I was curious as to what the philosophy was for setting the final volume level of the individual samples? Is it to normalize each sample so that it takes maximum advantage of 16 bits?
This may be naive, but I would like it if the samples were played back at the same relative level they were recorded at (at least once the library goes 24 bit.) That would seem to make realistic mixing much easier. Obviously, that would make it easier to realistly mix between sections of the orchestra (so that the solo oboe at forte doesn't sound louder than the trumpet at forte or the entire violin section as a hypothetical example.) But also, it would allow for more realistic dynamic tessituras. For example, high register trumpet or flute ought to be much louder than the lower registers of the same instrument, but that doesn't seem to be the case right now (maybe I'm wrong.) In VSL's defense, all sample libraries I have used to date seem to do this (and they've all been 16 bit libraries too, which may be why.) I think it would add much realism to move away from this though.
This may be naive, but I would like it if the samples were played back at the same relative level they were recorded at (at least once the library goes 24 bit.) That would seem to make realistic mixing much easier. Obviously, that would make it easier to realistly mix between sections of the orchestra (so that the solo oboe at forte doesn't sound louder than the trumpet at forte or the entire violin section as a hypothetical example.) But also, it would allow for more realistic dynamic tessituras. For example, high register trumpet or flute ought to be much louder than the lower registers of the same instrument, but that doesn't seem to be the case right now (maybe I'm wrong.) In VSL's defense, all sample libraries I have used to date seem to do this (and they've all been 16 bit libraries too, which may be why.) I think it would add much realism to move away from this though.