Bill, is it the methodology of the recording you don't like?
Without the $$$ for an anechoic chamber, ultra sensitive mike(s) and spectrum analyzer this seemed like a very sane way to come up with a comparison. I would
guess that most folks who are water cooling PCs, and who are interested in quiet operation, have at least one 1046 or 1048 in their history. The 1046s in particular are nearly inaudible. Yes, they shake a bit, and can resonate a case, but Roscal made an effort to eliminate/control this. Yes, there were a number of things not controlled for ( I can think of: differing weight/base area ratios - so differing vibration isolation; differing axes of rotation - so possible different "lobes" of high/low noise emanating - issue if microphone directionally sensitive; microphone (or recording equip't) response not "flat" across frequency ranges at low levels; possible transmission of sound from the inlet/outlet hoses - would have used the softest silicone I could have found) but they are all fairly minor. But my major question was, indeed, around the D4. I've never herd one, so I'm going on indirect evidence. You had mentioned that they were around 30dB and there isn't enough difference in volume between the 1046 and D4 to account for a 7 to 10 dB difference in sound levels - so maybe some sort of automatic gain was on (?) making this recording inappropriate for judging relative sound levels.
Roscal - thanks for going to all this work! You've got me working on my very rusty (haven't used it for 30+ years except to order beers on some carribean islands) French - and it wasn't very good to start with
And thanks for including a link to procooling. If nothing else you're making folks here a bit more sensitive to what non-English speaking folks go through trying to be part of
this community...