Study say No difference CD/SACD/DVD stereo streams

Anyone read the latest issue of AES? A year-long double blind study by Meyer and Moran with 554 people in the Boston Area (many from the Boston Audio Society) found that people were unable to discern when two channel digital streams were switched.

=> They can't identify high resolution audio from ordinary redbook CD stream!

This proves that either

1) CD is good enough for stereo playback
2) Boston audiophiles are not discerning

Will some people will say the test is meaningless as they can attest to having heard enormous differences themselves with their golden equipment and golden ears? If this is true then they should be able to prove this AES study and paper wrong in some sort of test...Meyer and Moran have thrown down the gauntlet to all those who claim to hear differences!

Note 1:
The authors readily admit that high resolution audio and the newer formats do have advantages in multi-channel and studio mixing applications. They also admit a lower noise floor is attainable with higher resolution formats (when volume is turned up very high and no music is playing)

Note 2:
This is the first test of this kind that I can recall. It may trigger lively debate for years to come. Obviously no audio equipment manufacturer will be pleased to see a result that claims 25 year-old technology is good enough!

I expect the audio rags to get hold of this and have a field day, imagine all those glowing reviewer recommendations and yet people can't even hear a difference.....lots of lively debate to come!

Remember you saw it first on A'gon.
7ac929ee be73 4853 8000 abc8d06023d4shadorne
The thing about redbook is, from first principles: it is an entirely appropriate/adequate format to capture the recorded event. 16 bit, 44.1kHz provides sufficient resoution to contain upwards of 95 dB dynamic range. We would be so lucky if cd's, sacd's, or dvd-a's had half that dynamic range on an average basis. It is not the process of encoding (digitizing and burning a cd layer) that is the issue. The music industry is listening to and responding with what the average human being wants: loud compressed formats. Unfortunately we are the extreme minority who want higher quality recordings.
I think only if we get the measured best (analyze the dynamic range) of redbook recordings compared to the best of pure DSD recordings would that test be more meaningful. If DSD/SACD happens to be better then perhaps it's in the post processing that manifests the performance benefit (ie the weak link of redbook is not the bits/sample rate)

I personally have several SACD's and lots more cd's. There are some SACD's that, well basically suck; otoh, there are cd's that sound about as good as my best sacd's (sheer presence/coherence and ease of sound) but are not avail on SACD. This seems to point the finger to recording technique as the limiting factor, not the "high res" format.
Unless God comes down from the heavens and charges the music industry leaders with the direction to fully realize the available dynamic range on redbook, we will remain where we are. It will probably get a whole lot worse (eg louder cd's) before there might ever be a return to high fidelity/wide dynamic range recordings.

A lot of people cannot tell the difference between a ten dollar bottle of wine and a two hundred dollar bottle in similar "studies". But many can and do. Can't make a sweeping generalization based on this, either.
It could prove that the test methodology is flawed.
Do you really expect anyone here to say that properly conducted tests have any significance? Do you honestly not expect many to say that digital is the devil's work and only analog can take you to heaven?
Can't make a sweeping generalization based on this

...if 554 people can't "hear" anything different doesn't that indicate something?

I bet that any ordinary person could taste a difference between two different wines (one at $200 and another at $20), although I grant you that they might not be able to say which was the $200 wine...but that is an additional level of detail that would require training.

The 554 people tested were not asked to identify what sounded best...just to discern a difference.
The thing about wine comparisons is that there is no "objective standard" by which to judge them--ie nothing equivalent to live music. And maybe wine is an acquired taste and lots of people just don't care after a certain price point...10-20 bucks. If audio IS the same as wine, we should all pack up our bags and go home. And surely, Boston Audio Society members care (about sound). Not to say the study isn't flawed, but not in this way.
Do you have a link to this study? Floyd Toole and Todd Welti of Harmon International have done many psycho acoustical studies, and the selection of subjects and a controlled environment are two very important aspects of their studies. I know a fair amount about their methods in doing subjective tests like this and they openly discuss the difficulties. I know nothing of Meyer and Moran, but it seems odd that if I read your (Meryer adn Moran's) conclusion correctly out of 554 people studied, none of them could discern a difference, or not a statistically significant number could hear a difference. Depending on how you selected your subjects, trained them, and the environment they were in that could be the case.

