there is a fundamental error that runs through all of your practice for crosstalk (and will be finely checked for distortion)..
c is to use these sweeps to measure f-r ... for something other than what they are made for .. clearly they are generally not treated for the rest ..
explains the evidence that its all these aberrant measurements of crosstalk.....
we encounter the same kind of thing with 1k ref level signals.. which are also essentially dedicated to.. the "level"
normal.
i gives you two "school cases" that I have just rechecked..
xg7001
if on sweep for "response in frequencies"... the results are catastrophic in crosstalk, you noticed it constantly..
if we use the two 1k refs that denon presents as "clean".. the results become very correct and coherent LvsR..
same tonearm cartrige
hungaroton:
1k ref 8cm "for level"... a huge imbalance like you come across so often...
but
offers an astonishing battery of tests dedicated to measuring crosstalk.. in the end, it's not common..
I have a 40db..and balanced! L>R R>L..
I fear that it is necessary to pass in general by dedicated tests and often on spot frequencies etc..(even the few disk sweeps that seem less inconsistent should be controlled via spot frequency etc.)
we are not in digital sweep practices etc.
Sorry