http://www.techtv.com/screensavers/howt ... 28,00.htmlI do not know why you folks have not posted a link to the article on techtv's site, about Yoshi's silent pc project .... he measured it at 8db above 100hz, and 14db below 100hz! Now, that is quite! Go check it out, if you have not heard about it yet... very interesting...
I had seen it. It's also a video, as things at TechTV generally are. Yoshi applies massive amounts of mediocre damping to a Lian-Li PC6070 case (shown in recommended cases, page 2), noise baffles for fan vents, and uses Vantec fans and 3-fan PSU with fantmates(?) to quiet what appears to be a P4 system.
In the end, he claims to MEASURE the noise in some kind of "quiet room" at Dolby Labs that he says has a background noise of 0-5 dB. Not dBA. With a "Larson Davis 824 sound-level meter" he claims readings of 8 dB above 120 Hz and 14 dB below 100Hz. He says it is completely inaudible at 3'.
Forgive my skepticism if it is misplaced, but the measurements don't add up for me, and the sketchiness of the project as he tells it is more annoying than enlightening. He says, for example, that he used Dynamat, Paxmate and "as many layers of Melamine foam as I could fit." What thickness? Profile? where? how?
1) a meter than can measure down to such levels is unbelievably expensive. The manufacturer actually does not seem to specify how low a noise level it can measure down to. In the video, Yoshi mentions something incredible like the meter being able to go down to -30 dB. The rep from B&K told me I'd have to spend at least $20,000 to get to 10 dBA. -30 DB???
2) a room with that low a background noise, 5 dB!, deserves much more than a passing mention as a "quiet room".
3) no details are given of real measuring conditions -- ie, distance, actual weighting scale used.
4) he does not provide complete details of his system -- not even whether wha kind of CPU it is and what speed it runs at.
5) The drive used is a Seagate Barracuda V mounted normally. Without decoupling, the drive will cause low frequency noise that is guaranteed to go right through all the damping. I can hear a a normally mounted Barracuda from 6 feet away without any problem in my room.
It's not that I think his PC is not quiet. He has layered on solution after solution to get to his quiet level. But to measure such a machine -- down to 8 dB!!? -- is so challenging a feat that I have a hard time believing someone with such sloppy language/description of the acoustics tasks could have done it correctly.
To me, the most interesting thing about this article is that someone could go so far to achieve silence in a PC and then write an article that does so little justice to that achievement.
But it probably comes with the medium -- it is TV, after all. Let's face it -- it is a commercial site designed for a middle of the road consumer audience. Still, it's interesting to see the concept of quiet computing being promoted. I am interested to see what SPCR members think of it.
A quiet room such as an anechoic chamber is said to be about 10-15dBA from what I read, then your computer has to be considerably lower to get a accurate reading. 