One possibility is that the DAC used could not take full advantage of the higher bit rate, the other possibility is that the RT time of the room was so long that any sound reproduced was muddied up in the room. A better control would have been to do the test with high resolution headphones--but again I have no idea how the test was conducted and what controls were used. If you have a link, I'm sure this would be a very interesting read.
I still love the time a number of years ago when a bunch of French wine judges got together in secret and decided to all vote for a French wine no matter what.

It was a blind test, but being experts they all agreed which wine was French and their favorite. Well, they blew it and all voted for a California wine. Their reaction was the same as when they messed with the Olympic skating. They all marked the wrong box by accident...(the same wrong box) and wanted to change their votes...
This proves that either

1) CD is good enough for stereo playback
2) Boston audiophiles are not discerning

3. or the test was in some way flawed

What player did they use? Whatever it was it would have to convert one of the formats to play them all. Wouldn't this affect the results?
Yes Herman,
It would be interesting to know all the details about how the test was set up.
Perhaps the results are valid, then again.......who knows? Ladies and gentlemen, the details please!
SACD is dead anyway.
I read that they used a $150 SACD player as their "reference." Anybody confirm or rebut this?
The test results don't surprise me nor does the "poor design" response. Why do you think that audio reviewers absolutely refuse to do blind tests or blind reviews? I think the current review system is far more flawed than this study may be.

The thing about wine comparisons is that there is no "objective standard" by which to judge them--ie nothing equivalent to live music.

Perhaps, but I've listened to a lot of live music and must say that it sounds like crap just as often or more than it sounds good. Poorly miked, bad room, WAY TOO LOUD, bad mix, outside noise, noisy crowd, etc. Where is the reference then? Yes, you can say that you know the sound of a certain instrument and the stereo does or does not reproduce it properly, but many things that audiophiles strive for at home doesn't exist in a live venue.
If you want details you can contact the writers at the Boston Audio Society or get a subscription to AES.

They did 554 tests (not 554 people) does not actually say if some people tried more than once to guess correctly. There were at least 60 individuals from teh Boston Audio Society and many many others. 256 trials wer ewith people aged 14 to 25 - the most likely age to have good hearing. They also included females in 48 trials. They used audiophile speakers including various pre and power amps and including electrostatics. A variety of venues were also used. They used a professional studio with professional monitors, a University professional studio with audio engineering students as well as "audiophile" home systems. People were permitted to use any genres they wanted to.

The researcher tried slice and dicing the data by female, male, age, and experience..all to no avail as no group could statistically tell when the 16 bit/44.1 KHZ A to D and D to A was in the loop.

Best result out of 554 trials was 8 out of 10 correct with one single user one single time....given the large number of test this can easily be explained by chance. There were two 7 out of 10 correct results. All in all it was fairly conclusive.

The basic setup was an SACD/DVD-A universal player either feeding analog directly to the speakers or else they added an A to D and D to A 16 bit/44.1 Khz processor in the loop between the analog signal and the various speakers.

I believe they used a Pioneer 563A DVD-A/SACD universal player which keeps the DSD output of the disc in DSD form without converting to PCM. The conversion to 16 bit and back was a Sony DTC-790 DAT machine (1996 consumer model).

The authors added that all DVD-A and SACD discs produced very high quality music....much much better than the average CD. However, these sounded just as good when passed through the 16 bit/44.1 Khz AtoD and DtoA loop.

Only discernable difference was a higher noise floor with the AtoD and Dto A in the loop....and this was only evident at much higher volumes with no music playing.

Music was played at around 85 DbA SPL and louder/softer in some cases.

Now you know nearly as much as I should really get the AES journal if you want more.

Both the testers are past Presidents of the Boston Audio Society and are obvioulsy passionate about audio.
Well, that's it then. There is no difference and it's all in my head. What a shock... fooling myself all this time. Well, at least I can recoup some cash by selling a bit of high-res gear. Look for it here before too many people read about this.

Thank God they didn't compare a vinyl stream with anything. No, no, if there's a DB study on that I don't want to know.

P.S. I know I won't be the only one to benefit from this study. Now the music biz can drop those money-losing high-res projects and concentrate on MP3s. I knew there was more to those than it appeared!
Well, I guess that's the nail in-the-coffin for SACD. Damn, I just started my collection yesterday. Didn't fare well with Beta years ago neither.
Do I understand that regardless of the particulars; things sounded the same when passed through a Sony?
To say that the processor was final to both signals and was more involved than even the presence of a passive "through"?
Do I understand that regardless of the particulars; things sounded the same when passed through a Sony?

