There should be an ASR measured hiss chart, as this is significant for active speakers.
Agreed. It's one measurement, 10 cm in front of tweeter. If existing mics should not adequately cover <<20 dB(A), we would have to come up with a solution for that. I would not be averse to unconventional solutions, like getting a common Rode NT1 calibrated (hmm... couldn't one even use the Klippel for that? Backwards, basically?) and using that in a distance of 20 cm to keep proximity effect at bay. That would still be netting a better SNR than any of Rode's small diaphragm models at half the distance, or even a Neumann KM184 for that matter - we are talking 4 dB(A) of self-noise vs. 16 or 13 dB(A) at best, so we can afford doubling the distance (-6 dB) to offset any geometric disadvantage.
Potential snags: (a) getting ambient noise down (being out in a field in the snow tends not to be an option for much of the year, and I'm not sure whether it snows a whole lot where Amir lives anyway), so construction of a mini anechoic chamber may prove necessary, and (b) level calibration of the whole chain, but that's a normal problem.
Hmm, I wonder whether one couldn't use two mics to tackle the ambient noise issue much like human hearing does - place them in an advantageous way so that processing can easily be applied, and aperture has effectively been doubled, so 3 dB SNR gain right there. (As an example of the kind of processing I'm thinking of, there's this neat Foobar2000 plugin called
foo_dsp_centercut that uses FFT magic to extract the exact center of a stereo signal, which cleans up stereo rips of noisy mono LPs substantially better than just extracting L+R. Isn't that basically beamforming for audio? There must be something out there on that, my almost decade-old Dell notebook's sound driver implements it for the built-in pair of mics.)
Hmm. I wonder whether the Klippel couldn't do that, too. Could you cover a whole conical segment entirely with as many positions as that takes and then average the results with suitable weighting / delay to obtain sort of a very directional "super mic"? I mean, the noise we're trying to measure isn't
coherent, so that could be a problem, but...
Sorry, that was rather more rambling than originally intended. On the FX50s themselves, I do wonder whether it wouldn't be useful to compute dispersion results for smaller effective distances than 1 m as well in order to illustrate the usefulness of coaxes in such scenarios. A coax should retain sensible dissipation at 0.5 m or even 0.3 m, while any non-coaxial speaker would eventually fall apart with a typical midrange dip sooner or later. It is quite obvious that e.g. the 305P or T5V (let alone their bigger siblings) aren't exactly optimized for such short distances.
@amirm, I do not remember offhand whether you were or were not keeping the raw data, as I do recall it being in the gigabytes?