I started using a network audio system in 2006, Sonos, and I still believe it is a fine system fundamentally. I Sonos Connect nodes (originally called Zoneplayers) to provide coax SPDIF digital signals to systems in every room, mostly from my kitchen Mac's hard drive (a 1Tb Fusion drive), sometimes making connections from analog sources in different rooms. All digital signals are uncompressed 44.1/16, which I've always believed to be fine and good. John Atkinson measured remarkably low jitter from a Sonos based wireless system in 2006. I actually eschew wireless and hard wire all my nodes with Cat-6a ethernet which permits attaching analog sources in 5 different rooms simultaneously for uncompressed forwarding. All nodes are individually wired to a central gigabit switch. I've often dreamed of DIY replacement for this, but it's almost unimaginable with my time constraints now. Newer network audio systems Never (as far as I know) permit attaching analog sources in different rooms. Sonos was the first and only to offer analog inputs and it was why I chose them instead of Squeezebox, though I missed the Squeezebox high rez capability. And most new audiophile computer audio systems are astronomically expensive, like $20,000. I bought Sonos Connect nodes like candy because they're only $395, small price for another room of fully 2-way connected audio.
It's very convenient, but has never played high resolution sources, so I maintained a collection of disc players to play high rez such as DVD-Audio discs (which are sometimes incredibly good), DAD/DVD-Video discs by Classic Audio, which are all fabulous, and SACD's, which I consider pop high rez (widely used if not the best system) but still indispensable. And HDCD's too, which are all among the best recordings I have, and the more I know about it the more I still think HDCD wasn't a bad idea.
Now I've even had my mind turned to the idea that disc players may be different and possibly better than computer network based systems. In any case, it's an excuse to play ordinary CD's on my Denon DVD-9000, which is just about as fun as clicking on things on a computer screen, plus it has class (and weight!). In most cases now I take the analog outputs from disc players and resample them to digital (for my crossovers and DSP) using a Lavry AD10. So, FWIW, I am preserving the sound of the different players, to the degree that it can be done this way (and it is my belief that the 96kHz/24bit sampling of the Lavry is extremely good and transparent enough to do this). I haven't seriously explored the "better sound" I get from playing discs as compared to network bits, but in a few cases I've noted the difference between taking the analog outputs and the digital outputs, and in almost every case I've preferred taking the analog. I consider my Lavry resampling to be as good as the best preamps, and probably better than most preamps. Digital sampling sounds impossible but has been incredibly well developed by our technological society.
As someone with collector tendency, however, somehow I wouldn't mind having EVERY classic disc player in the rack, so I could try them all and vary according to mood. Unfortunately my latest rack was only just set up a month ago and already out of space.
Anyway, I've long felt I wanted a Real SACD player, and not a Universal not particularly noted as having the best sound on SACD's like my Denon DVD-5900.
I've long scanned eBay for SCD-1's and SCD-XA777ES's, the two original and very similar super high end SACD players by Sony. To me, these represent REAL SACD, as Sony originally intended it. I don't necessarily consider these players to be any sort of ultimate for regular CD's, after all Sony was primarily hoping to sell the idea that SACD's were superior to CD's, so one would imagine these players pulling out all the stops for SACD but possibly falling a little short on CD (and some reviews bear this out). Some do say some later machines are actually better on SACD also, such as the much more cheaply built but still pricey XA-5400ES. Somehow that latter machine has never appealed to me much. Anyway, even if it did, I wouldn't want to spend that much as current used pricing for such a lightweight machine. (I could be tempted to spend that much for a heavyweight universal, such as an Esoteric UX-3.)
I especially loved the balanced outputs of the SCD-1, which would be fabulous into my Lavry. I also noted the especially clean noise floor of the SCD-1 in Stereophile review, and I attributed that in part to the balanced outputs (and multiple overbuilt power supplies a factor also).
But both those players had a very fancy but notoriously unreliable CD mechanism, the "fixed laser" mechanism. There doesn't seem to be any real advantage in making the CD spindle move instead of moving the laser. It just seems to be something rather difficult to accomplish, something that few companies would even attempt. Sony of course did claim it was better, but there are many machines people believe to be better transports than the SCD-1. To many observers, and to me, the "fixed laser" motor seems to be a marketing gimmick with no real advantages, and something that saddles the owner of these players with future repair costs. Sony never seems to care about that, in fact their history has been one of a very high level of planned obsolescence, including not making stuff to last very well beyond the warranty period. This is probably why Sony, which now seems to own just about everything, should be the most successful company ever but instead is barely keeping the lights on in their vast empire. What has happened is that generation after generation of new customers has left Sony behind after being burned one time too many. I've been among those burned many times, though I'd still been a big Sony customer until very recently (when they don't seem to make anything I want anymore).
Anyway, the 3rd Sony SACD player to have SACD electronics reportedly similar to the SCD-1 (said to be something like 75% the same) is the Sony DVP-9000ES. There is hardly a time on eBay anymore when there aren't a half dozen or more of those for sale. After years of scanning for them, I've gone ahead and purchased one with a newly replaced laser tested on multichannel SACD's--the hardest to play. This cost more than the current average price but I hope will be worth it.
The DVP-9000ES has a specially enclosed but ultimately ordinary DVD/SACD/CD mechanism as far as I can tell. The laser is shared with other units and plentifully available. I don't even know if it's all that special but people have said the 9000ES is a decent transport...perhaps as good as the SCD-1.
This machine will be my go-to machine for SACD's. For CD's I'll stick to the Denon DVD-9000, the differential 1704 based behemoth. For DVD-Audio and DVD-Video-Audio I'll use the Denon DVD-9000, my current Oppo BDP-95, or perhaps get a new BDP-105 for the living room.
I would have preferred to do DVD-Audio on my DVD-9000, but that feature is broken on my unit now. I could also repurpose my Onkyo RDV-1 as a DVD-Audio player and use something else, perhaps my Denon DVD-5000, as a supertweeter DAC.