The conclusion is that there was no discernable audible difference, yes.

To say that the processor was final to both signals and was more involved than even the presence of a passive "through"?

Yes the processor in SACD player (that took a DSD stream and converted to Analog) was always in the test - that is what read the High resolution SACD discs. The piece that was added or removed was the Sony DAT AtoD and DtoA 16 bit/44. KHz, between the SACD player and the preamp.

People could not tell when the Sony was in the loop or not in the loop.

The Sony DAT player was apparently selected because it has a good reputation for transparency and lack of any noticeable coloration.

Bear in mind tha this test does not negate the benefit of a high end player with a high end processor or with a tube output or some other does not imply that everything sounds the same by any means!!!

All it proves is that high resolution digital formats may not bring much in the way of audible improvement to two channel consumer stereo least not enough to be discernably audible to a fairly large community.

The possibility remains that some systems and some unusual listeners out there may hear something. If you can then it is time to step up to the plate and defend high resolution audio...devise some simple test to prove this AES paper wrong!!
Let's make that live acoustic music then. I should have been clearer. Pop concerts, any concerts miked over pa sound systems sound like sh*t to me.
This subject has been under discussion for well over a week on AA (prop head) and has included discussion by the authors of the paper. They state in this link - - that they did a single test with a high quality, but unnamed, player, and that the vast majority of tests were carried out with a Yamaha DVD-1500. The Yamaha is an inexpensive player selling at around $150 and cannot be expected to have the engineering quality of a good player, and especially not on all formats. At the very least, the tests should have been carried out in all cases with players competent enough to make any differences evident.
Thanks for the update, Flex. So this was a test which proved that in some situations many people can't hear a difference. It did not prove there is no audible difference in other situations.

What you prove depends almost entirely on what you want to prove and on how you do the measuring. If you want to prove there is a difference, and you believe there is one, consciously or unconsciously you set things up so that difference will stand out as much as possible. If you don't think there is a difference, or that it's very small, or you think few people will notice or care if there is one, you set things up another way. Like with a cheap multiformat player.

Your expectations determine what you are likely to discover.

Many of us have obtained results different from those of this study by using a different experimental setup: our listening rooms (or a good pair of headphones) and gear which was the best we could afford. The tests were perhaps not double blind ot even blind but IMHO their validity is not negligible.

Another spot where tests have proved to many of us that there is a difference between high-res and lower-res data streams is the demo room of an audio store.

Oh for heaven's sake. There is a difference, no there isn't, yes there is, no it's all in your head, no it isn't yes it is, you don't know how to listen, yer mom wears a jock strap, OK I'll devise a test to prove I'm right--what's the point except we love to argue and we'll fixate on most any topic?
Tobias, doesn't your assertion: *Your expectations determine what you are likely to discover.* work both ways?

I often hear this assertion when a study or test determines that people cannot hear differences in equipment. Doesn't it also hold true when you really want that big, beautiful, shiney, glowing piece of audio gear that everyone is raving about?
At the very least, the tests should have been carried out in all cases with players competent enough to make any differences evident

This is a great point!! Several people have raised the issue of cheap player and poor components invalidating things.

I would say this spells a huge opportunity for any first rate high-end SACD player manufacturer to prove to everyone that their player does just as Flex and Dopogue and others all suggest is likely!!!

If Moran and Meyer are wrong then it is really up to the industry to step up to the plate.

I would surely expect one of the many high end manufacturers to step forward ...they could even demand what other components be in the entire setup to ensure success, as the researchers have demonstrated they are fllexible. Of course it would take some time and a little money no doubt but think of the HUGE MARKETING opportunity to prove the worth of a high end player. Better still a non-result would at least give the manufacturer KUDOS for helping audiophile society in Boston and i the eyes of the community.

Am I being realistic here? Why wouldn't someone step forward?
"The thing about redbook is, from first principles: it is an entirely appropriate/adequate format to capture the recorded event..."

Hi Dpac, I'm quoting your comment because I have read somewhere that human auditory system has a 20 bits resolution, so in fact 16 bits is not the limit. Of course 20 bits of resolution is theoretical and absolute silence is a condition. In daily life we don't have absolute silence and those 4 bits are "drowned" into noise.

