MFSL Blind Faith – Ultradisc vs. Ultradisc 2

I'm not sure why you are saying I am criticizing you. All I did was reprint your words and let them speak for themselves. It is up to the reader to determine whether your opinions have credibility or not.

 

Dave, I will say what others apparently won't: when it comes to digital audio, you just don't get it. This isn't meant as an insult, and more as a statement of fact. I base this upon such things as the following (or, uh, preceding):

This question reveals that you fundamentally do not understand my position, and what I have argued time and time again. Of course a null test would register a change of .5db. A null test would register a change of .01dB (although the "Aftermath" would be comparatively slight). A null test checks to see if the contents of two samples are more or less identical. SOMETHING--an "audio content" something, in fact, and not mere white noise--is left over if one sample is the same, but louder than the other. If I paste an absurdly-low-volume recording of someone going "MEEPMEEPMEEP" under the song "Helter Skelter" on an otherwise-unadulterated "White Album" CD, it's unlikely that anybody would hear it, but a null test would reveal it.

This is a fallacy by omission. Those aren't the only options. Firstly, "lying" is such a strong word. I'm not accusing you of lying. Nor did I ever accuse Whitefang of lying. However, even if you're not "lying," you can definitely be "hearing things that aren't there," especially if you are convinced (as you apparently are) that you have "golden ears" and an "especially revealing system." People are very suggestible. It's not a sign of weakness, nor something shameful. Heck, I find that I "imagine" the sound of something differently depending on disc/package art. A bright colored background tends to get me to think of a song one way, a dark background another. I recently caught myself describing El Malo's third album (imaginatively entitled "III") as "bleak," despite the fact that the songs really aren't all that bleak at all--the artwork, on the other hand, is quite bleak indeed.

It's very hard, however, for the computer to "hear things," especially when it isn't "hearing" at all. I'm asking the computer "are the 1s and 0s that make up this sample the same?" If they are...well, the ball's in your court, I guess, as my options are exhausted.

You have to realize here that your position is fairly untenable scientifically. What you are saying is that observation trumps fact. The computer programs are not prone to failure, at least not in the way you describe. If the contents are the same, it's going to tell me. If they're not, it's not. This is pretty much a binary system (although there is occasionally a third option..i.e. "the contents are the same, but one is louder," which often has essentially the same results as option A), with the odds greatly stacked against the contents being identical.

Thus, there is another option: the contents are identical, and people hear things that aren't there (or, alternately, some vagueries of individual systems somehow impart a slightly different quality to things that are, in fact, the same at the source).

Look, we all make mistakes. I've made tons in my tenure. Lots of embarassing assertions...lots of "I think this is a little bit better" (with, say, the Repertoire "My Friend Jack," which turns out to be identical--and in fact the *source*--for most subsequent versions) which turned out to be unwarranted. But the key to this methodology is realizing that bias and assumption do not necessarily translate to truth. When I found out that the version of "My Friend Jack" on the Repertoire CD matched the one on the Sin-drome CD, I didn't say "That's ridiculous. The Repertoire one clearly sounds better. There must be something ELSE going on." I said "huh, isn't that interesting?" and moved on. At no point did I put myself above the deductive process.

[

"Proven by many to not hold water?" Excuse me? WHEN? I've never seen proof. In fact, if we're going to talk about "proofs"--and, consequently, facts--I do believe that *I* have facts on my side, and you have testimonials. I've seen people grasping at straws and implying a hilarious amount of god-in-the-gaps reasoning to the realm of digital audio. Tell ya what...you have a proof? Why don't you troop on over to James Randi and make $1,000,000.

This is, of course, also my reasoning for speaking of. Despite what you might think, people do not come into this debate biased on the side of objectivity. An authoritative voice is often very persuasive (once again, see White Fang).

I'm done here. I see this as the "definitive iteration" of this endless ****ing argument, and I will now shut up and crawl back under my rock until the rapture raises me on up to heaven, freeing me from these earthly compulsions.

 

Mr. Goodwin, you are right on the money once again.

 

I think you mean brutally revealing system. But seriously, your explanation is on the money. No one has accused anyone of being a liar. There is a reason why medical researchers use placebo drugs on some patients. You cannot trust humans to accurately gauge what they are experiencing.

Placebo:

a. A substance containing no medication and prescribed or given to reinforce a patient's expectation to get well.

b. An inactive substance or preparation used as a control in an experiment or test to determine the effectiveness of a medicinal drug.

