I read this test a while ago; well before I had come upon this site. And at the time I found it frustrating partly because it sets out to be a "scientific" test, and then in the end, it is so vague and inconclusive. I'm glad for the description of the method -- I'd bet a dime against a $1 that the meter that they were basing the whole thing on is the Radio Shack digital sound level meter. I own the analog RS meter (which is the same except for the meter/readout, AFAIK), and it is useful for some relative measurements, but it is hardly absolute.
Another hobby that I have since stopped pursuing as a hobby is music and hifi audio, and I used my cheap little RS meter to help tune the bass on my Snell Acoustics Type E II's. The microphone element in the RS meter is hardly a flat response item -- somewhere in this messy office where I am, there is/was a printout of the *known* errors, nevermind the individual error that is always present in mass produced anything -- including fans!
So, how does this relate to the subject at hand? Well, taken for what it is, the Directron test has some useful *relative* comparisons, but the gapping holes include no comparisons of CFM and the inclusion of thermally controlled fans (each with totally different thermal/speed curves!!) limit it. So, what does
it tell you? That some fans have more noise from whatever source, be it mechanical, bearings, turbulance, or just plain old moving air! That's it.
I think that any truly useful test would COMBINE these sorts of measurements with subjective tests. Say what? Yup, our ears are what are going to be the final judge -- because even if you bought the "quietest" fan, it might have a sound that drives you nuts. Or, it might move a totally inadequate amount of air. The bottom line is: can you hear it, and does the noises it makes "work" for you -- and does it keep your case cool enough?
BEFORE I had read the Directron test (I don't think it had been published), I had bought (from Directron) one of the (supposedly) 12dBA Papst 80mm fans and one of the Silencer 80mm jobs. (It's NOT Directron's fan, BTW. See the post above.) The Silencer made *slightly more noise than the Papst, but it also moved about twice as much air (based on my totally scientific hold-your-hand-in-front-of the-fan methodology
) and at the time, the Silencer was about half as expensive, to boot. So for me, the Silencer makes more sense, and is a much better value.
Both fans went into a "work" machine that was even too loud under the desk. The Papst went in the front as the intake, and the Silencer went in the back, and both screamers went into the box. The stock (5000RPM?) fan on the Athlon XP 1800+ (Palomino) was then the loudest thing in the case, followed by the 80GB Western Digital and the 350watt power supply. It was still somewhat noisy, but you could still think when it was running, and there was not so much the great relief when it got shut down anymore. And it remained perfectly stable, which is the point, really.