Toshiba and the rest of the Japanese armada are scrambling around with counterproposals based on a new dual-layer DVD. Classic Records is already releasing standard DVDs with 24-bit/96k audio (no video). The one thing audiophiles know for sure is that the venerable 16-bit Compact Disc won't be the be-all, end-all format much longer. Puts a damper on the high-end processor market, don't it?

In the past, audio magazines have talked about "digital obsolescence" in the sense that "alpha male" digital products got dethroned with frightening regularity; every month Stereophile seemed to anoint a new king-of- the-digital-hill. Looking back, I see that those were the good old days. Now we're talking real obsolescence. By next year at this time you may be looking at the hundreds or thousands of antiquated silver discs on your shelves and feeling very depressed. So why buy the latest, greatest processor to play them; wouldn't that make you just more depressed? Prozac doesn't grow trees (St. John's Wort does grow somewhere, but I'm not sure I want to know where). I wish I had a good crystal ball to see into the hi-fi future, but mine functions even less well than the Editor's BS detector or maybe the latest wire levitator. The next best way to predict the future is to look at the past. Although it seems like there have been more top digital dogs than you could shake a stick at, the technology has been relatively stable for the last three years. The Sonic Frontiers SFD-2 Mk-II processor that I just replaced with the Processor 3 was, shockingly, already in production over three and a half years ago. Has it really been that long? Most of the important circuit bits the HDCD decoder/filter chip, the UltraAnalog AES 21 input receiver, and the UltraAnalog dual D20400A DAC modules are stiII basically state-of-the-art items. The SFD-2 Mk-II's full differential circuitry was a "balancing act" pretty difficult to top.

Many of us in the club that originally started up this magazine, The Audiophile Society, feel that the now-discontinued BiDAT by Ed Meitner is, when used in balanced mode without a preamp, still as good sounding a digital processor as there is these days. The design is a few years old now, too (although Ed tweaked it constantly). Club President Arnie B. and the club's ace-audiophile The Brass Ear have run comparisons of it with most of the top contenders out there, and they continue to use it as a reference. I've compared two versions of it (but not the latest) to the SFD-2 Mk-II and Line 3 preamp combo, and felt that I could live with either. The Meitner unit had a somewhat cleaner bottom, while the Sonic stuff gave an airier top end, more apparent detail, and slightly greater dynamics. The point is that three-year-old processors can sound pretty damn good.

On the playback side we're running into the law of diminishing returns. Three years ago some high-end reviewers were telling us to stay out of the digital chase until things cooled down. Little did they know the Ice Age had begun. If you followed their advice, you simply denied yourself enhanced listening enjoyment. To be sure, there have been advances in jitter reduction now that manufacturers have gained experience measuring it (as opposed to just listening and poking and hoping for the best), but mostly it has been a period of stability while the pack caught up with the leaders. Now, Sonic wants to reassert itself and mush farther off into the digital frontier. To do it they've unleashed a new top DAC dog, the Processor 3. It provides: 1) improved sonics over the high standard set by the SFD-2 Mk-II; 2) a method for reducing jitter in the playback system to virtually nil, and 3) a modular design that should allow for relatively painless upgrades when improvements or new formats do come along.

Better Sound Now

Sonic Frontiers claims that in almost every "technical and performance parameter (both subjectively and objectively) the Processor 3 outperforms our current Class-A rated SFD-2 Mk-II." Many enhancements went into the new processor, the most obvious of which is that it's a two-box affair with a separate power supply. And that power supply is substantial, with a large, potted and encapsulated toroidal transformer (we're not talking wall-wart here). As with the top Sonic Frontiers preamps, the two boxes connect via a thick (and, thankfully, detachable) umbilical cord.

The other obvious change is that the analog circuit uses four rather than two output tubes (still 6922s). This lowers the output impedance to less than 45 ohms unbalanced), which is pretty incredible for a tube circuit. Then there are the little hidden enhancements: Arlon circuit board for the DAC module; Sanyo OS-CON electrolytic capacitors for DAC power supply regulation; discrete, outboard current-to-voltage stage modules for the DACS, and a direct-coupled output via d.c. servo. The direct coupling allows the Pro-3's low output impedance to be constant at all frequencies. Again, this is quite remarkable for a tube circuit, where output impedances typically rise enough to cause audibly non-linear behavior in the bass.

