Thread: Meyer Moran result debunked - again

Posts: 111
Page: prev 1 ... 5 6 7 8 9 ... 12 next

Post by audioholik October 1, 2010 (61 of 111)
Arnaldo said:

(2) Are new tests needed? Yes. And who is qualified to conduct these new tests? I'm certainly not, but neither are Meyer and Moran.

Pras and Guastavino, McGill University did a new test comparing 44.1 kHz vs. 88.2 kHz. The following is a direct citation from the abstract of their study:

"Sixteen expert listeners were asked to compare 3 versions (44.1kHz, 88.2kHz and the 88.2kHz version down-sampled to 44.1kHz) of 5 musical excerpts in a blind ABX task. Overall, participants were able to discriminate between files recorded at 88.2kHz and their 44.1kHz down-sampled version. Furthermore, for the orchestral excerpt, they were able to discriminate between files recorded at 88.2kHz and files recorded at 44.1kHz."

http://www.aes.org/e-lib/browse.cfm?elib=15398
Sampling Rate Discrimination: 44.1 kHz vs. 88.2 kHz (May 2010)

Post by DSD October 1, 2010 (62 of 111)
Torvi Chesky, ECM and Turtle recordings are PCM. Most newer Telarc SACDs are DSD recorded but not all, Telarc has also released many SACDs from 44.1kHz and 50kHz masters. If a Telarc SACD does not say PURE DSD it is not! I believe so far all Kimber Kable SACDs are DSD.

I wonder how accurate the scores were for actual real DSD recordings? If indeed any except Kimber Kable were used.

Post by audioholik October 1, 2010 (63 of 111)
torvi said:

What audioholic is trying to get across is that 44 and 88 kHz PCM can be told apart in a listening test, while 44 kHz PCM and DSD/SACD can not. RIght?

Nope, we can't make any definitive conclusions regarding 44.1kHz CD vs 2.8MHz DSD as the Pras/Guastavino study compared only 44.1 and 88.2kHz. We should remember however, that the SACD format has much better temporal resolution than lowly 88.2kHz pcm (used in the test), so the audible difference should be even more evident.

Post by audioholik October 1, 2010 (64 of 111)
torvi said:

Well, the test you quoted gave a positive result, but M&M test did not.

Why do you think this 44 versus 88 test is more valid than M&M test which was much much bigger?

The M&M test did not give a positive result because, as it was pointed out, the whole study was flawed.

Post by audioholik October 1, 2010 (65 of 111)
torvi said:

Also the part where new demo quality disk were used? How come there was no positive result even there?

It's simple - if an experiment is flawed, it produces flawed results, which then lead to wrong conclusions...

Post by audioholik October 1, 2010 (66 of 111)
torvi said:

What was the flaw when new demonstration quality disks were used?

The biggest flaw of the experiment was that no control experiment was carried out.

Post by TROLL-Buster October 1, 2010 (67 of 111)
torvi said:

What was the flaw when new demonstration quality disks were used?

By gosh you are stupid enough to be Petrus.
If the demo disc was of music of little interest to the listener plus they may be one of the poorer hi frequency listeners (AS WELL,) then interest wanes.They are very unlikely to be able to have a stored mental comparison of pure original sound.In some cases the demo material is just flashy and effective therefore It wouldnt prove much about general ultra hi frequencies.
The tests are on PEOPLE and prove variabilty of individual listening and observation skills.Absolutly sweet F All about much else.Select the better listeners and test them and you come up with totally different statistics.

Note for Cano.
I have plenty of high resolution digital recordings in excess of 30khz from way back early 1980s..
If you are going to make statements about digital recordings historically then I suggest you show some proof of your opinions.The Alan Parsons disc you refer to is also in this league.Pink Floyds bore is obviously not but I doubt if those who voted for it have ever heard PF actually Live.

Post by audioholik October 1, 2010 (68 of 111)
torvi said:

I am here to learn. What kind of control experiments you mean that they should have carried out?

16/44 vs 16/48

If a test subject scored x1 during the test and x2 during a control experiment we could test for a statistically significant difference between x1 and x2. Currently, the whole study is useless.

Post by audioholik October 1, 2010 (69 of 111)
torvi said:

What kind of science is that?

Applied statistics.

Post by david moran October 1, 2010 (70 of 111)
Clarification: My assumption about the McGill hi-rez 44/88 comparison paper was incorrect; an abstract and precis of it were reviewed by two readers before selection. So while it's not like a published AESJ paper, it's not an ordinary preprint either.

It will be interesting to see if the work gets published. There is quite savvy and informed discussion of it, including possible problems with its statistical methods, its hardware, and its 'inconclusivity', as well as author comment, at hydrogenaudio, a forum quite different from this one.

Page: prev 1 ... 5 6 7 8 9 ... 12 next

Closed