Blackness - how quiet does it need to be?

It may have been covered to some degree by your discussion in the
context of digital, but to me, i think it goes way beyond hiss- it is the
resolving power of the system to pluck out the nuances of the music or
voice against the background at low levels. The more 'absent' the
background, like ambient noise, the better you can hear all the little things
that are going on in the recording. (rough analogy would be trying to hear a
conversation in a noisy room, compared to quiet space). i think there is a
tendency to
turn up the volume to get more reality, and while that makes it 'louder,' it
doesn't make it more real sounding- in fact, sometimes the opposite
happens; the natural volume of the recorded performance is exceeded and
rather than sounding more lifelike, the system sounds like a sound
reproduction system. So, one 'test' (not really a test in the scientific sense,
but sort of a 'by thumb' way of getting the sense of a system) is how well it
resolves stuff at low volume. Not 'detail' per se, but the nuances, the
shading, the dynamics (differences between loud and soft) at a lower
overall volume level. if a system can do that well, i think it has a low 'noise
floor' in the sense that the music is presented with less ambient junk
around it and
stands out more clearly against the so-called 'black background.'
The point above about the 'natural volume' is a little different, and is just
that there seems to be a db level on playback where the recording just
sounds right, given how it was made, your system, room acoustics, the
type of music, etc. Kinda 'clicks in' just like getting the VTA on a tonearm
has a 'right spot' for a particular record.
And while the recording itself may have nothing to do with the noise floor of
the system, a system with a lower noise floor should be able to benefit
more from a good recording, if the above makes any sense.
my 2 centavos.

But tape hiss from an older recording is not really about the 'noise floor' of the system.

I read the Steve Hoffman piece, and my takeaway is as follows (but I'm still not sure this has anything to do with the 'noise floor' of a playback system):
1. given the original limits of record cutting and playback equipment at the time, stuff was heavily EQ'd and dynamics were often constricted, essentially 'gain riding' to bring up the level of the soft passages to overcome inherent noise in vinyl playback.
2. trying to remaster these older recordings on more modern, wider bandwidth equipment, including using differeent playback equipment than a circa mid-70's JBL monitors, means that some of the fiddling originally done with the tapes, described above, has to be 'undone.'
A.If I have that right, I'm still not sure what that has to do with the noise floor of the system- perhaps you are saying that these older recordings are going to sound noisier on a modern system, whether or not remastered. I certainly hear that on old RCA records, which had notoriously noisy surfaces (but gloriously natural sound in many cases).
B. If remastered, there is a danger that the life of the recording can be lost. I have found this- and i don't think i'm alone, in the case of a number of 'remasters'- the other day, i listened to a recent remaster of Janis Ian's 'Between the Lines,' it sounded sterile, lifeless. I pulled out a 1975 pressing that i bought recently, sealed, and the magic was there. Same experience with other audiophile warhorses, like Tea. The old UK pressings are far more natural sounding to me than the reissues or remasters, including the old, expensive UHQR.
C. Granted, I'm listening to vinyl only and don't have the technical expertise to address what happens in a digital processing environment, so I'm not extending my comments to that.
D. But having said all of the above, how does this relate to the noise floor of the system itself?

NoNoise, I'm not trying to be contentious. But, I don't think system noise
floor is entirely program material-dependent. If you had a very good
recording (let's stick with vinyl since that's what I'm more familiar with),
quiet pressing, good quality recording, etc., a system with a lower noise
floor is going to reveal more that is on that recording - more music -and all
the descriptors that come with good fidelity- dynamic shadings, tonality, etc.
with less ambient 'gunk' (my scientific term) from the system itself to
obscure what is on the record.
I base this on my experience in using horns with extreme efficiency- where
the sound of the electronics, the AC power, etc. is revealed with
ruthlessness unless the system is very, very quiet. I am raising this not to
advocate a position, or to say that horns or vinyl or whatever is better, but
just to focus specifically on the question of 'noise floor' of the system itself
and how it bears on musicality of home reproduction. And, as mentioned, i
judge it not from how 'noisy' the system is at steady state without a record
playing (although that can be relevant i guess) but how revealing the
system is at low volume with music playing.

BTW, i appreciate your response.

Ralph, i was summarizing what was said in that Hoffman article that Nonoise
posted a link to. Perhaps I summarized it inaccurately, to the extent that the
limits placed by the engineers were solely due to playback limitations, not to
limits in what could be cut into the groove in the first place. Here's one relevant
passage: "The one nice thing about LPs is that there's a certain set of rules and
regulations that you have to follow. You can't add a ton of high end and you
can't add a bunch of low thumping bass, because you just can't! Unless Einstein
was all wrong, there's just a certain type of groove that you can cut. And that
groove can only have so much of this and so much of that. In one way, that's a
good thing, because that keeps engineers honest. They can't screw around. On
the other hand, it's a pain in the butt. You want to get as much level as you can
on the record, so the music is louder than the surface noise of the record. So
you want to get it as loud as possible. But in order to preserve dynamics, you
need to make sure that you're not overmodulating anything. In the '50s and '
60s the engineers just reduced the dynamic range by using analog
compression, which is what is on most EQ cutting masters of master tapes. The
ones marked "master" are the ones that they used to actually cut the LPs, the
ones that have been dynamically compromised. So that's how they got away
with it. They kept the levels above the surface noise of the record just by
reducing the dynamics. In order for me to sleep at night, I want to use the
original master tape, which has dynamics. So then what do I do? It's just
compromise."

In another part of the article he discusses how he tried to 'undo' EQ from tapes
that were mixed over older equipment.