Ok well first, take a look at my just posted article for Audioholics. It goes into some of this:
https://www.audioholics.com/room-acoustics/accurate-microphone-or-ears
But more specifically, These room correction systems are far more sophisticated in how they derive their correction filters and the kinds of filters they use (other than ARC). The main guys behind Dirac and Trinnov have both authored technical papers that give a lot of insight into how their algorithm works. They should make clear that the use of minimum phase filters derived from basic in-room measurements with no special analysis or processing does not yield good results. John didn't design or intend for people to use REW to do this, he just didn't restrict it. I can't speak as to why, I'm sure he just wanted to provide a tool and let people use it how they wish. All the while providing guidance and suggestions on it's best or most optimal use. His suggestions have been to be very cautious about applying any EQ in the mid and high frequencies.
I think it is OK to rule out the use of EQ of this type (minimum phase) using basic in-room measurements. We know more than enough scientifically to know why that is a bad idea. What I would say is that we shouldn't rule out all forms of EQ or response correction through DSP based on in-room measurements. That is because the technical reasons for not doing it can be resolved if we use different filters and more sophisticated processing of the measurements to correct only the correctable. Especially as we move to a MIMO based system that truly allows us to actively correct the room's acoustics. It IS my understanding that Geddes (through a decade of having this debate) and Toole have primarily said they believe that this kind of room correction is a problem. Both have given their reasons, and they actually echo my own views mentioned in the Bi-Amp thread. It isn't that the sophisticated methods don't work as advertised, it's that its a band-aid to fix a problem that is far bigger and more important. That is, most speakers are poorly designed. The EQ cannot fix the design flaws and it seems that both Toole and Geddes feel this makes EQ a distraction. That it isn't helping the industry move in a better direction toward better-designed speakers. True...but I still think it is ok. I think we need these correction systems because very few people will totally dedicate a room to this hobby, very few will treat their rooms properly, and even the worlds most perfect speaker system cannot address the LF problem.
So to be clear, I'm not against room correction or full bandwidth correction. I'm against improper use of said correction. The most egregious misuse is of REW, where improper room measurements are taken and full bandwidth correction auto-calculated for minimum phase filters and output to something like a MiniDSP. This approach is not an intelligent correction method like DIRAC, Trinnov, etc. It uses dumb measurement which are incapable of discerning measurement artifacts from real problems, conflates the correctable with the uncorrectable, and applies a blanket fix.
The last thing I'll say about this is with regard to Dirac, my favorite of these systems. I have been a beta tester of DIRAC for close to a decade (if not a decade) and I have the best and most current version of it on my laptop. I use it often (including having tested the newest edition with updates to the algorithm) and I have a relationship with DIRAC for this purpose. With problematic rooms and speakers that have a less than desirable tonal balance but otherwise good behavior, I find DIRAC to be wonderful. With more ideal speakers and a dedicated treated room, I find it makes very little difference. In fact, I did a listening test with
@shadyJ as well as a couple friends who are die-hard DIRAC lovers (Separate events). All of them commented they were not so sure they could hear the differences when it is on or off and when they could, weren't sure it was an improvement (And I'm sure James can share, I was absolutely doing my best to bias them pro-Dirac). Geddes has told me in the past that this is exactly what he would expect to happen. That once the speaker is already good and the room issues are addressed, he wouldn't expect the correction to make an audible difference. Yet the difference in the measurements between the two (on and off) are profound! DIRAC has also introduced errors in the past and I don't know that most people would have caught them. I just knew what was going on and what to look for. I'm not sure if I'm allowed to share details of what happened, so I won't, but at least take my word that all of these systems have proven to sometimes make things worse. It might be an error rather than an inherent flaw in the approach, but that doesn't mean it's ok.
