Wanted: acoustic "noise" source
 

similar (in effect) to the one shown here;

McIntosh Reverberant Room

to use for measuring/calibrating "room response". Something "electrical" rather than mechanical would be fine, as long as the acoustic power remains uniform, and the directivity doesn't change too much . . .

Simple, portable and gotta be cheap . . .

That one is a squirrel cage fan. So why not do that?

I think a fullranger on each side of a cube would work pretty well, probably better then the fan on the high end - and EQ-able ;)

I actually have a genrad sound and vibration analyzer ;)

Not all "squirrel cage" fans are created equal (I have several, and they "sound" different). Finding the "right" one (or one with a "calibration curve") would be fine with me . . .

Too many resonances, probably, and I'm not sure how one would deal with the power response irregularity when they start beaming (in other words, how would one calibrate it?). The wind turbulence noise may show more spectral uniformity?

I'd like a tool . . . not another "project" . . .

Will be interesting to see if you can find a cheap tool like this that isn't a project.

I have to admit, I don't quite get what they were trying to achieve with the secondary noise source. Is it a reference floor that is extracted from the DUT output? We have far better ways to understand polar radiation.

If one had the noise source, then you would need to build the room. A far harder and more costly project.

Without a more clear explanation of *what* you don't understand it's hard to help. In the McIntosh lab the noise source provided a reference standard for power response measurements. What I'm looking for is that, and a way to (easily) measure the absorbtion profile of a room. Anything with a known power output spectrum would do, but a flat spectrum makes it easier, since the room can then be read directly with a spectrum analyzer (without having to apply a correction curve).

So, you are looking for basically a perfect omni-directional pink source and known power. Not sure one exists. Well, a fan is pretty clever for the 50's. Probably a bit more white than pink. Might explain some of Mac's speakers :) I am guessing they were trying to save the absurd cost of an anoeic chamber.

I believe the current thoughts are to use very short bursts that are less that the reverb time, and to gate the measurement. This basically eliminates the room from the DUT. I have never been a fan of broadband noise testing. About as useful as pure tone. I am sure others have their views and preferred methods. I am afraid the "easy" way we are all searching for still eludes us. It might be worthwhile to note, Joe D'Appolito gets buy with his basement and a few old surplus cotton mattresses.

Doesn't have to be "perfect" . . . that makes it a quest for the Holy Grail. "Known power" is more important that directionality (within reason, of course).

No, they were looking to directly measure power response, since "guestimating" it from averaging multi-axis measurements was even more tedious then than it is now.

That's fine for single axis measurement of loudspeakers, not so fine for power response measurement, and all but useless for measuring a room's absorbtion spectrum.