Does a transport make any difference? Well I don't know, but I imagine if the transport is temporally unstable to any degree, this is going to affect everything downstream to one degree or another. I suppose instability could be erased simply by reading an entire song into memory. FWIW, people I know had disliked memory-based players. At the risk of sounding technically illiterate, I actually don't know if any electronic systems have less jitter. I've very suspicious of all asynchrous digital systems like ethernet, and I have always thought USB to be subject to the same highly jittery phenomena but it might have a kind of "synchronous" mode, I don't understand that but I still think it's not fully synchronous.
One has to understand that even if things are buffered, reclocked, or whatever, at the end, those subsequent phases will ALWAYS have to adapt to the source...so the source matters. (And a jittery transport layer matters also, though entire-song delivery would seem to be safe, so where is it unsafe, I don't know.) It may be in principle that something based on a later clock can simply clock things out. But if _it_ has to wait longer or shorter, that may cause some unintended small difference, say from the heat in the circuitry for having to wait longer, or negotiate with the source for more data packets more often. So one can try to wall off noise in the time domain, but no such wall is ever perfect, computers don't really work on 1's and 0's, they work with electrons and semiconductor junctions, everything shifts a little, and for most work the absolute clock stability isn't important once you're off the chip and even perhaps there.
So I really don't know if semiconductor memory in a dedicated circuit, or any actual circuits, would abolish jitter any better than a spinning disc. I think it probably would, but I don't actually know. It could well be that if disc transports sound different than electronic sources, it's because the disc transports are expressing their character (of some kind) in the timing. And in this case, where it isn't captured by conventional measurements or ideas, I'm all for this kind of difference. I want to try all the different flavors of candy in the shop of my imagination!
So what looks good to me? Well, the VRDS Neo players from Esoteric, I'd love to have UX-1 or UX-3 because those are Universal Disc players, and that's ultimately what I need. Even if it isn't Sony, or even "real" DSD, it's close enough to be a Universal for me, high end mechanism and quad 1704's. These are the most heavyweight low resonance mechanism ever.
I'd love to have a CEC made belt transport, even in the Parasound 2000 version (which was reputedly based on CEC's cheapest model). You can't argue with top load and screw down, that's always been the best way IMO.
Pioneer's Stable Platter machines look interesting, but only a few use "the real" Stable Platter with heavy parts. These were incredibly high end machines I knew nothing about since I wasn't much following high end audio in the mid 1990's.
Sony did make a few excellent pieces of machinery in the 90's. I believe the XA5ES, one of Lampizator's favorites, is one of these. After that, and after the whiz bang SCD-1 and SCD-777ES, it's been pretty much all plastic. I think the key assembly of the mechanism of my 9000ES is all plastic, if not the entire thing. I recall my 507 is largely plastic. (I fixed the 507 mechanism replacing the quickly broken rubber band with a weight.)
Phillips really did more in the direction of engineering things to last (but then, sometimes didn't). Famously they made the CDM-1 mechanism and a large number of all metal derivatives.
The original Theta Data transport was based on a Phillips Laserdisc transport (basically, that's all it was, with thick panel added and tiny proprietary board added, which according to Lampizator, doesn't do anything useful, but he could be wrong, and price raised from $500 to $5000--US Engineering!).
Nevertheless, Lampizator has loved that transport and thinks there is magic in the much heavier mechanism and 30 times more powerful motor in a laser disc player. So....there's a wide field of interesting looking laserdisc players. Notably the top Pioneer players, that look to have excellent mechanism and digital output, but more research necessary. Meanwhile I find the top Pioneer Stable Platter players, a cheaper version of which was used in a Wadia player, most interesting. And I had not known at all about such things until now.
I do think that keeping the disc stable and clamped to a heavy disc on one side and tight clamp on the other is a most excellent idea. I have long liked serious looking players with user tightened weights. The SCD-1 appealed to me but now I see was not the giant I thought, but I still like the top load idea, Krell did some most excellent top load CD players and there may have been others. But it seems Pioneer was accomplishing this full disc clamping idea through a well engineered mechanism long ago, with just an ordinary (well, large) drawer.
For those with transport obsession (or is it transport lust?) one essential guide is a transport list, this is a good one, though I think I've seen others that are about the same in comprehensiveness, and nothing is ever truly complete.
Every time I look at these lists I find interesting factoids. For example, my Denon DVD-9000 (aka DVD-A1) uses a Hitachi HOP-1000. That rings true as I know it needs a Hitachi laser. The Esoteric X-01 and X-03 use variants of the VRDS-Neo and the UX models aren't even listed (but I think they use the corresponding transports). The Onkyo RDV-1 uses a Mitsumi PVR-202T. And, what I was just checking for today, the Parasound C/DP-1000, C/DP-2000, and C/BD-2000 (transport only) all use the same CEC transport, the CEC-Sanyo SF-P1 / SF-90. So guess which model is the easiest path to belt drive CD nirvana? In the Pioneers, the PD-75, PD-77, and PD-95 all use the ultimate "magnetic stable platter" PWY1004. CDM-1's can be obtained in a number of Proceed models. My old Sony 507ESD uses a KSS-151A. The later Sony 990 favored by a friend sometimes used a much cheaper looking KSS-240A, which is among the most widely used transports ever. The expensive but problematic SCD-1 mechanism is the KHS-180A. Teac (the parent company of Esoteric) makes some models that claim to have a VRDS mechanism like the big dog Esoteric's. But they don't, they have the VRDS CMK which looks like a barely warmed over Sony mechanism, which it appears to be, as the transport also includes the suffix KSS-151A--the same as my respectable but ancient CDP-507ESD. Interestingly enough, a number of well known Wadia models use the VRDS CMK KSS-151A transport also. BTW it was a broken spring, not a broken rubber band, that I replaced with a weight. The original spring had gotten too weak for the mechanism door to work.
Here's a discussion of the various top Pioneer CD players from the 1990's with Stable Platter. The biggest of the big dogs is PD-93, followed by PD-95, PD-S95, and PD-75 which all use the same Magnetic Stable Platter mechanism I listed above. And also the PD-91, PD-73, and PD-77. Other lesser Stable Platter machines include the PD-65, PD-54, PD-52. The Burr Brown PCM63K equipped models are the PD-93, the PD-73. Much as I enjoyed the discussion, it's critical to check back on the mechanism list to be sure what is what.