Sonic Frontiers has also made standard an "unauthorized" modification that many customers had clamored for: "Hot-rodding" the output level for non-HDCD discs. As you may know, Pacific Microsonics asks that CD processors using the HDCD decoder-filter chip (the PMD 100) attenuate the output of non-HDCD audio by 6 dB. The rationale is that the greater dynamic range of HDCD material leads to the average playback level sounding subjectively lower than with normal discs. Cynics think it's just a cheat to favor HDCD in side-by-side comparisons. In any event, the old SFD-2 Mk-II conformed to the Pacific Microsonics spec; a resistor was switched into the analog output circuit to cut the gain when HDCD processing wasn't active.

Of course, a resistor and a relay should be sonically transparent, but purists got all pale and weak at the very thought of naughty extra bits coming between them and their Calvins, so Sonic Frontiers said: Screw it; no monkey business at the output. The output level is now 6 volts balanced at all times versus the older processor's 3.5 volts balanced for non-HDCD and 7 volts balanced for HDCD. (To calculate unbalanced levels just reduce the above by half.) What's my verdict? Pacific Microsonics was perhaps right in the first place; some HDCD discs now sound a little too quiet. With the SFD-2 Mk-II, I never felt the urge to alter the playback level when switching to an HDCD disc. With the Processor 3, many HDCD discs, particularly those that are less compressed to begin with, now come up obviously too low in level. Sonic Frontiers has taken this into consideration by providing a jumper on the input p.c. board that can be moved (no soldering) into a position that engages the "HDCD-approved" 6-dB attenuation. (See the photos below.)

I don't know if this has anything to do with the "hot-rodded" output level, but I hear a trace more noise with the Processor 3 in the system. The noise floor is benign sounding and extremely low by any standards; you have to be within a foot of the speakers in a dead quiet room to hear it but the reduced output of the SFD-2 Mk-II made it just a little more impressive noise-wise. But let's face it, 6 volts is just more output than is usable or desirable when fed into a normal preamp. Admittedly, 6 volts will not clip the input of a Sonic Frontiers Line 3 preamp, but with some preamps the highish output level may put the volume control in a less than linear part of the range. Unless you are using a passive preamp, all Pro-3 owners should simply move the jumper and cut the output level.

OK, I had to find something I didn't like, otherwise what kind of critic would I be? Everything else about the Processor 3 is quite simply as good as it gets. The two very small sonic flaws of the SFD-2 Mk-II Ñ a slight lack of control in the bass and a bit of soundstage blurring or fuzziness Ñ are virtually eliminated in the Processor 3. Some of my reference CDs (Chris Isaak's Baja Sessions, Emmy Lou Harris' Cowgirl's Prayer, Eric Clapton's Unplugged, and Holly Cole's Don't Smoke in Bed) had a tendency in my system to sound over-ripe in the bass, particularly if played too loud. The Pro-3 didn't exactly transform my system, but it did make all my favorite discs sound better: Firmer, sharper, cleaner, and more rhythmically compelling.

I had the opportunity to compare the Pro-3 with a venerable (but newly updated to accommodate 96k processing) Theta Gen V. The SF Line-3 preamp allowed me to match levels and A/B switch between the two processors while running both off the same transport. The one fly in the ointment was that I didn't have two pairs of identical balanced cables for connection to the preamp; I thought that the audible differences between the cable brands swamped the differences between the processors. However, I feel fairly safe in concluding that the Sonic Frontiers unit now matches the Theta in terms of bass punch and dynamics. As an aside, if you think my Phil Jones-designed Reference-1 speakers don't have enough bass output to let me make that determination, then you haven't heard them in my room (measurably flat at the listening position to under 30 Hz, although 20 Hz is admittedly a no-can-do proposition). You may also think that since I use a tube amp, I don't know from bass. I beg to differ. The new Sonic Frontiers amps are killer in the bass. They are the amps that all tube-loving audiophiles should have been waiting for: Amps that give you the pretty, glowing bottles (and the pretty, glowing midrange that such bottles suggest), without all the usual audible and measurable tube colorations. The problem is that what too many tube-ophiles love about their amps are the very things Sonic Frontiers has banished from theirs: Massive second-harmonic distortion, tube noise, and frequency response nonlinearities caused by high output impedances. I've said it before and I'll say it again: Audiophiles live for distortion. Without it, there's so little to argue about.

