Looks like I'm going to have opportunity to do a side by side test on these. I forgot about a iLF218 that will be at the shop for a few days before going out on an install and I have a couple TH115's (8 ohm) in the shop already.
Just to keep the comparison fair, should I reduce input to the iLF218 by 3dB because it is a 4 ohm or parallel 2 TH115's for a 4 ohm load?
I think I'll try it both ways and see what happens. Will post results if I am able to acquire any meaningful data.
Sensitivity is one test (and remember the current model TH115 is 4 ohm) so it will have 3dB higher sensitivity.
However the max output remains the same. You just need a larger amp.
If comparing max output-consider using an amp that is capable of driving the 8 ohm load good enough.
I would suggest around 1500-2000 watts @ 8 ohms. You may have to bridge an amp to get that power-depending on what you have around.
Also remember that the damping factor goes in half when you bridge an amp.
You could get some freq response measurements without worrying about amp size.
I would HIGHLY suggest when doing freq response measurements to measure 1 cabinet-then move it out of the way-put the other cabinet where the first one was and then simply move the cable to the new cabinet. DO NOT touch any output/gain controls and remeasure.
Also be sure to move "the other" cabinet during the measurement a good distance away-since the cabinet not being measured can cause a notch in the response of the cabinet being measured.
I would place the mic on the ground 2M or so away. If you could do it outside-then further away would give a more representative response-as long as there are no reflections that would enter in to the measurement. Even then as long as the reflections were present for all cabinets in the measurement-then they would still be comparable.
Hence the reason for the cabinets to be in the same physical position for the measurement.
It is not important to get a calibrated response-since you are doing a comparative test. Just run it at a level that comfortable.