Articulations and Dynamics do NOT sound the same exported in other programs (from Musescore MIDI file)
As a classically-trained composer writing orchestral music exclusively, I noticed that what sounds great in Musescore, doesn't sound the same in professional "Hollywood grade" sound libraries, or other notation programs like Notion 6. In fact, after careful study of music theory and examining the scores of Wagner, Stravinsky, and Beethoven--I rely on a combination of playback sound and informed "hypothesis" based on what I've observed (i.e. correlating notation symbols with sound during live performances/recordings)--I know that what sounds good with a real orchestra doesn't always sound good in a scoring software, so I rely on theoretical knowledge rather than midi playback for what is "realistically-correct"!
Know that I do use volume dynamics (p, mf, ff, sfz) and slurs (for violin bowing), along with pizzicato, staccato, and all standard articulations one finds in late-romantic German and Russian music.
That said, I write all parts based on how it sounds in Musescore, and often it's a shock to hear it in programs that claim "real samples from London Symphony Orchestra"--because the tempo is slower therein, the tremolo sounds thick and sloppy instead of a crisp, dry fluttering, and recently someone rendered my symphony in a mix of Hollywood sound library and Logic and the woodwinds are disproportionately louder than brass and strings (a single Clarinet isn't supposed to dominate Horns and Trombones!!!). I wrote my 1 clarinet in piano or mezzo-piano (mp) with 2 horns and 2 trombones in MP or MF (a trumpet comes in somewhere too).
My question/concern is: I am unsure if exporting to more "realistic" sound libraries is desirable anymore, since the music only seems to sound "perfect" in Musescore. Additionally, how does one truly know what one is writing will sound as envisioned, considering these pieces never sound quite the same in every software? Which is the "more accurate" version? How can you really know until you attend an orchestra rehearsal?
I am strictly-traditional, and execute musical craft according to 19th century ideals, values, and methods. Music must always be done with real performers/orchestra--and I absolutely-refuse to use digital "mixing" and turning a bunch of knobs on a mixing console because that isn't a composer's job--that is the place of producers making electronic music who speak of "frequencies". We traditional souls spend our lives learning the theoretical dimensions to produce Visions, and our form of "mixing" is called writing Italian phrases above staffs to indicate louder or softer!!!
I use musescore to write/engrave, and I did consider making MP3 files and selling them as "digitized orchestral music" in case no orchestral would perform these compositions. I deleted Notion 6 because 1) to actually "score" music therein is the most convoluted, confusing series of clicks, commands, imputs (totally NOT expedient or user-friendly) and 2) the "real orchestra samples" sound like a worn-out 60's Mellotron!!!
So I decided to go the traditional way, however, I must be sure that what I'm hearing is how it will sound then and there--any way to know this in exact precision? Any way to rest assured all will sound as envisioned?
Comments
First of all: Computers are machines. Musescore is a software which works with computers. We cannot to expect it sounds like real human players.
Second: The dynamics and articulations we can get from Musescore are a soundfont file dependent issue. The Musescore format files are recording including those "effects". Unfortunately, the standard MIDI format files don't include those "effects". The Musescore file format can be only be read from Musescore. You have to use the "Exporting to MusicXML format file" option to exchange all those MIDI parameters. However, not all the MIDI software can read that MusicXML format files.
In reply to First of all: Computers are… by jotape1960
Ok I'll try that. Thanks
"That said, I write all parts based on how it sounds in Musescore, and often it's a shock to hear it in programs that claim "real samples from London Symphony Orchestra""
- What programs in particular do you mean?
In reply to "That said, I write all… by innerthought
Notion 6. As I said earlier, Uninstalled because it sounds like a cheap Mellotron or toy organ imitation.
Those who make soundfonts or equivalent playback libraries (vst, sfz, etc) either use a loudness preset of their own or set each instrument to be at the same decibel. Of course, neither one nor the other of these options sounds the same to everyone, and while it is great for some, it is broken for others.
Of course, this also depends on how, and how different software uses these banks.
For example, another software may sound the soundfont of MuseScore software in a different way.
But as for the intermediate dynamics settings: You also know that every human player in the orchestra adjusts their volume according to the others. So if the Strings section plays "mp" instead of "p" that day, the horn player, flutist and the rest will adjust their volumes accordingly.
Even though the highest and lowest volumes are taken into account when adjusting the soundfonts, the intermediate dynamics may not be what everyone wants. According to one, the level that is in the "p" dynamic may be in the "mp" dynamic according to the other. The quality, resonance, and other things of the instrument whose sample was recorded may also have influenced this.
Since there is no human factor that will adjust itself to other players here, we have to settle for the result that the software presents based on the recorded sample.
So how can we be sure that the loudness of the instruments presented to us in soundfont (when played together as an orchestra) is in a good mix?
You can test this by spending some (long) time on it.
Let's say you stick with MuseScore:
First, you need to open the empty orchestra template you will use, write long-lasting notes to all instruments, duplicate them as ff, mf and pp, and turn on the mixer and compare the loudness of each one. And then you can save it with a meaningful name and use it like a template (example: full_orchestra_my_mix.mscz).
For example: First you tune the String part between each other, then you tune the Woodwinds accordingly, and then the Brass and Percussions if any (Yes, I know, it takes a long time). But ultimately, you will only do it once in this software and this soundfont.
With a well-tuned soundfont, you usually don't need to play around with this setting. And I think Musescore's soundfont is set well too.
In other words, I don't think you need much of the above process in classical works (such as the works of Beethoven, Mozart, Haydn, Vivaldi).
However, it may be necessary to make a few minor adjustments according to the instruments used in the piece (using mixer). For example, some instruments may dominate others in a piece. However, the sound levels of the same instruments may be lower in another piece. You need to find a medium setting.
You can even see that some instruments are more dominant in the tutti section of the same piece, while the same instrument is lower than the others in the quieter parts.
I used about 16 classic works for testing for my own soundfont. However, while listening to one or the other, some instruments have loudness settings that I want to change, but how could I adjust this when the sound of the same instrument is higher than the average in one piece and low in the other?
What can be done in this case: You can go to workarounds like first removing the play checkbox of the real dynamics from the Inspector and then adding invisible dynamics to the parts of that instrument you deem necessary. But if you want to work like this, however, you will need to save the score written for a real orchestra and work on an second copy to use it as a demo recording.
What I prefer: I leave it as is.
And when a real orchestra performs your piece, it's all on the conductor's baton. Because the conductor will also set a level according to himself. For example: By the conductor: In rehearsals, the string section will be asked to play this part a little louder, or it will be that the clarinet player is requested to play a slightly louder solo here.
I don't know if what I've told you so far will work for you. Think of it for informational purposes.