When I first started trying to calibrate in the standard fashion I found problems: some of the usual sources were not reliable down near the subsonic, measurement/meters that were not reliable near the subsonic, and low frequency standing waves in the room that varied the reading on any measurement device within an eight or ten decibel range depending on where the measurement microphone was placed. This last problem means that there is no one point in the room that is accurate for all frequencies the subwoofer will be reproducing.
You might think me a bit nuts, but I’ll tell you what I did.
By a small margin, the more important place to have balance between normal loudspeakers and subwoofers is at the crossover point, so …
I moved my subwoofer to a location that seemed to result in a good response without being boomy. After calibrating the regular loudspeakers, I took one of my front main speakers and placed the woofer within several inches of the sub’s driver. I temporarily wired the main speaker with the leads reversed so that the main speaker would be 180° out of phase. I used a PC based Real Time Analyzer (even the free versions usually allow sine wave tone generation) to generate a simple sine wave at the crossover frequency I had chosen for the main speakers, in this case 80Hz, and fed that signal only to the main front channel I was using. With the processor’s sub level set at 0db (not zero output, a mid level output that is neither + or - ), I adjusted the subwoofer phase and level so that the main loudspeaker and the subwoofer’s acoustic output canceled each other out. I put the main loudspeaker back where it belonged and changed the wiring back to normal.
I now knew that at the crossover frequency, both the main speakers and subwoofer would each be contributing the same amount of acoustic energy into the room. This gave me a good starting point. The particular mix of the standing waves in the room would not mean that at every point I would hear the main and sub equally, but this should average out as I move about the room.
As a final check, I played back sections of music that had solo instruments, or sections that for a few moments brought forward certain individual instruments, that produce a mix of frequencies, some of which will be in the subwoofer’s range, some of which will be in the main speaker’s range. Examples: bass drums, tympani, string bass, electric bass. Knowing what each was supposed to sound like to my personal listening/interpretation system (ears/brain), I listened from various locations in the room to see if in each selection that one instrument in question sounded correct. If other instruments were playing at the same time, I tried to pay attention primarily to the one instrument. Was the low end too strong or too weak compared to the upper frequencies of the same instrument? Depending on my location, I would have to raise or lower the subwoofer level a bit. In the end the average subwoofer level was, in my situation, about –2db down from my ‘cancellation adjustment’ level. I think this was primarily due to the fact that, in their usual locations, the low-frequency ‘placement gain’ for the subwoofer, on the floor, was greater than for the main speakers resting above the floor between sitting and standing ear level.
While this method is certainly not definitive, it helped in my ‘real world’ application.