What's new

"Correct" use of non-divisi libraries

shireen

Member
Hello,
a very small newbie question:

As I understand it, If I write a 2 voices part on a Violins 1 patch of, say 8 violins, which doesn't have divisi, What I'm hearing is actually 16 violins? I mean, it makes sense :P

And if that's the case - while writing for the entire strings section, I should always write one note per patch?

Thank you!
 
Technically yes, playing 2 notes on an 8-violins patch means you're hearing 16 of them. But it doesn't mean you can't do it if it sounds good! Plus there are loads of overdub these days when they record sections several time to make them play different things at the same time.
 
You can write as many notes as you like if it sounds good to you. When you start playing bigger block chords it gets a little more problematic. Then, it's better to break it across multiple parts to get different timbres from different players. Also try to not copy and paste. These are two of the big reasons that your string section can sound like a pipe organ or "synthy".
 
Samples aren't that cut and dry since you're using fixed recordings.

If you divisi a single patch in my experience you don't get "16 vlns" You get an ensemble of 8 violins playing 2 different notes at the same time. In other words the sound of the recording doesn't magically change to an ensemble of 16 players doing divisi on a single patch.

But, since samples are already a compromise, if you use one patch to play two different notes, it's close enough. What you have to watch for is the build in volume. If you take a real 16 section violin and have them play 2 part divisi, it actually decrease the sound of each note. The actually volume of each part will be less while the overall volume of section remains the same ( not accounting for resonance produced by harmony which will make things seem fuller).

On the other hand if you take an 8 vln ensemble sample and you then double it up. You're actually increasing the overall volume of the patch and thus it will sound louder. There's a build in volume as you add another or other voices to the same patch.

So if you do use a single patch to play more than one note, in order to keep the balance of the ensemble you have to lower the volume of that patch while it performs the "divisi". So for real players it's an actual "divisi" or dividing of the section and for samples it's not really a divisi but a summing of 2 different recordings.
 
Technically yes, playing 2 notes on an 8-violins patch means you're hearing 16 of them. But it doesn't mean you can't do it if it sounds good! Plus there are loads of overdub these days when they record sections several time to make them play different things at the same time.
Nice explanation. Nice to see you here JJH.
Thx. After 3 or so years of not really thinking about samples, I've started to think about them again. So what better place to be than on VI-Control to catch up on the latest.
 
It depends largely on what you're trying to do, for me I don't layer things that an orchestra couldn't accomplish. That's just me though, the "everyone in the room" mindset! If you're coming at it from a place of overdubbing (which is super common practice) then I don't see an issue.

I will say it's very nearly unnoticeable in string writing unless you're doing full chords across multiple sections. Brass is where it sticks out the most I think.
 
Hello,
a very small newbie question:

As I understand it, If I write a 2 voices part on a Violins 1 patch of, say 8 violins, which doesn't have divisi, What I'm hearing is actually 16 violins? I mean, it makes sense :P

And if that's the case - while writing for the entire strings section, I should always write one note per patch?

Thank you!

Don't worry about this at all. If it sounds musical, keep moving.

A lot of pedantic nonsense has been written about this issue that gets people all tangled up in their heads with analysis instead of musicality. It's hard enough to write anything decent, so just worry about the notes.
 
I try to avoid using anymore than one note on any other instruments besides strings when doing mockups most of the time.
 
Don't restrict yourself when it comes to music. Unless it is going to be played live to an audience, there's no reason not to just do what sounds good. Multitracking has been a thing for decades, so if you happen to write something that sounds better with two, three or more passes of a full section just do it and you can multitrack the final thing/
 
I wouldn't worry about it. Most of the time, when you play even one note, and you move the modwheel, you are hearing double the amount of violins (or whatever instrument) anyway. :)

Which of course means, that if you play two notes, you are hearing 4x the violins. :)
 
... If I write a 2 voices part on a Violins 1 patch of, say 8 violins, which doesn't have divisi, What I'm hearing is actually 16 violins?

And if that's the case - while writing for the entire strings section, I should always write one note per patch?

Thank you!

Use the same violin patch in 2 different channels and give them different EQ settings. This can make wonders ... .
 
Don't worry about this at all. If it sounds musical, keep moving.

A lot of pedantic nonsense has been written about this issue that gets people all tangled up in their heads with analysis instead of musicality. It's hard enough to write anything decent, so just worry about the notes.
I've just stopped using notes entirely, nothing but the intention of notes
 
Splitting one patch into two doesn't make much difference, but if you get into really heavily divided string writing the sound does get a bit unnaturally mismatched, at least in my opinion. When strings divide you hear more detail among individual players, like individual attacks and vibrato. Might not be as big a difference when you're dividing from a group of fourteen strings to a group of seven, but if that group divides three ways or if all the string sections are dividing then I believe it does.
 
It doesn’t really matter when it comes to samples, none of this is real anyway :) And if you tell a real string section to play “non divisi”, like a double-stop, it won’t sound like twice as many violins, but twice as many notes, so it will just become a more full chord (or become pitchy).
 
Splitting one patch into two doesn't make much difference, but if you get into really heavily divided string writing the sound does get a bit unnaturally mismatched, at least in my opinion. When strings divide you hear more detail among individual players, like individual attacks and vibrato. Might not be as big a difference when you're dividing from a group of fourteen strings to a group of seven, but if that group divides three ways or if all the string sections are dividing then I believe it does.
This is correct. It really begins to matter after the "3rd" divide. Guys like Ravel did this and it thins out texture while thickening harmony. In the real world, strings are rarely a volume problem. One trumpet can blow away the violin section, so it's not so much "volume" as "texture".

Since most of my real-world orchestration work is for Symphony "pops", I do the exact opposite if I need a string line to stand out: I use vln1, vln2 and viola on the melody in unison, and often celli too an octave down. When there's a rhythm section and a brass section on the gig, strings shouldn't even have to show up. :) They won't be heard much, so I need as much volume as possible if they "have" to be heard on a melody.
 
And if that's the case - while writing for the entire strings section, I should always write one note per patch?

As has been mentioned, it's more complicated than this.

First of all, let your ears decide... (but let your ears be critical)

Second, in my opinion (or: to my ears) usually the vibrato is quite relevant.
Sounds with less vibrato work better (when multiplied by divisi playing), while samples with too much vibrato tend to produce a sound that's very artificial in a not very nice way.

Sometimes, when you use lots of this kind of "divisi" in a piece, so that it would require lots of players, it's helpful to attenuate the hi end a bit. Also the frequency areas that are getting harsh.
This suggests a larger space and bigger distances (where a huge ensemble normally would make sense...)
 
Top Bottom