Marantz SA8260 multi-channel SACD/CD player.

Manufacturer: Marantz America, Inc. 1100

Maplewood Dr. Itasca, IL 60143; 630/741-0229;

www.marantz.com

Price: $1,099

Source: Manufacturer loan

The first SACD/CD player I owned was a relatively low-cost, second-generation player made by Sony. It was a decent CD transport, but as far as SACD playback was concerned, it was a bit disappointing. Sure, I guess most folks could take advantage of its multi-channel ability, and even though I didn't have one of these systems it hardly mattered--I just hooked up two cables to the its outputs and was able to enjoy the stereo layer of the SACD. But I wasn't blown away by the sound as much as SACD's proponents said I would be. It just sounded like very good CD sound to me. And there were plenty of plain old red book CDs that sounded just as good as or better than the few SACDs I had at the time. Perhaps if I was listening to multi-channel sound ... but that's a moot point. What I wanted was to take advantage of SACD's two-channel sound and hear the incredible sound that everyone was talking about. Perhaps it was time to take a step beyond entry-level. Enter the Marantz SA8260.

The 8260 sports a full-sized black cabinet, which Marantz says, is made with two layers to minimize the effects of vibration. It uses a "high accuracy" SACD transport that they developed themselves, "audiophile grade" capacitors, what they call "advance Cirrus Logic D/A converters", and their proprietary current-feedback HDAM circuits that facilitate high-quality signal transfer. All the terminals are gold-plated.

On the rear panel of the 8260 are two pairs of analog RCA output, left and right marked front; and a pair of analog RCA outputs, left and right marked surround. There is also a pair of analog RCA outputs, one marked center, and one for the subwoofer. The manual says that if you are going to use the 8260 in a two-channel system you should only hook up the front left and right channels. There are both coaxial and optical digital outputs for the red book CD signal, and a filter that is switchable between standard and custom. This filter is provided to attenuate ultra-high frequencies so that the protection circuits of standard amplifiers are not activated. The presence of these ultra-high frequencies on SACDs, says Marantz, might damage normal amplifiers that are not designed to handle SACD playback. The 8260 is shipped with the switch in the standard position. When the unit is connected to a "next-generation amplifier that supports SACDs" and the switch is set to "custom," this will, according to the manual, ensure play with "a wider frequency range". I'm sorry to take up so much space about this feature; it just started me wondering why I never saw this feature on any other player before this one. Perhaps it's just built into the others without giving one the option to switch it on or off. There is also a pair of RCA outputs meant for a remote control if using the player in an all-Marantz system that has something called a D-bus function.

The front panel of the 8260 is a model of convenience. It has oversized easy-to-locate controls that are intelligently arranged, a large display; and has all the controls that are on the remote control. And lo and behold, on the front panel can be found a rarely seen, nearly extinct headphone jack and its volume control.

The review system consisted of an analog front end of a Basis Debut V turntable with Lyra Helikon SL low-output moving coil phono cartridge mounted on a Tri-Planar VI tonearm. The preamplifier was a Balanced Audio VK-3i, and it was hooked up to a Krell KAV-250a power amplifier. The speakers were Sound Lab Dynastats and a Velodyne HGS-15b subwoofer. All the hardware sat upon an Arcici Suspense rack, and all of it was plugged into PS Audio Power Plant AC regenerators, except the turntable and subwoofer which were plugged into a Change Lightspeed power conditioner. Cable was mostly Virtual Dynamics with some MIT. I also have on hand an Arcam DiVA CD192 CD player and a Music Hall Maverick CD/SACD player for comparison.

The first thing I noticed between the three players was that the Marantz was the only one that has to be told what format the disc is. Of course, the Arcam only plays CDs, so it doesn't matter--insert the disc, and it plays it. The Music Hall recognized what discs are inserted, but only plays them back as stereo, so even if the discs are multi-channel, it just automatically plays them as stereo. The Marantz is a multi-channel player with its default set to multi-channel. When a multi-channel disc is inserted, the player's display reads multi-channel; the disc starts to play--even when I only had the left and right outputs connected. I had to select the stereo mode from the remote after it read the disc, and the disc had to re-read and then start playing again. This took a while. The 8260 had to read each disc after it was inserted, display what it found, and then go into play mode. If it was interrupted, for example by my telling it to switch to stereo mode, it stopped, and had to start all over again. I don't want to make too much of a big deal about this, I eventually got used to it, but it was a bit less than convenient.