Question: Do transports even make a difference? Well if they don't, or if even if they do and systems based on USB or Ethernet are better, we could just kiss transports goodbye, except for the "fun" of putting discs in them, and for that we could just get the cheapest available unit (not sure of how many there are anymore, but tons of old classics on eBay). Except I personally get more satisfaction from putting my CD's in a big heavy expensive and rare player, especially if it was widely regarded as being the best.
For most of the time since CD was introduced, I haven't believed that transports make any difference, and I still don't necessarily believe they do.
As long as the bits are being decoded properly (and it is usually assumed they are, but I think periodic maintenance testing may be called for) the only difference the transport could make would be in the timing. Rather than emitting the digital samples at a perfectly stable rate of exactly 44.1kHz, it might vary slightly. This is called Jitter.
And this can (in principle) be measured directly or indirectly (it is usually measured indirectly by looking at the distortion spectrum), but few people do these measurements and publish them except for John Atkinson of Stereophile.
The issue that separates the objectivists from the subjectivists is not whether or not something like jitter is possible, it certainly exists! The issue that separates the two camps is whether jitter in actual digital systems is audible.
The best and just about the only objective research on that topic was done quite awhile ago and published in JAES. Basically they found that jitter because audible at levels about 10-100nsec, which is about 100-1000 times greater than is found in typical digital audio systems (say, 250psec).
So that's it, and I had pretty much accepted that view for the last 15 or so years, and even before that I highly doubted the audibility of jitter.
While I strongly doubt that there is this kind of difference between players, I still want the best one with the lowest jitter and the most substantial build quality and design quality, etc., just because I deserve it just as much as people who think, wrongly, that they need it.
Anyway, here's Lampizator's take on the audible differences.
I have this different theory that like looking at nice stuff, and then we enjoy the music more. But anyway, here's another discussion on the transport audibility. Many in this group are using a what is a pedestrian DVD player now (the Denon DVD-2900, which may have originally sold for $1000 in 2002 but now typically sells for about $100 used) as their CD transport if not their CD player, and more focusing on getting good (or at least interesting) DAC's. I would generally agree that the DAC is more important, but that brings up another thing--according to objectivists all DAC's should sound the same also, given distortion below 0.1%, except for NOS dacs which have distortion well above 1%.
My heart is warmed by seeing one guy in that group (and I know this is fairly popular) using a Sony DVP-9000ES as transport and/or CD player. Because that's the machine I will be getting tomorrow. To me, however, it doesn't look special either as CD player or as transport necessarily, I am buying it for the very serious 1st generation stereo SACD capability, said to be among the best. As a transport, it looks to use a commodity laser and similar mechanism to that in other DVD players.
To me, a CD transport should be a dedicated CD transport and go beyond the call of duty in every way possible, such as holding the disc flat and not flopping and shaking.
I have seen what most people have not, the insides of a player being adjusted, and it's not a pretty picture, and it does even lead you to wonder how the CD system can work at all.
Of course the bottom line is stability, error free reading, and lowest of the low jitter, and also low hum and noise (because even on the SPDIF output, that contributes to jitter also). By most reasoning, it doesn't matter how much metal the transport uses with the intent of reducing jitter, it's the actual jitter performance which matters, not the intent, expense, effort, or intelligence that went in to reducing jitter. The problem is that it's not so easy to measure jitter and errors, and it can't be done all the time until we retire the player many years later (if it's a good one, and possibly never if possible). So we look to surrogates of good performance, like "build quality", or fancy mechanisms which clamp discs tightly or are claimed to reduce vibration in some way (which could all be marketing hype, as it could have been for the fixed laser transports). The more effort and expense was put into making something better, the better we think it probably is.
So I laugh with Lampizator when he laughs at the real deal VRDS-NEO transports..."lots of heavy metal, if you call that engineering." But I also laugh at him, and me, for dissing all the cheap lightweight plastic transports anyway. Of course it should be understood that most intensive engineering actually goes into designing the plastic mechanisms. They take lots of engineering to do correctly, including perhaps such things as finite element analysis. With big heavy slabs of metal, you can just pile it on to your heart's and wallet's content. And if Esoteric is doing the piling for you, it's going to cost many kilo bucks.
(Lampizator is very inconsistent, BTW. On the one hand he liked the Theta Data transport because it's a big Laserdisc transport with big motor. Elsewhere he says smaller motors are better for good sound. And so on. He's unbelievably consistent in another way. Everything he modifies becomes the best thing ever. But he's fun to read and especially the dissections of CD players is interesting.)
Lampizator did an interesting dissection of the Audio Alchemy DDS Pro. According to him it is basically just a Pioneer PDS-502. And he think compared to standard mechanism, the cheap version of Stable Platter is all plastic, every significant part is plastic, including spindle and bearing and it wobbles like hell, and though it works fine, ordinary middle of the road Sony and Philips transports of the time were high end by comparison. This makes me feel good I got the PD-75 with the largely metal Magnetic Stable Platter which has features like a heavy metal platter and spindle, sapphire bearing, and a motor with far more torque than most players. Below, however, Stephen Sank has a different view, he recommends the Audio Alchemy to others even though he has one of the big dogs.
BTW, regarding the Denons, the transports are generally metal. The HOP 1000 mechanism in my DVD-9000 is all metal, looks solid and ok, but not overbuilt though, the cast metal early Philips CDM mechanisms looked far stronger (from pictures). And there isn't necessarily anything special wrt how the CD is gripped. I think that detail is too often overlooked, which is why I like the Pioneer stable platter idea. Even top loading CD mechanisms that require you to hold down the disc with a magnetic weight still have essentially let the CD flap in the breeze.
Did Pioneer lose the 1990's CD wars because Sony had better sound, as The Vintage claims, or better marketing. I think I'm more cynical.
A guy at DIYAudio did distortion spectra testing with a number of CD players and DAC's. Guess what, one of the best of the bunch was the PD-75. TDA 1541A looks pretty awful actually, and TDA1540 one of the very worst. PS1 looks horrible. 1702 is not good at all, 1704 better but not perfect. His favorite chip is a particular PCM56 he found, which adjusted looks just about perfect, but he notes that other PCM56's have been far worse. PCM63's in PD-8500 looks awful.
Strangely however, people find the bitstream chips in the PD-75 to sound particularly undynamic, and the TDA1541A is more popular than ever.
