Tuesday, April 25, 2017

Sony 9000ES

I still love the sound of my Sony 9000ES playing SACD's.  They are smooth sounding but plenty passionate dynamics as well.  CD's don't fare as well, they are rendered a bit too smooth generally.

[Update: I later discovered that since I bought it last year the 9000ES had the Audio Attenuator turned ON in the setup menu.  After changing that to OFF the output is now 10dB louder and precisely the same within 0.1dB as my Denon DVD-9000.  The Sony is now competitive on CD's with the Denon which is my current favorite, but the Sony still sounds slightly ethereal compared with the palpable and shake-the-earth DVD-9000.  I think in general I may prefer the Sony to the Onkyo RDV-1 on CD's, but the Onkyo output level is about 1dB lower and until I control for that any comparison will be unfair.  In any case, this change has hugely improved the 9000ES and now it seems I will not have to buy an expensive preamp to equalize the levels for nearly optimal playing if not scientific comparisons.  So I might put my money into a better main DAC instead.]

It remains my belief that the first 3 SACD players from Sony (SCD-1, SCD-777ES, and 9000ES) were extra special because they have the real 1-bit DAC (the so called Current Pulse DAC).  That was the end of the line for Sony DAC's, afterwards Sony decided simply to buy off-the-shelf DAC's from the likes of Burr Brown.

But these Current Pulse DAC's are special in that they represent the 1-bit format itself, unlike multibit Sigma Delta DACs which 'simulate' it.  I think this is what gives these early players a very important sense of dynamic realism which I've never heard from multibit Sigma Delta DACs.  Though 'feedback' remains part of even a 1-bit system, I think it has less to do and over a more limited space, a mere point inside the chip deals with the 1-bit realization of music, and using much more oversampling than sigma delta feedback.

But I may have erred in getting the 9000ES.  Back last year I figured the 9000ES would be more reliable, having a 'more conventional' drive.  But now I've discovered that it is the 9000ES, also having a unique disc spinner, that has more reliability issues, and nobody seems to have solved them.

As is now commonly known, simply replacing the laser on a 9000ES won't bring back the ability to read all SACD's.  And whatever functionality it does restore may not last for long.  Many blame the now 3rd party Chinese lasers (Sony no longer sells factory replacements...haven't for a long time actually).  But looking back in the reviews, people were having trouble and more trouble specifically with the 9000ES back to the beginning.

Another issues was that the 9000ES came out before there were any multichannel SACDS, or any multichannel hybrid SACD's.  When they did come out, many immediately found they would not play on 9000ES.  Or, as in my experience, that they would sometimes play and sometimes not.

I've been able to determine that my unit will play most hybrid multichannel discs WHEN warmed up.  And the best way to warm it up is to play a CD (or non-multichannel SACD) first.

I've tried other tricks too, such as leaving the drawer open after playing the last disc.  That works in keeping the unit "warm" for awhile, but not until the next day.  Similarly you can just leave the disc in the unit without turning anything off, and that will help keep it warm for about a day.  If you turn the unit fully off, it won't be long until it won't read another multichannel hybrid SACD off the bat, and you have to warm it up by playing a CD first.

Reading the hype about the "special" mechanism unique to this player, it seems that the disc enters the mechanism and then descends to a spindle which is lower than the SACD drawer.  This is all in the interest of making a more stable disc platform, with less vibration, which is also the goal of the "moving spindle" mechanism used in the SCD-1 and SCD-777ES.

Subsequently, in later generations of players, they both didn't seem to have so much trouble reading discs, despite far less complexity or intensity of design of the mechanism.

Was this just learning curve?  Did Sony figure out how to achieve the same goals (and even do better) with lighter simpler mechanisms?

I have a pet theory which I have no other evidence for than what I've just described.  Nobody else talks about this at all, but I have a idea.  Much of the original motivation for SACD was to make it impossible to make bit perfect copies that would play.  To that end, the SACD design includes "pit width modulation."  A conforming SACD player must detect the pit modulation, and if the pit modulation isn't correct, it will shut down playing as a "copyright violation."

My theory is that Sony lightened up on the pit modulation DRM after the first generation of players.  Later players aren't quite as picky about how the modulation is.  I wonder if they even examine it at all, though Sony has never said they stopped doing the pit modulation thing, considering all the gazillion mechanisms now made that read SACD's, I think they must have at least made it much easier.

So these first 3 players tried to do the pit modulation copyright thing to the max, and can barely be kept at a sufficient level of performance to keep on doing it.  But the outrageously costly moving spindle mechanism used in the SCD-1 may be able to do it just a tad better than the also specially engineered mechanism in the 9000ES.  So it's quite possible that SCD-1 is in the long run the more maintainable machine.  I shouldn't be telling you this, perhaps, because I still want one myself.

SCD-1's were notorious, in the beginning, for burning out lasers, but once that issue appeared to have been solved, they seem to have been more reliable than the 9000ES's.

It's perhaps not surprising that the PS Audio 1 bit DAC was inspired by the DAC in the 9000ES (and the other original players).  Just like them, it is a 10x oversampled 1-bit converter (operating at 640fS where fS is 44.1kHz).

Now I'm wondering if any other early generation machines used similar 1-bit converters, because sigma delta converters took over everything no much later.  Actually quite a few high end early players used 1704's in dual differential form.  Sony never licensed SACD on the Denon DVD 9000 which had dual differential 1704's, but they did license dual differential 1704's in the Esoteric UX-1, a few years later, and I suspect a number of other machines.  In those cases the 1-bit MUST be converted to PCM around 700kHz, because the 1704's max out around 768khz.

The Phillips SACD 1000 might have used a 1 bit converter like Sony's, but that mechanism was the most notoriously unreliable of all.  Phillips had a couple of downscaled models at that time, the 962 and 963 which might be worth investigating.  I haven't yet looked at the first SACD players from Pioneer.

No comments:

Post a Comment