Matthias:
I measured the frequency response at the STAX SRM-1 MK2 Pro's output:
Flat line, within <0.02dB 5Hz-20kHz, completely independent of "Level" (that's the label) control setting, even in the most critical -6dB position.
This measurement was done with Fluke 8060A, no risk, this thing is unbreakable.
A setup for spectrum analysis via the STAX SR L700 and a Bruel&Kjaer Artificial Ear 4153 did just show that neither setting has visible distortion components.
This measurement was limited by the low audio level related S/N ratio of just about 70dB, so only shows that nothing really bad (Sebastian: no obvious "quirks") is going on.
Then I repeated the audition under more controlled conditions
I built an adjustable "stopper" for the STAX level control.
This way I got exactly repeatable level settings with 50dB between full scale and dialed down.
Now I could do (sighted only) A/B comparison.
Everything else as described before.
Result: same as before.
Even better gripe on it now that I can fast and easy switch back and forth.
Too obvious difference to be just imagination.
I have to admit:
You have to be concentrated and relaxed.
Need a quiet and undisturbed surrounding.
The listening level is really low, ca. 70-75dB SPL Peak at max.
If I go up with levels only moderately the difference diminishes, I tried that.
My current approach for an explanation:
When I dial down ADI-2 Pro to -54dB a lot of the musical component are, of course, way below that level, specially with the music program I choose.
Lot's of space between the notes where you can hear sound decays, ambience and reverb, all kinds of micro-details.
I'd estimate these being down to -100dBfs or lower with ADI-2 Pro's -54dB setting.
Might be the DAC chips resolution isn't sufficient any more down there.
The OPA's might not distort at low level (I'd even question that for push-pull output stages), the DAC certaily does if it has just a few bits left to work with.
I have to admit this all is just a humble attempt to find an explanation.
I'm thinking about a way to measure that.
Might be I need to revoke and adapt my measurement system for capacitor distortions that can analyze distortions down to -155dB.
I'll have to find something to deal with the very low levels.
Sebastian:
Your statements about stuff "known to be quirky" made without personal experience and, quasi quoting without naming the source, doesn't help much.
There is no "common knowledge" in the audiophile community, it's 99% mumbo-jumbo without underlying facts.
People throwing around technical talk without having any technical insight.
Red cables sound warm, blue cables sound cold ... you got what I mean.
I'm trying to find evidence for what I can clearly and repeatably hear.
I know I am scratching on a base that Matthias an his team has established, and I don't do this lighthearted.
I'm an objectivist, have the knowledge and the equipment, and will do my best to do exactly that: finding objective reasons for what I hear.
You as a colleague, are 1000% welcome to chime in with whatever idea you have, but I'd kindly ask you to do me the favor and leave out hearsay please.