|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
69.54.140.123
In Reply to: RE: I'm with you... posted by John Elison on October 01, 2016 at 18:44:51
In September 2007 the Audio Engineering Society published the results of a year-long trial, in which a range of subjects including professional recording engineers were asked to discern the difference between SACD and a compact disc audio (44.1 kHz/16 bit) conversion of the same source material under double blind test conditions. Out of 554 trials, there were 276 correct answers, a 49.8% success rate corresponding almost exactly to the 50% that would have been expected by chance guessing alone
Follow Ups:
The Meyer Moran tests have been widely criticized by thoughtful audio professionals on the basis of numerous biases and faults in the M&M methodology. There have also been other AES papers which differ considerably in their results. Meyer himself in response to valid criticism also said years later that his study had not really been a scientific one.
You could stop reading about any one else's critiques or other completely different tests which reached different conclusions, and just stick with M&M's results because they support your own mindset. Flawed tests happen all the time in many fields. Usually, that is not discovered until well after publication. It is a process called peer review. So, it is important to read and understand all of that, not just cherry pick the published result that you like.
It also happens years ago that I exchanged some correspondence with Moran. Yes, he attempted to defend the tests which bore his name, though I found his defense weak, myself. But, he told me he was also in the process of buying an SACD player for himself at the time. Apparently, he was not totally convinced of his own conclusions himself.
Do you actually know the methodology and layered assumptions behind the hilarious Rube Goldberg-esque Meyer and Moron study?
A far better way than using a $300 Pioneer player and content not established to truly be high resolution is to use the approach that recording engineers do: compare low and high resolution output taken from the same live source. A British study did just that and came to a different conclusion.
Pras - Guastavino AES study
Thanks for that. It's complicated reading, but worth it. Kind of goes with what I was saying about always attributing "different" with better, but it's really quite a dense article. It's got lots of different analysis, and findings. It's not just 1 story, like "Yes, tobacco causes cancer"
you can locate lots of in depth discussion of the methodology outside of this AA thread.Fundamentally, the study lacked any notion of controls. SACD content was chosen without any consideration as to whether or not it actually contained high resolution material. There were many systems used, much of it substandard including a $250 Pioneer universal player that JA found to have poor performance. They merely assumed that the ABX box was not blending the signals since most use common grounds. Why not just compare a live recording using two different resolutions as others have done?
There's a frequent poster here named Tony Lauck who makes high resolution recordings and happens to have been a college roommate with Meyer. He says while Meyer was a wealthy kid, he had crappy audio gear. :)
Edits: 10/04/16
It's also important to note that a lot of these "studies" have yet to be fully replicated. Usually in medicine for instance, 1 study is how things begin, then further studies build on, refute, or enhance it.
Saying 1 study proved or didn't prove X, Y or Z is usually difficult to back up. At best we can try to find multiple sources of information that overlap at some point or points we care about.
Like, how we decide what normal human temperature is. Every now and then that has to be revisited. :)
Another might be the Fletcher-Munsen curves. If not done already, it may be time for more research, to understand how they are holding up.
Best,
Erik
While the 1930s era research results are likely valid today for understanding the perception of various frequencies, I've always found compensation circuits to sound unnatural.
We perceive distance as a function of level. If I'm hearing an orchestra with a bass drum in row C, it will have greater level and impact than if I'm sitting in the back. Boosting the low end to *fix* the effect of distance has the effect (to me) of selectively moving parts of the orchestra forward while the rest of the players remain in place. Huh?
If you want more bass impact, get closer, i.e. turn up the level!
We don't always listen the same way. There are times when I'd like to listen at low volume but still hear it all.
Having those convenience features would be nice if they work well. I liked the "Loudness" knobs in the old Denon receivers a great deal.
Best,
Erik
Not to sidetrack the discussion, but the Fletcher-Munson stuff holds up pretty well from what I have seen. Other studies may disagree somewhat on the actual curves, but the concept has pretty well stood he test of time.
