Conference hips::uk_audioo

    re .0.....
    
    First of all you need to accept that no matter how good either medium
    is, CD and LP will always sound different.
    
    IMO a lot of the sonic aberrations of CD are down to the bits-is-bits
    philosophhy which proclaims that nothing can go wrong in the digital
    domain if the data is preserved, therefore given a halfway competent
    DAC the sound should be as near the master tape as it's possible to
    get. 
    
    The first stumbling block is the CD production master tape; it
    isn't just a digital reel-to-reel version of the analogue master tape,
    as you'd need more tape drives than Barclays Bank to read the data, but
    is usually a Umatic video tape. If you've ever viewed one of these
    after it's had several plays via a good monitor you'll see how many
    opportunities there are for the CD master tape to start becoming
    useless. Data errors can be corrected to a degree, of course; but how
    many uncorrectable errors do the CD mastering plants allow before they
    throw out a tape and use a new one? Your guess is as good as mine, but
    you can bet your life it'll be lots.
    
    Then there's the CD mastering process itself. CDs are pressed using
    stampers, just like records, and just like LP stampers, CD stampers
    wear out. Apart from obvious data errors creeping in, the leading and
    trailing edges of the pits are crucial because the CD player needs to
    synchronise its clock with that used in the recording process, and this
    is implied by being embedded in the data rather  than a discrete
    seperate signal. Producing the stampers is the expensive part of the
    process, so not only are some plants tempted to cut corners in stampeer
    production, but they may well use them well after they should have
    been binned. Part of this is no doubt due to the way the industry is
    organised these days; producing LP masters has always been a black art,
    needing special skills all the way from the cutting lathe to acrylic
    production. It was usually done in house and cost the major labels a
    fortune - and probably still does. CD manufacture is done by specialist 
    automated companies, and for a long time now there's been over capacity;
    consequently the mastering plants  have had their margins squeezed by
    the labels, screwing the lowest price out of them. So the temptation
    (or imperative!) to cut costs by getting the most out of each stamper.
    
    Then of course you have problem of clock instability in the CD player
    itself. The complexity of a music-derived signal is such that correct
    interpretation of timing is essential, but only a few CD players have
    clocks good enough. Those who like CD point to those of us who don't
    and say we're addicted to LP's droopy frequency response, but for me
    that's tosh. Spectral aberrations I can live with/get used to, but
    it's the inability of the CD players I've heard (not many, I'll grant
    you, but some fairly state-of-the-art, upmarket models) to play music
    in time that really spoils CD. Intense multitrack recordings conceal
    this to a degree, as does some classical music, but small
    ensembles/groups or solo recordings show up the CD medium very badly -
    to my ears in my system, anyway.
    
    In my view, CD is suffering the same problems LP did in the 70s - high
    costs sqeezing margins causing quality to be sacrificed while prices
    stay up. Conversely, a lot of LP mastering/pressing is done by 
    specialists now, like The Exchange, and Porky Peckham's, and some of
    the stuff they produce is exquisite - using quality electronics, direct
    metal mastering, quality-controlled electroplating, LP has never been 
    better since CD came out!
    
    Going back to your two grand burning a hole in your pocket, my advice
    would be to avoid CD until it's sorted out. Spend it on a CD player and
    you'll need a new recorded music collection too. Spend it on
    Ekossing/Trioka-ing your Linn, and you'll get one for free, as it'll be
    just like replacing all your records.
    
    Believe me, I've been there ;-)
    
    Dave

    aggree with .1
    
    Have been living with Arcam Delta 170/Black Box CD & LP12/Ekos/Trokia
    for some time.
    
    (through Naim 72/250 / Isobariks).
    
    I Would like to hear ANY CD that can sound better than record. If it
    happens I will replace arcam so I can play my CD only (Grr ;-{) Zappa
    Records.
    
    Rob.

RE: .0;

Funny that DDA sounds really good. I've heard the Tracy Chapman LP sound
much better than record (demoed by Linn at Bristol HiFi show), and demonstrable
to a lesser degree on my record.

