Minimum Phase

[OCA]

I have never been a big fan of what Dirac does. It might improve very problematic systems where speakers are placed very badly, etc. but with an already descent set up, the correction it applies to high frequencies (-6dB cut at above 16kHz in your case) is textbook sound degradation especially in the absence of any frequency dependent windowing. That frequency has a wavelength of 2cm and I wonder how it decided the magnitude response will not change around the LP and what happens to the peak energy time of the signal after a filter is applied.

Your room has modes just like every other and that dip at 70 Hz is probably a wall surface about 1.25 away from the subwoofer. Ignoring the way Dirac bravely boosted it, there's not much one can do about it by moving subwoofers unless you are using an active MIMO solution. But the peaks at 20 Hz and around 100Hz can be perfectly corrected and it will help the clarity of the rest of the frequency band IMO.

 

[ernestcarl]

EDIT: for the sub a total inversion is totally possible. I strongly advice correcting mains and sub separately. you can also create the perfect crossover this way

Yes, though it probably isn't always absolutely necessary or practical to do it that way. Now, important to note is that depending on the path length difference between the sub(s) and mains -- and the room acoustics e.g. boundary interference -- one will get either better or worse summed speaker alignments the further one moves away from the central MLP.

 

[gnarly]

I have never been a big fan of what Dirac does.

I have no opinion about Dirac itself, as I don't know how it works.
Until i can measure some line-level electrical transfer functions of the correction files it makes, I think there are too many variables in play to make any valid assessment..

And I must admit, I'm not expecting any significant listening improvements from it.
I got it simply to learn for myself how a leading "automated" room correction program works.

Also must admit, I'm not big on the idea of DRC room correction to begin with. I've looked deeply into Acourate, Audiolense, and DRC.
I think they get overly minitua bound with their impulse correction implementations.....often missing the much bigger pictures of how valid are the measurements being corrected, and exactly what in the measurements can be corrected and what can't.

I'm pretty sure when, or rather if, I do get into DRC...it will be a roll-my-own. I've do so much work on quasi-anechoic DIY speaker measurement and tunings, I feel confident I can figure out out to make a room transfer function, and then decide what and how much to correct.

Probably the real holdback with DRC for me is, I just don't care about optimizing sweet spot listening.
I'm constantly trying out new DIY speaker designs, sometimes mono, sometimes stereo, sometime even three channel LCR.
Plus outdoor listening will forever rule for ultimate SQ, no matter how great the room gets, ime/imso.

For now, I'm quite happy building the best large unity/synergies i can in terms of quasi-anechoic response, and then making any "room corrections" either via placements, acoustic treatments, and a combination of fixed EQs. And I love using real-time EQ as needed to adjust tracks' tonality vagaries.
Dirac is pure curiosity...

 

[Keith_W]

I have never been a big fan of what Dirac does. It might improve very problematic systems where speakers are placed very badly, etc. but with an already descent set up, the correction it applies to high frequencies (-6dB cut at above 16kHz in your case) is textbook sound degradation especially in the absence of any frequency dependent windowing.

Is that true, that Dirac does not apply a freq dependent window? That's surprising, I thought it would be a pretty basic function to include.

 

[ernestcarl]

Is that true, that Dirac does not apply a freq dependent window? That's surprising, I thought it would be a pretty basic function to include.

They most certainly do apply their own windowing algorithm... However, you don't get a lot of options unlike Trinnov, for example. Even then, the best semi-automated integration and system calibration still involves manual adjustments. Full manual "room correction" easily makes sense in smaller systems (e.g. stereo), but as channel count goes up, the task gets harder to do without any kind of help or assistance.