One interesting thing I have noticed in this ongoing debate. I haven't seen anyone write that they think a UD2 is superior to a UD1. I think it has something to do with the collector mentality many of us have on this board.

 

With either, my attitude is "prove it." I know that many folks have much better ears and better stereos than mine, but I am skeptical by nature and I cannot simply believe that a person hears differences simply because they say so and believe so. If a person claims to hear differences between two CDs that are digitally identical but has not demonstrated their ability to do so in a controlled, objective setting, I tend to be skeptical and to suspect it may be listener bias that is influencing what they think they hear. Human perception can be colored by so many psychological and situational factors. I've found that sometimes a CD will sound different to me tonally when I'm driving at night than it did during the day. Since it's the exact same CD, I'm pretty sure there are psychological, perceptual reasons for what I hear rather than objective ones. If there are any objective studies in which a person has been able to consistently distinguish between two digitally identical discs I would love to hear about them.

A question... if one computer (a PC doing a null test) is not able to distinguish differences between two CDs, how is it that another computer (your CD player) would be able to distinguish differences and then replicate them upon playback for your ears to hear? It doesn't make sense to me.

 

Nor am I attempting to insult you, but it is the computer programs and those who believe strictly in them that don't get it. Here I quote the most sensible explanation for hearing what I do...

I may be more inclined to believe in these programs 100% if I had not heard repeated differences first and had not done the actual listening. The programs I've heard are missing something, don't know what it is, that humans are hearing.

OK, fair enough. If I were hearing something that was not there how do you explain others with entirely different systems hearing the same thing? These are not phantom things David where I am being fooled repeatedly time and time again. I've never said that you've called anyone a liar. I've never said ever that I have golden ears. Apparently my system is very revealing and always has been. I discovered that back on the old DCC Forums when Steve told us all to check for a back-splash echo off of a snare drum hit on a CCR track and I reported back describing exactly what I heard. (Can't recall which right now)

Listening observation trumps computer programs that can not distinguish as well as the human ear for some inherent reason with regards to audio sound. I don't pretend to even understand why.

If this were the case you are saying that several mid to high end systems, which I don't believe for a second BTW, are adding coloration's to only some CDs. Also, it would effect even the UD1's if this were the case.

Actually with Aja it was more wow, no kidding! Then immediate acceptance. With MFSLs it's a different situation entirely where I have checked and re-checked and re-checked and... Even blind-folded with my Mrs. trying to fool me.

Did you miss Gary's post? He's not the only one. You never will see it, you have to hear it. David, my offer goes out to everyone including you. Come on by with both if you can make your way here and we'll spend a few hours together listening. PM me, my door's always open. James Randi's challenge has no way to win the money. If there were, I would be there for certain.

 

I just have to say something here; I don't even know why.

Forum members:

If someone claims they hear something, you are never going to be able to convince them to "unhear" it, unless they are some kind of lemming.

I gotta say this whole argument sounds a lot like "Who's got a bigger pecker?" talk. Who cares? I'm as passionate about this stuff as anyone, but I realize when I've talked myself to death and no null test is going to convince anybody of anything, regardless of whether I have air-tight proof to back me up or not. At some point, you just have to give up.

The only thing I'm curious about is: Has anyone ever taken two Ultradiscs or Ultradisc IIs or any two identical discs and compared THEM? There are some folks that say NO two discs sound the same, no matter if it's a gold disc or whatever. Anyone?

 

The computer isn't listening to anything! It is looking at the data on the disc. Imagine if the data was not encoded onto the CD but instead was printed on paper, just a long list of 1's and 0's. If you compared the output of a UD1 and a UD2 in this way, and confirmed with your own eyes that the pages were identical, would you then claim that your eyes were not able to "distinguish as well as the human ear for some inherent reason", and that the pages were actually different?

 

One could still claim that the 0's and 1's look sharper and clearer, with more detail on paper I than on paper II.

 

I can't believe this debate is still ongoing. Personally, I fall into the camp of Mr. Goodwin and others. However, I accept that there are those 'audiophiles' among us that will never believe what I and the entire scientific and medical community hold as fact based on actual, scientifically sound data that has been established over years and years and years. They will always be inclined to accept conjecture, coincidence, and anecdotal testimonials as 'proof', rather than accept anything objective.

