Music in the Round #8

For months now, I've been beating the drum for full-range center-channel speakers, to reproduce recordings with a true center-channel signal. There are many reasons for this.

First, if you have a refrigerator-sized, big-screen TV in the middle, there's no way your main L/R pair can create decent centerfill without help (and even then, don't expect much).

Second, even without that monster in the middle, redistributing the center channel to the left and right speakers will, of necessity, introduce phase problems that can't be corrected with level adjustments or speaker repositioning.

Third, to achieve a continuity of spectral balance across the front, the center-channel speaker needs to speak with the same voice as the L/Rs, and no dedicated center-channel design can do better than a good imitation, with or without a TV in the way. Even an identical sibling of the main speakers will still have a small harmonic imbalance, due simply to the unique position of the center speaker—but that's nowhere near as great as with dissimilar speakers.

The only way for me to get as uncompromised an audio setup as possible was to give my TV the old heave-ho. So I popped for a new wall-mounted Fujitsu 50" plasma display and cleared the way for a center speaker. To make the video stuff as unobtrusive as possible, the guys from Sound Ideas, in Armonk, New York, installed a pair of Panamax 3-Bay Max In-Wall cable panels ( one behind the display, the other behind my equipment rack. Between those boxes, all of my AudioQuest/CinemaQuest signal and power cables ran inside the wall. The Panamaxes also provide surge protection for each of the signal lines.

To enhance the performance of all the audio and video equipment, dedicated AC power was taken from the breaker box via JPS Labs' 10-gauge Power AC In-Wall cable ( to Wattgate 381 receptacles ( And to protect my investments, I added a BrickWall PW8R15-AUD series-mode surge protector and filter (

RCA Living Stereo on three-channel SACD
Finally, the decks were cleared for a third Paradigm Studio/60 front loudspeaker—in the center (footnote 1). I carefully repositioned the three fronts in a strictly symmetrical array: all speakers equidistant from the listening position.

Now, how best to prove to myself (and to you) that all this cost and effort were worth it? Sure, 5.1-, 5.0-, and 4.0-channel recordings sounded better balanced and more seamless across the front—but I needed a test that would provide a sharp and definite contrast between two channels and three. It came fast.

Late this summer, BMG Classics announced the fall release on SACD of remasterings of 20 original RCA Living Stereo albums. I don't have to tell you how prized and respected these recordings have been, from their first release on LSC vinyl half a century ago to their various incarnations on CD, including JVC's XRCD productions and vinyl reissues by Classic, Chesky, Mobile Fidelity, and other labels. These recordings' musical and audiophile significance are rivaled only by the contemporaneous Mercury Living Presence series, which themselves are due to be released soon as three-channel hybrid SACDs.

The remastering was done in Boston at Soundmirror by John Newton, Philip Nedel, and Mark Donahue. Transcription from the original master tapes was on a customized Studer A80 VU transport (with five-channel Aria Reference electronics and two- and three-track head blocks) directly to DSD by dCS converters with no additional equalization or signal processing, in order to ensure close fidelity to the original producers' intents. The only exception to those direct transfers was for short segments where old tape edits had detached some oxide, thus creating brief dropouts. John Newton told me that in these cases, identical segments from an unedited "B" (backup) master tape were substituted in the DSD domain.

Here's the neatest part: Many of these revered Living Stereo recordings were originally made as three-channel recordings, mixed down to two for commercial release. The SACDs will carry the original three-channel recordings, where available, on their multichannel tracks. If the production master is two-channel, the SACD will have a two-channel DSD track derived directly from it, along with a CD-compatible (16-bit/44.1kHz) PCM track taken from the DSD transfer. If the production master is three-channel, the SACD will contain the direct transfer of that on its multichannel DSD track, a 3:2 mixdown on the stereo DSD track, and a 16/44.1 two-channel track, taken from the latter, on the hybrid layer. Whew!

Vindication for all
So eager was I to hear the Living Stereo SACDs that I hounded BMG for advance copies, and Stereophile was rewarded with a set of check discs of the first 10 releases (see Sidebar, "Recordings In the Round"). Because I didn't have that list of which discs were made from three-channel masters, I popped them, one by one, into my Sony SCD-XA777ES SACD player, to see how many would light up the player's multichannel indicator. Despite its later appearance on BMG's list of "two-channel" masters, the Sony recognized as multichannel Jascha Heifetz's recording of the Beethoven and Mendelssohn violin concertos, and Fritz Reiner's of Bartók's Concerto for Orchestra. I then gathered up all 10 discs, along with some earlier editions of the same recordings, both CD and XRCD, and took them all off to my multichannel system.

I had, with uncharacteristic self-control, planned to pair each Living Stereo SACD with an earlier CD version and to listen first to the CD, then switch to the SACD and listen first to the new two-channel CD track, then the DSD two-channel track, and finally to the three-channel DSD track (footnote 2). My self-control lasted only through the second disc (Van Cliburn). JVC's XRCD24 version (JM-XR24004) of this has brilliantly clear, punchy sound with good lateral soundstaging and decent depth. Like most XRCDs in my experience, it's a bit forward, but other than that, it was quite similar to the SACD's "Red Book" layer. I doubt if the XRCD was more detailed, although its more brilliant balance inclined me to infer just that.