The Sonic Frontiers amplifiers sound consistent from system to system, and that sound can be characterized as neutral and colorless. To me, that's the highest praise, but again, I'm not sure other tube addicts will applaud it. Take that word "colorless."

That's the adjective some will use to describe the Sonic Frontiers amp line-up. To the single-ended triode/horn loudspeaker aficionado, it means boring. Trust me, this amp is anything but boring. It's clean and dynamic, with bass as ballsy as all get out.

Back to the Pro-3. The improvements in its bass can probably be chalked up to the reduced output impedance. The rest? Who knows. (Maybe just new tubes, though I don't think so.) Whatever SF did made a difference; the Processor 3 offers digital sound at an unsurpassed level of quality. Unsurpassed, that is, save for a Pro-3 being fed a digital signal from the Sonic Frontiers Transport-3 via the I²S bus.

Even Better Sound Now

Sonic Frontiers processors have traditionally accommodated every type of digital input: Toslink, glass optical, BNC coax, RCA coax and AES-EBU balanced. The Pro-3 adds a new one: The I²S-enhanced, and this additional digital input may well prove to be the most important improvement over the SFD-2 Mk-II. The I2S-enhanced interface is said to virtually eliminate jitter induced by the Sony/Philips Digital Interface (S/PDIF).

Because the S/PDIF mixes the clock and the audio data (the advantage being a simple, one-conductor cable hook-up), the recovered clock signal will contain jitter related to the audio signal. According to UltraAnalog, "there will be timing errors in the [D/A] conversion that will result in voltage errors in the recovered analog signal that are related to the audio signal itself, thereby resulting in jitter-induced distortion." The I²S-enhanced interface puts the clock and the audio data on separate conductors, which "allows the master clock to be generated in the receiving product with the transmitting product operating as a slave." UltraAnalog claims a 6:1 reduction in jitter over the best case implementations of S/PDIF; improvements are more typically 10:1. If it sounds like we're being led into the digital promised land on the strength of this new I² S-enhanced cable, be aware that a nasty little disagreement has erupted between Moses and God, and it seems that God does not want Moses to enter the promised land. The part of God is being played by UltraAnalog's Richard Powers; for this performance, Kevin Halverson is our Moses.

Halverson first proposed an I2S interface using a 13W3 cable in early 1997. The 13W3 is a computer-type cable used in professional video production for controlling video decks and transmitting video signals. Halverson showed that it was ideally suited to digital audio communications between transport and processor. The UltraAnalog guys thought Kevin's proposal was swell, then thought some more and decided it could be sweller. They made some changes that resulted in what we call a "fail-yuh to cuh-munecate" between Powers and Cool Hand Kevin. Each insists that his interface is superior; each claims to be on the side of the angels. Halverson pointedly denies the charge of a "not-invented-here ego trip."

Halverson is the man behind Muse, and it seems he is going ahead with his standard, while UltraAnalog is proceeding with theirs. The two are not compatible; a Muse transport with an I²S output will not plug into the I²S-enhanced input on the Pro-3. Bummer. UltraAnalog is, to my mind, the Intel of high-end digital audio: UA chips are the heart of products from ClassŽ, Krell, Mark Levinson, Parasound, Spectral, and Sonic Frontiers. What would be the point of flouting UltraAnalog?

