Pardon the disorganized thoughts here:

I'll tell ya what I did for my template and Synchron Stage MIRpro settings...

I imported a couple of modern classical recordings (not film scores, because they "cheat" the levels on things like the harps, celeste, string shorts, etc) and then did mockups of those excerpts.  I think one was Nimrod by Elgar, Beethoven 3rd opening, some tutti chords from Mahler 5, stuff like that...

I made sure that the velocity cross fade of the instruments were at their proper levels, not all jacked to 127 just so I could hear them.  But where the manuals said the layers were marked.

THEN I went to the faders in VEPro and matched them to the recordings.

One thing that will help is making sure you find the honky resonances in the MIRpro EQ at the bottom of the instrument profile on the right hand side.

Then there is balancing the instruments in MIR with the Synchronized or Synchron libraries.  I used those as a guide as well for balance, usually starting with the "Classic Surr to Stereo" preset and then matching the VI libraries to that (so long as it was also close to those modern orchestra recordings).

I can't remember where the FF chords landed on my stereo bus, but I'm sure it was in that -6 range.  I almost always end up automating the stereo bus fader on super soft stuff (like that Elgar variation).  And in my experience, the lower levels are even with the loud among the instruments.

Then there is the aspect of color and tone intensity.

Watching rehearsals of great conductors, they stop the string section all the time when they lose the intensity on soft passages.  For this, I lower the expression, but even if marked pp, I'm using a mf velocity crossfade level.  That shouldn't effect the levels except if you want to bring out a group, like the horns.  Instead of automating each expression fader (4, 6, 8?!) I just grab the horn bus and ride it a little bit to help with swells or whatever.

So yeah... faders are always in flux, and those natural volume levels are a great place to start, but I think they can be tailored to your personal programming tendencies.

OH! important thing... also be prepared for some CC24 filter automation.  Sometimes you need loud trumpet or horn, but not so brassy.. a nice "ohhh" tone. Use the filter.  Also, with high strings/woodwinds all up in the stratosphere, use the filter a bit on the strings and suddenly you'll get a better balance of woodwinds (not to mention not letting your strings get all icy).  So maybe as you are checking the mf balance of the orchestra in the upper range, before adjusting the fader, try the filter.

Adhere to orchestration principles.  If you have 3 flutes on middle C, playing as loud as they can, and a cello melody in tenor clef with some french horn chords gluing things together, those flutes should not stick out.  If they do, your flutes are probably too loud across the board.

Sam, in answer to your question about changing overall volume, it does not change the objectively-measurable relative balance you've already set up. The non-linearity of decibel-scale faders is perfectly ([ed.] ... erm, well that's in theory; in practice it's "more or less") innocent in this regard, although it's helpful and generally good practice to keep the faders in the zone where you get best dB per inch resolution.

That said, there is indeed something to watch out for when changing overall level radically, because a mix balance may appear to change subjectively as a consequence. It all depends on the characteristic pitch and timbre of the different instruments in the mix, and the fact that the pitch-response of our ears is different at different levels of loudness. Therein lies a part of the mix engineer's craft and art. And since I'm not a pro mix engineer I'll just say use your ears and your common sense, and try out your mix on several different kinds of speakers and headphones, and at different listening volumes.

BTW there are lots of tutorial videos on YouTube about setting up "Gain-Staging" prior to actually doing a mix. You may find some relevant and very useful practical tips there.

Stephen:

Originally Posted by: stephen limbaugh

Pardon the disorganized thoughts here:

I hereby pardon you 

Originally Posted by: stephen limbaugh

I imported a couple of modern classical recordings (not film scores, because they "cheat" the levels on things like the harps, celeste, string shorts, etc) and then did mockups of those excerpts.  I think one was Nimrod by Elgar, Beethoven 3rd opening, some tutti chords from Mahler 5, stuff like that...

This is fantastic advice.  Thank you!  I will do this.

Originally Posted by: stephen limbaugh

I can't remember where the FF chords landed on my stereo bus, but I'm sure it was in that -6 range.  I almost always end up automating the stereo bus fader on super soft stuff (like that Elgar variation).  And in my experience, the lower levels are even with the loud among the instruments.

I'm a little unclear as to what you mean.  Do you mean to say that - in the recordings you referenced - it sounded like that super soft passages were brought up on the stereo bus and so you did the same?  If so, this makes a lot of sense, since what I'm essentially worried about is that in order to preserve a serious dynamic range yet avoid clipping, the distance in volume between the loud and the soft parts will be so great that the soft parts are just too quiet.  Buuut...if none of the loud stuff is playing, raising the stereo bus (albeit subtly so as not to betray the intended dynamic range too much) circumvents that problem.  Is that a correct understanding?

Originally Posted by: stephen limbaugh

Adhere to orchestration principles.  If you have 3 flutes on middle C, playing as loud as they can, and a cello melody in tenor clef with some french horn chords gluing things together, those flutes should not stick out.  If they do, your flutes are probably too loud across the board.

How accurate, would you say, are Rimsky-Korsakov's approximations?  A childhood of bad EDM has left my ears having a ways to go towards intuitive understanding of orchestral balance, so any training wheels help... 

And, of course, thanks so much for all your help!  Your advice is gold.

- Sam

Macker:

Originally Posted by: Macker

Sam, in answer to your question about changing overall volume, it does not change the objectively-measurable relative balance you've already set up. The non-linearity of decibel-scale faders is perfectly innocent in this regard, although it's helpful and generally good practice to keep the faders in the zone where you get best dB per inch resolution.

Wow, you just saved me a lot of time worrying about a non-issue!  For some reason, I was stuck on this idea that changing an entire section's volume meant shift-clicking every fader.  I seem to have forgotten that buses/folders exist.    Thank you for clearing that up!

- Sam