Know the style you’re working in
This one goes hand in hand with the previous topic. You need to know how individual instruments and families work, but you also need an understanding of how they all interact to form the typical orchestral sound. This is less straightforward than familiarizing yourself with single instruments, and you will not find as many helpful youtube vids to guide you.
Arranging for orchestra is a vast subject and music scholars have been writing lofty tomes about it for two centuries. Going into the details is beyond the scope of this article, plus that I’m still learning myself. I’m just going to touch on a few general things that people often get wrong or have no idea about when starting out.
It’s a frustrating feeling, knowing that you’re using all right instruments — of excellent quality, to boot — and the end result still sounds wrong and fake somehow. In most cases this is a symptom of not being quite clear on how things work and fit together. On their own these seemingly trivial issues may not detract from realism and believability in a major way, but put together they add up to create that rarely desired “fake synth orchestra” sound.
Voicings and range
One major thing is that newbies tend to treat instrument sections like synth patches. Kind of like, “Okay, I have an instrument section here. Let’s use it to play standard root-third-fifth chords on my keyboard. That should do it.”
Um, no. It won’t do it, because first of all, one single note features all players of a section. Add a second note, and the number of players has suddenly doubled! If you’re going to play three-part harmony or chords in the trumpets, you should use three solo trumpets instead of sections samples or it will end up sounding like you have a small army of trumpet players in there.
Secondly, different sections voice chords and harmonies in different ways. While it’s comfy to play root-third-fifth chords, or inversions thereof, you should make sure that this is something that the section in question actually could, or would, play. Perhaps the instruments in a section almost always play in unison? In the case of strings, the different sections are sometimes divided into smaller subgroups (divisi) when the number of notes in a particular part exceeds five, but this is far from the norm. Most often the strings will be playing monophonic lines, one for each section.
It’s also important to avoid dense arrangements where you have a lot of stuff playing similar things in the same octave. Orchestras have a HUGE range and you shouldn’t be afraid to take advantage of this. Unless a cluttered, dense sound is what you’re after for some reason, always make sure to space things out as much as possible. If a part is getting in the way of another more important part, transpose it up or down an octave or two. If it ends up in a range that this section can’t handle, move the part over to a different section that can handle it. Suddenly you will have freed up space for the important part, by reducing the number of things going on in the same frequency range.
Another common newbie mistake is lack of variation. Listen closely to a classical piece or an orchestral movie soundtrack and you will quickly realize that there’s new things happening all the time. Not necessarily new musical ideas — in fact there’s often a lot of repetition, but you don’t really perceive it as repetition because the arrangement and instrumentation is constantly changing. Sometimes it’s softer, sometimes it’s louder. Sometimes it’s faster, sometimes it’s slower. And all along the parts played are moving between different sections and families, with little embellishments and counterpoints and accents weaving in and out here and there, to add even more interest.
I read an interview with a composer once who said that in order to maintain the right amount of variation in your music you should imagine that the listener is a 7 year-old child with ADD, or something along those lines. I don’t remember who it was, and he was most probably not entirely serious. Still, this thinking is not far off the mark. I would say as a simplified rule of thumb, if something in your piece starts over without at least some form of variation happening, it’s going to sound predictable and boring.
I know this might sound like a crapload of work but it’s important to remember that variation doesn’t always have to be large and radical. Most of the time you can get away with subtle changes, like a thematic melody moving over to a different section, in a different octave, or simply doubling the melody an octave higher or lower when it starts over, while the rest of the arrangement remains largely the same. Sooner or later you’re going to have to do some radical shifts though, to keep things interesting. Like groups of instruments trading places, e.g. strings providing accompaniment and brass handling the melodic part -> brass providing accompaniment and strings playing melody. Or any variation thereof.
A word of caution though: introducing new musical ideas is not variation per se and should not be used as a way of adding interest and making your music less repetitive. This one’s difficult, and to this day I struggle with it myself.
When you’ve spent half a day writing a part you tend to get a feeling of, “hey, I need to move on to something new here!”. When in most cases what you should be doing is repeat the part, though with variation, because the time it takes to compose something versus the time it takes to listen to it creates a state of cognitive dissonance. The part you wrote might actually be only 20 seconds long, and suddenly moving into something completely new will sound rushed and unexpected. Present a theme, establish it through varied repetition and let it sink into the listener’s subconscious. Then, and only then, should you move on to something new. But don’t forget to return to the themes you have established previously!