It really looks as though people prefer distortion. But perhaps the guy above didn't do comprehensive enough measurements. It turns out The Audio Critic did a review of the PD-75 when it was introduced and found very clean highs but excess distortion (still at levels believed inaudible) in the bass. That does suggest something wrong with the audio section on this unit, but not likely the transport, as if the transport were adding excess jitter that would be as much or more evident in the highs.
In that review, the reviewer said he saw no advantage in putting the disc upright (or clamping it onto a felt platter, though he didn't mention that) except that less dirt would fall on the laser.
Unfortunately it appears that some time after its introduction, while much loved, the Stable Platter transport became infamous for dropping lasers onto CD's, or falling apart during shipping. Famous equipment repair and modification authority Stephen Sank advises strongly against PD-75/91/93/95 for that reason and one other. He himself uses the PDR-09, which features the same fantastic mechanism but with a much better laser. PDR-09 is even more high priced and less attainable than PD-93. For the rest of us pedestrians, he suggests the PD-S06, which has the more plasticky PD-65 mechanism but with a very good clock, and PCM1702's, very highly regarded for sound. He says it makes a great transport, but you'll want to use it as a player because it sounds so good.
Well guess what there's a PDR-09 for sale on ebay right now which I could have budgeted for and might get someday. The PD-S06 is widely available on European sites according to HiFiShark but not on US eBay.
Lampizator also likes the PD-S06, though he started out with his usual complaint about the supposedly-great "stable platter" mechanism being nothing but a plastic job like any cheap player. But he doesn't inform us this is hardly the only "stable platter" the real stable platter machines are the Magnetic Stable Platter big dogs like PD-75 and PDR-09. If he looked at one of those he wouldn't be able to make his joke. So here we we see even in Lampizator as with all other reviewers that we never get the full story.
Rethinking the ideal transport issue, it may be that the master clock within it may be the most important concern, as both Sank and Lampizator suggest. Whatever the mechanism does, it ultimately feeds bits faster or slower into some kind of memory buffer. The memory buffer drives the spindle motor servo to run faster or slower to keep the buffer at some middle level. Only if there are errors or buffer under run is the mechanism an issue (according to some). The clock ultimately clocks everything from the buffer into the internal DAC and SPDIF output. So it's the clock, ultimately, which counts most (or entirely, according to some).
I argue that given a particular clock mounted to a particular machine, the drive also matters as the drive speeds up and slows down, changing the rate of information being added to the buffer. As you get close to one end of the buffer or the other, I'm concerned about some kind of quantum time compressing effect which most engineers wouldn't be concerned about. Also there is the unavoidable fact that different amounts of current will flow in the motor, possibly upsetting the clock.
So even in my far out view, the clock is still the most important, but if the clocks are otherwise equal, then the mechanism could still have some effect.
But how good are clocks to begin with? I have read in fact that the designers of the CD standards already knew that the clocks were going to be very important, and they set standards, with two levels of excellence. The first is 50 ppm, and the second is 1000ppm. Presumably any high end CD player or transport would far exceed the first standard, and would that be audible? I would think it would be already quite good enough, particularly with an "Elite" or "ES" player, good enough to not require a clock to have jitter far less than would be audible.
At this time I haven't figured out how to convert from these "ppm" specifications to the usual "picoseconds of jitter" specification. And then how often do we actually know the jitter specifications, most manufactures don't give them and they are highly dependent on the equipment setup. According to many, if your biggest concern is jitter your best approach is to buy the best possible "single box" player you can get, which only has analog outputs. Then there won't be any issue with jitter caused by the transmission of digital signals outside the box, which is one of the biggest concerns in modern audio systems, at least in part because we use (even in AES digital) single wire (or differential, which is only slightly better) signals which combine the clock AND the data and therefore lead to the possibility of the highly undesirable data-influenced jitter.
One thing, unless I can determine that my PD-75 has far less jitter than anything else I have, I may not be happy with it until it has a sufficiently upgraded clock. I'm realizing now that my Onkyo RDV-1 has a special Apogee sourced clock and maybe I should use that player as a transport (as well as using it for DVD-Audio…for which it has PCM 1704's).
In fact, I'd been thinking is that perhaps all a person needs is a memory player with a good clock. This sounds like it would be cheap to make, but the only one I know of is the $4000 PS Audio Memoryplayer.
If it has a good clock, and the servo is strong enough to keep the internal buffer near midpoint, then all a good transport needs to do is provide data reliably. But how do we know it's reliable? Sadly few players have ever provided any way of knowing. One of these was the Cambridge Audio CD3, which had a Disc Fault light (I would assume those show the hard errors that cause fill-in, not the soft errors that are fixed by CRC, we don't really need to know the soft errors, but I think what would be best is a counter that shows both types for the current disc being played). The CD3 looks to be a nice unit with a Philips CDM1 Mk2 mechanism. A transport version CD3M was also made and now priced many times higher at resale. The companion DAC unit used 4 1541A's.
It's very convenient, but has never played high resolution sources, so I maintained a collection of disc players to play high rez such as DVD-Audio discs (which are sometimes incredibly good), DAD/DVD-Video discs by Classic Audio, which are all fabulous, and SACD's, which I consider pop high rez (widely used if not the best system) but still indispensable. And HDCD's too, which are all among the best recordings I have, and the more I know about it the more I still think HDCD wasn't a bad idea.
Are Disc Transports better than Computer-network systems?
Now I've even had my mind turned to the idea that disc players may be different and possibly better than computer network based systems. In any case, it's an excuse to play ordinary CD's on my Denon DVD-9000, which is just about as fun as clicking on things on a computer screen, plus it has class (and weight!). In most cases now I take the analog outputs from disc players and resample them to digital (for my crossovers and DSP) using a Lavry AD10. So, FWIW, I am preserving the sound of the different players, to the degree that it can be done this way (and it is my belief that the 96kHz/24bit sampling of the Lavry is extremely good and transparent enough to do this). I haven't seriously explored the "better sound" I get from playing discs as compared to network bits, but in a few cases I've noted the difference between taking the analog outputs and the digital outputs, and in almost every case I've preferred taking the analog. I consider my Lavry resampling to be as good as the best preamps, and probably better than most preamps. Digital sampling sounds impossible but has been incredibly well developed by our technological society.