You know what? Who cares? Lost in this whole debate is the fact that, ultimately, it is the music and the pleasure that it brings that matters. So what if someone gets more pleasure from their listening experience by grand rituals and 'golden-ears' training? So be it. I don't care if someone feels they hear more 'detail' from their system if they jump up and down on one leg blindfolded with one hand on their remote control, the other on their album sleeve of Pet Sounds and their speaker cables run through their belt loops covered in aluminium while a tub of water sits on top of each speaker. As long as we can all enjoy the music and the sound, that's all that matters - at least in the context of these forums.

Debating this is akin to debating the existence of God.

 

I want to to clarify one thing.

It is entirely possible that UD discs (pressed in Japan) create a different sound on certain systems than the corresponding UDII disc, even if they are digitally identical. The reason is that we can't examine the physical distribution of the data over the disc. Theoretically, it should not make any difference, since any (however miniscule) mastering differences would have resulted in a digital difference. But in practice, we don't really know how CD players react to different pressing methods. Jitter is one element that is not measurable when analyzing a CD, but it could affect the sound.

So we don't claim that it is impossible that a UD sounds different than a digitally identical UDII.

What makes this debate a bit heated is that Dave and others claim that mastering differences could be missed by computer programs. And that shows a lack of understanding digital audio. Is it a big deal? Probably no, because we all have made mistakes at times. But many posters have pointed that out, this time and in the past. Dave is an authority with a big and well-sorted collection, and I trust him when it comes to sound quality questions. But in this particular subject, I wish he would acknowledge his error.

Again, I am not saying that people are wrong when they hear a difference between a UD disc and a UDII disc. All I am saying is that any audible difference must come from jitter or a similar phenomenon that has to do with the physical characteristics of the disc. If the discs are digitally identical (which is provable), then mastering differences are not possible.

 

Jitter can be measured with a CDA2000 analyser, or if you want to try this at home, get a Plextor Premium drive for your computer a get the Plextools software. But - this won't necessarily tell you if the same jitter occurs on your CD player.

Otherwise, excellent post Andreas; the penultimate paragraph especially hits the nail on the head with regard to this kind of thread.

 

Exactly. As I said before, I've never verified whether or not there is any difference between a UDI or a UDII of ANY MFSL title. However, it seems to me that relying on a computer program that can ONLY tell you if the 1s and 0s are the same is suspect. For example, is

11 0 111 01 00 1101101 0 1 0101 equivalent to
1 10 1 1101 00 11 011 010 10101 ????

In terms of computer science, the answer is, "It depends on what the data represents, and how it is read." In the case of a purely binary system of single bits which are being considered separate and apart from anything else, the two series of numbers are identical. If individual groups of numbers are treated as words, the two series of numbers represent two different things, even though they contain the same bits and the same number of words.

In this situation (digital bits that are being converted to an analog waveform), we can assume that we are dealing with a binary system of single bits, but no one really knows if the positioning of the bits is significant or not. I (and many in both the scientific and recording communities) feel that it probably is, although there is not yet any 100%conclusive scientific evidence either way. It is possible that it makes no difference at all, and it is also possible that the repositioning of a single bit by a few thousandths of a micron could alter the derived analog signal in a way that is undetectible to present equipment. When talking about CDS you have to consider two things: (1) As DVD audio has demonstrated, 16 bits is not enough to reproduce the original analog signal in the first place. (2) The derived audio signal is only an approximation of the original signal ANY time you enter the digital domain.

The truth is, our understanding of analog/digital conversion is still in infancy. Therefore, we really don't know what these computer programs that analyze the 1s and 0s are telling us.
For example, let's assume that 11 0 111 01 00 1101101 0 1 0101 and 1 10 1 1101 00 11 011 010 10101 are both digital representations of an elephant. I DO trust present computer programs enough to believe them when they tell me each of these represents an elephant. However, I doubt any program in existence today is capable of determining if each of these series of numbers represents EXACTLY the SAME elephant because (1) We have no way of knowing, at this point in time, how well our digital approximation of an elephant represents a REAL elephant, or if our conversion algorithms are as efficient as they need to be - or even if our conversion algorithms are CORRECT, or if the equipment we are using to extract the converted data is doing its job correctly (and based on past experience, to believe that suddenly we have reached the all-time and forever pinacle of digital sound reproduction is absurd. And, (2) Any results obtained from ANY program are open to interpretation, even if the interpretation centers only on whether or not the result set is valid - because it might not be.

Therefore, to point to results obtained from what in a few years will probably be considered an archaic program that produced primitive results (at best) as conclusive evidence of ANYTHING isn't very scientific. It's just another way for one person to try to strengthen their position over another. It is no more valid than another person saying "but I hear a difference."

