Elegy for String Orchestra

[Brett Harrington]

I made an account here to get some feedback for my most recent piece, anything would be very helpful!

 

https://youtu.be/N9Sk-SqFkpY?si=9wW_owO0aiIfRhfo

[Steven Linden]

I enjoyed this quite a bit. I like the late-Romantic harmonies and pathos. I think my favourite thing here is the way you make those key changes. Others have commented that it feels repetitive, and I sort of half agree with this. Certainly, it's a small amount of melodic material to sustain an eleven-minute piece, but personally I did not find that it got boring. The repetitive, almost trance-like nature of the material is, I think, part of its charm.

There were just two moments toward the end that I found less effective. First, the chromatically descending passage at m. 118 felt a little out of keeping with the style of the harmony in the rest of the piece. And second, I'm not sure I was convinced by the out of sync crescendi and diminuendi in the different parts starting at m. 133. But maybe that's just the limitation of the Musescore playback - I like the idea of the different parts fading in and out relative to each other, but the sound of it just wasn't quite effective for me.

It's a shame Musescore does that thing where the change in dynamics don't line up exactly with the beginning of a new note - I'd love to hear this done with a better VST, or a real orchestra!

[HS Teoh]

One thing I've always done w.r.t. coaxing the computer to produce a better audio, is to have two scores: (1) the actual "official" version of the score, the one you'll show would-be performers, and (2) an unofficial copy of the official score meant only for the computer, where you pull out all the stops and do whatever hacks it takes it get the computer to do your bidding. Weird dynamics, fake notes, hidden background instruments, you name it -- to get the audio to sound more real. Stuff you'd never want to show anybody for its ugliness.

Could be one way of going farther if you're financially tight for investing in a "real" VST. (Actually I'd do the same thing even with a real VST -- some things really are meant for computer consumption only, which would either be redundant or downright ridiculous to write for a real player.)

[David]

quite a number of people have this approach -- or indeed simply use a DAW for a mock-up which is totally independent from the score. For me it's too much like hard work -- in general the score is only there to help a listener to follow the music so it has to be reasonably legible but that's about it. On the rare occasions so far when something has actually been performed, I put more effort in, particularly with pesky accidentals and of course page turns in the parts which no notation software seems to be able to completely work out on it own.

[HS Teoh]

It depends on your approach to composition -- many composers these days start directly in a DAW and record what amounts to on-the-spot improvisation; the score is merely an afterthought. I work the opposite way: I hear the music in my head and write it down in score, *then* render the audio to confirm that I've written it down correctly.

I actually use a slightly more developed approach: since I use Lilypond for writing scores, which has a programmatic kind of input syntax, it's possible to combine both the "real" score and the intended-for-midi-only score into one. It does make the input somewhat cluttered when there's an abundance of midi-related hacks in the same place, but at least it guarantees that the two won't inadvertently go out-of-sync, which usually is the source of a lot of the extra work involved with this approach. But doing it this way is somewhat specific to Lilypond; for users of other software this may not be a possible approach and you end up having to maintain two separate scores, which does become cumbersome.