I had the house to myself and the day off work, so I dove in deep with Dirac today. Here how the Flex fits into my system:
View attachment 183429
After first spending a bunch of time comparing crossover points and listening to them, I settled on the LR 24dB/oct at 80 Hz for crossovers. I also put a BW 24dB/oct high pass filter on my subs at 18 Hz. They are only 8" and there is no point wasting any effort down there. The first test was whether DSPing the subs makes any difference. Here are my subs with and without EQ:
View attachment 183426
And the pic below is what dirac corrected each one to. Ha! EQing the subs was a waste of time other than for learning. This is freaking impressive! I think it's fair to say any differences in this graph are from me moving the mic during the different measurements and process. Dirac apparantly knows what it wants to do with my setup in my room and willl do it no matter what I did to my subs:
View attachment 183427
This is the uncorrected respose vs the default correction that Dirac does with my system from nothing but crossovers set. I think it's quite impressive and it's quite a bit better than I was able to do on my own given only 10 bands of EQ per channel in the Flex:
View attachment 183430
But Dirac lets you go further if you want. By making quite a few corrections to the target curve, I was able to get a flatter response than I was ever expecting to pull off in my room. This pic shows the adjustments to the target curve in Dirac:
View attachment 183431
Every single point on the curve was confirmed and dialed in over dozens of REW measurements. You don't have to dial in any numbers, just ad points and drag. You can zoom in super close if you need also. I think the interface is pretty slick. The comparison below shows how much I was able to dial it in on my own on top of Dirac:
View attachment 183432
And because comparisons are fun, below is the original uncorrected response from the image above compared to the final corrected response I was able to get in Dirac alone. I have zero EQ settings in the miniDSP. It sounds fantastic! Dirac was 100% worth the $200. Zero question at all in my mind.
View attachment 183438
Yep, and this is definitely as loud as I'd ever listen on this system. I didn't test louder because I have no interest in damaging my subs for cool graphs.
View attachment 183439
I spent many hours on this today, but the final result was generated by:
1. Set delay/levels/crossovers in miniDSP plugin
2. Run Dirac, export correction default curve to Flex
3. Measure in REW and see what it did
4. Adjust target curve, save to Flex, then repeat between this and step 3 until happy
If I had just done that from the start, less than 2 hours.