Thursday, September 29, 2016

You Can't Handle The Truth

The Truth is that CD players can correct small reading errors perfectly (these are in categories C1 and C2).  Larger errors (category CU), they conceal by interpolation from the surrounding good data.  Finally, only the largest errors cause the player to skip or stop completely.

The only way I know about to see counts of C1, C2, and CU is to have a Plextor CD ROM drive and the software with it.  The Plextor software had a feature to scan the disc for these errors.  Of course they are also limitations on the part of the cdrom drive hardware to some degree, but if you have a Plextor drive it's probably about the best drive anyway.

Outside of Plextor, no software or device I know of can do this.

What's especially troublesome to me is this "conceal" possibility (and, btw, "conceal" is exactly what it is technically called, I discovered after reading many documents).

I'd like to know whenever my player is "concealing" read errors.  In that case, I'm not getting the "perfect sound" of 16 bits.  I'm getting something just a bit better than garbage, at least for a tiny moment in time.  Listening to a CD, especially a old/dirty/hybrid/poorly-written CD which maybe very important to me, I have no way of knowing if a large portion of what I'm listening to is fill-in.

At that time, it might be a good time to:

a) trash the disc (or return it) and buy a new one
b) try to clean the disc
c) try to clean the laser with a laser cleaning disc (I've seen the turbo ones with teeny fins recommended, and not the brush kind, and never use liquids)
d) service the player, such as by a technician who will disassemble the player, clean the laser with alcohol and check the electronics
e) buy another player, perhaps newer than the antiques I have
f) switch to playing downloads
g) decide never to listen to recorded music again

Some early CD players, notably those using Philips transports and IC's, had a Disc Fault light which would possibly light under CU error conditions.  Or possibly it would take something worse, similar to that which would make the player skip or stop or not play at all (such as if you put the disc in upside down).  I have looked at the manuals of a few CD player models having these lights, and they were not at all clear.  None discussed CD error correction (always perfect) or concealing (always some kind of educated guess).  Such players include:

Cambridge CD3
Marantz CD74

Now it's a big bother to have to watch the screen where the Disc Fault light is during an entire CD.  It would be nicer to have a counter.  Then you could just look when you were taking the disc out and see how many errors there were on THAT disc.  That's what I think all large or reference CD players should have.

Perhaps, however, various companies involved would rather not even have you think about such things.  Perhaps, legitimately perhaps, there would be endless complaints (this machine gives me a CU when my old one didn't) about hardware and software.  Hardware companies would certainly rather not have complaints about their hardware.  And some hardware companies are also major software (in this case, actual factory recorded CD's) owners or sellers, who would get complaints both ways.

And, perhaps legitimately, even CU's are no big deal.  You could probably blind test a disc with 1000 CU's and nobody would tell the difference.  So why am I even worried, when I can "hear" the difference between my old bitstream, my old PCM, and my newer sigma delta?

Well, because I can't be sure of anything I hear, and having an unknown number of errors, any of which might be ameliorated by some corrective action I could take...well, it bothers me as much as anything.  It may not be clearly audible, but the other things I think I hear probably aren't there either.

BTW, after years of cursing DSD/SACD I am delighting in the sound of my newly acquired minty DVP-9000ES playing SACD's, and I was never even aware of how many SACD's I actually have (such as the old RCA Red Seal and Mercury Living Presence hybrids.  Sadly it is one (just one so far) of the Mercury hybrids which won't play on my 9000ES despite new laser.  When then makes you wonder about SACD that play many errors are being concealed?  It is said that the SACD format is automatically concealing because of the nature of sigma delta.  Anything bit lost is no more than any other, no concealing scheme can do much better than that.  I'm thinking there may be some dispersion on the disc also as there is with CD, so the loss caused by a large gap has only a slight effect on a larger area.

Perhaps the format is even more loss proof, like the CDROM and DVD formats, which use an additional layer of correction.

