Here's an interesting paper to read:


http://www.prismsound.com/m_r_downloads/cdinvest.pdf

The blind test there says that no one was able to perceive a difference, not even the two "golden ears" that tested it.

Barry Diament whom I am sure you have all heard of is involved in a thread at Steve Hoffman's forums claiming he hears a difference. I never have.

Claiming is one thing. Proving is another, especially if there are no problems with the copying process.

I absolutely agree however he's one of the most respected mastering engineers and he's done testing compared to the master tapes, a luxury most do not have.

As they say in Korea, even the granddaddy of the most respected mastering engineer in history is susceptible to observer bias as well.

I've never heard about him; did he ABX'ed it?

Ivan.

Here's the thread:

http://www.stevehoffman.tv/forums/showthread.php?t=83595


Here's his website:


http://www.barrydiamentaudio.com

He's written many papers AGAINST the compression/loudness wars.

Credits:


Some mastering credits:

AC/DC
Aerosmith
Arditti String Quartet
Atlantic Soul Classics
Bad Company
Anita Baker
Tony Banks
Tommy Bolin
Edie Brickell & New Bohemians
Jackson Browne
Hiram Bullock
C.S.Angels
Call
Cars
Coasters
Phil Collins
Jayne Cortez
Jim Croce
Crosby, Stills & Nash (& Young)
Jim Crow
Cult
D*Note
Danzig
Divinyls
Dokken
Dream Command
Dreamtime
Eagles
Emerson, Lake & Palmer
Enya
Kevin Eubanks
Fears For Art
Firm
Roberta Flack
Steve Forbert
David Foster
Frankie Goes To Hollywood
Robert Gass
Genesis
J. Geils Band
Georgie Satellites
Guns N' Roses
Steve Hancoff
Hanoi Rocks
Hoodoo Gurus
Amhad Jamal
Freedy Johnston
Jolly Boys
Grace Jones
Juster
Rusty Kershaw
Kitaro
Eartha Kitt
Led Zeppelin
Alvin Lee
Julian Lennon
Bob Marley (ENTIRE CATALOG)
Meatloaf
Medicine Band
Mike & The Mechanics
Milo Z
Modern English
John Mooney
Mötley Crüe
Ivan Neville
Anthony Newman
Gary Nicholson
Stevie Nicks
Nico
Nu Shooz
Ted Nugent
Of Cabbages & Kings
Onyan Art
Robert Palmer
Robert Plant
Rascals
Ratt
Otis Redding
Rise Robots Rise
Linda Ronstadt
Serious Pilgrim
Shadowland
Simply Red
Slayer
John David Souther
State Of Being
Steps Ahead
Subdudes
Sun and the Moon
Tesla
Dr. Fiorella Tirenzi
Pete Townshend
U2
Underworld
Danny Wilde
Work Of Art
Yes
Warren Zevon


Some of the labels:

Atlantic
Bar None
BMG
Bola Press
Caravan of Dreams
Chrysalis
Def American
EastWest
Elektra
Fibre
Geffen
Island
Mercury
Midnight Fantasy
Mode
Musical Heritage Society
Music Masters
Polygram
Second Hearing
Sire
Soundkeeper Recordings
Spring Hill
Sword In The Stone
Triple X
Tuff Gong
TVT
Warner Brothers
Zoo

That doesn't mean a thing. Judging from his posts, the tests were sighted; flawed comparison.

The thread sounds like somebody cut out the TOS#8 violations from HA and compiled it into a separate thread.

I really had a hard time to refrain myself from swearing and really wonder how people with such reputation could argue like that (or, the other way round, how people with opinions like that can get such a reputation)

oh - my - god.

I guess he is good producer, but still living in analog domain (you know, turntables sounds better than CDs), as I've read in the pdf here.

Ivan.

The fact is - assuming your rip is correct - that a good burn of the copy (e.g. on TY media) can indeed show a lower BLER and better jitter values (e.g. L3 amplitude) than the industrially-produced original.

A low-quality standalone cd drive which shows problems reading some rather 'marginal' original discs might show a positive reaction (sound-quality-wise) when being fed with the high-quality copy.

But I seriously doubt that anyone could successfully ABX between original and copy in this scenario ... either the original or the standalone device would have to be of really horrible quality to make positive ABXing possible ... which would render correct ripping on a computer drive difficult, too.

Night Rain, please use a codebox when posting long lists. It is all the more annoying that the lists you posted have no bearing on the subject. If anything, they prove that even a crackpot audiophile can have a resumé three pages long. Though he doesn't particularly impress me as one. Note his wording: he never goes out on the limb and claim something as a fact, preferring extrapolations and expressions like "...which suggests to me...". Either he is humble enough, or smart enough to know his limits.

JeanLuc, some of that Diament's prudent minciness must have rubbed off on you. :P I'm sure you know the facts as well as anyone: if there are audible differences between the pressed original and the CD-R copy, they will manifest themselves as skips or pops, crackle or some such artifact - not the better "focus", fine detail and delineation of space BS Barry is talking about in that thread.

