Deep_333 wrote: "Why do you guys pursue a flat response when you buy a subwoofer?...
The minute you put that flat sub in your room and take some
measurements, it is anything but flat
(it's a rollercoaster with all kinds of peaks/nulls etc, EQ to the rescue)..."
Deep, I come at this situation from the other side of the fence; I make subwoofers.
At the risk of overgeneralizing, I see the room as typically doing two things to the response of a subwoofer:
First, the usually room imparts a gently rising trend to the subwoofer's output as we go down in frequency. This is because as we go lower and lower in frequency, the room's surfaces become closer and closer to the sub in terms of wavelengths. So we get progressively more and more approximately in-phase reinforcement from the room surfaces as we go down in frequency. This trend has been called "room gain" (which may or may not be technically correct), and 3 dB per octave below about 80 Hz is an approximation which has been suggested by a couple of different researchers (Martin Colloms and a woofer designer whose name slips my mind at the moment)... the exact figure of course depending on the room's acoustics as well as the subwoofer and listener locations within the room.
The second thing that happens is, room interaction imposes that roller-coaster peak-and-dip pattern you mentioned. The specifics of how the subwoofer's output is altered by the room once again depend on the room's acoustics, as well as the subwoofer and listener locations within the room.
Imo the first issue can be addressed by designing the subwoofer to have a native frequency response which slopes gently downwards by the approximate inverse of "typical" room gain, or 3 about dB per octave. I'm not saying this is the only way of addressing this issue, but imo it results in a reasonably good starting point.
The second issue (rollercoaster in-room response) presents an interesting challenge. Opinions vary on how to address it.
I think "flat frequency response" is a reasonable target for a subwoofer system, as long as we're talking about the ACTUAL IN-ROOM response. I do not think a "flat" frequency response which ignores the room's effects is the ideal starting point.
Duke