|
Not only does it do 192, it does it so well, I've finally jumped off the Good Ship Lavry...
The ULN8 is the ONLY device that has made me seriously question Dan's position (not the 60K thing, but the practical 96K = sufficient thing). Previously, in cases where I like 192K better, it was a clear "enhancement" of the source, not accuracy. I could reliably pick out the 192 source from the "lesser" source blind, though usually BOTH could be distinguished from the input blind, indicating an improvement unrelated to fidelity. The 8 is definitely accurate: if you switch between any analog source and a round trip through the 8 @ 192 in the analog domain, you may well not be able to detect which is which. But it's also "better": at lower rates it gets a lot easier to tell. After I measured and established nothing was broken, I realized there was a difference: better filters, far out of the audio band at 192, just sound closer to the source in current implementations. The sound of everthing passing through gets progressively tighter and better, all the way to the top.
Maybe MH has gotten timing/jitter to the point of irrelevancy, eliminating the timing issues (at 192K a converter with low jitter at 48 will have 4X more)? Maybe it's the phase-linearity and power linearity of it's DC coupled design (both flat from 0Hz - Nyquist)? Who knows? Not me. I just know what my ears tell me.
Upsampling 44.1 sources to 192 for processing yields measurably better results. And there lies my heresy. I've developed a new bad habit of mastering at 192 of late. It eats resources, and bugs me intellectually, but sound is sound, and soundBlade+ULN8 sound better at 192. I get paid to master with my ears, not my calculator, so here I am.
-d-
|