Music in the Round #33

Welcome to the wonderful world of firmware and software updates. With almost every audio device now microprocessor-controlled, and the tasks to be performed increasingly complex, it's not surprising that "finished" products leave the factory only to be stymied by reasonable but unpredicted user practices. It's not that we're all becoming beta testers, but we are contributing to the intelligent evolution of product capabilities.

I first came across this with Denon's AVR-4806 A/V receiver and DV-3910 universal player. Way back then, in 2005, the procedure was still clumsy, requiring the burning of CD-Rs from Web downloads or, as in the case of the otherwise spiffy Bel Canto Pre6, chassis surgery and a bit of machine code! Many machines remain un-updated because users can't access the firmware, and some require users to sign a document absolving the manufacturer of responsibility if the update turns the device into a costly brick. While Integra requires such a signature, many, including me, have updated the DTC-9.8 preamplifier-processor (which I reviewed in January) without problems. Even better, some new products can be connected directly to the Internet and update themselves, just like the major software applications on your computer. I think that's great.

The Anthem Statement D2 A/V processor that I wrote about in September can still be updated only via its RS-232 serial port, rather than USB or Ethernet, but the process is painless via a PC's serial port or a USB-to-serial adapter. I have incrementally updated my D2's firmware from v1.29 to v1.33, and Anthem Room Correction (ARC) from v1.1 to v1.2.5. Every update has gone smoothly and without error, and each has either improved the effectiveness of ARC and/or cured nitpicky little issues. So when I see Web forum postings bemoaning this process, and screaming that all these updates indicate manufacturer incompetence in releasing unfinished products, I say that I'd rather have the current version than still be stuck with a perfectly functional v1.0 whatever when others are buying the v2.2 Improved.

Over the past two months I've been enjoying the Statement D2 in my system and playing with Anthem's almost constant stream of updates. Although my D2 worked fine out of the box, Anthem seems almost obsessive in their responsiveness to user comments. There's no need for all users to constantly update their systems if they experience no problems, but some of us are almost addicted: "Hey, I never noticed that bug, but boy, am I happy it's fixed!"

Currently, with no issues in need of attention, my D2 with ARC works and sounds just dandy. I can access two stored sound configurations, for music and for movies, with the push of a button or two. The first includes a tiny correction of my room's gain (a midbass bump), and the response is corrected up to 20kHz. The second configuration includes a larger correction and, as recommended by ARC, is corrected up to 5kHz. Each is very satisfying in its particular application, but is it as good as I can get? How does it stand up to Audyssey, especially with the latter's standalone Sound Equalizer and Pro software?

How to get two into one?
Comparisons were not that simple. I could run the balanced output of the Anthem Statement D2 directly to my Bryston 9B-SST power amplifier, and route the D2's unbalanced outputs via the Audyssey Sound Equalizer (SEQ) to the Bryston's RCA jacks. Unfortunately, these two routes would result in a huge (15dB) difference in gain: the balanced output benefits from an inherent advantage of 6dB, and the unbalanced route is subjected to a 9dB loss in passing through the SEQ. Moreover, the D2 would be using different output stages, and the unbalanced route would entail additional A/D/A conversions (though I did find these innocuous when I reviewed the SEQ in the March 2007 issue). I suppose I could have used six Y-connectors on the D2's RCA outputs to feed both directly to one input on the Zektor MAS7.1 source switcher and, via the SEQ, to another input. But I didn't have a bunch of Y-connectors.

My solution was to run the D2's RCA outputs via the SEQ and from there to the Bryston amp, using the SEQ's Bypass button to engage or disengage it. Not so simple, though. For ARC mode, I had to bump up the D2's gain to compensate for the loss through the SEQ. In Audyssey mode, I had to turn off the added gain and ARC, as well as adjust the D2's individual channel gain and crossover settings to accommodate Audyssey's recommendations.

Having run ARC with eight microphone positions, I tried to repeat those exact eight positions with the Audyssey. Amazingly, the speaker distances measured by the ARC and the Audyssey differed by only inches. After completing the Audyssey process, I stored all the requisite settings in the SEQ and D2, and had the D2 store the SEQ results as the Installer Default, the ARC results as the User Default. Thus, to go from Audyssey to ARC, all I had to do was restore the User Default and hit the SEQ's Bypass button.