The second thing I noticed was the huge sound that emanated from this player on just about every SACD that I played through this player. The bass wasn't the only the thing that made the sound huge, but the images were huge, and the overall sound was huge. There was a weight to its sound that was more than just and increase in bass or and extended frequency response. There was a heft to the sound that was very appealing. And it was great to finally hear treble come from silver discs that had a purity that was unheard of in my listening room. It was more than just that the graininess that I was used to was gone, but the treble took a more lifelike presence and each instrument was more separated in the large soundstage. On regular CDs the highest treble seems to completely fill the highest reaches of the CD's sound, on an SACD on this player there seems to be room above these sounds, giving the treble a more relaxed, almost spacious feel. This aided in placing the treble within its own soundspace, and aided in the perception of greater imaging. The midrange imaged with a solidity and at the same time, almost creaminess that was more analog sounding than the treble. I was impressed.

On the SACD of Paavo Jarvi conducting his own arrangement of experts from the ballet Romeo and Juliet on Telarc, I once commented on the weird treble that emanated from this CD. It's a shame that that when I was listening to the CD layer I could not hear what beautiful sound came was contained on the SACD layer. What a difference it was playing this SACD through the Marantz. This is a disc of reference quality, and it is not a bad version of Romeo and Juliet, either. The treble that I once thought of as a one-note quality was a colorful, textured, sonic picture of the actual sonic event. Cymbals had an abundance of air around them, and the dimensions of the rear of the stage and the sound of the hall were clearly audible because of the crystalline ambience surrounding the sound of the highs. The midrange was almost analog-like in that it made it possible to make out each instrument and each section of instruments in the orchestra, as if I could picture each musician contributing to the score.

When I was listening to the Rolling Stones Let it Bleed SACD through the Marantz, it was without a doubt just that much closer to hearing what was on the master tape. I've written before (and so have some others) in these pages how excellent these ABKCO releases are, and I used as musical examples in equipment reviews the CD layer of them before. The SACD layer is a revelation, and if your goal is to hear exactly what Mick, Keith, and company laid down in the studio way back when, these SACDs are the ticket. Through the Marantz, I felt I was definitely getting my money's worth.

But it's one thing to sound like a master tape, but quite another to sound like real instruments playing in the spaces in which they were recorded. The 8260 managed to accomplish this feat that I value more than just about any other player I've heard. If you've read any of my reviews before you'd know that I feel that this trait is more important than any of the other audiophile traits that are casually thrown around in reviews. For example, if it has a great soundstage, some reviewers seem to believe that it hardly matters if the instruments at the rear of that described expanse sound nothing like the actual instruments that were recorded. To me, it does matter.

On the Rolling Stones album mentioned above, there are plenty of instruments besides electric bass and guitars. There is percussion galore, harmonica, keyboards, and of course vocals. Each instrument occupied its own space in the recording and either entered the listening room or was clearly audible within the space in which it was recorded. And they sounded alive, for lack of any other term.

So is the Marantz SA8260 perfect, demanding you rush out and buy one today? That depends. I spoke earlier of its operational quirks, like its pausing to electronically read a disc for what seems like an awfully long time, and the need to manually switch the player from multi-channel to stereo if one wants to play a format that the player doesn't choose by default. But it was quite easy to get used to the procedure, and didn't seem like to much of a chore after a while. But the biggest obstacle in me giving the 8260 an outright recommendation is that its red book CD performance wasn't nearly as good as its SACD performance. I don't mean that I'm disappointed that CDs didn't sound as good as SACDs. Of course they didn't. It's just that the CD performance was bettered by every player I had in my system of late and the ones on hand for this review. Its SACD sound was great, though, and if one already has a good sounding CD player I just wouldn't get rid of it after buying the 8260. If you are looking for a player that excels in both formats, you'll have to look elsewhere.

I don't mean to imply that CDs sounded bad through the 8260, because they didn't. It was a perfectly acceptable CD player. It is just that the Music Hall player that has the same specifications as the Marantz unit sound much, much better on all CDs. And this difference in sound was immediately apparent and pretty easy to spot. If you think that all well built CD players sound the same, pay no attention to me, but I could definitely hear the difference.

Still, I would not hesitate for one minute recommending the Marantz SA8260 as an SACD player. It is packed with features, sounds incredible, is built like a tank, and worked flawlessly throughout the rather long review period. I reluctantly returned it to Marantz at the end of that review period.

--TL

Dr. David Rich on the Marantz SA8260: The Marantz SA8260 offers a pleasant surprise on the analog side with all-discrete signal processing. The operational amplifier that is driven differentially from the DAC uses a single voltage gain stage, called a folded cascode, to provide wide bandwidth and good settling time. Both are critical for handling the one-bit output streams of an SACD player. The absence of second voltage gain stage in this discrete op amp reduces total open loop gain available to reduce distortion; without measurement equipment at T$S, one can only guess whether there is a distortion penalty. The self-biased JFET complementary source-follower to the right of the folded cascode is of traditional design and is the simplest approach to building a complementary unity voltage gain output stage. Although they saved money on the back end, they based the differential pair with a real current source, not just a resistor.