As someone with collector tendency, however, somehow I wouldn't mind having EVERY classic disc player in the rack, so I could try them all and vary according to mood. Unfortunately my latest rack was only just set up a month ago and already out of space.
Anyway, I've long felt I wanted a Real SACD player, and not a Universal not particularly noted as having the best sound on SACD's like my Denon DVD-5900.
I've long scanned eBay for SCD-1's and SCD-XA777ES's, the two original and very similar super high end SACD players by Sony. To me, these represent REAL SACD, as Sony originally intended it. I don't necessarily consider these players to be any sort of ultimate for regular CD's, after all Sony was primarily hoping to sell the idea that SACD's were superior to CD's, so one would imagine these players pulling out all the stops for SACD but possibly falling a little short on CD (and some reviews bear this out). Some do say some later machines are actually better on SACD also, such as the much more cheaply built but still pricey XA-5400ES. Somehow that latter machine has never appealed to me much. Anyway, even if it did, I wouldn't want to spend that much as current used pricing for such a lightweight machine. (I could be tempted to spend that much for a heavyweight universal, such as an Esoteric UX-3.)
I especially loved the balanced outputs of the SCD-1, which would be fabulous into my Lavry. I also noted the especially clean noise floor of the SCD-1 in Stereophile review, and I attributed that in part to the balanced outputs (and multiple overbuilt power supplies a factor also).
But both those players had a very fancy but notoriously unreliable CD mechanism, the "fixed laser" mechanism. There doesn't seem to be any real advantage in making the CD spindle move instead of moving the laser. It just seems to be something rather difficult to accomplish, something that few companies would even attempt. Sony of course did claim it was better, but there are many machines people believe to be better transports than the SCD-1. To many observers, and to me, the "fixed laser" motor seems to be a marketing gimmick with no real advantages, and something that saddles the owner of these players with future repair costs. Sony never seems to care about that, in fact their history has been one of a very high level of planned obsolescence, including not making stuff to last very well beyond the warranty period. This is probably why Sony, which now seems to own just about everything, should be the most successful company ever but instead is barely keeping the lights on in their vast empire. What has happened is that generation after generation of new customers has left Sony behind after being burned one time too many. I've been among those burned many times, though I'd still been a big Sony customer until very recently (when they don't seem to make anything I want anymore).
Anyway, the 3rd Sony SACD player to have SACD electronics reportedly similar to the SCD-1 (said to be something like 75% the same) is the Sony DVP-9000ES. There is hardly a time on eBay anymore when there aren't a half dozen or more of those for sale. After years of scanning for them, I've gone ahead and purchased one with a newly replaced laser tested on multichannel SACD's--the hardest to play. This cost more than the current average price but I hope will be worth it.
The DVP-9000ES has a specially enclosed but ultimately ordinary DVD/SACD/CD mechanism as far as I can tell. The laser is shared with other units and plentifully available. I don't even know if it's all that special but people have said the 9000ES is a decent transport...perhaps as good as the SCD-1.
This machine will be my go-to machine for SACD's. For CD's I'll stick to the Denon DVD-9000, the differential 1704 based behemoth. For DVD-Audio and DVD-Video-Audio I'll use the Denon DVD-9000, my current Oppo BDP-95, or perhaps get a new BDP-105 for the living room.
I would have preferred to do DVD-Audio on my DVD-9000, but that feature is broken on my unit now. I could also repurpose my Onkyo RDV-1 as a DVD-Audio player and use something else, perhaps my Denon DVD-5000, as a supertweeter DAC.
Does a transport make any difference? Well I don't know, but I imagine if the transport is temporally unstable to any degree, this is going to affect everything downstream to one degree or another. I suppose instability could be erased simply by reading an entire song into memory. FWIW, people I know had disliked memory-based players. At the risk of sounding technically illiterate, I actually don't know if any electronic systems have less jitter. I've very suspicious of all asynchrous digital systems like ethernet, and I have always thought USB to be subject to the same highly jittery phenomena but it might have a kind of "synchronous" mode, I don't understand that but I still think it's not fully synchronous.
One has to understand that even if things are buffered, reclocked, or whatever, at the end, those subsequent phases will ALWAYS have to adapt to the source...so the source matters. (And a jittery transport layer matters also, though entire-song delivery would seem to be safe, so where is it unsafe, I don't know.) It may be in principle that something based on a later clock can simply clock things out. But if _it_ has to wait longer or shorter, that may cause some unintended small difference, say from the heat in the circuitry for having to wait longer, or negotiate with the source for more data packets more often. So one can try to wall off noise in the time domain, but no such wall is ever perfect, computers don't really work on 1's and 0's, they work with electrons and semiconductor junctions, everything shifts a little, and for most work the absolute clock stability isn't important once you're off the chip and even perhaps there.
So I really don't know if semiconductor memory in a dedicated circuit, or any actual circuits, would abolish jitter any better than a spinning disc. I think it probably would, but I don't actually know. It could well be that if disc transports sound different than electronic sources, it's because the disc transports are expressing their character (of some kind) in the timing. And in this case, where it isn't captured by conventional measurements or ideas, I'm all for this kind of difference. I want to try all the different flavors of candy in the shop of my imagination!
So what looks good to me? Well, the VRDS Neo players from Esoteric, I'd love to have UX-1 or UX-3 because those are Universal Disc players, and that's ultimately what I need. Even if it isn't Sony, or even "real" DSD, it's close enough to be a Universal for me, high end mechanism and quad 1704's. These are the most heavyweight low resonance mechanism ever.
I'd love to have a CEC made belt transport, even in the Parasound 2000 version (which was reputedly based on CEC's cheapest model). You can't argue with top load and screw down, that's always been the best way IMO.
Pioneer's Stable Platter machines look interesting, but only a few use "the real" Stable Platter with heavy parts. These were incredibly high end machines I knew nothing about since I wasn't much following high end audio in the mid 1990's.
Sony did make a few excellent pieces of machinery in the 90's. I believe the XA5ES, one of Lampizator's favorites, is one of these. After that, and after the whiz bang SCD-1 and SCD-777ES, it's been pretty much all plastic. I think the key assembly of the mechanism of my 9000ES is all plastic, if not the entire thing. I recall my 507 is largely plastic. (I fixed the 507 mechanism replacing the quickly broken rubber band with a weight.)
