I seem to recall - only vaguely - reading somewhere in VSL literature what size the audio 'word' is that's being streamed from VSL Player outputs in VEP7 to the receiving devices in the master DAW. But I can't remember what it was and I can't find it anywhere now.
Is it 24-bit integer, 32-bit floating, or something else?
At the very least I need to know if it's more than 16 bit integer (though I think it must be). This is because I want to set up all VSL players in VEP7 at a nominal static gain and leave them at that setting permanently; then I'll do all non-timbral level conditioning and controlling (including Expression level, instrument levels, etc) within the confines of Logic. Of course I don't want to exceed the intrinsic maximum audio level anywhere in VEP7 (including within VSL Players in VEP7), so knowing the audio word format (in terms of its numerical range) is crucial.
[In the idea I'm pursuing here, I need purely objective numerical hygiene for audio in VEP7 while the crafting of expressive and other subjective dynamics during audio tracking and mixdown is kept locally within Logic as much as possible. Hence the only audio level changes I want in VEP7 are those which necessarily result from velocity changes, velocity-layer X-fading (which is the only level change I need Logic to manage remotely), recorded-sample dynamics. or baked-in sample level differences. If the audio word size is at least 24 bit integer, or perhaps 32-bit floating, etc, then I don't have to worry much about peak signal levels going quite far 'into the red', nor average signal levels falling quite far below what the meters can display, within VEP7 and its inserted VSL Players. I won't be using any level-dependent Fx plugins inside VEP7.
If this all sounds a bit like laziness, well, perhaps I'd admit to favouring economy of effort in this case, lol. But my main rationale is, I'm determined to avoid fixing door handles onto door handles fixed onto door handles ....]