I never questions the principle of the curves, at all, but most of these studies have some sampling bias. I'm glad it's holding up!What I mean to question is how the study was done vs. the population you are trying to learn about. Do women have the same curves? Children? How about Hispanics vs. Latvians? Do their F-M curves look the same? Without testing I have no idea. :)
This is the sort of thing that isn't usually considered on the first studies, but we gain more knowledge of as time progresses. The first studies are usually done with whoever is at hand, and often in colleges that's young men and women of whatever the predominant genetic stock is there, so projecting their response to drugs, or hearing, onto the general population may or may not be accurate.
Best,
Erik
Edits: 10/04/16
Okay so I changed player software and got .dsf files working.For anyone unfamiliar with .dsf files they are basically a direct SACD DSD rip. To me this type of file sounds AT LEAST as good as the .flac files containing 16/44 sound.
Specifically I listened to "Pink Floyd-Dark Side of the Moon" and "Eagles-Hotel California". Compared to their CD sourced counterparts they seemed to sound more like the vinyl versions of each, minus the crackle, groove noise, inner groove loss of resolution, etc. That being said I could not hear a huge difference difference between the two and I could imagine easily failing an ABX test comparing them.
Looking at the SACD catalog I am disappointed to see that virtually all releases are classical, jazz, and 60s-80s pop/rock. My interests only include the third one and nothing from those time period has really great holographic imaging or inner detail. Its not like I'm going to find "Hellyeah", "System of a Down", or anything current on SACD. The fact remains that SACD did not replace CD because it failed to wow anybody in a way that CDs didn't. I know someone said that SACD is more aimed at audiophiles, which I will agree are few in number, but what about the vast number of tube amplifiers that are mostly only of interest to that group? I would argue that there are more makes and models of tube amplifiers in existence that there are albums that have been released in SACD or any other high resolution format for that matter. BluRay Audio? Really?
The fact remains that SACD did not replace CD because it failed to wow anybody in a way that CDs didn't.
It failed because it wasn't widely supported by the music industry and became a more expensive niche product.
I would argue that there are more makes and models of tube amplifiers in existence that there are albums that have been released in SACD or any other high resolution format for that matter.
And you would be mistaken. HD Tracks has 5890 titles on 24/96 alone . Acoustic Sounds offers 1635 SACD titles.
.
Of course, that may not matter if you don't mind paying hundreds of dollars per OOP title! ;-)
Wouldn't it be nice if all of those titles were downloadable?
I understand the whole kiss-your-elbow-rip-to-ISO thing, but it would be nice in today's computer based market if that content could be more widely available.
Even though I'm inclined to ascribe these results to chance (49.8% is TOO close) I'd really like ALL of the 1/2 who 'got it right' to take ANOTHER battery of tests.
Repeat the test with the OTHER half and AGAIN compare results.
If their is some way to tag each result with a 'confidence level', that would be even better if they were RIGHT.
If they can 'repeat' their original performance, even maybe 30% or 40% of 'em, I'd have to start 'believing'.
Too much is never enough
Well, if you believe the test is valid, it simply proves that hi-res sounds the same as low-res. In other words, it disproves your stipulation that low-res sounds better than hi-res because the test shows you really can't hear the difference. On the other hand, if you believe you can hear the difference, then you're simply wrong that low-res sounds better than hi-res because I can hear the difference, too; and I know hi-res sounds better! ;-)Good luck,
John Elison
Edits: 10/04/16
. . . half the listeners in the test got the results correct. So perhaps this result means that some listeners (approximately half) can't hear the difference between hi-rez and CD-rez, while others (also approximately half - in this case the other half) can. I'd like to hear a lot more details about that test before I would be comfortable concluding anything.
Having said all this, I've heard some VERY well recorded CD's out there! ;-)
I don't think anyone got the test results correct. The test showed that no listener had a success rate better than about 50%, which means nobody could prove they could hear the difference.
I'm afraid the article will require me to take a closer look, but as I read it, they go through a number of findings, not just 1 conclusion.
One that I thought was interesting was that trained listeners could hear a difference, but they guessed which was the high rez version incorrectly.
Best,
Erik
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: