Compare & contrast: Dirac and ARC

Is anyone here who's used recent versions of both Dirac and Anthem Room Correction (ARC) and who would care to compare them? I'm specifically interested in ease of use and quality of final results, especially when using multiple subwoofers.

I've been using ARC for several years on an Anthem STR preamp, and it has been easy to use. It has pretty good flexibility and has produced good results. At the same time, there have been a few cases where results have been inexplicable or inconsistent and a few where I wish I could tweak something that Anthem doesn't allow tweaking.

I've used Dirac in a much simpler system with a miniDSP SHD and found it more cumbersome to use, but with more flexibility than ARC for user tweaks. I understand that ease of use depends on the manufacturer, and that many Dirac implementations are easier to use than miniDSP's.

I can't compare the two DRC systems very well because the audio systems and rooms are so different, it would be pointless. And I can't take the miniDSP out of our TV system, even temporarily, or the rest of the family would kill me.

So I'll appreciate comments from those familiar with both.

@JStewart - Thanks for the link to that video. It is helpful. It's interesting to hear Poes state that all these algorithms sometimes give really terrible results. That's surely been my experience in using DRC for 20 years, but I've never heard anyone else say it so plainly.

The comments on Dirac ART are also interesting (that it often doesn't work). Multi-sub optimization is a tricky problem, and we are just at the beginning of trying to figure it out. I hope it gets a lot better in my lifetime.

Sonnie said:

all depends on your room and setup as to whether it will work for you.

Click to expand...

Thanks, Sonnie. I find ARC pretty good with 2 subs, less good with 4 unless I gang them to mono and do a lot of manual work beforehand. (The shortcomings are what I mention in the next paragraph.)

I'm glad to hear you've had good results from ART. In my view, DRC (by which I mean digital room correction in general) has been stuck in a simplistic paradigm for far too long, in which it has paid little or no attention to speaker interactions (including subs). That's not because anything thinks interactions are unimportant, but because the problem is far more difficult to solve than flattening each speaker individually.

Would you care to comment on the type of room and setup in which you would expect ART to be successful?

JStewart said:

I hope it’s not long before DRC can learn from it’s results and improve on them.
It would also be very helpful to express the results as differences from standards, much like color calibrations are for a display, so that “good” would be objectively determined.

Click to expand...

Yes, a learning algorithm might have promise. It's an interesting idea.

Regarding standards, I like more objectivity, but I wonder how that would be possible. The difference between monitor calibration and sound is that monitor calibration takes place at the monitor -- we care only that it is producing the correct tones. Sound calibration has a more difficult objective, to get good sound in some sense at the listening chair. That opens the question of what "good" is and over what area one wants to correct. Is it better to use as little DSP as possible, to avoid odd effects at other seats? To optimize for one seat only? To optimize across space, which necessarily involves compromise at the center seat? How much attention should be paid to phase? How about the balance between flat measured FR and decay time? Those are the questions that come to mind immediately. I imagine people can come up with others.

I do hope that we will be able to get to a more objective place soon, one that is objectively excellent on a system basis, not just a speaker-by-speaker basis. The programs that seem to promise this are Dirac's ART and Trinnov's Waveforming. Do you agree, or maybe I've missed something? (I expect Anthem to come up with something similar before long.)

The urgent need is to move beyond the single-speaker model of optimization. I am constantly frustrated (e.g.) by ARC Genesis's inability to use speaker A to fill a dip in speaker B's response, instead of trying to flatten an unfillable dip in speaker B.

Sonnie said:

it [ART] sounds better than DLBC in my room, which has always sounded very good.

Click to expand...

Thanks, again, Sonnie. Good info. I hope to be able to try at least DLBC and possibly ART soon.