Phillips really did more in the direction of engineering things to last (but then, sometimes didn't). Famously they made the CDM-1 mechanism and a large number of all metal derivatives.
The original Theta Data transport was based on a Phillips Laserdisc transport (basically, that's all it was, with thick panel added and tiny proprietary board added, which according to Lampizator, doesn't do anything useful, but he could be wrong, and price raised from $500 to $5000--US Engineering!).
Nevertheless, Lampizator has loved that transport and thinks there is magic in the much heavier mechanism and 30 times more powerful motor in a laser disc player. So....there's a wide field of interesting looking laserdisc players. Notably the top Pioneer players, that look to have excellent mechanism and digital output, but more research necessary. Meanwhile I find the top Pioneer Stable Platter players, a cheaper version of which was used in a Wadia player, most interesting. And I had not known at all about such things until now.
I do think that keeping the disc stable and clamped to a heavy disc on one side and tight clamp on the other is a most excellent idea. I have long liked serious looking players with user tightened weights. The SCD-1 appealed to me but now I see was not the giant I thought, but I still like the top load idea, Krell did some most excellent top load CD players and there may have been others. But it seems Pioneer was accomplishing this full disc clamping idea through a well engineered mechanism long ago, with just an ordinary (well, large) drawer.
For those with transport obsession (or is it transport lust?) one essential guide is a transport list, this is a good one, though I think I've seen others that are about the same in comprehensiveness, and nothing is ever truly complete.
Every time I look at these lists I find interesting factoids. For example, my Denon DVD-9000 (aka DVD-A1) uses a Hitachi HOP-1000. That rings true as I know it needs a Hitachi laser. The Esoteric X-01 and X-03 use variants of the VRDS-Neo and the UX models aren't even listed (but I think they use the corresponding transports). The Onkyo RDV-1 uses a Mitsumi PVR-202T. And, what I was just checking for today, the Parasound C/DP-1000, C/DP-2000, and C/BD-2000 (transport only) all use the same CEC transport, the CEC-Sanyo SF-P1 / SF-90. So guess which model is the easiest path to belt drive CD nirvana? In the Pioneers, the PD-75, PD-77, and PD-95 all use the ultimate "magnetic stable platter" PWY1004. CDM-1's can be obtained in a number of Proceed models. My old Sony 507ESD uses a KSS-151A. The later Sony 990 favored by a friend sometimes used a much cheaper looking KSS-240A, which is among the most widely used transports ever. The expensive but problematic SCD-1 mechanism is the KHS-180A. Teac (the parent company of Esoteric) makes some models that claim to have a VRDS mechanism like the big dog Esoteric's. But they don't, they have the VRDS CMK which looks like a barely warmed over Sony mechanism, which it appears to be, as the transport also includes the suffix KSS-151A--the same as my respectable but ancient CDP-507ESD. Interestingly enough, a number of well known Wadia models use the VRDS CMK KSS-151A transport also. BTW it was a broken spring, not a broken rubber band, that I replaced with a weight. The original spring had gotten too weak for the mechanism door to work.
Here's a discussion of the various top Pioneer CD players from the 1990's with Stable Platter. The biggest of the big dogs is PD-93, followed by PD-95, PD-S95, and PD-75 which all use the same Magnetic Stable Platter mechanism I listed above. And also the PD-91, PD-73, and PD-77. Other lesser Stable Platter machines include the PD-65, PD-54, PD-52. The Burr Brown PCM63K equipped models are the PD-93, the PD-73. Much as I enjoyed the discussion, it's critical to check back on the mechanism list to be sure what is what.
Question: Do transports even make a difference? Well if they don't, or if even if they do and systems based on USB or Ethernet are better, we could just kiss transports goodbye, except for the "fun" of putting discs in them, and for that we could just get the cheapest available unit (not sure of how many there are anymore, but tons of old classics on eBay). Except I personally get more satisfaction from putting my CD's in a big heavy expensive and rare player, especially if it was widely regarded as being the best.
For most of the time since CD was introduced, I haven't believed that transports make any difference, and I still don't necessarily believe they do.
As long as the bits are being decoded properly (and it is usually assumed they are, but I think periodic maintenance testing may be called for) the only difference the transport could make would be in the timing. Rather than emitting the digital samples at a perfectly stable rate of exactly 44.1kHz, it might vary slightly. This is called Jitter.
And this can (in principle) be measured directly or indirectly (it is usually measured indirectly by looking at the distortion spectrum), but few people do these measurements and publish them except for John Atkinson of Stereophile.
The issue that separates the objectivists from the subjectivists is not whether or not something like jitter is possible, it certainly exists! The issue that separates the two camps is whether jitter in actual digital systems is audible.
The best and just about the only objective research on that topic was done quite awhile ago and published in JAES. Basically they found that jitter because audible at levels about 10-100nsec, which is about 100-1000 times greater than is found in typical digital audio systems (say, 250psec).
So that's it, and I had pretty much accepted that view for the last 15 or so years, and even before that I highly doubted the audibility of jitter.
While I strongly doubt that there is this kind of difference between players, I still want the best one with the lowest jitter and the most substantial build quality and design quality, etc., just because I deserve it just as much as people who think, wrongly, that they need it.
Anyway, here's Lampizator's take on the audible differences.
I have this different theory that like looking at nice stuff, and then we enjoy the music more. But anyway, here's another discussion on the transport audibility. Many in this group are using a what is a pedestrian DVD player now (the Denon DVD-2900, which may have originally sold for $1000 in 2002 but now typically sells for about $100 used) as their CD transport if not their CD player, and more focusing on getting good (or at least interesting) DAC's. I would generally agree that the DAC is more important, but that brings up another thing--according to objectivists all DAC's should sound the same also, given distortion below 0.1%, except for NOS dacs which have distortion well above 1%.
My heart is warmed by seeing one guy in that group (and I know this is fairly popular) using a Sony DVP-9000ES as transport and/or CD player. Because that's the machine I will be getting tomorrow. To me, however, it doesn't look special either as CD player or as transport necessarily, I am buying it for the very serious 1st generation stereo SACD capability, said to be among the best. As a transport, it looks to use a commodity laser and similar mechanism to that in other DVD players.
