As brought up in the EVGA NU Audio Pro thread, this card's line out has been tested with a whole bunch of opamps, including a V6 Classic, results are linked here:
https://forums.evga.com/EVGA-NU-Audio-Pro-Certified-Opamp-Rolling-List-m3027421.aspx
This line stage appears to be of moderate impedance, low kOhms is my estimate, with a modest +/-5 V supply.
Unfortunately I can only describe the performance of the V6 Classic as
"epic fail". Noise is not the worst of the bunch but only of the level of a dated NJM4565 (~10 nV/sqrt(Hz)), while distortion at 1 kHz is something like 15 dB worse than even the most mediocre IC opamp tested, and over 30 dB worse than the best.
Now Burson rates the V6 Vivid higher in "Transparency" and "Details", so its performance may be better, and I suspect higher supply rails would also help matters, but on the whole I'm afraid it's just kinda hopeless.
The datasheet for both suggests an open-loop gain of 70 dB with a bandwidth of 46 kHz, so its GBW (52 MHz) isn't actually that low... OLG is just maxed out at a moderate level in the entire audio band, a bit similar to some video / high-speed opamps. Slew rate (36...49 V/µs) actually is more than decent, too, even by the stricter standards of JFET input parts. So this part follows the old "if you can't make distortion completely inaudible, make it flat across the audio band" philosophy. Things may have looked substantially more competitive in a 19/20 kHz IMD test. But whether it would have been able to get close to e.g. the LM4562, I really have my doubts...
If they are still using a single-stage folded cascode architecture, that would explain a number of things... most discrete implementations of folded cascodes seem to be utterly disappointing compared to their IC counterparts. (
*cough* Marantz HDAM modules *cough*)
In sum it's not surprising that the
product page for these is audiophoolery galore, you certainly couldn't market them on the strength of their measurements...