Moving on to the two-channel DSD track, I found it as meticulous as the XRCD, but with a more relaxed tonal balance and a slightly more distant perspective. The three-channel track, however, was killer, and significantly better than the other three options. In three channels, the orchestra's breadth and depth were greater, with no diminution of centerfill. In fact, individual instruments were much more localizable in both dimensions. Coupled to that was a surprising solidity in the piano, here afforded the advantage of being the focus of the center channel. That said, the piano was not recorded in mono—indeed, switching out the center channel left the piano only slightly but still noticeably reduced in weight and power. Alternatively, the piano maintained its presence when only the center channel was played, but lost much of the supportive ambience.

The three-channel track of Cliburn's performance of Tchaikovsky's Piano Concerto 1 and Rachmaninoff's Concerto 2 were so remarkably more revealing and satisfying that I telescoped my listening plans: I only sampled the two-channel alternatives, then, as quickly as I could, reveled in the three-channel tracks.

Jumping next to A Hi-Fi Spectacular, aka Charles Munch and the Boston Symphony performing Saint-Saëns' Symphony 3, "Organ," and works by Debussy and Ibert, I was even more impressed. Again, the three-channel track had a slightly more distant orchestral sound than on earlier versions, but its depth distinctions were excellent. Remarkably, the soundstage width was much greater; high-frequency sounds, such as the fff cymbals, were smoother and more realistic without losing their edge, and the overall weight and integrity of the orchestra approached that of the best modern discrete multichannel recordings. The organ was simply monstrous, in the best possible way.

Sure, three channels means a 50% increase in bass drivers, but my three Studio/60s have only six 6.5" woofers among them, so that doesn't tell the whole story. I could take any of the two-channel versions and set the Outlaw ICBM to have the Paradigm Servo-15 subwoofer double the main speakers in the sub-40Hz range—that only made it louder, and not as discrete and defined. What I think the three-channel recording demonstrated was that one cannot simply mix the center-channel feed with other differently spaced signals without creating some degree of phase cancellation, and that the effect is most noticeable in the bass. Bass nut that I am, I also fed the low bass, from the center channel only, to the Servo-15 and, with a little bit of trimming, achieved a low-frequency epiphany while sacrificing nothing from the main channels.

One after the other, each Living Stereo SACD demonstrated the same hierarchy of preference. But when I went through each recording—from XRCD to CD layer to two-channel DSD to three-channel DSD—there were some surprises. The expected improvement in the presence and firmness of Heifetz's violin in the Beethoven concerto seemed lacking. The Denon DV-5900 universal player indicated "5.1," which it does whenever it plays a multichannel track, though none of these SACDs includes a surround or LFE signal. Putting my ear to the center speaker, I heard only a very-low-level hiss/buzz. Of course, BMG's press release said this one was only in two channels; still, this was a three-channel track. The answer is that the Mendelssohn Concerto on the same disc is in three channels, and most gloriously so. Not since the mono LP of Vitali's Chaconne have I been so knocked out by the slashing clarity of Heifetz's Guarnerius del Gesu violin.

I soon found myself skipping directly to the multichannel track on each disc and testing to see if it was in three or two channels. As with the Heifetz disc, the results varied from selection to selection. For example, the sainted Reiner/CSO performance of Bartók's Concerto for Orchestra is two-channel, as is the Music for Strings, Percussion, and Celeste; the rest of the disc is—surprise!—in marvelous three-channel.

Mussorgsky-Ravel's Pictures at an Exhibition is three-channel, as is Night On Bald Mountain and most of the other Russian pieces on the disc—but, in the middle, Tchaikovsky's Marche Slave is in two-channel. But I won't carp—the entire disc is spectacular. Pictures is particularly outstanding, with orchestral power and instrumental delineation to compete with anything recorded in the decades since that 1957 session. The timpani and bass drum are so good, and so much superior to those on CD and LP, that they might make Telarc envious. In my September column (p.69), I compared several multichannel recordings of the Mussorgsky-Ravel Pictures; the Reiner/CSO now goes to the top of that list.

Overall, the first 10 releases in BMG's RCA Living Stereo SACD series are overwhelming successes. I eagerly await the next 10, as well as the comparable series of Mercury Living Presence SACDs.

Next time in the Round
In December I'll be writing about three pieces of hardware: the Zektor MAS3 multichannel input selector, the Auralex SubDude bass isolator, and Arcam's DV-79 universal player. See you around!

Footnote 1: I will be reporting on the Paradigms in the December Stereophile.

Footnote 2: I never thought I'd be discussing vinyl in this column—I sold my SQ decoder some years ago. Nonetheless, it seemed appropriate to compare some of these releases with their LP ancestors. Of the three selections I pulled from the stacks, it was apparent that the two-channel SACDs had preserved the LPs' general tonal properties and soundstage spaces. If anything, the SACDs were a little less brilliant than the LPs I had (Cliburn/Rachmaninoff, LSC-2601; Heifetz/Beethoven, LSC-1992). However, the margin was so small that the difference on the Reiner/Bartók Concerto was notable with my 1S/1S pressing of LSC-2374—but the two-channel SACD sounded like a dead-on clone of my 2S/2S pressing. I leave the rest to those with a more abiding interest in vinyl.