To me, a CD transport should be a dedicated CD transport and go beyond the call of duty in every way possible, such as holding the disc flat and not flopping and shaking.
I have seen what most people have not, the insides of a player being adjusted, and it's not a pretty picture, and it does even lead you to wonder how the CD system can work at all.
Of course the bottom line is stability, error free reading, and lowest of the low jitter, and also low hum and noise (because even on the SPDIF output, that contributes to jitter also). By most reasoning, it doesn't matter how much metal the transport uses with the intent of reducing jitter, it's the actual jitter performance which matters, not the intent, expense, effort, or intelligence that went in to reducing jitter. The problem is that it's not so easy to measure jitter and errors, and it can't be done all the time until we retire the player many years later (if it's a good one, and possibly never if possible). So we look to surrogates of good performance, like "build quality", or fancy mechanisms which clamp discs tightly or are claimed to reduce vibration in some way (which could all be marketing hype, as it could have been for the fixed laser transports). The more effort and expense was put into making something better, the better we think it probably is.
So I laugh with Lampizator when he laughs at the real deal VRDS-NEO transports..."lots of heavy metal, if you call that engineering." But I also laugh at him, and me, for dissing all the cheap lightweight plastic transports anyway. Of course it should be understood that most intensive engineering actually goes into designing the plastic mechanisms. They take lots of engineering to do correctly, including perhaps such things as finite element analysis. With big heavy slabs of metal, you can just pile it on to your heart's and wallet's content. And if Esoteric is doing the piling for you, it's going to cost many kilo bucks.
(Lampizator is very inconsistent, BTW. On the one hand he liked the Theta Data transport because it's a big Laserdisc transport with big motor. Elsewhere he says smaller motors are better for good sound. And so on. He's unbelievably consistent in another way. Everything he modifies becomes the best thing ever. But he's fun to read and especially the dissections of CD players is interesting.)
Lampizator did an interesting dissection of the Audio Alchemy DDS Pro. According to him it is basically just a Pioneer PDS-502. And he think compared to standard mechanism, the cheap version of Stable Platter is all plastic, every significant part is plastic, including spindle and bearing and it wobbles like hell, and though it works fine, ordinary middle of the road Sony and Philips transports of the time were high end by comparison. This makes me feel good I got the PD-75 with the largely metal Magnetic Stable Platter which has features like a heavy metal platter and spindle, sapphire bearing, and a motor with far more torque than most players. Below, however, Stephen Sank has a different view, he recommends the Audio Alchemy to others even though he has one of the big dogs.
BTW, regarding the Denons, the transports are generally metal. The HOP 1000 mechanism in my DVD-9000 is all metal, looks solid and ok, but not overbuilt though, the cast metal early Philips CDM mechanisms looked far stronger (from pictures). And there isn't necessarily anything special wrt how the CD is gripped. I think that detail is too often overlooked, which is why I like the Pioneer stable platter idea. Even top loading CD mechanisms that require you to hold down the disc with a magnetic weight still have essentially let the CD flap in the breeze.
Did Pioneer lose the 1990's CD wars because Sony had better sound, as The Vintage claims, or better marketing. I think I'm more cynical.
A guy at DIYAudio did distortion spectra testing with a number of CD players and DAC's. Guess what, one of the best of the bunch was the PD-75. TDA 1541A looks pretty awful actually, and TDA1540 one of the very worst. PS1 looks horrible. 1702 is not good at all, 1704 better but not perfect. His favorite chip is a particular PCM56 he found, which adjusted looks just about perfect, but he notes that other PCM56's have been far worse. PCM63's in PD-8500 looks awful.
Strangely however, people find the bitstream chips in the PD-75 to sound particularly undynamic, and the TDA1541A is more popular than ever.
It really looks as though people prefer distortion. But perhaps the guy above didn't do comprehensive enough measurements. It turns out The Audio Critic did a review of the PD-75 when it was introduced and found very clean highs but excess distortion (still at levels believed inaudible) in the bass. That does suggest something wrong with the audio section on this unit, but not likely the transport, as if the transport were adding excess jitter that would be as much or more evident in the highs.
In that review, the reviewer said he saw no advantage in putting the disc upright (or clamping it onto a felt platter, though he didn't mention that) except that less dirt would fall on the laser.
Unfortunately it appears that some time after its introduction, while much loved, the Stable Platter transport became infamous for dropping lasers onto CD's, or falling apart during shipping. Famous equipment repair and modification authority Stephen Sank advises strongly against PD-75/91/93/95 for that reason and one other. He himself uses the PDR-09, which features the same fantastic mechanism but with a much better laser. PDR-09 is even more high priced and less attainable than PD-93. For the rest of us pedestrians, he suggests the PD-S06, which has the more plasticky PD-65 mechanism but with a very good clock, and PCM1702's, very highly regarded for sound. He says it makes a great transport, but you'll want to use it as a player because it sounds so good.
Well guess what there's a PDR-09 for sale on ebay right now which I could have budgeted for and might get someday. The PD-S06 is widely available on European sites according to HiFiShark but not on US eBay.
Lampizator also likes the PD-S06, though he started out with his usual complaint about the supposedly-great "stable platter" mechanism being nothing but a plastic job like any cheap player. But he doesn't inform us this is hardly the only "stable platter" the real stable platter machines are the Magnetic Stable Platter big dogs like PD-75 and PDR-09. If he looked at one of those he wouldn't be able to make his joke. So here we we see even in Lampizator as with all other reviewers that we never get the full story.
Is it All in the Clock?
Rethinking the ideal transport issue, it may be that the master clock within it may be the most important concern, as both Sank and Lampizator suggest. Whatever the mechanism does, it ultimately feeds bits faster or slower into some kind of memory buffer. The memory buffer drives the spindle motor servo to run faster or slower to keep the buffer at some middle level. Only if there are errors or buffer under run is the mechanism an issue (according to some). The clock ultimately clocks everything from the buffer into the internal DAC and SPDIF output. So it's the clock, ultimately, which counts most (or entirely, according to some).