Just a correction to add. It seems that Meyer and Moran did two or perhaps even more tests. One in 2004 with the equipment I mentioned above ( details of which are on the Boston Audio Society website) and a new study, as referred to in teh recent AES journal.

B. Meyer says he used

The A/D/A link is not a secret; it was an HHB CDR 850, a very highly regarded pro CD recorder that some really fussy engineers have used as their main A/D for acoustic sessions.

Not the Sony device they used in an earlier trial in 2004.

It is confusing because they don't list the actual equipment used in the AES journal. If there is any criticism then this is a big one. Naturally with such a controversial conclusion everyone would like to know the nitty gritty.

I quote Meyer stating

In the meantime, we are planning to make available the details we left out for the purposes of clarity and brevity. That handout/email should be ready this coming week if anyone here is interested.


The above high-gain test was done first with an inexpensive Pioneer player, which was plenty quiet enough to reveal the difference; the test exposed a small but audible low-level nonlinearity in its left channel decoder. We tried a $2000 Sony player, which sounded clean, and wound up doing the great majority of our tests with a Yamaha DVD-1500.

If anyone gets the details please post it here as I am just as curious as anyone. I apologise if the previous details I posted are here are incorrect but the details were not given readily in the AES Journal.
Tomcy6, of course you are right, it does indeed work both ways.

A well-designed experiment defines expectations in the initial stages. As for the expectations which are not brought to consciousness and allowed for, as I understand it peer review is meant to deal with them.
Maybe audiophiles have an inflated view of a human's sense of hearing. I have never seen a zoologist of medical professional familiar with the subject weigh in on one of these debates, but I have always been under the impression that human's have less developed hearing in certain parts of the frequency range to optimize us for hunting large prey. Hope this doesn't offend any creationists...
The audience on Audiogon is definitely skewed. A while back, there was debate as to whether the new audio formats were as significant to sound quality as the new HDTV formats were to image quality. I didn't see any debate at all. Any guy off the street can discern a huge difference between HDTV and NTSC but it takes a pretty sophisticated listener to hear the difference between CD and the more hi-res audio formats. This study even calls into question the ability of sophisticated listeners to discern the difference.
I wonder why the difference in picture quality between video formats seems so obvious, but the difference in audio is so debatable.

Not that I am debating whether the difference exists; I have owned one and listend to a few SACD players and I have heard an appreciable difference. It just wasn't big enough, and I wasn't into digital enough, for me to want to replace my software collection.

It would be lovely to speculate on the difference between time-display media and space-display media ( they're all time-based though ). I could wander through what it takes to discern differences in detail over time compared with detail in space... but never mind. It would just be speculation.

As hard experience I offer that of my oldest friend. He got a big iPod, ripped all he could to it and used it as a primary source for about four months. He carried it with him and used earphones and he plugged it into his ( inexpensive ) home system using a Monster miniplug-to-twin RCA cable.

He was initially delighted with the system. Now, though, he tells me that it doesn't satisfy him any more. It doesn't sound as good as CD, and he wonders if ripping at a higher bit rate would make a difference.

( I said it wouldn't make the kind of difference he wanted, if what he wanted was Redbook resolution. )

So it took him--a non-audiophile whose system cost less than a grand at some place like Best Buy--four months of constant listening, but eventually he heard a considerable difference in resolution between two media.
Put an excellent CD in an outstanding CDP and put a poor SACD in a mediocre player, the CD will sound better.
My point being, we could manipulate the test method to provide any results we desire.
Unfortunately, we audiophiles are probably the only ones who care about this anyway.
There seems to be a lot of talk about how the study could be flawed due to the equipment used, the quality of the recording on the CD, and other factors, but these complaints seem to be pretty insignificant in evaluating the study. As Tobias hinted, there are differences that most people should be able to recognize right off the bat if they are truly significant differences.

For those that would argue it's a question of the resolution capability of the equipment, would you also argue that SACD's and DVD-A's are only for those wealthy enough to afford such equipment?

Dusty, thanks for reading my post. Perhaps I wasn't straightforward enough. I did not mean to hint that most people should be able to recognize significant differences right off the bat.

I meant to infer from my friend's experience that there are very significant differences between media of different resolution, even on mass-market quality systems, but that many or most non-audiophiles are not able to recognize these right off the bat. However, even for people in this group, the difference becomes evident after a certain period of exposure.