There is a point, but it's an obscure one. Halverson is also the man behind the technical implementation of the new 24-bit/96-kHz DVD audio discs released by Classic Records. Several high-end manufacturers are supporting this Ônew" audio standard (it's not really new since the specs were part of the DVD standard all along) by building DVD players specifically capable of handling the higher bit/sampling rate data. Ayre, Bel Canto, Resolution and Theta are among them. Ayre's Charlie Hanson and Resolution's Jeff Kalt indicate that they'll back Halverson's unenhanced I² S. Politically, they've formed a group that needs to pool information, so it's understandable that they would stick together on this.

On a more substantive level, however, some of Halverson's supporters have been saying that the I2S-enhanced interface won't work with 96-kHz data. That's a dangerous position, as they are most likely wrong. UltraAnalog's white paper states clearly that it will work at multiple sampling rates, including the imminent 88-kHz or 96-kHz sample rates." Sonic Frontiers says it'll work. The engineers at UA who developed it say it'll work. Richard Powers says it'll work. Does God make mistakes? Is a bear Catholic? Halverson himself merely states that the enhanced interface won't work "in a master/slave interchange" on the DVD-based discs that he's had a hand in mastering. As an audiophile hobbyist, I think this little battle is fascinating; as a consumer I feel it's unfortunate nonsense: Bad for us, bad for dealers, bad for the industry. To utilize the new I2S enhanced digital interface, the consumer has two choices: Either SF's new reference Transport-3 (the mate to the Processor-3) or alternatively, Sonic Frontiers now has an upgrade going for their SFT-1 or SFCD-1 CD player; adding the I2S-enhanced upgrade connector will cost $699 in U.S. dollars (I²S cable AND return shipping included). Not bad, given that the price of the new, top-loading, wowy-zowy Transport 3 is a semi-staggering $6,999. I've got the upgrade, and as happy as I was with the Pro-3 before, I'm even more taken with it now that I've heard the true capabilities of the I²S-enhanced interface.

What's the sound of less jitter? I would describe the sound as tighter, clearer, more open, less laid back (in both a dynamics and soundstage sense). As with most things in audio (and maybe in life), accuracy and truth won't necessarily make you happy. A system that's already lean and bright might sound worse using the I²S. If you really want to hear what's on a given CD, however, this is the way to go.

Because I've grown increasingly aware of the "junk science" nature of high-end audio, with too many products that are just expensive placebos for the desperately deluded, I felt some sort of blind, controlled listening would be necessary to confirm that I was hearing something other than my own wishful thinking. I first tried to have my five-year-old son do the A/B switching for me, but between his trips to the potty and complaints that he was missing "Power Rangers in Space," I had to abandon the experiment.

Thankfully, Audiophile Society stalwart Ross Wagner agreed to come over and play test rat. He was a little anxious as we began, even turning down the glass of wine I offered him to steady his nerves. He wanted to be at his best, lest I conclude he "can't hear worth shit." (His words.) I first played "Two Hearts" by Chris lsaak from The Baja Sessions (HDCD encoded). I did the switching for Ross, then he did it for me. Initially, we identified which connection method was in use (glass-optical or I2S-enhanced), then we did a trial run of single-blind switching. Afterwards, we discussed what we heard, on the theory that we I could teach each other what things to listen for. Ross thought he heard soundstaging differences, with the I²S-enhanced being more forward. I also thought the I²S-enhanced made for a more "up front" sound, but to me the difference was more analogous to the frequency response variances in a concert hall. Move forward a few rows in the audience, and the sound will be fuller, brighter, less warm, more extended, and seemingly more full range.

Ross provided the CDs for the blind A/B switching that I then subjected him to: The ubiquitous Muddy Waters Folk Singer on Mo-Fi; Black Light Syndrome from the "lesser" supergroup Bozzio Levin Stevens, and Mary Kay Mann & Janet Jackson Biely's Heartsounds, a lovely sounding private-label CD with vocals, flute and Celtic harp (available by calling 610/388-7523).

