Thanks for that info @Jane. Good to know, and it explains the file that @David B made.
Here's the spectragram and audio for my recording F, VSL F, then my recording G, VSL G.
You can see that VSL does not support silent key, because the second and fourth impulses have the bright lines toward the bottom while the first and third don't. Basically, in my real F308 recordings, the open string is unstruck and not vibrating prior to the staccato F/G as in the VSL ones.
Here are my thoughts:
VSL's SR is different than the real piano. The first three-five lower harmonics are correct, but the upper ones are not the same volume. This is exactly the same problem with my SR model in my piano libraries.
But what IS correct, however, is the overall SR volume. David's example is slightly louder than my recordings. And remember, the SR that I'm showing is THE ACTUAL INSTRUMENT with mics INSIDE the piano. You can't get louder SR than what I recorded. This indicates to me that people's expectations of resonant SR turned up with "halo effect" is actually hyper-real and not like a real instrument. Again, what sounds nice to people may not actually be what a real instrument can do.
For me it's important to remember the purpose of SR in a piano library. For some, it's that "halo effect", which in the synth industry is basically called a pad. Its a second instrument (sometimes strings) that enhance the sustain effect of a piano library. But for me, the purpose of SR is purely to create more randomness in the sound. When certain harmonic key combinations are struck, the strings give a slightly more brilliant sound than non-consonant combinations. And I mean slightly/subtly. Not like a pad.
But that's just me.