The “Water Drop” and the “Bessel curve”…
Well it is quite amassing discovers not the fact itself, it is well known, but the amplitude with which the fact demonstrates itself.
I run the “Water Drop” at 12K second order but not against the tweeter transformer but against an 8R voltage divider. So the filter sees a more of less purely restive load (with a primary of the transformer kick in at 3.0H few octaves below). The filter does a Bessel curve and it sounds phenomenal but here is the catch… moving a cap or a coil out of calculation that makes a perfect Bessel Q, even within 10% do affect sound. Sure it is easy to re-align the driver, and I did it many times, but I never was able to get in the “off Bessel Q” the proper sound. Interesting that when the filter in the exact Bessel setting, then regardless off crossover point the driver demonstrate much less sensitivity to timing errors. Furthermore… and this is truly astonishing: the driver in and out of Bessel curve has completely different integrateability with MF channel. Out of Bessel curve the driver should be dialed superbly precise and ½ db makes it “too hot” of “not hot enough”. When the filter is the Bessel curve then the volume control on the tweeter become WAY less critical and even adding 8dB (!!!!) on the tweets dos not sound irritating. It is truly amassing how sensitive this thing is….
 ôèëüòðå äëÿ EV-350 – êîíäåðû Electrocube «Polypropylene/Foil" type “950B” 0,1ìô õ 400V, 270VAC ±10%. Part number 950B1E104K.