Right then - I have the results, but frankly I'm at a loss to explain some of them.
I'll detail in full the test conditions - because hopefully somebody will spot the flaw I must have made to produce these results. Clearly, I have no ability to measure each mics self-noise in any quantitive way, so following Dave's suggestion, I took a DMX terminator, and replace the 680Ohm resistor with a 150Ohm. In my studio, I set the mic pre-amp in my Tascam 16/4 to the setting I would normally use for fairly close in VO type work, which is around 60% rotation of the knob and never a problem, noise wise, even with the least sensitive mic I have - the Shure SM7B.
I then plugged an XLR cable into the other studio, where there is no machinery or noise generating items. I set out a towel, centre of the room to place the mics on, one after the other, and set up a video camera on a mic stand. There is a very low level noise source in the room - an almost unperceptable low frequency rumble from the air handling, which I forgot to turn off and I think I can hear in some of the clips, but not in others.
The test procedure was as follows. Starting with the line terminator, I put Cubase into record, and the same with the video camera. Then, one by one, without turning off phantom, I simply disconnected and reconnected each mic. I announced each mics name from around two feet away, again - hardly audible I discovered on revising the recording. Each mic would have a loud click or pop as the XLR went in, then some handling noises as I positioned it on the towel, then a quiet announcement of which one it was. I then stayed as quiet as I could for a few seconds, then repeated the process with the next mic. NO adjustments to the recording at all.
On the conclusion of the tests I brought into cubase the video file and slid it to line it up against the audio recording. The audio from the camera was not used.
Then, clip by clip, I edited out the clicks, and handling noises, and my announcement - just leaving a short recorded silence. I then Moved these all up to each other producing a noise floor that goes up and down - and is produced by the microphone - the preamp and system noise remaining constant throughout.
I could see on the screen a few had defined raised noise floors, but at this time I didn't;t do anything about it. I then exported this audio track to a file on disk. All this took place at 44.1KHz, 32 bit sampling - which is the current template I'm working with for some CD products.
This file I then brought back into Cubase and normalised it to 0dB, making the noisiest microphone virtually full scale, and then the quieter ones able to have a reading taken from the Cubase LUFS meter I now use lots, thanks to this forum.
I tabled the results and they are as follows:
150 ohm resistor -42.2
shure sm57 1970's -35.8
shure sm57 modern -38.2
shure sm58 -40.2
shure beta 58 -23.1
shure beta 57 -41.3
Beyer M201 -40.2
shure sm86 -33.8
thomann sc-100 -21
samson c-01 -16.2
oktava mk-309 -18.8
chinese mc-100 - 10.6
akg 414 -2.8
at 2020 -22.2
akg d112 -38.6
akg190 -32.9
thoman condenser -21.7
chinese mc10 -0.1
shure sm7b -39.4
electrovoice re320 -37.7
None of these mics cause me any issues when I use them, but I'm at a loss to make sense of these results. The second noisiest mic is the AKG 414 which makes no sense, but I re-ran the test and got the same results.
There MUST be a functional and basic flaw in this test, but I cannot think what it is?
The video showing what it looks like is
The second half is the sequence without the normalisation.
There are a few other issues. There appears to be problems with at least a couple. The Shure Beta58 when the gain is ramped up in the normalisation process brings with it some strange noises. Some of the noisy mics have constant noise, but others it changes in volume and texture?
I'm hoping that some of you can find the flaw in the test process, or if not can maybe replicate the results with their kit.
My thought process was simple - if I found the noisiest clip, then increasing this to maximum level by adding gain in the normalisation process would raise EVERY clip by the same amount. This worked as expected on most, but the noisiest mic I believe to be actually faulty - you can hear it's noise when used at normal recording distances and gain - BUT the AKG 414, the second noisiest mic in the test is NOT a noisy mic - but where is this noise coming from? The dynamics are all less sensitive and have low noise figures, bar one. The two Shure 57s with nearly 40 years between them are close but not identical. The quietest Shure condenser the 87 does pretty well compared to others.
Is the test flawed in a way that I haven't spotted? Clearly, they all have different sensitivities, but the gain setting on the preamp was where I expect give or take a bit. What has gone wrong? Clearly my expectation that the differences would be small were totally wrong - I really didn't expect these results at all. Mics I figured would be less good aren't and ones I really like doing less well.
I really need somebody else to do something similar and see if the results follow the same trend.
I very strange set of results - the conclusion just differs from the expected!
The noise for each one is different - which is interesting.
Sorry for the on screen watermark - I thought I had the full version and turned out to be a demo!