mike_mccuehow to differentiate between the distortion you are committing with intention to distortion that would occur without intention when you use a setting such as 0.1ms in a hard limiter.

 
You can differentiate with the Nyquist frequency.

.

Aliasing.
 
But for a lookahead peak limiter/loudness maximizer thingy oversampling does help let it see intersample peaks. [edit] And I believe most of those kinds of plugs already do oversample - I think it's more general purpose comps set for aggressive limiting that might have a problem. [/edit]

.

Just to further the unexplained possibilities.
Sonarluminesce remains an unexplained phenomena.
https://www.youtube.com/watch?v=LWO93G-zLZ0

Um... as both a physicist and nuclear engineer I could not get too deep into that video before the comment on temperature made me close it. Heat transfer is universal, and a "bubble" being neutrally bouyant in a fluid is far fetched (especially with changing temperature). Be wary of "science" such as this one... "cold fusion" began a similar way and we all know how that one ended.

You could very well be right. There are a few of the same
and the intro debases the actual documentary.
 
But the heat is an implosion (blue) and not radiant,(red).
 
"Sonar"luminous is blue light produced by sound vibration above audible range.
 to see it reproduced more easily 
 https://www.youtube.com/watch?v=HuDyy4MTosA

I believe another benefit of using 96khz is your midi VST audio recording latency goes down because it's based off your audio clock cycles...I believe that to be correct. -Shawn (I use 44khz just for space reasons)

For the same buffer size the latency is lower at higher sample rates. The reason is very simple.
Assume you are using a buffer size of 64 samples.
At 48K this corresponds to a latency of 64000 / 48000 = 1.3 msec
At 96K this same buffer corresponds to a latency of 64000 / 96000 = 0.6 msec which is half that at 48K 

shawn@trustmedia.tv
I believe another benefit of using 96khz is your midi VST audio recording latency goes down because it's based off your audio clock cycles...I believe that to be correct. -Shawn (I use 44khz just for space reasons)

That's true in theory, and often in practice but not always. If 96 stresses out the computer too much, you may need to increase the number of sample buffers, so you sort of end back where you started.