Bill,
Well written, and i think you've hit the key in sample performance.
Orchestration.
We're used to hearing the classics in particular from the POV of 'standard' orchestration. (Generalisation) But that orchestration is a reflection of live instruments with individual interpretations of how the music should be played, no matter how fine the difference between one player and another in the same section. Listening to live recordings will always highlight a basoon player who is slightly sharp in intonation, or a clarinet player who interprets the first beat of a bar a fraction later than someone else, just to give two examples.
Yet in sample performance, it is the same player inputting each line or phrase, with their own interpretations of where 'one' is, or the mength of each note in a phrase.
We can play in 'live' each instrument of the orchestra, and the human element is apparent, with the exception of sections, notably strings. And string sections contain more subtle variations than any other section due to (a) the number of players, and (b) their interpretations.
If we were to be objective about sample performance, we would use a section player sample, and play the same line according to the number of players in the section. And i did say a section player sample, as opposed to a solo instrument, whose sampled sound is more expressive.
I think this is the important key to the question of sampled strings, and with experimentation could produce the effect we are all looking for, or at least closer to the real thing than we are now.
If i had a wish for the VSL team, it would be to record section string instruments in pairs, reflecting a different seating position. I would then play in the number of pairs required to fill a section. Some would think this was a lot more work, but the end result would require a lot less manipulation, so I think the balance would be maintained.
As for three days to go.............
Regards,
Alex.