I argue that given a particular clock mounted to a particular machine, the drive also matters as the drive speeds up and slows down, changing the rate of information being added to the buffer. As you get close to one end of the buffer or the other, I'm concerned about some kind of quantum time compressing effect which most engineers wouldn't be concerned about. Also there is the unavoidable fact that different amounts of current will flow in the motor, possibly upsetting the clock.
So even in my far out view, the clock is still the most important, but if the clocks are otherwise equal, then the mechanism could still have some effect.
But how good are clocks to begin with? I have read in fact that the designers of the CD standards already knew that the clocks were going to be very important, and they set standards, with two levels of excellence. The first is 50 ppm, and the second is 1000ppm. Presumably any high end CD player or transport would far exceed the first standard, and would that be audible? I would think it would be already quite good enough, particularly with an "Elite" or "ES" player, good enough to not require a clock to have jitter far less than would be audible.
At this time I haven't figured out how to convert from these "ppm" specifications to the usual "picoseconds of jitter" specification. And then how often do we actually know the jitter specifications, most manufactures don't give them and they are highly dependent on the equipment setup. According to many, if your biggest concern is jitter your best approach is to buy the best possible "single box" player you can get, which only has analog outputs. Then there won't be any issue with jitter caused by the transmission of digital signals outside the box, which is one of the biggest concerns in modern audio systems, at least in part because we use (even in AES digital) single wire (or differential, which is only slightly better) signals which combine the clock AND the data and therefore lead to the possibility of the highly undesirable data-influenced jitter.
One thing, unless I can determine that my PD-75 has far less jitter than anything else I have, I may not be happy with it until it has a sufficiently upgraded clock. I'm realizing now that my Onkyo RDV-1 has a special Apogee sourced clock and maybe I should use that player as a transport (as well as using it for DVD-Audio…for which it has PCM 1704's).
In fact, I'd been thinking is that perhaps all a person needs is a memory player with a good clock. This sounds like it would be cheap to make, but the only one I know of is the $4000 PS Audio Memoryplayer.
If it has a good clock, and the servo is strong enough to keep the internal buffer near midpoint, then all a good transport needs to do is provide data reliably. But how do we know it's reliable? Sadly few players have ever provided any way of knowing. One of these was the Cambridge Audio CD3, which had a Disc Fault light (I would assume those show the hard errors that cause fill-in, not the soft errors that are fixed by CRC, we don't really need to know the soft errors, but I think what would be best is a counter that shows both types for the current disc being played). The CD3 looks to be a nice unit with a Philips CDM1 Mk2 mechanism. A transport version CD3M was also made and now priced many times higher at resale. The companion DAC unit used 4 1541A's.
Measuring Jitter
When I started this blog, I had hoped to do objective measurements more frequently than, well, the sort of rambling I had ended up usually doing. But it seems I have things I need to say even without doing measurements.
But I'd sure love to check out the jitter of different players and/or player configurations. NwAvGuy, the guy who did excellent audio blogging for a few years then disappeared, gives the details here.
But I'd sure love to check out the jitter of different players and/or player configurations. NwAvGuy, the guy who did excellent audio blogging for a few years then disappeared, gives the details here.
Other Info on Stable Platter players
This report gives far more details than the short review in The Audio Critic on the PD-75. He measures THD as 0.004% from 2-20Hz, and 0.0035 at 1kHz. That's excellent, and not worth worrying about, however it is as much as 9dB worse than the CD format is capable of (around 0.0015), so maybe that's in line with what The Audio Critic is reporting. However reading this review about other design features and aspects of technical performance makes it sound like the PD-95 is one of the best players ever. I'm not so sure about that, but it's comforting to know about all the good design features and aspects of performance.
Actually the above report may simply have been cribbed from the Stereo Review review of the PD-95 published in September 1991, as shown in this collection of Pioneer reviews.
The Audio Critic review of the PD-75 complained that the distortion rose to -78dB in one channel in the mid bass, and -83dB in the other. What are these numbers expressed in percent?
-78dB is 0.0125% distortion about 4 times more than claimed by Stereo Review (they claimed 20-20k at 0.004%)
-83dB is 0.071% distortionn, about twice what Stereo Review claimed.
BTW, I think the best you can do with CD is THD -96dB, which would be:
0.00158%
though I'm not quite sure, I've also seen -98dB, which would be:
0.00126%
In any case I believe The Audio Critic when they say that the levels of distortion produced by PD-75 would not be audible, and they gave this player the "silver" award after noting in disappointment that the PD-73 had been better and winner of previous tests. But the reviewer preferred to buy the Sony CDP-779ES. He had to excuse himself, because there is no reason to buy such a pricey player (it was $1900) for audible performance alone, on that score a $20 CD player would be sufficient according to his endless preaching. But he liked how the 779ES was made, and he said that for $12,000 he'd pass, but for $1900 he could afford and wanted the undeniable quality in excess of need. I believe this was Aczel speaking, in a later set of reviews it was David Rich who bought the successor CDP-707ES, essentially the same as the 779ES. Or vice versa.
The Audio Critic review of the PD-75 complained that the distortion rose to -78dB in one channel in the mid bass, and -83dB in the other. What are these numbers expressed in percent?
-78dB is 0.0125% distortion about 4 times more than claimed by Stereo Review (they claimed 20-20k at 0.004%)
-83dB is 0.071% distortionn, about twice what Stereo Review claimed.
BTW, I think the best you can do with CD is THD -96dB, which would be:
0.00158%
though I'm not quite sure, I've also seen -98dB, which would be:
0.00126%
In any case I believe The Audio Critic when they say that the levels of distortion produced by PD-75 would not be audible, and they gave this player the "silver" award after noting in disappointment that the PD-73 had been better and winner of previous tests. But the reviewer preferred to buy the Sony CDP-779ES. He had to excuse himself, because there is no reason to buy such a pricey player (it was $1900) for audible performance alone, on that score a $20 CD player would be sufficient according to his endless preaching. But he liked how the 779ES was made, and he said that for $12,000 he'd pass, but for $1900 he could afford and wanted the undeniable quality in excess of need. I believe this was Aczel speaking, in a later set of reviews it was David Rich who bought the successor CDP-707ES, essentially the same as the 779ES. Or vice versa.
No comments:
Post a Comment