"
It is common knowledge that low frequencies (below about 100Hz) are non-directional,"
That is not a true statement. I know why people make that erroneous generalization but that doesn't make it true.
It is true that the statement they have made is commonly repeated... but it is an erroneous statement.
That is why they back pedal and attempt to explain further why placement does actually matter.
Here's what got my curiosity sparked:
If you are testing for the right and the sub is on the left ARC will build it's fuzzy logic, convolution file package and attempt to correct everything, however it has no practical way to control the actual driver discretely so what ever solution is employed will effect the entire system before the cross over.
That's the part I find fascinating... what happens to the common spectral range that is coming out of both the sub and the mains? Crossovers have a range of overlap and so there is some portion of the sound that is coming out of both the sub and the mains. It's the same sound... coming from 2 places.
I'm imagining a scenario where the speaker on the right is getting info that includes compensation for sound that is actually emanating from the left and it all gets generalized into a single "solution".
An extreme example might be if you swapped your speaker placement e.g move the left one to the right and the right to the left but didn't tell ARC. It can probably work it's magic and make compensations to make the best of that circumstance.
It seems, to me, like a compensation system could be even more effective if it treated each and every discrete driver AFTER the crossover rather than treating speaker systems as systems.
Just wondering out loud.
best regards,
mike