The question is the same as it relates to null tests. These tests may be valid when performed on analog source material, but we still have no gaurantee that, AT PRESENT, any null test we have today is 100% valid in the digital domain - because we have no way to know if we are TRULY measuring the results correctly, or if listening is involved, if our hearing is acute enough to determine if any difference related to bit position in a null test conducted on digital source material is relevant. It might show up only as small bits of white noise, or be otherwise undetectable.

Sorry, but until these questions are answered BY THE SCIENTIFIC COMMUNIT, and each of our present tests is analized and tested in a scientific manner based on FULL knowledge of the digital problem set (which we do not yet have), pointing to ANY "scientific" test of digital source material as conclusive proof of ANYTHING is just another brand of SNAKE OIL. There is no proven validity or reliabilty. We don't even know if we're testing the right things.

 

That is the most ridiculous argument I have ever heard. If a null test is performed and shows that the significant (i.e. 'audible') digital data is identical there is no debate. Saying otherwise is just a refusal to accept scientifically, and mathematically proven fact. It has nothing to do with stregthening one's position. There is only one position in this argument that has any science behind it and speculation about where a pit is placed on a CD is not it. Period. That is an argument to justify a position that has little or no scientific or mathematical legitimacy. One that is based purely in speculation and pop-science theory. I'm sorry, but that is the truth whether you choose to believe it or not.

I could speculate that maybe if the moon really were made of cheese that the oceans would be pink instead of blue, but that doesn't it make it so.

 

What GoldenBoy said

 

Since opinions on this matter apparently differ (and it doesn't look like either side can be "convinced" of the opposite), I would like you to vote:

http://www.stevehoffman.tv/forums/showthread.php?t=56628

Have fun! I am VERY CURIOUS about the results.

Roland

 

I was looking for an article which I can't find on ABX audio testing. It was a fairly long time ago, around the same time copying CD's gained popularity. It involved record companies conducting the tests to confirm their fears. I'm doing this from memory, but it stated how the first groups of tests were invalid. It generally talked about chances, odds, environmental factors when doing blind comparisons of identical and/or slightly altered sources. Even knowing the person that was doing the "switching" has an effect in that you have a deeper knowledge of how that person thinks. The point was that an ABCX (C and B are identical CDR's) were needed to reduce "chances, odds and luck" and other factors. When a third CD was introduced into the equation the tests revealed that the listeners picked the "right" disc 30 percent of the time. Which statistically meant there was no difference. Just a little info for those conducting tests.

 

C'mon guys. IMHO, a discussion is entered into in the hope of possibly learning new things or changing your opinion. Has anyone changed their opinion? It seems not. Lots of sniping at each other from the trenches. Let's agree to differ and move on.
Blind Faith MoFi (UD1 or UD2 - I have a UD2) is much better than the Deluxe Edition or the original release. Listen to Ginger's tom drum dynamics in Do What You Like - squashed on the DE, there in full sound on the MoFi.

 

Which is exactly what you're doing, in effect, since your test depends on knowing what aspects of the digital data are "significant (i.e. 'audible')" and our understandings of human hearing, digital conversion, et. al, are so primitive at present that there is no way you can possibly have that information unless it was handed to you on a stone tablet by the almighty. There is, at present, no way you can scientifically demonstrate that pit position doesn't matter - it hasn't been proven either way, and there are extensive (complicated scientific) arguments on both sides. (Do you ever read any of this stuff?) Your entire theory only holds up if the bits on the disc exist in a vaccuum, which they do not.

From the purely scientific viewpoint you claim to operate from, neither side of this argument has a scientific leg to stand on. If you can't see that, and continue to insist that my (perfectly rational) argument is ridiculous or absurd (reductio ad absurdium is NOT a scientific argument), then there is little point in debating this subject with you ever again.

And THIS response comes from someone who is willing to admit that YOUR argument MAY be correct, even if I do not think it is possible for you to prove it at this point in human history. Maintaining the type of attitude you seem to be presenting here accomplishes nothing. It does not encourage healthy debate, it doesn't prove anything, and can only serve to alienate potential members of your OWN choir.

 

Whatever. If you refuse to accept scientific data, then so be it. There's no reasoning with types like you, because any proof offered up only draws out an argument about the mystical, indefinable human abilities drawn out of the ether that may or may not exist. You ask for these abilities to be disproven by science, but science does not exist to 'disprove' anything.

