First, repeated disclaimer : I am not affiliated with anybody - on this thread or otherwise.
It is a classic situation when a long time established status quo is being seriously challenged - similar to CD vs SACD wars that flushed much of the dirty laundry on either side for the public to see for the first time - which is good.
It is being presented in somewhat - to say the least - questionable way by Eric65 . It is perfectly clear to anyone who knows what it is all about that technical/electrical background of Eric65 is not exactly his forte. Furthermore, it is that his (or mine for that matter ) native language is not English.
On the last oscillograme posted by Eric65 is not a SQUARE wave of 10 kHz, but TRIANGLE wave ( source - link provided by Eric65 ) at that frequency driving a much above normal load - 470 pF, an equivalent of approx 4 pairs of Stax headphones. It is next to perfect. So what KG wrote regarding transformer output inductance resonating with the capacitive load "around 10 kHz " when reproducing "square wave" is WAY off and misleading - but yes, if that had actually been the case, resonance at 10 kHz might look very similar to that picture. The response of 10 kHz SQUARE wave at the same level using the same load is also shown - and although not perfect, it IS better than most direct drive amps can provide using SINGLE Stax headphone or load approx 120 or so pF.
What I would like to see are the true unretouched measurements of all competing amps into known specified load at precisely specified frequencies, levels and distortion limits. Particularly is of interest the power response limit - or frequency to which is the amplifier still capable of driving the load to full voltage as attained at low frequencies < 1kHz. But take care - any output transformers HAVE to be rated for CONTINOUOS output power under such conditions, or else damage or complete failure of the output transformer WILL occur. Obviously, direct drive amplifiers share no such limitations - but in a well designed transformer it should also not be an issue.
If a set of meaningful tests for testing amplifiers for electrostatic transducers could be devised and agreed upon, there would be no waxing either way possible anymore. The amp that performs better in real world conditions wins - end of story. This is NOT comparable to amps for dynamic phones, where amp can exceed the driving needs of the load ad libitum, micro minute differences can be make or break; compared to those are electrostatic amps barely alive, and can not afford to care about make up, toenails, eyelashes, etc - to merely survive driving the nasty purely capacitive load WELL is already extremely close to the win.
I do not like the word "shill" ( which I learned of first on head-fi ...) - with behaviour of Eric65 being a hair thin line from it, it is a situation one can not be exactly comfortable with.
The same can be said regarding the misleading (mis)interpreatation by KG regarding the 10 kHz oscillogram as posted above by Eric65.
I am perfectly aware here is money/bussiness interest at stake, jostling for position in the marketplace, but please try to be fair - me personally would far more prefer "he/they is/are good, please allow me/us the opportunity to show I/we are better" than trying to badmouth ( in this case "badpicture" ) the other party or use (consciuosly or unconsciuously) misleading interpretations. To my knowledge to this day, the only openly published measurements of "anything driving electrostatics" were DIY designs of Mr. Heyser and Mr. Sanders in 70s in The Audio Amateur ( later reprinted in The Speaker Builder ) magazine. Everything else, at least what I am familiar with, was hear>say>claim>prove-me-wrong-if- you-can; and even that mostly on subjective basis. To highlight only the pros and remain completely silent regarding the cons does not qualify for open publication in my book.
I am the last to say that subjective evaluations are meaningless, but if a device does not measure at least "sound", it is not a good device no matter how much might I like it subjectively. Problem with ESLs is high voltage : the probes that are specified for voltages required and still having low capacitance/extended frequency response do exist, but are ANYTHING but cheap and common even in highly specialized labs. And yes, here the necessary precautions regarding lethally dangerous voltages/current are to be repeated: in order to measure anything, one has to gain access to exactly what should NEVER be touched. Regardless if it is coming from a direct drive or transformer coupled amp. All this means I have NEVER seen an objective test with measurements of a commercially available device designed to drive ESLs in any commercially available magazine - only circumstantial evidence, if you will, of acoustic output of ESLs being driven by said device.
I think it is about time to "pour some pure wine" ( as we say around here ) in these murky waters ESLs are forced to swim due to the technical difficulties that result from the purely capacitive load characteristics of the transducers themselves. But whatever one might wish do : this is not + - 15 V preamplifier - so use this caution seriously.