L+R measurement questions

[FairlyGreatContent]

Thank you for this amazing tool! I've been playing with it the last few days and watching a bunch of YouTube videos, and so now have questions about my stereo speaker measurements (monitor bookshelf speakers).
When measured individually, each speaker appears to be fine, but then when I measure them together L+R, I'm getting these spikes above 300 Hz. The first graph shows each speaker measured individually, and the second one has the L+R overlay in orange. Is this a candidate for room treatment or Dirac?
Front wall 18 feet
Side walls 30 feet (right wall has 10 feet sliding glass patio window at first reflection point but covered by thin roller shades)
speakers 4 feet away from front wall and 4.5 ft from sidewalls, and 9ft equilateral triangle to listening position.
Lots of sofa furniture and giant 10 x 14 area rug in front of speakers.
I also attached the mdat file

 

[FairlyGreatContent]

oops, just realized I cut off the dB scale

 

[FairlyGreatContent]

And does this RT60 measurement mean that my room has too many reflections since i'm at 600m or above between 1k and 5k? I've read that I should be under 500m, ideally, correct?

 

[FairlyGreatContent]

wow I just threw in some 3 inch thick broadband absorption at sidewall 1st reflection points and completely eliminated the spikes. This is L+R measurement with the absorption panel (rock wool).
So is the only way to solve the dip at 600 to add a subwoofer?

 

[John Mulcahy]

When speakers are measured together differences in the distances from speaker to mic causes comb filtering (areas where the time difference in the signals from the speakers corresponds to 180 degrees phase shift, causing cancellations). The RT60 looks a little high, but not unduly so for the size of the room. Dips tend to be sensitive to measuring and speaker position. A sub shouldn't be doing anything at 600 Hz.

 

[FairlyGreatContent]

When speakers are measured together differences in the distances from speaker to mic causes comb filtering (areas where the time difference in the signals from the speakers corresponds to 180 degrees phase shift, causing cancellations). The RT60 looks a little high, but not unduly so for the size of the room. Dips tend to be sensitive to measuring and speaker position. A sub shouldn't be doing anything at 600 Hz.

Thanks John, I'll continue to play with speaker positioning. So in actual listening, do we hear these comb filtering effects from L+R response graph, or is that an artifact unique to the microphone and I should just make individual L/R measurements then average them?

 

[John Mulcahy]

The summation that occurs at the microphone also occurs at our ears, but it is a normal part of hearing multiple sources. Smoothing is the most straightforward way to reduce the visual clutter of comb filtering on the graph.