Gotcha, SY. Thanks for the cited text.
And, yes, the central point of burden of proof has not been lifted.
And, yes, the central point of burden of proof has not been lifted.
Also be aware that lots of people with a rather dim understanding of sensory testing use "ABX" incorrectly as a general term for an ears-only test, regardless of the format.
I've been down these roads. Just like I've been down the digital road. Remember, I have worked in the audio manufacturing industry since '87. Here is what I'm saying in my head right now: "I'm getting this preaching from a guy who thinks a master recording from a late 80's DAT tape is a great reference".
In the same sense that a raven is a writing desk. I think you're using words that you don't understand.
Is this some kind of appeal to authority i.e. your circle of recording engineer friends? Why are these random facts to be taken on face value, I never heard (knowingly) a late 80's DAT tape. This seems rather obscure.
20yr. or so ago Bruce Jackson founder of Apogee emailed me to tell me he built a custom mixing board for Fleetwood Mac's HBO unplugged special around the AD744. 8-legs good?
Robert is having a dig at the fact that SY (and I ) like the redbird CD, which was recorded on DAT which to Robert's super ears is a horribly flawed and awful sounding format. Of course Robert has no evidence for this, just his personal anecdotes. in 10 years he never bothered collecting this as it was so obvious to his ears. shrug. Redbird is still a darned good CD and would buy more from these sessions if they are ever released.
Aside: best hifi show demo I ever heard (I know not hard with the dross out there) was Dave Wilson playing DAT tapes. If that makes me as deaf as Dave then so be it. But it says nothing about DAT quality or Wilson speakers of course as a single data point.
Aside: best hifi show demo I ever heard (I know not hard with the dross out there) was Dave Wilson playing DAT tapes. If that makes me as deaf as Dave then so be it. But it says nothing about DAT quality or Wilson speakers of course as a single data point.
And this is what you call Nousaine's 'disassembly of Leventhal's argument'? He states "How about sensitivity? Les Leventhal makes his entire fairness case around the idea that subtle differences may only be present 60-80% of the time during the tests." Oh dear! He obviously doesn't understand what is being tested as this statement is a testament to. It's not that "differences may only be present 60-80% of the time during the tests" - it's that the differences may only be perceived 60% of the time due to the nature of auditory perception, blind testing, working memory, etc, etc. This major blunder, which he repeats throughout his 'disassembly' immediately highlights that he doesn't understand the basics of sensory testing.
It also repeats the usual idiocy seen in audio forums
"A casual survey of any of the underground magazines shows that audiophiles typically find it fairly easy to perceive differences. Leventhal implies that p may be a low value when there is nothing in the audiophile position to support such a notion. Read any decent "audiophile" review and draw your own conclusion as to the value of p inherent in their position."
The often seen statement - that audiophiles claim 'night & day differences so therefore differences should jump out & slap them across the face in a blind test'. Totally bogus & again so lacking in any understanding of psychoacoustics as to be unbelievable. He demonstrates such a lack of knowledge about sensory testing that I question you citing him?
I've seen Nousaine's tenuous grasp of experimental procedure before in his trying to prove how flawed sighted listening is compared to blind listening. He gave two identical boxes to audiophiles (one box had a certain amount of inbuilt distortion) for them to identify the boxes. He then brought some people together, trained them in the sound of the distortion & they did blind testing of the two boxes.
Do you think this test scenario shows that he has a grasp of experimental procedure?
Got a link to those experimental results which contradict the statistics presented in Leventhal's analysis?
Last edited:
I had always assumed that, to be elected regional VP of AES you generally needed a better grasp of the subject than a regular internet bloke. Clearly I was wrong. Sadly Tom is no longer around to defend himself.
Just goes to show. When you see him say this you really must wonder
"A low sensitivity value of, say, 0.6 for p suggests that for every 10 trials only 6 real trials occur. Thus one must increase the sample size to add enough real trials to avoid Type 2 error. A low-sensitivity test of 16 trials is only a 10-trial test under these conditions. If the differences are only present on 60% of all the program material available, and if your material is chosen from a random sample, then the sensitivity issue might apply."
I mean really? He couldn't be any more wrong! And even more telling is that he doesn't know he is so wrong or he would correct this jibberish - it's on his own site.
Indeed they don't, and if one doesn't understand what words like "statistics" means, then you can conclude that data and evidence regarding the topic at hand are not likely to be offered, nor will you be likely to have a useful and intelligent discussion.😀
You are saying he should have changed what he wrote but just couldn't as circumstances overtook him. You see the errors I highlighted.
That article on his website from which I quote was from 2005 & I believe he died in 2014. Was he very debilitated for 9 years?
I see you can use the bold button on this website. And that you can rant. Neither of which suggests you are more clued up than Tom was.
Last edited:
Who was it accused me of 'magical thinking' just a few days ago. Ain't nowt as queer as folk.
...this ABX/testing/statistics stuff is so incredibly off topic, could you guys just go get a room somewhere, or start a thread on this topic and keep that discussion IN that thread?? Just maybe, it's like a VIRUS, it seems to infect far too many otherwise useful topics.
- Status
- Not open for further replies.