I also think it quite possible that a double-blind comparison using the same performance on 1) a well-recorded Redbook CD and 2) on an ordinary SACD, both played on the Yamaha unit cited above ( which I have not heard ) or on a mass-market CD/DVD player ( of which I have three )... might turn out inconclusive.

Perhaps an inconclusive result might be presumed to support the contention that there is little or no audible difference.

Still, the potential flaws in the test seem multiple: inability to identify source does not mean there is no audible difference ; short-term exposure is not a reliable indicator of long-term experience ; the home audio experience is not focused on telling differences between media but on something else ( enjoying the music ? ) ; mass-market players have flaws which affect their sonic output and which may not be allowed for in testing ; there are differences in the quality of recordings no matter what media are used.

However I agree that the last two may well be less important than the first three in evaluating the study.

Sorry to have gone on so long. Finally, I'm afraid I don't understand your final point, Dusty. How does personal wealth affect the outcome of this study?
Bottom line, most consumers could care less about SACD or DVD-A. Neither has anything to do with their lives. The cost difference between higher resolution media and standard CDs is not justified for them. The sonic differences don't matter, the CD is "good enough" for them. The $99 CDP is more than satisfactory for most. If the mass market doesn't adapt the SACD and DVD-A, and it doesn't look like they will, say bye-bye!
These formats are or will soon be relegated to a niche market, people who care about sound quality, "audiophiles". The "audiophile" market share will not be enough to sustain the viability of SACD or DVD-A thus they are both doomed to fail.
Plus, have you noticed an undercurrent of distrust? A lot of consumers feel that the higher resolution formats are just a conspiracy to FORCE everyone to repurchase their music yet another time.
None of this has anything to do with sound quality but rather it's a question of mass market acceptance. The convenience factor of CD over LP, not the audio differences although much touted at the time, was the driving force that guaranteed the success of the CD.
Those of us who love music and machinery to optimally reproduce it have different concerns. Unfortunately or sadly we do not drive the market.
The point I'm making in the end of my post is that (sticking with the same example) if you want to see HD television, you need a TV with the ability to show those extra lines of resolution, and the same goes for speakers and high definition sound; that is to say, the difference in quality from one media to the next will not be as apparent if the equipment cannot output the differential of quality from said media.

If the differential between redbook and SACD is such a small amount, or the resolution so high of the latter that only certain equipement (read: probably expensive) can output it, then SACD would be reserved for the wealthy who can afford such equipment.

Personal wealth wouldn't affect the outcome of the study, but if the outcome (i.e. no difference being heard) is attributed by some to the equipment being used, then the end result of that outcome is that only certain equipment would lend itself to SACD. That equipment being expensive.
I went to a Tweeter, when the Sony "777" 5 disc Sacd player came out. They were running it through a Adcom 5800 into SFaber book shelf speakers (had reasonable bass). They played a few discs, first in SACD then in CD. Their purpose was to sell SACD machines. It was on the same unit to play both medea. I heard a difference. I don't know that there couldn't have been a "fix" in, but I heard a big difference. i didn't buy as I thought my system at home sounded as good. Ignorance sticker applied, my amp was certainly nicer than the Adcom and I had nicer cables throughout. But i don't know what the Sony would have sounded out in my system. Full disclosure over. But as they presented it, for their purposes, there was a massive difference between formats on the same machine.

There can be differences between the CD audio and SACD audio on a disk - it depends on the care and objective of the manufacturer. There is a well known analysis of Pink Floyd DSOTM SACD that questions why the CD audio is harsh sounding, compressed and clipped and yet the SACD audio is fine. (Nobody knows why the Audio CD was made deliberately poorer in quality ....perhaps they were just lazy with that bit....some suspect a conspiracy to show that SACD is much better sounding)

In any case, to add to the discussion, here are more details on the gear used in the AES Journal tests Boston Audio Society Explanation

There is a list of excellent audiophile music at the bottom....
Fantastic job, Shadorne. This is very enlightening information, thank you very much for the link.
OK, So why would anyone who has spent the effort to develop superior sounding components bother with such an obviously meaningless test?
OK, So why would anyone who has spent the effort to develop superior sounding components bother with such an obviously meaningless test?