Crunch time
I was disappointed by the A/B face-off: both sounded very good, both sounded much better than with no EQ, and, surprise, they sounded pretty much the same. Back and forth I went, becoming increasingly frustrated that I couldn't choose a winner. Was the Audyssey ever so slightly tauter in the bottom end? Was the ARC better integrated in the upper bass? Did the Audyssey have a smoother, sweeter treble, or was the ARC more open and transparent? These and many other impressions passed through my mind as I grasped at fleeting minutiae.

Because I was unable to pick a winner by ear, I figured it was time to call in the measurement corps. There is no shortage of tools for this, Room EQ Wizard being the most common recommendation, but, lazy as I am, I opted for a newer, handier tool: the XTZ Room Analyzer, which runs on Windows machines only. For all tests, the pink-noise or swept signals were sent to the analog L/R inputs of the D2, which I used for bass management to reroute the bass to the subwoofer. (Results with the subwoofer input were reasonably similar but, I think, less representative of real-world performance.) First, I used the pink-noise–based real-time analyzer module to confirm that both Audyssey and ARC were providing similar gain levels, and relatively flat and remarkably similar spectral responses from 16Hz to 20kHz.

Switching over to XTZ's room-mode analysis module, I measured at the first three mike positions used for both the ARC and Audyssey. The XTZ figures show frequency on the vertical axis, time on the horizontal, and indicate sound intensity with color, the scale for which is at the right. With both EQs bypassed, the display in fig.1 shows lots of energy lasting long beyond 50 milliseconds (vertical cursor mark) below 80Hz. Note the intensity of the energy at about 60ms from 125 down to 55Hz, and especially the foci of energy at 41 and 23Hz, both of which extend out to almost 200ms. The 41Hz mode is probably related to the room width, and the 23Hz mode corresponds to an oblique room dimension. Not surprisingly, this correlated with a thick, undefined low bass, and while organ-pedal notes had imposing weight, their tonality suffered.

Fig.1, XTZ room-mode analysis, no equalization.

A single button-push put ARC online, and immediately I could hear the improved definition in the organ pedals; if anything, that increased definition focused the energy such that they had, subjectively, even greater weight. What did XTZ say about that? Well, fig.2 shows that, with ARC, all the really high-energy nodes dropped below the 50ms line from its upper-frequency limit of 250Hz down to about 32Hz. A very-low-energy plume is seen at about 70Hz and a stronger one at 29Hz, but the 41Hz node has been greatly reduced in amplitude, and the original 23Hz node is completely quashed. All to the good. In addition to the improvement in transient definition at the extreme bottom, the crossover from the subwoofer to the L/R speakers, set for 40Hz, became inaudible, even though the sub is at the rear of the listening room—something I ascribe to the general reduction in extended energy in the crossover region. Overall, I think the measurements confirm what I heard from the ARC.

Fig.2, XTZ room-mode analysis, Anthem Room Correction.

Switching to the Audyssey SEQ made little audible change, but XTZ revealed something new. In each of the previous analyses, XTZ had provided a list of the identified modes and some suggested equalization. With Audyssey, what I thought was an error message popped up: "No modes detected!" Indeed, fig.3 shows that almost all energy extending past 50ms lies at very low levels, and the only deviation from silence in that range was that same, very-low-level plume at 70Hz seen with the ARC (footnote 1). So the Audyssey measured better. Of course, having seen these charts, I then tried to hear what they show, but I can't say that the Audyssey actually sounded better at the bottom end with music. I could hear a difference with recordings of artillery, but I'm grateful that I have no real-world reference for such sounds that would permit me to say that one rendition was more accurate than the other.

Fig.3, XTZ room-mode analysis, Audyssey equalization.

So the residual difference between the Audyssey SEQ and Anthem's ARC was not at the bottom end but in the midrange and treble, where the impressions accumulated through extended listening suggested a tiny but consistent difference. For music, I preferred the silky smoothness and harmonic integration of the Audyssey correction; for movies and video, the honest immediacy of the ARC. But I grasp at straws here. Each confirmed the postulate of the other: In addition to competent acoustical design and treatment, room equalization should be a part of every audio/visual system.

X