The operational amplifier that forms the filter and output drive stages has, at its foundation, the same folded-cascode front end, but then embellishes upon the design with a second, highly innovative, voltage gain stage followed by a standard bipolar-based complementary emitter-follower topology for the output stage. One is left to wonder--if they used the standard, high performance topology in this op amp, why did they use the simple JFET stage in the first op amp?

The second voltage gain stage is a complementary cascode using bipolar Wilson current mirrors. The current gain for the circuit is supplied from a complementary common source amplifier using JFETs. The engineers' choice of topology for the second gain stage is puzzling to me, because I do not see how this topology will yield lower closed-loop distortion than a traditional discrete op amp topology. Such complex topologies are favored by some expensive high-end equipment. Indeed, some very high-end stuff contains some very low-end integrated operational amplifiers. In contrast, the SA8260 engineers, working with discrete components, included a local loop around the second voltage gain stage in a current mode feedback configuration. Clearly, the engineers are shooting for some signal-processing property not considered in the design of a standard op amp that they think will improve sound quality.

I opine that the special circuitry in the SA8260 does not improve the unit's sound on a double-blind test. If talk of double-blind testing turns you green, then the complex SA8260 circuitry may do something for you. Again, you can pay a lot more and get less ultra high-end fairy dusting of the analog stages. High-end circuit designers may scoff at the electrolytic bypass cap on the output of the analog stage (no other caps in the signal path) or the absence of a relay for the analog mute function (inexpensive bipolar devices are used instead).

The unit has no balanced outputs and all the active circuitry is single-ended. I would prefer balanced signal processing, an approach taken by the DAC manufacturer on the performance demonstration board they designed.

The power supply is discrete with three transistors and a zener diode reference, consistent with the high-end paradigm. Analog supplies are at +/-12V. $200 CD players use integrated regulators on supplies as small as +/- 2.5V. The analog power supply is powered on a separate transformer winding. For a stereo unit, I would complain that the supplies are not dual mono, but with six channels, I think asking for 12 regulators and filters may be too much at this price point. The DACs lack the polished treatment, as does the crystal oscillator. The +5V supply comes from a integrated regulator on a transformer winding that also supplies power to other parts of the CD player.

The Crystal CS4397 DAC is a relatively new design, although the Cirrus website ironically does not recommend it for new designs. The replacement part is the CS4398, which offers better distortion performance and a digital volume control ahead of the DAC. I do not have pricing for the CS4397, but the CS4398 sells for $4.30 when you purchase 10,000. Considering this unit uses three CS4397 ICs, the unit's list price comes into focus. The CS4397 uses the latest multi-bit Delta Sigma modulator technology. The DAC has no guaranteed specs. Full-scale distortion plus noise at I kHz is slightly higher than 16 bits, but improves to 19 bits when the signal level is reduced by 20 dB. Dynamic range is also at the 19-bit level. The chip has a DSD pass-through mode direct to the analog filters. In this mode, dynamic range is unchanged, though the THD + noise is reduced by half a bit. The combined frequency response of the analog and digital filters is -0.02 - +0.015. The digital filter alone is +/- 0.001. Stop-band attenuation is -102 dB at 24 kHz for the 44 kHz mode; the chip has no slow filter mode. Moving up to 96 kHz, the digital filter is down 0.1 dB at 43 kHz and the stop band attenuation if 82 dB down at 55 kHz. Nice performance, but no end to the means, since the SA8260 does not support DVDA playback.

In the DSD mode, only the analog filter is operational. In-band frequency response is specified at no more than -0.013. From graphs on the data sheet, the filter appears to be first-order. The discrete circuitry in the SA8260 adds three orders of filtering to keep the sharply rising noise floor of direct DSD signals from swamping the output of the CD player. Needless to say, analog filtering is a non-issue for DVD-A running at 96 kHz. I fail to see any advantages for the DSD stream, and much prefer the approach used in some universal players where DSD is converted to PCM in the digital domain and out-of-band noise is masked. Of course, conversion in the digital domain is anathema to SACD purists.

Finally I point at the excellent performance of the internal CMOS op amp used in the Crystal CS4397 for the analog filter stage. One is left to wonder whether this op amp removes whatever property of the music signal that Marantz struggles to maintain in its complex back-end circuitry, but the ultra-high-end moves in strange ways.

Passing DSD to an AV receiver requires deluxe receivers, so the analog output of the SACD player is the only means to hear DSD reproduced with 19-bit dynamic range at anything approaching a sensible price. If you somehow think DSD sounds better, then you also need to buy into the concept that some property of the music signal that cannot be measured using standard test equipment is somehow preserved in the DSD process. Whatever that magic thing is, the SA8260 appears to be designed to pass it to the output jacks. Remember, using traditional measurement techniques, DVD-A at 96 kHz exceeds DSD by a significant margin. I find the inability of the SA8260 to play DVD-A a fatal shortcoming.

--DAR