RE: .1;

I've never tried the comparison of small ensembles on CD & LP. I'll have to
give it a go.

Mark.

    Re: .1
    
    I'm somewhat amazed that given all the problems mentioned in .1 about
    CD mastering and production is even possible. If it's not possible
    to get enough correct data on the disk for audio, then CD-ROM must be an
    impossibility! Just like the bumble bee! :-) :-) :-) :-)

	Here here!. If there are those who prefer the muddy qualities
	of vinyl discs so be it, but they do tend to lay it on a bit don't
	they.

	-John

Two points:

- The CD-ROMS are kept in plastic cases to protect them.

- They use error correcting similar to BACKUP I think.

Even with that, I couldn't install DECdecision from them.

Mark.

    The DECUS CD's were withdrawn because of mastering errors. Any attempts
    to put more that 500Mbyte on them caused data to be stored too close to
    the edge, which results in vastly increased error rates.
    
    Also the CD device driver can read the disk MANY times to try and get a
    error free (well correctable) read.  This isn't possible for CD players
    because the player only reads ahead by a short period of time and hence
    any re-reading is limited.
             
    Also CD-ROM doesn't suffer from the timing problems Dave mentioned (I
    think).
    
    Rik (who-would-love-the-conveniance-of-a-CD-if-only-they-sounded-ok)

re .4:

Hey, I'm perfectly happy with the CD-ROM on my VAXstation.  After all, I don't
care exactly when it gives me my data, as long as it does so within the next
second or so.  Which it usually does.  But I care rather a lot if my next music
sample is a little late or early.  Quite how late or early it can be before I
start objecting I don't know, but I'll be surprised if it's much.

Since (as far I'm aware) time is continuous, not discrete, I reckon CD players
are analogue devices.  Hmmm.

Nick

    >> Since (as far I'm aware) time is continuous, not discrete, I reckon CD
    >> players are analogue devices.  Hmmm.

    Sorry,
    	But, as far as I'm aware, time is discrete. I think the quanta for
    time is about 10E-22s but I can't remember for sure. This is discussed
    in ``A Cartoon History of Time'' - the same people who do the New-
    Scientist cartoons. It also discusses time travel and black holes. But
    alas absolutely nothing on the pros and cons of CD reproduction.

    	Rik.

    re several back......
    
    *Please* read .1 properly. I did not assert that you can't get all the
    data onto a CD. All I said was that a Umatic cartridge was used because
    the number of reel-to-reel tapes would be large, given the amount of
    data required to encode music correctly to the current standard.
    
     Comparing getting data off a CD-ROM with correctly reading a music CD
    and reproducing a reasonable facsimile of the original performance is
    misleadingly reductionist.Data doesn't need to be read in real time, 
    or converted to analogue, therefore the timing relationship between
    succesive words is not so critical. "Music data", on the other hand,
    needs to be presented to the inputs of the DAC in exactly the same time
    relationship to that with which it was encoded in order to preserve
    the original note shape which contains critical timing information. If
    CD mastering is such that the master synchronising clock derived from
    the data shifts beyond certain limits, or the clock within the CD
    player that's trying to get in sync drifts of it's own account, then
    data will *not* be presented to the DAC with the correct time
    relationship. To assert that just because the data has been read
    accuratley, or read and corrected, therefore the sound will be of
    constant quality is a gross oversimplification, and one which has been
    used to mislead punters since CD came out; I am surprised so many
    people in the industry we are in have been taken in by it. Then of
    course there are the mass-market/transistor radio quality supporting
    electronics, which compound the felony.
    
    CD when mastered and played back as well as it can be sounds very good
    indeed; but mostly only the specialist labels produce disks of
    appropriate quality, and currently only players at the mid-to-high end
    from specialist manufacturers are designed to tight enough specs to
    take advantage of these superb disks. Just like LP, except the problems
    with LP replay a better known, are largely overcome.
    
    Dave