Originally Posted by notnyt
Except it kind of sucks with XT32 from what I'm seeing.
The first graph here is raw response of a sealed system compared to Audyssey XT32, the second graph is the same system after being EQ'd MiniDSP, then running XT32. I had better results with standard XT, displayed in the third graph. Fortunately, I have a MiniDSP which I used to EQ my subs flat before running audyssey which presented me with some better results, shown in the second graph. Audyssey really had to stretch for this one
Yellow is after Audyssey XT32, green is raw response without Audyssey.
Blue is MiniDSP only, red is MiniDSP then Audyssey XT32
This is an older graph from before I upgraded my receiver with Denon to support XT32. Yellow is raw, green is after running Audyssey XT
I think your results are not difficult to explain. In the first graph you do not see Audyssey raising the sub level at the low end because Audyssey does not EQ to flat below the -3 db point. Your -3 db point is fairly obvious in this graph. Keep in mind that Audyssey is used by thousands of users and has to be safe from a liability standpoint. Forcing higher output below the - 3db point could destroy a subwoofer driver that does not have the Xmax to achieve flat below that.
In the second graph there was not much for XT32 to do, so it didn't do much.
In the third graph, the results are as expected. It seems no -3 db point was found. The rise at the low end remains because I think XT's does not correct that low.
Though the -3 db point is fairly obvious in your first graph, the reference for the -3db point is determined through a fairly complex look at the sub output over its operating range. Hence in most cases it is very difficult to eyeball the - 3db point as identified by Audyssey.
I've found that shifting the mic locations within the listening bubble can significantly affect where the - 3db point is identified. I've had my system (two 15" drivers per side) left with a -3 db point of 28 Hz and then be flat to 20 Hz by just moving the mics a bit or doing less locations (with XT32).
Another consideration is that Audyssey is trying to provide the best listening experience across the listening bubble. Hence the result at any one position is not a good indication of what Audyssey is doing. Additionally, if you mic, say, 10 locations in the listening bubble you must test exactly those same ten locations and integrate the results to assess exactly what Audyssey is doing. This alone explains most complaints about Audyssey not achieving flat response. A truly flat response at multiple listening (or mic'd) locations is not physically possible in a credible theater, home or otherwise.
Tell us, how big is your listening bubble relative to the room size, how many mic locations did you use for the Audyssey setup, and how was your own testing done (number of positions, location relative to Audyssey mic locations, etc.)?
And, of course, your results are far from problematic with or without Audyssey. Your subs are equidistant and it appears you have an excellent room and probably some good room treatments. You haven't left much for Audyssey to help with. Any help Audyssey provides is icing on the cake and is probably limited to spreading any remaining ill effects across the listening bubble so there is no bad seat (though no perfect seat either).