As the relatively informal test began, Ross covered his ass by stating that the sonic differences between the AT&T and I²S-enhanced connections were subtle, and that he didn't think he'd be able to identify which was which. I smiled knowingly, or maybe stupidly, because I shouldn't have smiled at all Ñ experimenter bias!

Anyway, by the end of the Muddy Waters cut Ross correctly identified which was which. He remarked that with the I²S-enhanced Muddy's voice seemed more projected, less chesty and muffled. About halfway through the next selection he cut short my switching and correctly made the ID. About 30 seconds into the third trial he stopped me again, correctly picked the I²S-enhanced and said, "It's just better." Not just different; better.

(By the way, the I²S-enhanced input on my early sample of the Pro-3 didn't work. A minor parts problem was corrected on later production runs, but the review processor needed a resistor and choke mod on the I²S input board to make it functional. My soldering prowess is such that if I don't burn down the house, I consider it a triumph, so I enlisted Arnie's help. We put our heads together, figured out which parts went where, and Arnie did a very neat soldering job... but the Pro-3's I²S input still didn't work. Back it went to Sonic Frontiers. Cold solder joints was the diagnosis. Well, at least we didn't burn down the house.)

Still Better Sound Later

Frankly, I've become inured to high-end manufacturers' claims of "upgradability." Except for a few Theta and Wadia products, it's always a case of "you can't get there from here." Each new product development brings a host of small but significant changes. Go to the dealer and inquire about an upgrade, the upgrade he promised would be do-able when you bought the thing six months earlier and you're told the modifications are so extensive that it's cheaper to sell the old player and buy a new one. But what if new formats make your old high-end processor or player virtually worthless in the used market? Like an 8-track recorder.

If any processor will prove to be upgradeable, it'll be this one. That's no guarantee, but the modular construction makes board swapping fairly easy. The display is already capable of showing 88.2- and 96-kHz sampling rates, and when Pacific Microsonics delivers its new PMD-200 (slated for Fall 1998), the Processor 3 can then (presumably) be made compatible with the 96-kHz DVD-standard discs now out from Classic and Chesky (if, big if, you have a DVD player that outputs 24/96 data).

Gazing deeply into my crystal ball, I see only one potential fly in the audio ointment: Multi-channel recording. With a five-or six-channel format (like Sony's DSD, which contains six channels and uses 1-bit rather than ladder DAC processing), there's no getting around the fact that the Pro-3 ain't gonna know what to do with it. Theoretically, new digital boards internally along with an outboard multi-channel processor could get things cooking with DSD gas, thereby salvaging most of your capital in the Pro-3.

Of course, at that point a digital processor may be the least of your problems; you'll need six speakers, three stereo amps, a new preamp and a new a transport. Oy!

You Can Just Do It

So will future digital formats make a mockery of any investment in current digital processors? In the case of the Sonic Frontiers Processor-3, probably not.

More importantly, if you want the CDs in your collection to sound their best, the Sonic Frontiers unit probably can't be bettered for the foreseeable future.

NOTES Sonic Frontiers Processor 3, $6,999 Sonic Frontiers International, 3535 Laird Road, Unit #16, Mississauga, Ontario, Canada, L5L 5Y7; Phone: 905/828-4575; Fax: 905/828-4585; Website: http://www.sonicfrontiers.com; E-mail Address: SFI@sonicfrontiers.com. Reference System Sonic Frontiers 3, SFD-2 MkII and Museatex/Meitner BiDAT digital processors; Sonic Frontiers SFT-1 transport; Sonic Frontiers Line 3 preamp; Sonic Frontiers Power-2, Balanced Audio Technology VK-500, and Amherst A-2000 amps; Phil Jones' Reference-1 speakers; Esoteric Audio Artus interconnects; Discovery and Esoteric Audio speaker wire; Bright Star Big Foot bases with Little Rocks or VPI bricks for additional damping on some components; ASC SuperTraps, tube traps, flat panels and RPG Skyline diffusors; Audioquest, TDK and Radio Shack ferrite noise filters; Sims Navcom feet; Audioquest Ultraconnect cleaner, and DBX 14/10 analyzer.