https://www.youtube.com/watch?v=qdfWWg1FOyQ
 
I noticed that there is tons of noise at the highest available frequences using my Audient.
 
Is this normal? It seemed to be a totally cheap approach to hardware design. I thought all ADCs would apply a low-pass before conversion. To skip it seems really cheap.
 
Maybe the noise is from something else? Why would the Audient show tons of noise and the RME not for the same Mics?
 
Do you think my guess as to the reason for the noise might be correct?
 
Is this a common thing to do on ADCs that go up to 96? Avoid the cost of the low-pass filter all together? And instead of a Low-Pass filter for 44.1 and 48, just give every other bit from the 96 conversion?
 
I don't know what I've detected, but I'm pretty afraid that the fold-over artifacts might be in the audible range at least in part.
 
My thoughts on this started on this thread...
http://forum.cakewalk.com/Me-with-my-first-Ribbon-MXL-R77-compared-to-Shure-SM57-m3227996.aspx#3228391
 
I created a new thread because it really isn't about comparing the mics.

In this SOS article...
http://www.soundonsound.com/sos/aug14/articles/audient-asp880.htm
 
I found this image...

SOS
This graph shows the total harmonic distortion level produced by the A‑D converter relative to its digital output level. The THD figure rises gradually from ‑120dBFS with increasing output level, as would be expected, before jumping about 12dB as the signal level exceeds ‑10dBFS. This is a recognised characteristic of this family of converter chips and, although probably inaudible, highlights the benefits of maintaining headroom when recording!

 
What's interesting to me is that the amount of noise at 40K rises and falls with the signal level from the Mic. To me, this suggests something interesting and I don't know of SOS tested for it. Basically, if there is no LP Filter before the DAC, there may be artifacts in the audible range. It's a rather tricking thing for me to replicate in my home studio to prove, but just as with Craig Anderton's demonstration of Synths bounced at different sample rates sounding significantly different, I think this could actually be a short-coming in the interface.
 
I suppose it's all conjecture. Someone out there knows but not me. And if I can't do a home test to reliably show anything I'm not sure it's worth discussion so I'm going to pack up my gear and go make a tape.