McIntosh MX-121 - any thoughts? - Page 9

That's an alarming number of eqs (every boost or attenuation is damaging) to apply to an audio signal.

I think the question, "150 or 121," is brand new and very few have had to ask it. The 150 was purchased by many because it was the only game in the town of McIntosh High End, not because it was a preference to the 121, which is weeks old. I suspect that we could count on two hands the number of people who've auditioned both side-by-side and made a decision.

I've come to learn that it's entirely true. That forum (audioaficionado) is a mess, and I now understand why it's a virtual ghost town filled with ultra-cautious threads. Others have warned me about it. To call it a modern version of 1984 would be an understatement. There's an unnerving amount of fear within those walls and nothing of substance is allowed. This is why Admins need to allow a certain level of distance between themselves, their personal feelings, and the range of content provided by the diverse members of the forum if they wish to be successful. If the proprietor weren't independently wealthy, and addicted to the ego trips the site facilitates, I don't believe it would have lasted this long. I'm thankful for AVS, etc.

Even digitally. Phase shifts are present with each subtractive dip, and especially with an additive boost. I know of no eq (even those which claim phase accuracy) which doesn't render artifacts beyond the pure intent of the eq (more or less of a particular frequency and its surrounding neighbors).

I'd agree that minor digital equalization can be quite unnoticeable, but some of the rooms these processors are being installed in will need correction of up to 20 or 30 db. That's wickedly destructive. I don't know how aggressive these correction algorithms will get, but if they're not fully "correcting" the problem then what's the point?

In some cases where a room is too problematic, only a combination of software and acoustic treatments will provide good results. As always, the less you need to do in software, and the more you can do with treatments, the better.

I still haven't touched Audessey on this 121, and I'm trying not to. I'm not against correction (in fact I just gave an interview to a manufacturer whose DSP products I'm using in a studio installation), but I think of it as last resort. Speaker placement first, then treatment, then software if necessary.

Because cutting distortion is better than boosting it. The result would be less of the problem instead of more of it.

I do realize that we're dealing with the minutia (though I've truly seen room correction cuts of nearly 30db, which is enormous), but at this level of quality the minutia is where we often spend our time.

I actually told the forum that I was permanently checking out after that post, though I was in no way surprised by the fact that he deleted that explanation, banned me, then took a cheap shot. It's his forum - he can create any atmosphere he chooses.

If you feel that I was supporting a swindler then you, like a few folks there, completely misunderstood my point, and entirely ignored the later ones. No matter. I'm happy to put the Orwellian drama village behind me.

I was only responding to your questions, Kal. I initially mentioned boosts vs cuts parenthetically -- I don't actually have much passion for the subject. I've been expanding upon it when prompted.

The shortcomings of room correction are, in my opinion, relevant. Audyssey, etc., have been a dominant part of this discussion of the 121. Some have suggested: the more points of correction the better. I happen to disagree. It's quality vs. quantity. The fewer the better.

But I agree, it's not a very interesting side trip from the main road.

I agree 100%.

We're in no way referring to time alignment. Phase shifting is an unwanted side effect of the process of equalization which creates harmful distortion. This isn't debatable and it exists in both the analog and digital domains without exception, even with our best modern digital eqs.

I'm not sure why this subject is being shellacked with an ethereal layer of mysticism -- it's not at all complicated. Equalization causes artifacts. Period. We can mask the problem with vague terminology but we can't hide the effects. Companies have spent millions and millions trying.

And just to reiterate what I'd mentioned last week but is recently lost: in most cases, the benefits of Audyssey (or similar DSP) far outweigh any problems inherent to equalization. The option to correct a +15db mode at 70Hz is a wonderful thing. The bit of added distortion you'll gain in the process is negligible by comparison. In a perfect world, acoustic treatment would alleviate the problem, and the need for any DSP band-aids.

I understand what you're suggesting, but I was referring to a signal path free from equalization, or any type of processing. There would be no "additional" artifacts beyond what the signal path is injecting. Add an eq... you have additional artifacts.

That's a more complicated way of saying that equalization corrects for the ill effects of an unclean signal path. Certainly that's the primary purpose of eq in a playback system, and I think we all understand its potential chore as a counterbalance for impure - we could say problematic - electronics, or even personal preferences (in the form of adding treble or bass to one's taste), but here we're talking about compensation for the ill effects introduced by the acoustic properties of the listening environment. It's a bit like having a broken leg, installing a cast, and suggesting that the cast allows the patient to run a marathon as efficiently as a person without a cast. The sum of +1 and -1 is NOT zero because we don't have a way of subtracting a perfectly inverted value for +1. So our total is 0.0234, so to speak.

Anyway, room correction schemes, such as Audyssey, emit test sweeps through much of the same paths followed by our input sources (within the unit, but also cables, amps, speakers, etc.), so many of the issues inherent within the signal path will be present within the sweeps themselves, and thus attempts will be made to compensate for them by Audyssey's corrective filtering, introducing, once again, unwanted artifacts in the form of additional distortion. It's trying to fix acoustic AND electric deviations from a flat line.

This is another way of saying that the more a signal path is faulted, the more heavy lifting Audyssey will have to do. The trouble with that equation is that the less expensive units have the "dirtier" paths, accompanied by the less effective versions of room correction algorithms, and therefor a higher level of distortion.

Once again, this is swimming in a great pool of minutia, and it's hardly worth making these types of waves, especially in an MX-121 opinion thread. It feels like we're halving an onion into infinitesimally thin pieces and there's no taste left. We've lost sight of the recipe.