What was your power amplifier(s) and speakers again? (Also, sources - how many balanced / unbalanced?)
As a user, you are interested in two things only:
1. Will I have audible hiss?
2. Will distortion remain inaudible all the way to, say, 110 dB SPL?
(That is, with frequency response already out of the way, of course.)
Unfortunately, JA's test protocol for SNR measurement is outdated / incomplete and hinges on some assumptions that are not generally true for amplifiers with balanced inputs. (Basically, it relies on the input going more or less straight to the volume pot, as often done in traditional hi-fi.) His results will only give you a
worst-case estimate:
108 dB(A) re: 1 V output for the Benchmark is likely to result in inaudible noise in many scenarios.
92 dB(A) re: 1 V output for the Bryston is likely to be quite audible.
That said, have you ever met anyone who was listening with the volume cranked all the way up? That's what this test is doing.
What you actually want to know is more along the lines of: What's the output noise level when e.g. 2 V of input will produce 1 V of output? Or 100 mV? (Much like the way that power or integrated amplifier SNR is also being tested @ 1 W / 8 ohm, 50 mW / 4 ohm or 5 W / 4 ohm levels.)
Unfortunately, we don't know. The
briskness noted for the Bryston's volume pot
may indicate a two-stage volume control, and noise levels are likely to drop into unproblematic terrain. Still, the specs say -108 dB relative to full output, whatever that may be - it's not specified. This generally is way beyond 1 Vrms, so could result in some truly mediocre noise levels - I find that kind of hard to believe though. Man, I hate it when stuff doesn't have a decent set of specs. The bloody thing costs over 6000€, is that really too much to ask for?
What I can tell you right away, however, is that anything that manages 137 dB of SNR is pushing the physical limits. I'm assuming that's with the Benchmark's gain turned down to 0 dB (and may be purely synthetic, i.e. maximum output divided by minimum output noise), it's a bit hard to achieve otherwise. In any case you'll probably have to worry about your power amp's noise levels first. Running some Klipschorns or even more sensitive speakers should not be an issue.
There's only one thing about the Benchmark's results that I didn't like - 600 ohm load driving is not reference grade and could definitely be better.
One thing to consider: We're talking just preamp / source selection now. No room EQ or anything. Where is that supposed to be happening, preferably for all sources as well? A super high spec preamp is all good and well but you definitely shouldn't be ignoring the elephant in the, erm,
room. That's something plainly audible, unlike whether distortion is 90, 100, or 120 dB down.