Sorry if it came across as disagreeing. I don't think we are :) I just had a couple of points to add which you actually do already touch on...\n1-Take in room measurements before and after audyssey. Make sure to either average several measurements in a similiar configuaration as your audyssey measurements, or even better in my opinion use the MMM (moving microphone method) of your listening position. A lot of those small variations at higher frequencies go away the more measurements you take.\n2-Looking at speaker measurements and correcting the speakers using detailed measurements (such as from Harman, or Audiosciencereview) might be the best way to do correction. I'm not currently doing this, but this is what @Jon AA is talking about. According to tool as noted in the image, the series of bumps that are in all of the measurements are the ones that can be EQ'd.\n3-And I just wanted to point out to compare your before measurements with what Audyssey measures. Audyssey will do boosts\/cuts based on what the Audyssey measured (even if those measurements aren't accuate). I had a situation where Audyssey boosted high frequencies when I wanted Audyssey to cut. It did this because the origianal measuement didn't seem to be very accurate based on my MMM REW measurements. As you have already mentioned it is essential to measure what Audyssey has done. \nSorry if it sounded like I was disagreeing with you. Audyssey does exactly what you tell it to relative to what it is measures and is really great at it. I think people can fine tune their results for sure using what you have described.\n\n\nThank you for clarifying this. In fact, I have no problem disagreeing or agreeing to disagree with knowledgeable and open minded people like you and Jon AA, but I am merely curious about if there is something I missed or got lost in the back and forth, before agreeing or disagreeing. Based on my hundreds of REW plots, I also recognize that Audyssey, and Dirac Live (I must emphasize I was using the trial versions only) were unable to "correct" my speakers in room response at my MMP to better than 9 dB p-p with no smoothing in the 15-200 Hz range, let alone the higher frequencies. \n\nI can say, and must emphasize that their (at least Audyssey for sure) could do the correction to within 2 dB p-p for the same range (in my room) if fed with the actual results measured by REW\/Umik-1. So I would assume that logically speaking, Audyssey's algorithm to create the FIR based minimum phase filters are likely not the bottleneck, but the mic and the measurement\/data collection part is. Do you follow my logic? Or you have another "logical" explanation.\n\nNow that I am happy with +\/- 1.2 dB 20-110 Hz, I am going to start working on editing it slope up towards 15 dB, by perhaps 2 dB, and then will on the FL\/FR to extend the flattened part to about 300 Hz.