and what I've seen in others.

As far as where people go to school, I'm more interested on their actual ability and output, not their pedigree. After all, our president went to Yale and he's an imbecile. You'll find that if you have money, you can go to whatever college you want, regardless of actual ability. It would be foolish to base one's determination on how skilled someone is by their choice of college... the product is what's important.

As far as where people go to school, I'm more interested on their actual ability and output, not their pedigree. After all, our president went to Yale and he's an imbecile. You'll find that if you have money, you can go to whatever college you want, regardless of actual ability. It would be foolish to base one's determination on how skilled someone is by their choice of college... the product is what's important.

You just compared apples and oranges man. Especially in music, your skill and your work is what gets you into a school, not money. I could throw money all over the world and that's not going to help me get into Yale or any other school anymore. Your comment really off base.

I just get a very small melodic fragment and develop it within a form and within a piece.