BOC: WOTT mish mosh, remastering..

[Paul Mather]

There's been plenty chatter about CD remastering lately, so I thought
I'd chip in my 2p...

Craig Shipley writes:

> No insult meant, Bryan, but just upping the sampling rate or the bit
> density won't do a damn bit of good if the original source sounds like
> crap! Stero Review did a bit (naughty boy! No puns!) on the 20-bit gold
> CD's that Columbia was/is reissuing. Their finding was that the biggest
> improvement came from remastering the music for the CD medium from the
> master tapes. The 24k gold and 20-bit mastering contributes (the 24k gold
> is probably contributing more to profit$ than to sound) somewhat, but not
> as much as that lil' ol' remastering job. A little bit of sonic tweaking
> can help (aka Sonic Solutions or No Noise), but it must be used carefully
> or you can throw the baby out with the bath water! Like, getting rid of
> most, if not all, of the tape hiss on the first BOC...

> Ideally, the BOC back-catalog deserves the re-mastering effort, along with
> the restoration of the original liner notes/photos. 20-bit mastering would
> be nice (hell, start mastering _all_ CD's with the 20-bit thingamabobby;
> can't cost any more to do it than normal 16-bit. And just how many of you
> have a 20-bit CD player, anyhoo???)

Well, probably quite a few people now have "18 bit" CDs, but I don't
know if we're up to 20 yet.

Craig is right.  The most significant improvement in the quality of
reissues stems from the treatment of the master tapes.  If your source
is crap, you gwine get crap come out the other end.  Remember the
basic audiophile tenet: improve closest to source!

Regarding "20 bit" technology, etc., well, I don't know how much this
is just to impress us plebs.  I'm not fancy educated or nuthin', but I
do know that for the current CD format both the sample size *and* the
sampling rate is fixed (16 bit and 44.1KHz respectively).  So, like
Craig says, don't be looking for 20 bit samples to appear on your CDs
any time soon.

What "20 bit" mastering is telling you is that *during the mastering
process*, they work at a higher sampling resolution.  When comitted to
CD, however, these 20 bit samples will have to be squeezed back into a
16 bit range.  Personally, I would think that it would be NORMAL to
work not only in a higher resolution sample size, but also at a higher
sampling RATE when remastering.  (But then I'm just an information
theorist at heart.)  Sampling rate is just as important as sample
resolution; you can have 1024 bit samples if you want, but if you
don't sample often enough, you're not going to be able to accurately
recreate the original signal, no matter how many bits each sample has.

Which brings me back to 20 bit CDs...  Recall what a CD is trying to
achieve during playback: it is trying to reconstruct an analogue
signal from a sampled digital representation.  How well it can do this
depends essentially upon a) the sampling rate, b) the sample
resolution, and c) the digital-to-analogue converter.  It is possible
(think about it) to achieve a >16 bit *effective* resolution out of
your CD player from only 16 bit samples.  (This is what the ballyhoo
about Bitstream is all about.)  It all depends upon the accuracy of
your DAC.  In fact, I seem to recall that elementary sampling theory
states that so long as your sampling rate is twice that of your
highest frequency, you can **in theory** recreate the original signal
perfectly.  **In practice**, however, digital-to-analogue converters
are not good enough to interpolate that well.

Btw, this "gold CD" lark is a joke, right??  I presume it is touted as
lowering the bit error rate (BER)?  (Like the famous "green CD pens.")
Or are they still pushing that old scare tactic of "the lacquer that
ate my CDs..." horror story?  I don't think BER is a very significant
concern nowadays.  I mean, the 8-to-14 and Cross-Interleaved
Reed-Solomon error-correcting codes used for CD data is plenty ammo
against a troublesome BER.  In fact, the thing that most people forget
in their praise of CD is that once the data leave the DAC, it's the
quality of the *analogue* stages of the CD player that are going to
govern the ultimate sound quality.  (Which is a way of saying, "that's
right, not all CD players sound alike.")  If your CD player has cheapo
analogue components, probably the degradation due to noise will do far
more damage to the signal than a poor DAC will ever manage.

Cheers,

Paul.

obHiFi: Napalm Death, _Scum_

e-mail: paul at csgrad.cs.vt.edu                    A stranger in a strange land.