Ever heard of the Dr. Floyd E Toole...he is a great believer in ABX Double Blind testing of speakers...he is an engineer that designs and tests speakers (works for Harmon now). Tests might not be so meaningless if they help to provide guidence for meaningful design choices or where to spend your money for maximum value.
My 2 cents: I actually think this is a great recommendation for SACD. Redbook is a very mature technology, while SACD is still in its beginning stages. Redbook has improved greatly over the years. (Remember the days before jitter was understood?). If the big players commit to it, I'm confident SACD could become truly remarkable. The market will decide, of course, as it should. However, I think there are enough of us fanatics to keep a few labels involved in the medium to improve it over time. Hope so, anyway.
Happy listening.
Shadorne, I was referring to the particular study under discussion, which was obviously flawed. Just as obviously, Revels testing bears fruit.

The word you first used to describe the test was "meaningless" but now you say "flawed".

I suspect you have something important and meaningful to add but perhaps you could elaborate a little.
Thank you Shadorne. I will clarify. I feel the results of
Meyer/Moran tests are 'meaningless' -to me- insofar as increasing any knowledge or opinion I have related to the audibility of differences between Redbook CD sound and the sound of the other 'higher resolution' competing formats. I therefore feel that others who have examined the sound of the formats and reached their own conclusions are unlikely to change their opinions based on the results of the tests as I understand them. I did not have anything further to add to this topic; but, your encouragement to clarify my position has caused me to further reflect on the issue. Unfortunately I also feel the results of these tests do nothing to encourage those of us who would like to enjoy the best sound possible, if only for listening to favorite types of music, that manufacturers will be very interested in persuing such improved playback ability for the hitherto relatively ubiquitous CD. If members of specialty clubs-supposedly interested in 'superior' sound reproduction- are unable to design and perform a study to reveal differences in competing formats(or unwilling to put the effort required); is there really an adaquate market for superior playback devices? If I were investing-which by definition involves earning a profit- I would take this result into consideration. There you go, the other side of the coin; and now the result is no longer meaningless but is in fact-as my "smoker" buddy says 'another nail in the coffin'. Pity.
I have a collection of about 80 SACD's and listen to both RBCD's and SACD's frequently. The "sound" of SACD's is not strikingly different from CD's. I agree with the original posted observation, most people probably could not distinguish between them. The biggest difference I notice between formats is in sound-stage depth. A well-recorded SACD is WAY better in this regard. But most casual listeners never pay attention to sound stage. Most sound systems aren't set up to create a sound stage. In fact, I think most casual listeners find a holographic sound stage distracting. It forces an emersion in the experience of listening that people don't want. People like to have music on in the background while they work, talk with friends, do homework, exercise. These same people are very available to be emersed in the experience of watching TV however.
With the quality of programing on TV how could anyone help but be immersed in the experience? 8^)
Both items 1 and 2 could be true or both could be false. It really depends on the source material and how it was handled. Many of the SACDs appear to be mastered from the same source as CD. Not much care is used during the editing and mastering.

A good example of what SACD can sound like are the newer RCA Living Stereo SACDs that are being done by Sound Mirror; not by BMG. These discs are quite a bit better than their Living Stereo CD counterparts issued by BMG.

I routinely record live to 2 high res 2 track digital or high speed half track analog. All digital editing is done in the high res format. The last thing that is done is conversion to 16 bit digital for Red book CD. I can also author various Resolution levels of DVD-A to disc, from 24/96 all the way up to 24/192. While the 16 Bit CD version sounds very good, the higher res DVD-A formats do have more ambience and more of an analog sound.

To really do a fair comparison though, the 16 bit tracks need to be written to a DVD-A as 16 bit tracks, while the high res tracks are written as high res tracks. Now the same laser head is on the DVD-A player is reading both data streams and one additional variable is removed from the signal path.

And I will also say I too have been disappointed with the poor quality of some of the commercially released SACDs I have purchased.
Pascalini, perhaps you could tell us why the tests are flawed. Your logic appears to be that the tests are flawed because they failed to demonstrate an audible difference between the formats.
it really depends on the source material and how it was handled.

They made sure the source took DSD and converted to analog...the CD or 16 bit 44.1 KHz was an A to D and D to A loop that was added....the idea was to see if people could hear the degradation.

The researchers stated that people could hear the higher noise floor of the "CD loop" when the volume was turned way up but otherwise (with various music) they could not tell.

There is a list of the SACD they used on one of the links I gave...typical high quality audiophile stuff Chesky etc