Anyway, I still love CD, concealment and all.  I think mostly people rarely hear the real faults caused by the CD system-as-a-whole when they complain about how a CD sounds.  The system is nearly perfect, extremely transparent, even at 44.1/16.  There are only slight differences you can make while still being faithful to high fidelity reproduction, and not intentional sound doctoring (which is ok too, if that works for you, as it does even for Mark Levinson with his virtual Cello Pallette, or did for Peter Walker--by all accounts one of the greatest audio engineers ever who designed the Quad speakers and amplifiers) which somehow, I never do.

I recall when I used potentiometers to set volume I always had to readjust the balance for every recording.  It never occurred to me that almost all of the balance twiddling I had to do resulted from the non-accurate tracking of the preamp.  Now that was with my otherwise beloved Aragon 28k, because I was way ahead of this as early as 1979 when I had a sealed stepped attenuator, same as was used in the GAS Thaedra.  Every step was perfectly accurate.  I continued to use that in various systems until the 1990's, then began to realize that this was not necessarily the best design, having a 25k impedance, for driving cables.  But, somehow, that didn't sink in and I continued using the volume control of a Citation One preamp, even higher impedance.  Then came the Aragon, and even that, though it measured about the lowest distortion I've ever seen, essentially perfect, using RMAA and my Juli@ card, I always thought it sounded dark.  And the problem with the volume and balance, it was never right!  I think this must not have had P&G as some think.  Anyway, I have loved digital controls from then on, starting with the Classe CP-35 which I strongly and strangely preferred to the Aragon (but unarguably the digital controls had perfect balance tracking, much appreciated), to the digital controls of the Tact which I have used since 2006 mostly as just a digital preamp, what I still feel is one of the most useful things (a big digital selector with volume trim and more) but hardly anyone makes.  It has 0.1dB volume settings, and balance which is never needed, and polarity but it's a bit inconvenient to switch polarity back and forth since they added parametric EQ feature, which I can't use unless I'm using Tact Room Correction, which I mostly haven't because I don't like it messing with the highs, a feature they never dropped from my 2.0 version.  Anway, it could be better, and I need to think about making something like what is needed, but the 2 Tact units I got during the Tact 2.0 factory blowout have served me very well, and almost irreplaceable so, for 10 years.

Anyway, LP's are far worse than that.  But most of the time, with CD's, you just put it on and it sounds perfect, or if not exactly perfect it would not be proveably discernable to anyone.

But, again, that's assuming that there isn't lots and lots of filling going on.  And you never know, because there is no indicator on all but a teeny proportion of players, and even on them there isn't such a convenience as a counter.  And the best of all would be a counter not just for CU but C1 and C2 also, but possibly most systems don't make that information available.  But just along the same lines that audiophiles want a bigger power supply than absolutely necessary, some think having a larger motor drive the CD spindle is better (that's far out, but of course part of getting a PD-75), knowing about C1 and C2 would help you see patterns, trends, etc.  Know if a cleaning might be good.  And, if they even knew, audiophiles would keep those numbers down.

Instead, in at least some cases, they applied CD tweaks which caused more errors or even breakdowns.

But there you go, does the tweak manufacturer want you to know that?  Certainly not, and he is going to badmouth the manufacturer that lets it happen, so a conspiracy of silence ensues.

This company makes a test disc that permits testing of CD player performance in correcting errors.

A good resource on SPDIF is:


I got my Pioneer PD-75, hooked it up, and it sounded so good, I'm not sure I care about the truth anymore.  This is the best sound source I've ever had.  On CD, it totally blows the Sony DVP-9000ES away, perhaps no surprise.  On hybrid SACD's, it depends, sometimes the greater low end authority, wherever that helps such as in Rock or Organ music that's not too abrasive, the Pioneer still wins.  The Sony characteristically has an Etherial sound, which I consider the sonic characteristic of 1-bit systems, including DSD.  Yin.  The Pioneer (which uses bitstream PCM and minimum filtering) has a Yang sound, perhaps even more than slightly macho at times, reveling in shaking the walls.  It does that, and is still the #2 sweetest sounding source in my collection, following the Sony.  It can create sounds I've never heard that still sound correct.

No comments:

Post a Comment