There's a lot of truth to Inexplicable's comments. Human hearing perception is highly adaptive, which explains why many of us get sucked into perpetual tinkering for ever-reduced noise. Of course, for some of us, it becomes sport or even a livelihood.
I realize that that I have a semi-instinctive prioritized list of less & more desirable PC noises -- that I apply half-consciously when working on noisy machines that people sometimes bring to me for help with.
From worse to better:
1) intermittent production of any of the top 3-4 noises below
2) screaming small fans with bad bearings
3) screeching, whining hard drives
4) whine/scream & dynamic "chattering" of coils in PSUs, motherboards & VGA -- mostly from coils, sometimes capacitors
5) rattling/humming case panels
6) loud crackly grinding HDD seek/write noise
7) CRT monitor whine
8 ) whooshing of air turbulence from fans
This last item is the least annoying of all PC generated noises, IMO, and can simulate pink noise closely if it is at the right level & frequency balance. The tricks are:
- min 80mm fan diameter, 120mm is best
- no whistling or high pitched noise from fan blade edges (ie, rpm must be low enough)
- no whistling or high pitched noise from vent obtructions (no grills & smoth vent openings help)
- no rumbing or humming caused by panel resonances (more prevalent with aluminum; decoupling fans & HDDs helps)
- ~30 dBA or less
In other words, if a computer makes 30 dBA of mostly smooth air turbulence noise, with a minimum of other noise components, I would probably be perfectly OK with it in my room, even though it is a much high SPL than I am used to (well under 25 dBA). I'm willing to bet this is true of a lot of folks who come to SPCR in search of relief from PC noise.
EDIT: BTW, Sam, this is too funny & apropos:
One way to regain some perspective on whether all the silencing effort was 'worth it' is to go and listen to a friend's computer. I did so the other day; it sounded like the end of the world.