LOL ... usual alarm words for HA members ... makes you either start to uncontrollably laugh or throw up ...

Ok they've convinced me. I don't know what further evidence you could possibly ask for.

So basically he has absolutely no evidence, a lot of questionable ideas that completely undermine his credibility and is esentially claiming the impossible?

Unless I'm missing something, I'm going to assume hes wrong until he can prove otherwise.

Are you blind? He did mastering for AC/DC and Aerosmith! [/sarcasm]

Not necessarily (although you are certainly right for most kind of really defective discs) ... I was rather thinking about CD drive error concealment techniques which are hard to spot most of the times.

JeanLuc is of course right.

There is a theoretical possibility of a cd-r sounding better than a really badly pressed original audio cd.

However, this is mostly theoretical.

CD-discs are an error prone medium.

Errors are detected and fixed by EDC/ECC.

Not all errors can be fixed always.

The errors that cannot be fixed, the player will try to conceal.

There are various concealment techniques: sample&hold, null, linear extrapolation, non-linear extrapolation.

Almost all concealment results in loss of original signal (on signal theory level).

Some of this loss of signal could theoretically be audible under certain specific conditions.

If error distribution is random, the most likely effects are transient distortion during peaks or just lower s/n ratio due to random noise.

A good CD-R reader is better in EDC/ECC than a normal audio player.

Furthermore a really good CD-R burn can be several orders of magnitude better in it's readability (read error count) than a sub-par pressed audio cd.

Hence, a good CD-R reader/burner can make a 'better' (on error quantity level) copy of a bad audio cd.

Now, whether the possible non-fixable and badly concealed errors are audible and under what type of material / conditions remains highly debatable.

I have not seen clear evidence that such errors are easily/reliably detectable under standard double blind listening conditions.

Errors usually need to be more clumped and break even the error concealment mechanism of players (i.e. result in skips and pops), before they can be reliably detected. That's my experience anyway and I see no literature that contradicts this experience.

However, the theoretical possibility remains and the scientific mind remains open to this possibility while at the same time acknowledging it's very low likelihood.

I highly doubt anyone could tell the difference between a cd hot of the press and perfectly burned duplicate. That's like saying you can hear the difference between FLAC and WavPack

All of these points are of course true, but back in the real world, my experience is that genuine uncorrectable errors are very rare. A few months ago I completed ripping my entire CD collection for Squeezebox use (about 1100 discs), and although I didn't keep a detailed record, I'd say that there were less than a dozen CDs that had any uncorrectable errors at all.


My experience is that when a concealed error is audible, it usually takes the form of a low-level "tick" rather than any kind of distortion or noise in the traditional sense. (Of course, a tick *is* noise, but not what we generally mean by the term).

Interestingly, I've just ripped a brand new (visually pristine) CD which had lots of uncorrectable errors towards the end. Using a Plextor PX712A, with either EAC or Plextools, there was no way to avoid getting a stream of "ticks" in the result. So I took it to an audio CD player to see how it would sound (with the intention of recording the SPDIF output), but the ticks were still present. In desparation, I tried switching *off* audio error detection in Plextools, and got a tick-free rip at high speed. I suspect that using EAC in burst mode would have achieved the same result. From this, I conclude that the PX712A has a very good error concealment algorithm (better than the audio player I tried) when ripping in burst mode, but when asked to report on C2 errors, it would appear that it doesn't deploy error concealment (which makes sense in retrospect).


Well, it seems my Plextor certainly has better error *concealment* then the audio player I tried. I have no idea whether its error *correction* is any better. And this is just one anecdotal data point. I can't see why in principle an audio CD player need be any worse than a CDR drive. Indeed, since the audio player reads at 1x speed, the number of read errors can be expected to be lower than with a typical CDR drive, which will operate at higher speeds.

I have just tried ripping the CD with EAC in burst mode. It *still* had the audible ticks. So I tried synchronised/fast mode - still got the ticks. I tried various combinations of the options within secure mode, and still got the ticks. So it seems that perhaps the Plextor drive only switches on its error concealment in response to some special instruction from Plextools. Not really that relevant to this thread, but an interesting observation that I thought was worth posting.

I did a DAE test on PX-716A. It has virtually non existant concealment. It had not too bad error correction though. Highest spikes were around -12 dB(A) in the DAE test. I cannot bother to find the links now, but I made a number of posts on my disappointment with Plextor Premium and PX-716A. I never layed my hands on PX-712A though.

Triza

So what could be happening here is that inferior CD transports may be more sensitive to badly pressed CDs, in the sense that their concealment algorithms may not be up to par, or that high disc jitter could trip them up. I could definitely see this happening although it's hard to find concrete proof. Audiophile CD player manufacturers may or may not be using the best transports. Audiophiles can easily mistake increased sensitivity to errors to mean increased sensitivity in general. (Allegedly, modern audiophile MC cartridges are far more sensitive to pops and ticks than 1970s-era mass produced, high end MM designs.)