Originally Posted by m-fine416
GAIN match the subs. If you are lucky the peaks and nulls from room interations will partially off set.
Yes, if you're lucky...but remember, it works both ways. If level matched, both subs might have different gain settings and so energy "at the sub" might vary, but who cares if the end result equates to a smoother in-room response ?
THEN use EQ to get an even smoother response. The easiest method would be one EQ curve for the signal so both subs get the same signal and have the same gain.
One can still have similar enough gain with equal levels with maximum coupling power which would yield lowest distortion and EQ would further help your cause. That is what can be achieved.
Besides, what level are you matching?
You are making sure that both subwoofers are putting out equal levels at the listening position, or, if you're smart about it, you'll reduce the levels of each individual sub so that the sum results in less distortion.
Whether there is a peak in the response is irrelevant as that peak will still exist if you gained matched the same setup. You can EQ the result, cut down on the peak...use the coupling you've gained, coupling that normally would give you a 5-6 dB output advantage and, well, reduce the level to create a flat calibration level and you'll have very, very low distortion.
Unless all your subs have perfectly flat in room response at their respective locations, the entire idea of level matching is just tom foolery. You will end up level matching the peaks which may be different magnitudes at different frequencies.
Those same issues will still exist regardless of what method you use. See, your argument has full of tiny holes too. With gain matching, one sub is invariably going to have less output so frequency response of both is going to be "tom foolery" (not sure what that means) even though you would like to think that the energy "at the sub" is the same.
You'll still have peaks to contend with...and nulls, and all the ugly stuff you normally deal with. But you have to get around the frequency response anomalies, like comb-filtering. So you see, your method is not any better really.
EQ the subs before level matching and you lose the benefits of offsetting room interactions and you are highly likely to find the response is far from flat when you turn both subs on.
A flat response is so blase isn't it ? Not everyone wants a ruler flat curve anyway because a flat curve doesn't mean that it's perceptually flat but measurably flat which is more like deception if you think about it. The two don't coincide with each other. I, myself, prefer more of a house curve down low and flatter higher up in the mid-bass region.