Last week had some interesting discussions on the BASS-list regarding dipole placement and the pitfalls of backwave/backwall reflections. Since I had access to a big space without any furniture, I tried a couple of placements options. The backwall in this particular place is flat painted drywall over studs, no other wall treatments.
The red trace below is a 21 inch wide test baffle at 6 feet, perpendicular, from the backwall. Microphone was placed at 2 meter distance, 38" off the floor. The white trace is the same baffle/driver combo at 4 feet from the backwall and same microphone placement. The comb filter action of the echoes is exaggerated since I kept the measurement conditions the same, but that was the idea. The yellow plot shows what happens when the dipole is put 2 feet away, still perpendicular, from the backwall. Take a look at how "nicely" the comb correlates with the previous measurement at twice the distance.
The brown trace shows what happens if the baffle is angled with respect to the backwall. The distance of the RD75 in the baffle is still 2 feet from the backwall, the baffle however is angled at 60 degrees. Microphone was placed on axis at 3.2 meter to simulate a similar listening spot. This fix is convincing, the dips are pretty much halved in magnitude. The only real problem dip is at 300 Hz since it is wide. Unfortunately I was not in a position to apply diffusing wall covering to verify further reduction in dip magnitude. (The gain on the brown trace was reduced with 6dB as to not interfere with the other traces on the plot.)
BTW, angling the baffle with respect to reflective surfaces in the back, is also recommended at larger distances from the wall than the 2 foot shown here.
Measurements at 16k MLS with 4096 FFT on a 12 millisecond time interval. Smoothing set at 1/12 octave.