For your information, because I despise it when people attribute comments or opinions to me that I have never put forward, I have never been unwilling to accept that people are hearing differences. I just refuse to simply accept that there is an actual difference as opposed to a perceived difference just because someone with 'golden ears' who has no real understanding of how digital audio works says it's so.

BTW, I do read many scientific papers on the subject, having studied audio and acoustic theory formally. Unfortunately, the rules on these forums prohibit one from delving too far into the subject.

 

...and now, I take my leave from this thread, because it has become pointless and tiresome.

 

I am very surprised how this thread has gone.... I understand quite clearly, and I do not need to be convinced, that two identical WAV files cannot cannot have different mastering. Being identical means that the digital information is the same. The mastering MUST be the same.

But the post above is the only one that addresses a point that should be clear to all. What you see when a WAV file is extracted is the theoretical sound that a CD COULD make in a perfect world (with perfect DAC's). The sound you hear when you play a CD through your speakers will be different specifically for the reasons listed in the post above. It is not understood perfectly how different aspects of a disk (pressing quality, make up of the disk, scratches, etc) would color the sound, but it is quite logical that it would have some effect. The pitfall is to think of a CD as a digital only animal. But we cannot hear 1's and 0's, so analog HAS to come into play when we listen to a CD. That is why even when the digital files on a UDI and UDII are exactly the same, we cannot say with 100% certainty that they will sound exactly the same. We cannot even say with certainty that the same UDI CD will sound the same every time out. Granted these differences may be slight or imperceptible, but they are theoretically there.

 

That is part of the point I was trying to make. JUST comparing a bunch of 1s and 0s, and ignoring everything else in the chain is like a bind man feeling the elephants' trunk and saying that the elephant is like a snake. He might be giving a good description of the trunk, but he's ignoring the rest of the elephant. Not very scientific, IMO. In order to prove conclusively that two digital files should sound exactly the same on all systems, you would have to prove:

1) That both files contain IDENTICAL digital information (1s and 0s). Relatively simple today.
2) That the discs that carriy each file are 100% identical (a practical impossibility), or, barring that, you would have to prove that minor variations in disc thickness, pit location, pit depth, etc. are mathematically irrelevant, i.e, you would have to prove that your basic error-correction algorithms are perfect IN A PERFECT WORLD. Proving that they are "good enough" doesn't cut it. Not simple, and probably impossible.
3) If you were able to prove that your basic error-correction algorithms are perfect IN A PERFECT WORLD, you would have to prove that the implementation is perfect in the DACs/players you are using for the study, i.e., that BOTH discs you are testing ALWAYS produce the EXACT same analog waveform under all conditions in all players you are using. Unlikely, since given the current state of manufacturing there isn't even a gaurantee that a SINGLE disc will sound EXACTLY the same on two "identical" players from the same manufacturer.

Proving (1) is easy. however, since proving (2) or (3) is clearly impossible given our present state of technology, if all you have to start with is (1), in order to prove that two digital files should sound exactly the same on all systems you would have to prove that (2) and (3) are irrelevant, i.e., that from a psych-acoustic point of view, NO ONE on the planet can hear any difference if (1) is true. Also impossible.

Therefore, I assert that (1) in and of itself is not proof of anything.

On the other side of the coin, in order to prove that (1) is NOT relevant in and of itself, you would have to prove that (2) and (3) are clearly relevant. This can't be done for the same reasons that proponents of the "(1) matters" theory can't prove that (2) and (3) AREN'T relevant.

The only other option would be to prove that least ONE person on the planet could correctly ALWAYS identify each disc correctly in valid listening and testing conditions. Since this has never been done either, I also assert that arguing the scientific merits of either side of this case are absurd at this point in human history, and that ANY argument advanced by EITHER side is strictly OPINION, from a scientific point of view.

However, (1) by itself should be adequate to indicate if two discs result from different MASTERINGS - although there is no gaurantee test (1) could determine if one disc came from the ORIGINAL master and the otrher came from a digital copy. So once again, (2) and (3) would enter into play in the case of UDIs/UDIIs, because it IS likely that at least SOME of the UDII vesions of UDI discs were derived from a different set of glass masters made from backup tapes.

 

I agreed with you until right at the end. You COULD create a CD "needle drop" of the UDI and UDII and compare the two WAV's that they would create. I think that would be a more definitive test than comparing the ripped WAV files. Of course it would be subject to coloring by the system that is used, but assuming that for the most part the system would color everything evenly, then that would give you a clue as to whether there are differences or not. Maybe not a definitive clue, but a better clue at least.