Sub 1+2 raw actually looks pretty good. That null - while deep - is extremely narrow and won't be very audible.
When say reversed polarity do you mean you set the phase to 180 on one of the subs? Those two aren't the same things, actually.
If you are experimenting with the phase control, you might want to try 15-degree increments to see if somewhere between 0 and 180 looks better.
Thanks so much, Ed, for making time to reply. By Inverted Polarity, I'm referring to the "Invert Polarity" button in the MiniDSP. I was doing this because from the reading that I've done, my understanding is that, when using the Impulse graph in REW to time-align subs, we wanted to achieve two things.... first, is to make sure that the first bump in each of the impulse graphs points up (indicating the positive polarity), and second, we want to align/ overlap the tallest peaks (the ones that reach 100 on the verticle scale). In order for me to do that, I had to use the MiniDSP to first invert the polarity of sub 2, then I added delay to that same sub by measuring the distance between the two tallest peaks, and using that distance as the delay amount. But as you can see from the graphs in my first post, that strategy didn't work so well for me... and maybe I'm doing something wrong because as you pointed out, the graph for Sub1 + Sub2 RAW looks better than the tweaked one, and in that RAW graph the first bump for Sub2 is pointing down, which I understood to mean that the polarity was inverted.
So I've gone back to the drawing board, and this time I've also changed the location of Sub1, from the front left corner to the middle of that front wall. Sub2 is still in the rear right corner. As for the tweaks, I decided to use the first small bump of each Impulse graph to determine the delay as opposed to using the tallest peaks. I also didn't invert the polarity of sub2 via the MiniDSP.
So after calculating a delay of .7ms using the difference between the two small initial bumps of the impulse graph for each sub (using the Overlay window), I entered that delay into the MiniDSP and took more measurements at the MLP, as well as at the other 4 seats.
After taking measurements at all 5 seats, I used REW again to average them out, and then I used the EQ filter feature in REW to flatten out what I believe to peak in the Averaged graph. I then imported the calculated EQ filters into the MiniDSP and took more measurements. The one "Peak" which I did not attempt to EQ was the one around 62hz because I felt that doing so would create a bigger null in that frequency for the 2 rear seats, so I compromised.
I've attached pics of the new measurements and would love your feedback. I'd also like to know if I'm correct in assuming that anything that falls above the SPL level measured at around the 20hz mark in my graphs should be considered a peak and if anything below that SPL should be considered a null. I ask this because that is the SPL level that I used as the target when setting up the EQ filters.
If you notice, the Sub1 graph for the original front corner placement looks beefier than the new location at the front, mid-wall. I wonder if I should have kept it there and just done the time alignment again but this time just using the initial smaller bump in the impulse graph?
I'm all ears.../forum/images/smilies/biggrin.gif